{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9e-05,
  "eval_steps": 500,
  "global_step": 120000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00013107190604765774,
      "grad_norm": 1.6560418605804443,
      "learning_rate": 5.399999999999999e-06,
      "loss": 10.4207,
      "step": 10
    },
    {
      "epoch": 0.0002621438120953155,
      "grad_norm": 2.0435309410095215,
      "learning_rate": 1.14e-05,
      "loss": 10.3202,
      "step": 20
    },
    {
      "epoch": 0.00039321571814297324,
      "grad_norm": 1.6376142501831055,
      "learning_rate": 1.74e-05,
      "loss": 10.104,
      "step": 30
    },
    {
      "epoch": 0.000524287624190631,
      "grad_norm": 1.467596411705017,
      "learning_rate": 2.34e-05,
      "loss": 9.894,
      "step": 40
    },
    {
      "epoch": 0.0006553595302382887,
      "grad_norm": 1.453574538230896,
      "learning_rate": 2.94e-05,
      "loss": 9.7455,
      "step": 50
    },
    {
      "epoch": 0.0007864314362859465,
      "grad_norm": 1.4320564270019531,
      "learning_rate": 3.539999999999999e-05,
      "loss": 9.6131,
      "step": 60
    },
    {
      "epoch": 0.0009175033423336042,
      "grad_norm": 1.4820325374603271,
      "learning_rate": 4.14e-05,
      "loss": 9.4517,
      "step": 70
    },
    {
      "epoch": 0.001048575248381262,
      "grad_norm": 1.4840964078903198,
      "learning_rate": 4.7399999999999993e-05,
      "loss": 9.2702,
      "step": 80
    },
    {
      "epoch": 0.0011796471544289198,
      "grad_norm": 1.404575228691101,
      "learning_rate": 5.339999999999999e-05,
      "loss": 9.0683,
      "step": 90
    },
    {
      "epoch": 0.0013107190604765774,
      "grad_norm": 1.39352285861969,
      "learning_rate": 5.94e-05,
      "loss": 8.8432,
      "step": 100
    },
    {
      "epoch": 0.0014417909665242353,
      "grad_norm": 1.5529464483261108,
      "learning_rate": 6.539999999999999e-05,
      "loss": 8.5904,
      "step": 110
    },
    {
      "epoch": 0.001572862872571893,
      "grad_norm": 1.4607466459274292,
      "learning_rate": 7.139999999999999e-05,
      "loss": 8.3305,
      "step": 120
    },
    {
      "epoch": 0.0017039347786195506,
      "grad_norm": 1.279509425163269,
      "learning_rate": 7.74e-05,
      "loss": 8.1079,
      "step": 130
    },
    {
      "epoch": 0.0018350066846672085,
      "grad_norm": 1.2591545581817627,
      "learning_rate": 8.34e-05,
      "loss": 7.8698,
      "step": 140
    },
    {
      "epoch": 0.0019660785907148664,
      "grad_norm": 1.019737958908081,
      "learning_rate": 8.939999999999999e-05,
      "loss": 7.6731,
      "step": 150
    },
    {
      "epoch": 0.002097150496762524,
      "grad_norm": 0.8820710778236389,
      "learning_rate": 9.539999999999999e-05,
      "loss": 7.4882,
      "step": 160
    },
    {
      "epoch": 0.0022282224028101817,
      "grad_norm": 0.7616180181503296,
      "learning_rate": 0.0001014,
      "loss": 7.3478,
      "step": 170
    },
    {
      "epoch": 0.0023592943088578395,
      "grad_norm": 0.8487536907196045,
      "learning_rate": 0.00010739999999999998,
      "loss": 7.228,
      "step": 180
    },
    {
      "epoch": 0.002490366214905497,
      "grad_norm": 0.7819269299507141,
      "learning_rate": 0.00011339999999999999,
      "loss": 7.0934,
      "step": 190
    },
    {
      "epoch": 0.002621438120953155,
      "grad_norm": 0.8089203238487244,
      "learning_rate": 0.0001194,
      "loss": 7.0225,
      "step": 200
    },
    {
      "epoch": 0.0027525100270008127,
      "grad_norm": 0.6102705597877502,
      "learning_rate": 0.00012539999999999999,
      "loss": 6.9299,
      "step": 210
    },
    {
      "epoch": 0.0028835819330484706,
      "grad_norm": 0.8053275346755981,
      "learning_rate": 0.0001314,
      "loss": 6.8061,
      "step": 220
    },
    {
      "epoch": 0.003014653839096128,
      "grad_norm": 0.853755533695221,
      "learning_rate": 0.0001374,
      "loss": 6.7241,
      "step": 230
    },
    {
      "epoch": 0.003145725745143786,
      "grad_norm": 0.845130443572998,
      "learning_rate": 0.0001434,
      "loss": 6.6139,
      "step": 240
    },
    {
      "epoch": 0.0032767976511914438,
      "grad_norm": 0.9666972160339355,
      "learning_rate": 0.0001494,
      "loss": 6.5051,
      "step": 250
    },
    {
      "epoch": 0.0034078695572391012,
      "grad_norm": 0.8451218008995056,
      "learning_rate": 0.00015539999999999998,
      "loss": 6.429,
      "step": 260
    },
    {
      "epoch": 0.003538941463286759,
      "grad_norm": 1.2488940954208374,
      "learning_rate": 0.0001614,
      "loss": 6.3691,
      "step": 270
    },
    {
      "epoch": 0.003670013369334417,
      "grad_norm": 1.445700764656067,
      "learning_rate": 0.0001674,
      "loss": 6.3267,
      "step": 280
    },
    {
      "epoch": 0.0038010852753820744,
      "grad_norm": 0.8115171194076538,
      "learning_rate": 0.00017339999999999996,
      "loss": 6.2537,
      "step": 290
    },
    {
      "epoch": 0.003932157181429733,
      "grad_norm": 1.0047732591629028,
      "learning_rate": 0.00017939999999999997,
      "loss": 6.2275,
      "step": 300
    },
    {
      "epoch": 0.00406322908747739,
      "grad_norm": 0.8865781426429749,
      "learning_rate": 0.00018539999999999998,
      "loss": 6.1809,
      "step": 310
    },
    {
      "epoch": 0.004194300993525048,
      "grad_norm": 0.8600600361824036,
      "learning_rate": 0.0001914,
      "loss": 6.0634,
      "step": 320
    },
    {
      "epoch": 0.0043253728995727055,
      "grad_norm": 0.9079091548919678,
      "learning_rate": 0.0001974,
      "loss": 6.0005,
      "step": 330
    },
    {
      "epoch": 0.004456444805620363,
      "grad_norm": 0.9107650518417358,
      "learning_rate": 0.00020339999999999998,
      "loss": 5.9971,
      "step": 340
    },
    {
      "epoch": 0.004587516711668021,
      "grad_norm": 0.9232202768325806,
      "learning_rate": 0.00020939999999999997,
      "loss": 5.9575,
      "step": 350
    },
    {
      "epoch": 0.004718588617715679,
      "grad_norm": 1.1863384246826172,
      "learning_rate": 0.00021539999999999998,
      "loss": 5.905,
      "step": 360
    },
    {
      "epoch": 0.004849660523763337,
      "grad_norm": 1.2496181726455688,
      "learning_rate": 0.0002214,
      "loss": 5.8543,
      "step": 370
    },
    {
      "epoch": 0.004980732429810994,
      "grad_norm": 1.1276811361312866,
      "learning_rate": 0.00022739999999999997,
      "loss": 5.8504,
      "step": 380
    },
    {
      "epoch": 0.005111804335858652,
      "grad_norm": 0.952456533908844,
      "learning_rate": 0.00023339999999999998,
      "loss": 5.7879,
      "step": 390
    },
    {
      "epoch": 0.00524287624190631,
      "grad_norm": 1.126177430152893,
      "learning_rate": 0.0002394,
      "loss": 5.7604,
      "step": 400
    },
    {
      "epoch": 0.005373948147953968,
      "grad_norm": 0.8310980200767517,
      "learning_rate": 0.00024539999999999995,
      "loss": 5.7287,
      "step": 410
    },
    {
      "epoch": 0.0055050200540016254,
      "grad_norm": 0.9914957880973816,
      "learning_rate": 0.0002514,
      "loss": 5.6483,
      "step": 420
    },
    {
      "epoch": 0.005636091960049283,
      "grad_norm": 1.1346724033355713,
      "learning_rate": 0.00025739999999999997,
      "loss": 5.612,
      "step": 430
    },
    {
      "epoch": 0.005767163866096941,
      "grad_norm": 0.9475547671318054,
      "learning_rate": 0.00026339999999999995,
      "loss": 5.6072,
      "step": 440
    },
    {
      "epoch": 0.005898235772144598,
      "grad_norm": 1.1116561889648438,
      "learning_rate": 0.0002694,
      "loss": 5.5409,
      "step": 450
    },
    {
      "epoch": 0.006029307678192256,
      "grad_norm": 0.9825686812400818,
      "learning_rate": 0.00027539999999999997,
      "loss": 5.5056,
      "step": 460
    },
    {
      "epoch": 0.006160379584239914,
      "grad_norm": 1.2326000928878784,
      "learning_rate": 0.00028139999999999996,
      "loss": 5.4939,
      "step": 470
    },
    {
      "epoch": 0.006291451490287572,
      "grad_norm": 1.417091965675354,
      "learning_rate": 0.00028739999999999994,
      "loss": 5.4776,
      "step": 480
    },
    {
      "epoch": 0.00642252339633523,
      "grad_norm": 1.6037111282348633,
      "learning_rate": 0.0002934,
      "loss": 5.4743,
      "step": 490
    },
    {
      "epoch": 0.0065535953023828876,
      "grad_norm": 0.9632703065872192,
      "learning_rate": 0.00029939999999999996,
      "loss": 5.4157,
      "step": 500
    },
    {
      "epoch": 0.0065535953023828876,
      "eval_accuracy": 0.19562903225806452,
      "eval_loss": 5.403072357177734,
      "eval_runtime": 57.8092,
      "eval_samples_per_second": 34.597,
      "eval_steps_per_second": 0.554,
      "step": 500
    },
    {
      "epoch": 0.0066846672084305446,
      "grad_norm": 0.8948795199394226,
      "learning_rate": 0.00029999998956298354,
      "loss": 5.3457,
      "step": 510
    },
    {
      "epoch": 0.0068157391144782024,
      "grad_norm": 0.8844414949417114,
      "learning_rate": 0.00029999995348441004,
      "loss": 5.3317,
      "step": 520
    },
    {
      "epoch": 0.00694681102052586,
      "grad_norm": 0.9219761490821838,
      "learning_rate": 0.0002999998916354336,
      "loss": 5.3415,
      "step": 530
    },
    {
      "epoch": 0.007077882926573518,
      "grad_norm": 1.1252576112747192,
      "learning_rate": 0.0002999998040160649,
      "loss": 5.2779,
      "step": 540
    },
    {
      "epoch": 0.007208954832621176,
      "grad_norm": 0.9040408730506897,
      "learning_rate": 0.00029999969062631895,
      "loss": 5.2327,
      "step": 550
    },
    {
      "epoch": 0.007340026738668834,
      "grad_norm": 1.3966660499572754,
      "learning_rate": 0.0002999995514662152,
      "loss": 5.2914,
      "step": 560
    },
    {
      "epoch": 0.007471098644716492,
      "grad_norm": 1.0932937860488892,
      "learning_rate": 0.00029999938653577763,
      "loss": 5.2425,
      "step": 570
    },
    {
      "epoch": 0.007602170550764149,
      "grad_norm": 0.7051756381988525,
      "learning_rate": 0.0002999991958350346,
      "loss": 5.1788,
      "step": 580
    },
    {
      "epoch": 0.007733242456811807,
      "grad_norm": 0.8152607083320618,
      "learning_rate": 0.00029999897936401875,
      "loss": 5.1735,
      "step": 590
    },
    {
      "epoch": 0.007864314362859465,
      "grad_norm": 1.2139556407928467,
      "learning_rate": 0.0002999987371227674,
      "loss": 5.1026,
      "step": 600
    },
    {
      "epoch": 0.007995386268907122,
      "grad_norm": 0.8268054127693176,
      "learning_rate": 0.000299998469111322,
      "loss": 5.1691,
      "step": 610
    },
    {
      "epoch": 0.00812645817495478,
      "grad_norm": 0.9063141345977783,
      "learning_rate": 0.0002999981753297288,
      "loss": 5.1422,
      "step": 620
    },
    {
      "epoch": 0.008257530081002438,
      "grad_norm": 0.6071163415908813,
      "learning_rate": 0.00029999785577803815,
      "loss": 5.0912,
      "step": 630
    },
    {
      "epoch": 0.008388601987050095,
      "grad_norm": 0.8092164993286133,
      "learning_rate": 0.00029999751045630497,
      "loss": 5.0714,
      "step": 640
    },
    {
      "epoch": 0.008519673893097754,
      "grad_norm": 0.8723686933517456,
      "learning_rate": 0.00029999713936458864,
      "loss": 5.0811,
      "step": 650
    },
    {
      "epoch": 0.008650745799145411,
      "grad_norm": 0.8179002404212952,
      "learning_rate": 0.0002999967425029528,
      "loss": 5.0274,
      "step": 660
    },
    {
      "epoch": 0.00878181770519307,
      "grad_norm": 0.9187056422233582,
      "learning_rate": 0.0002999963198714657,
      "loss": 5.0565,
      "step": 670
    },
    {
      "epoch": 0.008912889611240727,
      "grad_norm": 0.783937394618988,
      "learning_rate": 0.00029999587147019993,
      "loss": 4.9994,
      "step": 680
    },
    {
      "epoch": 0.009043961517288384,
      "grad_norm": 0.9057349562644958,
      "learning_rate": 0.0002999953972992326,
      "loss": 4.9691,
      "step": 690
    },
    {
      "epoch": 0.009175033423336042,
      "grad_norm": 0.7294702529907227,
      "learning_rate": 0.0002999948973586451,
      "loss": 4.9699,
      "step": 700
    },
    {
      "epoch": 0.0093061053293837,
      "grad_norm": 0.8682928681373596,
      "learning_rate": 0.00029999437164852336,
      "loss": 4.9394,
      "step": 710
    },
    {
      "epoch": 0.009437177235431358,
      "grad_norm": 0.8431923389434814,
      "learning_rate": 0.0002999938201689576,
      "loss": 4.9238,
      "step": 720
    },
    {
      "epoch": 0.009568249141479015,
      "grad_norm": 0.7977835536003113,
      "learning_rate": 0.0002999932429200427,
      "loss": 4.8885,
      "step": 730
    },
    {
      "epoch": 0.009699321047526674,
      "grad_norm": 0.7827666401863098,
      "learning_rate": 0.0002999926399018778,
      "loss": 4.9131,
      "step": 740
    },
    {
      "epoch": 0.009830392953574331,
      "grad_norm": 0.6938887238502502,
      "learning_rate": 0.00029999201111456647,
      "loss": 4.8798,
      "step": 750
    },
    {
      "epoch": 0.009961464859621988,
      "grad_norm": 0.6832767128944397,
      "learning_rate": 0.00029999135655821665,
      "loss": 4.8498,
      "step": 760
    },
    {
      "epoch": 0.010092536765669647,
      "grad_norm": 0.7694419622421265,
      "learning_rate": 0.000299990676232941,
      "loss": 4.8329,
      "step": 770
    },
    {
      "epoch": 0.010223608671717304,
      "grad_norm": 0.7725641131401062,
      "learning_rate": 0.0002999899701388562,
      "loss": 4.881,
      "step": 780
    },
    {
      "epoch": 0.010354680577764962,
      "grad_norm": 0.7625282406806946,
      "learning_rate": 0.0002999892382760837,
      "loss": 4.8014,
      "step": 790
    },
    {
      "epoch": 0.01048575248381262,
      "grad_norm": 0.7446967959403992,
      "learning_rate": 0.0002999884806447492,
      "loss": 4.8109,
      "step": 800
    },
    {
      "epoch": 0.010616824389860278,
      "grad_norm": 0.6720506548881531,
      "learning_rate": 0.0002999876972449828,
      "loss": 4.8184,
      "step": 810
    },
    {
      "epoch": 0.010747896295907935,
      "grad_norm": 0.6993610858917236,
      "learning_rate": 0.0002999868880769192,
      "loss": 4.828,
      "step": 820
    },
    {
      "epoch": 0.010878968201955592,
      "grad_norm": 0.6835932731628418,
      "learning_rate": 0.0002999860531406973,
      "loss": 4.788,
      "step": 830
    },
    {
      "epoch": 0.011010040108003251,
      "grad_norm": 0.7491495013237,
      "learning_rate": 0.00029998519243646066,
      "loss": 4.7912,
      "step": 840
    },
    {
      "epoch": 0.011141112014050908,
      "grad_norm": 1.084183931350708,
      "learning_rate": 0.00029998430596435703,
      "loss": 4.8187,
      "step": 850
    },
    {
      "epoch": 0.011272183920098567,
      "grad_norm": 0.6859904527664185,
      "learning_rate": 0.0002999833937245388,
      "loss": 4.7582,
      "step": 860
    },
    {
      "epoch": 0.011403255826146224,
      "grad_norm": 0.6288307309150696,
      "learning_rate": 0.00029998245571716264,
      "loss": 4.7405,
      "step": 870
    },
    {
      "epoch": 0.011534327732193882,
      "grad_norm": 0.6571672558784485,
      "learning_rate": 0.00029998149194238976,
      "loss": 4.717,
      "step": 880
    },
    {
      "epoch": 0.01166539963824154,
      "grad_norm": 0.912996768951416,
      "learning_rate": 0.0002999805024003857,
      "loss": 4.7275,
      "step": 890
    },
    {
      "epoch": 0.011796471544289196,
      "grad_norm": 0.7243077158927917,
      "learning_rate": 0.0002999794870913204,
      "loss": 4.732,
      "step": 900
    },
    {
      "epoch": 0.011927543450336855,
      "grad_norm": 0.7728539109230042,
      "learning_rate": 0.00029997844601536847,
      "loss": 4.7086,
      "step": 910
    },
    {
      "epoch": 0.012058615356384512,
      "grad_norm": 0.8156994581222534,
      "learning_rate": 0.00029997737917270863,
      "loss": 4.702,
      "step": 920
    },
    {
      "epoch": 0.012189687262432171,
      "grad_norm": 1.1948306560516357,
      "learning_rate": 0.0002999762865635242,
      "loss": 4.7172,
      "step": 930
    },
    {
      "epoch": 0.012320759168479828,
      "grad_norm": 0.8796295523643494,
      "learning_rate": 0.00029997516818800287,
      "loss": 4.7105,
      "step": 940
    },
    {
      "epoch": 0.012451831074527487,
      "grad_norm": 0.9050677418708801,
      "learning_rate": 0.00029997402404633686,
      "loss": 4.7024,
      "step": 950
    },
    {
      "epoch": 0.012582902980575144,
      "grad_norm": 0.8355486392974854,
      "learning_rate": 0.0002999728541387226,
      "loss": 4.6435,
      "step": 960
    },
    {
      "epoch": 0.0127139748866228,
      "grad_norm": 0.7414559721946716,
      "learning_rate": 0.00029997165846536123,
      "loss": 4.6639,
      "step": 970
    },
    {
      "epoch": 0.01284504679267046,
      "grad_norm": 0.6685224175453186,
      "learning_rate": 0.0002999704370264581,
      "loss": 4.6355,
      "step": 980
    },
    {
      "epoch": 0.012976118698718116,
      "grad_norm": 0.6414394974708557,
      "learning_rate": 0.000299969189822223,
      "loss": 4.6286,
      "step": 990
    },
    {
      "epoch": 0.013107190604765775,
      "grad_norm": 0.7062374949455261,
      "learning_rate": 0.0002999679168528703,
      "loss": 4.5662,
      "step": 1000
    },
    {
      "epoch": 0.013107190604765775,
      "eval_accuracy": 0.25405767350928643,
      "eval_loss": 4.621243953704834,
      "eval_runtime": 58.8237,
      "eval_samples_per_second": 34.0,
      "eval_steps_per_second": 0.544,
      "step": 1000
    },
    {
      "epoch": 0.013238262510813432,
      "grad_norm": 0.8039721250534058,
      "learning_rate": 0.0002999666181186187,
      "loss": 4.6065,
      "step": 1010
    },
    {
      "epoch": 0.013369334416861089,
      "grad_norm": 0.8114719390869141,
      "learning_rate": 0.00029996529361969124,
      "loss": 4.5794,
      "step": 1020
    },
    {
      "epoch": 0.013500406322908748,
      "grad_norm": 0.6977509260177612,
      "learning_rate": 0.0002999639433563156,
      "loss": 4.5599,
      "step": 1030
    },
    {
      "epoch": 0.013631478228956405,
      "grad_norm": 0.627560019493103,
      "learning_rate": 0.0002999625673287236,
      "loss": 4.5624,
      "step": 1040
    },
    {
      "epoch": 0.013762550135004064,
      "grad_norm": 0.8633142113685608,
      "learning_rate": 0.0002999611655371517,
      "loss": 4.5373,
      "step": 1050
    },
    {
      "epoch": 0.01389362204105172,
      "grad_norm": 0.6539198160171509,
      "learning_rate": 0.0002999597379818408,
      "loss": 4.5184,
      "step": 1060
    },
    {
      "epoch": 0.01402469394709938,
      "grad_norm": 0.6654855012893677,
      "learning_rate": 0.00029995828466303616,
      "loss": 4.5056,
      "step": 1070
    },
    {
      "epoch": 0.014155765853147036,
      "grad_norm": 0.7834869623184204,
      "learning_rate": 0.00029995680558098743,
      "loss": 4.4808,
      "step": 1080
    },
    {
      "epoch": 0.014286837759194693,
      "grad_norm": 0.6177117824554443,
      "learning_rate": 0.00029995530073594865,
      "loss": 4.5236,
      "step": 1090
    },
    {
      "epoch": 0.014417909665242352,
      "grad_norm": 0.6060917973518372,
      "learning_rate": 0.0002999537701281784,
      "loss": 4.4809,
      "step": 1100
    },
    {
      "epoch": 0.014548981571290009,
      "grad_norm": 0.7016774415969849,
      "learning_rate": 0.00029995221375793966,
      "loss": 4.4836,
      "step": 1110
    },
    {
      "epoch": 0.014680053477337668,
      "grad_norm": 0.599354088306427,
      "learning_rate": 0.0002999506316254999,
      "loss": 4.4418,
      "step": 1120
    },
    {
      "epoch": 0.014811125383385325,
      "grad_norm": 0.6830163598060608,
      "learning_rate": 0.00029994902373113076,
      "loss": 4.471,
      "step": 1130
    },
    {
      "epoch": 0.014942197289432984,
      "grad_norm": 0.8255661725997925,
      "learning_rate": 0.0002999473900751086,
      "loss": 4.4555,
      "step": 1140
    },
    {
      "epoch": 0.01507326919548064,
      "grad_norm": 0.6182892322540283,
      "learning_rate": 0.0002999457306577141,
      "loss": 4.4326,
      "step": 1150
    },
    {
      "epoch": 0.015204341101528298,
      "grad_norm": 0.7591822743415833,
      "learning_rate": 0.0002999440454792323,
      "loss": 4.4681,
      "step": 1160
    },
    {
      "epoch": 0.015335413007575956,
      "grad_norm": 0.6211231350898743,
      "learning_rate": 0.00029994233453995274,
      "loss": 4.4271,
      "step": 1170
    },
    {
      "epoch": 0.015466484913623613,
      "grad_norm": 0.6285683512687683,
      "learning_rate": 0.0002999405978401693,
      "loss": 4.4079,
      "step": 1180
    },
    {
      "epoch": 0.015597556819671272,
      "grad_norm": 0.7309046983718872,
      "learning_rate": 0.0002999388353801805,
      "loss": 4.3955,
      "step": 1190
    },
    {
      "epoch": 0.01572862872571893,
      "grad_norm": 0.9915863871574402,
      "learning_rate": 0.0002999370471602889,
      "loss": 4.457,
      "step": 1200
    },
    {
      "epoch": 0.015859700631766588,
      "grad_norm": 0.6344852447509766,
      "learning_rate": 0.00029993523318080196,
      "loss": 4.3828,
      "step": 1210
    },
    {
      "epoch": 0.015990772537814245,
      "grad_norm": 0.8440383672714233,
      "learning_rate": 0.00029993339344203124,
      "loss": 4.4353,
      "step": 1220
    },
    {
      "epoch": 0.016121844443861902,
      "grad_norm": 0.6266127228736877,
      "learning_rate": 0.00029993152794429275,
      "loss": 4.3454,
      "step": 1230
    },
    {
      "epoch": 0.01625291634990956,
      "grad_norm": 0.7234435081481934,
      "learning_rate": 0.0002999296366879071,
      "loss": 4.3882,
      "step": 1240
    },
    {
      "epoch": 0.01638398825595722,
      "grad_norm": 0.7095479369163513,
      "learning_rate": 0.0002999277196731991,
      "loss": 4.3914,
      "step": 1250
    },
    {
      "epoch": 0.016515060162004876,
      "grad_norm": 0.6068082451820374,
      "learning_rate": 0.00029992577690049813,
      "loss": 4.3475,
      "step": 1260
    },
    {
      "epoch": 0.016646132068052533,
      "grad_norm": 0.6494437456130981,
      "learning_rate": 0.00029992380837013803,
      "loss": 4.3192,
      "step": 1270
    },
    {
      "epoch": 0.01677720397410019,
      "grad_norm": 0.745951235294342,
      "learning_rate": 0.0002999218140824569,
      "loss": 4.3327,
      "step": 1280
    },
    {
      "epoch": 0.01690827588014785,
      "grad_norm": 0.7726365923881531,
      "learning_rate": 0.0002999197940377975,
      "loss": 4.3504,
      "step": 1290
    },
    {
      "epoch": 0.017039347786195508,
      "grad_norm": 1.0547600984573364,
      "learning_rate": 0.00029991774823650674,
      "loss": 4.2528,
      "step": 1300
    },
    {
      "epoch": 0.017170419692243165,
      "grad_norm": 0.7189738750457764,
      "learning_rate": 0.0002999156766789361,
      "loss": 4.3318,
      "step": 1310
    },
    {
      "epoch": 0.017301491598290822,
      "grad_norm": 0.8451581597328186,
      "learning_rate": 0.00029991357936544165,
      "loss": 4.3366,
      "step": 1320
    },
    {
      "epoch": 0.01743256350433848,
      "grad_norm": 0.6054122447967529,
      "learning_rate": 0.00029991145629638347,
      "loss": 4.3053,
      "step": 1330
    },
    {
      "epoch": 0.01756363541038614,
      "grad_norm": 0.6151678562164307,
      "learning_rate": 0.0002999093074721265,
      "loss": 4.294,
      "step": 1340
    },
    {
      "epoch": 0.017694707316433796,
      "grad_norm": 0.765765368938446,
      "learning_rate": 0.0002999071328930398,
      "loss": 4.3105,
      "step": 1350
    },
    {
      "epoch": 0.017825779222481453,
      "grad_norm": 0.5758262276649475,
      "learning_rate": 0.00029990493255949705,
      "loss": 4.3056,
      "step": 1360
    },
    {
      "epoch": 0.01795685112852911,
      "grad_norm": 0.6724629402160645,
      "learning_rate": 0.0002999027064718763,
      "loss": 4.3166,
      "step": 1370
    },
    {
      "epoch": 0.018087923034576767,
      "grad_norm": 0.7258772253990173,
      "learning_rate": 0.00029990045463055984,
      "loss": 4.272,
      "step": 1380
    },
    {
      "epoch": 0.018218994940624428,
      "grad_norm": 0.6036768555641174,
      "learning_rate": 0.0002998981770359346,
      "loss": 4.243,
      "step": 1390
    },
    {
      "epoch": 0.018350066846672085,
      "grad_norm": 0.6966549158096313,
      "learning_rate": 0.00029989587368839203,
      "loss": 4.2308,
      "step": 1400
    },
    {
      "epoch": 0.018481138752719742,
      "grad_norm": 0.6581029295921326,
      "learning_rate": 0.0002998935445883277,
      "loss": 4.2566,
      "step": 1410
    },
    {
      "epoch": 0.0186122106587674,
      "grad_norm": 0.544323205947876,
      "learning_rate": 0.0002998911897361418,
      "loss": 4.2474,
      "step": 1420
    },
    {
      "epoch": 0.018743282564815056,
      "grad_norm": 1.222697138786316,
      "learning_rate": 0.00029988880913223886,
      "loss": 4.2386,
      "step": 1430
    },
    {
      "epoch": 0.018874354470862716,
      "grad_norm": 0.6340961456298828,
      "learning_rate": 0.0002998864027770279,
      "loss": 4.2244,
      "step": 1440
    },
    {
      "epoch": 0.019005426376910373,
      "grad_norm": 0.6256402134895325,
      "learning_rate": 0.00029988397067092236,
      "loss": 4.2268,
      "step": 1450
    },
    {
      "epoch": 0.01913649828295803,
      "grad_norm": 0.7098168730735779,
      "learning_rate": 0.00029988151281434014,
      "loss": 4.2329,
      "step": 1460
    },
    {
      "epoch": 0.019267570189005687,
      "grad_norm": 0.5391781330108643,
      "learning_rate": 0.0002998790292077033,
      "loss": 4.2303,
      "step": 1470
    },
    {
      "epoch": 0.019398642095053348,
      "grad_norm": 0.7200047969818115,
      "learning_rate": 0.0002998765198514388,
      "loss": 4.2195,
      "step": 1480
    },
    {
      "epoch": 0.019529714001101005,
      "grad_norm": 0.5863329172134399,
      "learning_rate": 0.0002998739847459775,
      "loss": 4.2144,
      "step": 1490
    },
    {
      "epoch": 0.019660785907148662,
      "grad_norm": 0.7362582087516785,
      "learning_rate": 0.00029987142389175514,
      "loss": 4.1845,
      "step": 1500
    },
    {
      "epoch": 0.019660785907148662,
      "eval_accuracy": 0.3005537634408602,
      "eval_loss": 4.20095682144165,
      "eval_runtime": 58.9457,
      "eval_samples_per_second": 33.93,
      "eval_steps_per_second": 0.543,
      "step": 1500
    },
    {
      "epoch": 0.01979185781319632,
      "grad_norm": 0.5994053483009338,
      "learning_rate": 0.00029986883728921155,
      "loss": 4.1936,
      "step": 1510
    },
    {
      "epoch": 0.019922929719243976,
      "grad_norm": 0.7042246460914612,
      "learning_rate": 0.00029986622493879115,
      "loss": 4.2199,
      "step": 1520
    },
    {
      "epoch": 0.020054001625291636,
      "grad_norm": 0.7159391045570374,
      "learning_rate": 0.0002998635868409428,
      "loss": 4.1633,
      "step": 1530
    },
    {
      "epoch": 0.020185073531339293,
      "grad_norm": 0.5796597003936768,
      "learning_rate": 0.00029986092299611974,
      "loss": 4.1542,
      "step": 1540
    },
    {
      "epoch": 0.02031614543738695,
      "grad_norm": 0.6809489130973816,
      "learning_rate": 0.00029985823340477946,
      "loss": 4.2332,
      "step": 1550
    },
    {
      "epoch": 0.020447217343434607,
      "grad_norm": 0.6476874947547913,
      "learning_rate": 0.00029985551806738426,
      "loss": 4.1559,
      "step": 1560
    },
    {
      "epoch": 0.020578289249482264,
      "grad_norm": 0.6709153652191162,
      "learning_rate": 0.0002998527769844005,
      "loss": 4.1671,
      "step": 1570
    },
    {
      "epoch": 0.020709361155529925,
      "grad_norm": 0.6842948198318481,
      "learning_rate": 0.0002998500101562992,
      "loss": 4.165,
      "step": 1580
    },
    {
      "epoch": 0.020840433061577582,
      "grad_norm": 0.7655808925628662,
      "learning_rate": 0.0002998472175835556,
      "loss": 4.1697,
      "step": 1590
    },
    {
      "epoch": 0.02097150496762524,
      "grad_norm": 0.6268697381019592,
      "learning_rate": 0.00029984439926664954,
      "loss": 4.1435,
      "step": 1600
    },
    {
      "epoch": 0.021102576873672896,
      "grad_norm": 0.5964300632476807,
      "learning_rate": 0.0002998415552060653,
      "loss": 4.1148,
      "step": 1610
    },
    {
      "epoch": 0.021233648779720556,
      "grad_norm": 0.6608432531356812,
      "learning_rate": 0.0002998386854022912,
      "loss": 4.1628,
      "step": 1620
    },
    {
      "epoch": 0.021364720685768213,
      "grad_norm": 0.6031919717788696,
      "learning_rate": 0.00029983578985582065,
      "loss": 4.1342,
      "step": 1630
    },
    {
      "epoch": 0.02149579259181587,
      "grad_norm": 0.5980697870254517,
      "learning_rate": 0.00029983286856715086,
      "loss": 4.135,
      "step": 1640
    },
    {
      "epoch": 0.021626864497863527,
      "grad_norm": 0.5951604247093201,
      "learning_rate": 0.0002998299215367839,
      "loss": 4.1217,
      "step": 1650
    },
    {
      "epoch": 0.021757936403911184,
      "grad_norm": 0.5948967337608337,
      "learning_rate": 0.00029982694876522583,
      "loss": 4.1134,
      "step": 1660
    },
    {
      "epoch": 0.021889008309958845,
      "grad_norm": 0.7004464268684387,
      "learning_rate": 0.00029982395025298763,
      "loss": 4.0987,
      "step": 1670
    },
    {
      "epoch": 0.022020080216006502,
      "grad_norm": 0.5953840017318726,
      "learning_rate": 0.00029982092600058433,
      "loss": 4.1089,
      "step": 1680
    },
    {
      "epoch": 0.02215115212205416,
      "grad_norm": 0.6414102911949158,
      "learning_rate": 0.0002998178760085355,
      "loss": 4.1247,
      "step": 1690
    },
    {
      "epoch": 0.022282224028101816,
      "grad_norm": 1.0331190824508667,
      "learning_rate": 0.0002998148002773652,
      "loss": 4.1159,
      "step": 1700
    },
    {
      "epoch": 0.022413295934149473,
      "grad_norm": 0.637819230556488,
      "learning_rate": 0.00029981169880760177,
      "loss": 4.0995,
      "step": 1710
    },
    {
      "epoch": 0.022544367840197133,
      "grad_norm": 0.5960279703140259,
      "learning_rate": 0.0002998085715997781,
      "loss": 4.142,
      "step": 1720
    },
    {
      "epoch": 0.02267543974624479,
      "grad_norm": 0.5699455738067627,
      "learning_rate": 0.00029980541865443143,
      "loss": 4.0956,
      "step": 1730
    },
    {
      "epoch": 0.022806511652292447,
      "grad_norm": 0.5006260275840759,
      "learning_rate": 0.00029980223997210355,
      "loss": 4.0827,
      "step": 1740
    },
    {
      "epoch": 0.022937583558340104,
      "grad_norm": 0.5959834456443787,
      "learning_rate": 0.00029979903555334034,
      "loss": 4.1421,
      "step": 1750
    },
    {
      "epoch": 0.023068655464387765,
      "grad_norm": 0.6544867157936096,
      "learning_rate": 0.00029979580539869256,
      "loss": 4.1301,
      "step": 1760
    },
    {
      "epoch": 0.023199727370435422,
      "grad_norm": 0.5601758360862732,
      "learning_rate": 0.00029979254950871507,
      "loss": 4.0581,
      "step": 1770
    },
    {
      "epoch": 0.02333079927648308,
      "grad_norm": 0.5569722652435303,
      "learning_rate": 0.0002997892678839672,
      "loss": 4.0894,
      "step": 1780
    },
    {
      "epoch": 0.023461871182530736,
      "grad_norm": 0.5775192975997925,
      "learning_rate": 0.0002997859605250128,
      "loss": 4.0453,
      "step": 1790
    },
    {
      "epoch": 0.023592943088578393,
      "grad_norm": 0.646247923374176,
      "learning_rate": 0.00029978262743241994,
      "loss": 4.0645,
      "step": 1800
    },
    {
      "epoch": 0.023724014994626053,
      "grad_norm": 0.6028469800949097,
      "learning_rate": 0.00029977926860676147,
      "loss": 4.0875,
      "step": 1810
    },
    {
      "epoch": 0.02385508690067371,
      "grad_norm": 0.6670252680778503,
      "learning_rate": 0.0002997758840486144,
      "loss": 4.0397,
      "step": 1820
    },
    {
      "epoch": 0.023986158806721367,
      "grad_norm": 0.545161247253418,
      "learning_rate": 0.00029977247375856007,
      "loss": 4.0583,
      "step": 1830
    },
    {
      "epoch": 0.024117230712769024,
      "grad_norm": 0.6030794978141785,
      "learning_rate": 0.0002997690377371845,
      "loss": 4.0268,
      "step": 1840
    },
    {
      "epoch": 0.02424830261881668,
      "grad_norm": 0.5466348528862,
      "learning_rate": 0.00029976557598507795,
      "loss": 4.0668,
      "step": 1850
    },
    {
      "epoch": 0.024379374524864342,
      "grad_norm": 0.5249160528182983,
      "learning_rate": 0.0002997620885028352,
      "loss": 4.0349,
      "step": 1860
    },
    {
      "epoch": 0.024510446430912,
      "grad_norm": 0.6327731609344482,
      "learning_rate": 0.00029975857529105537,
      "loss": 4.0491,
      "step": 1870
    },
    {
      "epoch": 0.024641518336959656,
      "grad_norm": 0.5931009650230408,
      "learning_rate": 0.0002997550363503421,
      "loss": 4.0192,
      "step": 1880
    },
    {
      "epoch": 0.024772590243007313,
      "grad_norm": 0.5834700465202332,
      "learning_rate": 0.00029975147168130335,
      "loss": 4.0235,
      "step": 1890
    },
    {
      "epoch": 0.024903662149054973,
      "grad_norm": 0.6450287699699402,
      "learning_rate": 0.00029974788128455147,
      "loss": 4.0416,
      "step": 1900
    },
    {
      "epoch": 0.02503473405510263,
      "grad_norm": 0.5849606990814209,
      "learning_rate": 0.0002997442651607034,
      "loss": 4.109,
      "step": 1910
    },
    {
      "epoch": 0.025165805961150287,
      "grad_norm": 0.5572938919067383,
      "learning_rate": 0.00029974062331038036,
      "loss": 4.0382,
      "step": 1920
    },
    {
      "epoch": 0.025296877867197944,
      "grad_norm": 0.6405125856399536,
      "learning_rate": 0.0002997369557342081,
      "loss": 4.068,
      "step": 1930
    },
    {
      "epoch": 0.0254279497732456,
      "grad_norm": 0.6164118647575378,
      "learning_rate": 0.0002997332624328166,
      "loss": 4.057,
      "step": 1940
    },
    {
      "epoch": 0.025559021679293262,
      "grad_norm": 0.639493465423584,
      "learning_rate": 0.0002997295434068404,
      "loss": 4.0102,
      "step": 1950
    },
    {
      "epoch": 0.02569009358534092,
      "grad_norm": 0.6389025449752808,
      "learning_rate": 0.0002997257986569185,
      "loss": 3.9931,
      "step": 1960
    },
    {
      "epoch": 0.025821165491388576,
      "grad_norm": 0.6529097557067871,
      "learning_rate": 0.00029972202818369424,
      "loss": 3.9955,
      "step": 1970
    },
    {
      "epoch": 0.025952237397436233,
      "grad_norm": 0.5967011451721191,
      "learning_rate": 0.00029971823198781544,
      "loss": 4.037,
      "step": 1980
    },
    {
      "epoch": 0.02608330930348389,
      "grad_norm": 0.5777824521064758,
      "learning_rate": 0.0002997144100699342,
      "loss": 4.0128,
      "step": 1990
    },
    {
      "epoch": 0.02621438120953155,
      "grad_norm": 0.5869829058647156,
      "learning_rate": 0.00029971056243070717,
      "loss": 3.9776,
      "step": 2000
    },
    {
      "epoch": 0.02621438120953155,
      "eval_accuracy": 0.32058993157380256,
      "eval_loss": 4.007896423339844,
      "eval_runtime": 58.0871,
      "eval_samples_per_second": 34.431,
      "eval_steps_per_second": 0.551,
      "step": 2000
    },
    {
      "epoch": 0.026345453115579207,
      "grad_norm": 0.5861641764640808,
      "learning_rate": 0.0002997066890707954,
      "loss": 4.024,
      "step": 2010
    },
    {
      "epoch": 0.026476525021626864,
      "grad_norm": 0.5712520480155945,
      "learning_rate": 0.00029970278999086434,
      "loss": 3.9853,
      "step": 2020
    },
    {
      "epoch": 0.02660759692767452,
      "grad_norm": 0.5917181372642517,
      "learning_rate": 0.0002996988651915839,
      "loss": 4.0515,
      "step": 2030
    },
    {
      "epoch": 0.026738668833722178,
      "grad_norm": 0.5557841062545776,
      "learning_rate": 0.0002996949146736283,
      "loss": 3.9697,
      "step": 2040
    },
    {
      "epoch": 0.02686974073976984,
      "grad_norm": 0.5299800634384155,
      "learning_rate": 0.00029969093843767633,
      "loss": 3.968,
      "step": 2050
    },
    {
      "epoch": 0.027000812645817496,
      "grad_norm": 0.5593048930168152,
      "learning_rate": 0.00029968693648441103,
      "loss": 3.9931,
      "step": 2060
    },
    {
      "epoch": 0.027131884551865153,
      "grad_norm": 0.9013563394546509,
      "learning_rate": 0.00029968290881452,
      "loss": 3.9912,
      "step": 2070
    },
    {
      "epoch": 0.02726295645791281,
      "grad_norm": 0.5617152452468872,
      "learning_rate": 0.0002996788554286952,
      "loss": 3.9847,
      "step": 2080
    },
    {
      "epoch": 0.02739402836396047,
      "grad_norm": 0.5254120826721191,
      "learning_rate": 0.00029967477632763303,
      "loss": 4.0109,
      "step": 2090
    },
    {
      "epoch": 0.027525100270008127,
      "grad_norm": 0.8099215030670166,
      "learning_rate": 0.0002996706715120343,
      "loss": 3.9829,
      "step": 2100
    },
    {
      "epoch": 0.027656172176055784,
      "grad_norm": 0.8291321396827698,
      "learning_rate": 0.0002996665409826041,
      "loss": 4.0092,
      "step": 2110
    },
    {
      "epoch": 0.02778724408210344,
      "grad_norm": 0.5507436990737915,
      "learning_rate": 0.00029966238474005223,
      "loss": 3.973,
      "step": 2120
    },
    {
      "epoch": 0.027918315988151098,
      "grad_norm": 0.5797262191772461,
      "learning_rate": 0.00029965820278509266,
      "loss": 4.0038,
      "step": 2130
    },
    {
      "epoch": 0.02804938789419876,
      "grad_norm": 0.648042619228363,
      "learning_rate": 0.0002996539951184439,
      "loss": 3.9693,
      "step": 2140
    },
    {
      "epoch": 0.028180459800246416,
      "grad_norm": 0.5793509483337402,
      "learning_rate": 0.00029964976174082876,
      "loss": 3.9715,
      "step": 2150
    },
    {
      "epoch": 0.028311531706294073,
      "grad_norm": 0.5391641855239868,
      "learning_rate": 0.0002996455026529746,
      "loss": 3.9618,
      "step": 2160
    },
    {
      "epoch": 0.02844260361234173,
      "grad_norm": 0.6265619993209839,
      "learning_rate": 0.00029964121785561323,
      "loss": 4.0136,
      "step": 2170
    },
    {
      "epoch": 0.028573675518389387,
      "grad_norm": 0.5064778923988342,
      "learning_rate": 0.0002996369073494807,
      "loss": 3.9579,
      "step": 2180
    },
    {
      "epoch": 0.028704747424437047,
      "grad_norm": 0.5632731914520264,
      "learning_rate": 0.0002996325711353175,
      "loss": 3.9353,
      "step": 2190
    },
    {
      "epoch": 0.028835819330484704,
      "grad_norm": 0.5410996079444885,
      "learning_rate": 0.0002996282092138687,
      "loss": 3.9648,
      "step": 2200
    },
    {
      "epoch": 0.02896689123653236,
      "grad_norm": 0.6270562410354614,
      "learning_rate": 0.0002996238215858837,
      "loss": 3.9735,
      "step": 2210
    },
    {
      "epoch": 0.029097963142580018,
      "grad_norm": 0.6164698004722595,
      "learning_rate": 0.00029961940825211624,
      "loss": 3.9404,
      "step": 2220
    },
    {
      "epoch": 0.02922903504862768,
      "grad_norm": 0.6305201649665833,
      "learning_rate": 0.0002996149692133246,
      "loss": 3.9279,
      "step": 2230
    },
    {
      "epoch": 0.029360106954675336,
      "grad_norm": 0.5270501971244812,
      "learning_rate": 0.0002996105044702714,
      "loss": 3.9948,
      "step": 2240
    },
    {
      "epoch": 0.029491178860722993,
      "grad_norm": 0.6147605180740356,
      "learning_rate": 0.00029960601402372365,
      "loss": 3.9699,
      "step": 2250
    },
    {
      "epoch": 0.02962225076677065,
      "grad_norm": 0.5281668901443481,
      "learning_rate": 0.0002996014978744529,
      "loss": 3.907,
      "step": 2260
    },
    {
      "epoch": 0.029753322672818307,
      "grad_norm": 0.5528964400291443,
      "learning_rate": 0.000299596956023235,
      "loss": 3.9235,
      "step": 2270
    },
    {
      "epoch": 0.029884394578865967,
      "grad_norm": 0.6090617775917053,
      "learning_rate": 0.00029959238847085016,
      "loss": 3.962,
      "step": 2280
    },
    {
      "epoch": 0.030015466484913624,
      "grad_norm": 0.5461755990982056,
      "learning_rate": 0.0002995877952180833,
      "loss": 3.947,
      "step": 2290
    },
    {
      "epoch": 0.03014653839096128,
      "grad_norm": 0.6361566781997681,
      "learning_rate": 0.0002995831762657234,
      "loss": 3.9239,
      "step": 2300
    },
    {
      "epoch": 0.030277610297008938,
      "grad_norm": 0.6482072472572327,
      "learning_rate": 0.00029957853161456405,
      "loss": 3.9674,
      "step": 2310
    },
    {
      "epoch": 0.030408682203056595,
      "grad_norm": 0.7600134015083313,
      "learning_rate": 0.0002995738612654032,
      "loss": 3.9919,
      "step": 2320
    },
    {
      "epoch": 0.030539754109104256,
      "grad_norm": 0.5228272676467896,
      "learning_rate": 0.00029956916521904327,
      "loss": 3.9599,
      "step": 2330
    },
    {
      "epoch": 0.030670826015151913,
      "grad_norm": 0.5676136016845703,
      "learning_rate": 0.000299564443476291,
      "loss": 3.9015,
      "step": 2340
    },
    {
      "epoch": 0.03080189792119957,
      "grad_norm": 0.5721566081047058,
      "learning_rate": 0.00029955969603795765,
      "loss": 3.9654,
      "step": 2350
    },
    {
      "epoch": 0.030932969827247227,
      "grad_norm": 0.628048837184906,
      "learning_rate": 0.0002995549229048588,
      "loss": 4.0491,
      "step": 2360
    },
    {
      "epoch": 0.031064041733294887,
      "grad_norm": 0.6172999739646912,
      "learning_rate": 0.00029955012407781453,
      "loss": 3.9253,
      "step": 2370
    },
    {
      "epoch": 0.031195113639342544,
      "grad_norm": 0.6174818277359009,
      "learning_rate": 0.00029954529955764927,
      "loss": 3.9231,
      "step": 2380
    },
    {
      "epoch": 0.0313261855453902,
      "grad_norm": 0.6109817028045654,
      "learning_rate": 0.00029954044934519186,
      "loss": 3.9392,
      "step": 2390
    },
    {
      "epoch": 0.03145725745143786,
      "grad_norm": 0.5473556518554688,
      "learning_rate": 0.0002995355734412756,
      "loss": 3.8852,
      "step": 2400
    },
    {
      "epoch": 0.03158832935748552,
      "grad_norm": 0.7217162847518921,
      "learning_rate": 0.0002995306718467382,
      "loss": 3.8951,
      "step": 2410
    },
    {
      "epoch": 0.031719401263533176,
      "grad_norm": 0.682639479637146,
      "learning_rate": 0.00029952574456242173,
      "loss": 3.8864,
      "step": 2420
    },
    {
      "epoch": 0.03185047316958083,
      "grad_norm": 0.553237795829773,
      "learning_rate": 0.0002995207915891728,
      "loss": 3.8997,
      "step": 2430
    },
    {
      "epoch": 0.03198154507562849,
      "grad_norm": 0.5129098296165466,
      "learning_rate": 0.00029951581292784216,
      "loss": 3.888,
      "step": 2440
    },
    {
      "epoch": 0.03211261698167615,
      "grad_norm": 0.5928903818130493,
      "learning_rate": 0.00029951080857928535,
      "loss": 3.9103,
      "step": 2450
    },
    {
      "epoch": 0.032243688887723804,
      "grad_norm": 0.6485490202903748,
      "learning_rate": 0.00029950577854436205,
      "loss": 3.8896,
      "step": 2460
    },
    {
      "epoch": 0.03237476079377146,
      "grad_norm": 0.5979528427124023,
      "learning_rate": 0.00029950072282393646,
      "loss": 3.915,
      "step": 2470
    },
    {
      "epoch": 0.03250583269981912,
      "grad_norm": 0.6064060926437378,
      "learning_rate": 0.00029949564141887715,
      "loss": 3.8717,
      "step": 2480
    },
    {
      "epoch": 0.03263690460586678,
      "grad_norm": 0.6553047895431519,
      "learning_rate": 0.0002994905343300571,
      "loss": 3.9202,
      "step": 2490
    },
    {
      "epoch": 0.03276797651191444,
      "grad_norm": 0.525600790977478,
      "learning_rate": 0.0002994854015583537,
      "loss": 3.9043,
      "step": 2500
    },
    {
      "epoch": 0.03276797651191444,
      "eval_accuracy": 0.3318724340175953,
      "eval_loss": 3.897535800933838,
      "eval_runtime": 59.3923,
      "eval_samples_per_second": 33.674,
      "eval_steps_per_second": 0.539,
      "step": 2500
    },
    {
      "epoch": 0.032899048417962096,
      "grad_norm": 0.5005170106887817,
      "learning_rate": 0.00029948024310464885,
      "loss": 3.8796,
      "step": 2510
    },
    {
      "epoch": 0.03303012032400975,
      "grad_norm": 0.4958856403827667,
      "learning_rate": 0.00029947505896982875,
      "loss": 3.907,
      "step": 2520
    },
    {
      "epoch": 0.03316119223005741,
      "grad_norm": 0.6923332810401917,
      "learning_rate": 0.00029946984915478405,
      "loss": 3.94,
      "step": 2530
    },
    {
      "epoch": 0.03329226413610507,
      "grad_norm": 0.5524294376373291,
      "learning_rate": 0.0002994646136604098,
      "loss": 3.8984,
      "step": 2540
    },
    {
      "epoch": 0.033423336042152724,
      "grad_norm": 0.5918364524841309,
      "learning_rate": 0.00029945935248760553,
      "loss": 3.8516,
      "step": 2550
    },
    {
      "epoch": 0.03355440794820038,
      "grad_norm": 0.6486288905143738,
      "learning_rate": 0.000299454065637275,
      "loss": 3.8729,
      "step": 2560
    },
    {
      "epoch": 0.03368547985424804,
      "grad_norm": 0.5548475980758667,
      "learning_rate": 0.0002994487531103266,
      "loss": 3.8983,
      "step": 2570
    },
    {
      "epoch": 0.0338165517602957,
      "grad_norm": 0.5621161460876465,
      "learning_rate": 0.0002994434149076731,
      "loss": 3.9149,
      "step": 2580
    },
    {
      "epoch": 0.03394762366634336,
      "grad_norm": 0.5826847553253174,
      "learning_rate": 0.0002994380510302314,
      "loss": 3.8814,
      "step": 2590
    },
    {
      "epoch": 0.034078695572391016,
      "grad_norm": 0.5493956804275513,
      "learning_rate": 0.0002994326614789232,
      "loss": 3.8695,
      "step": 2600
    },
    {
      "epoch": 0.03420976747843867,
      "grad_norm": 0.5203653573989868,
      "learning_rate": 0.0002994272462546745,
      "loss": 3.8835,
      "step": 2610
    },
    {
      "epoch": 0.03434083938448633,
      "grad_norm": 0.5431103706359863,
      "learning_rate": 0.00029942180535841545,
      "loss": 3.8672,
      "step": 2620
    },
    {
      "epoch": 0.03447191129053399,
      "grad_norm": 0.5847226977348328,
      "learning_rate": 0.00029941633879108096,
      "loss": 3.8442,
      "step": 2630
    },
    {
      "epoch": 0.034602983196581644,
      "grad_norm": 0.5371077060699463,
      "learning_rate": 0.0002994108465536101,
      "loss": 3.8546,
      "step": 2640
    },
    {
      "epoch": 0.0347340551026293,
      "grad_norm": 0.5218356847763062,
      "learning_rate": 0.0002994053286469466,
      "loss": 3.9067,
      "step": 2650
    },
    {
      "epoch": 0.03486512700867696,
      "grad_norm": 0.5628876090049744,
      "learning_rate": 0.0002993997850720383,
      "loss": 3.8591,
      "step": 2660
    },
    {
      "epoch": 0.034996198914724615,
      "grad_norm": 0.6259694695472717,
      "learning_rate": 0.00029939421582983764,
      "loss": 3.8985,
      "step": 2670
    },
    {
      "epoch": 0.03512727082077228,
      "grad_norm": 0.5161134004592896,
      "learning_rate": 0.0002993886209213015,
      "loss": 3.92,
      "step": 2680
    },
    {
      "epoch": 0.035258342726819936,
      "grad_norm": 0.5075753927230835,
      "learning_rate": 0.000299383000347391,
      "loss": 3.8775,
      "step": 2690
    },
    {
      "epoch": 0.03538941463286759,
      "grad_norm": 0.5651357173919678,
      "learning_rate": 0.0002993773541090719,
      "loss": 3.8937,
      "step": 2700
    },
    {
      "epoch": 0.03552048653891525,
      "grad_norm": 0.5567218661308289,
      "learning_rate": 0.00029937168220731407,
      "loss": 3.8718,
      "step": 2710
    },
    {
      "epoch": 0.03565155844496291,
      "grad_norm": 0.5376849174499512,
      "learning_rate": 0.00029936598464309216,
      "loss": 3.8877,
      "step": 2720
    },
    {
      "epoch": 0.035782630351010564,
      "grad_norm": 0.569720447063446,
      "learning_rate": 0.00029936026141738484,
      "loss": 3.8273,
      "step": 2730
    },
    {
      "epoch": 0.03591370225705822,
      "grad_norm": 0.5735238194465637,
      "learning_rate": 0.0002993545125311754,
      "loss": 3.8489,
      "step": 2740
    },
    {
      "epoch": 0.03604477416310588,
      "grad_norm": 0.521174967288971,
      "learning_rate": 0.0002993487379854517,
      "loss": 3.8618,
      "step": 2750
    },
    {
      "epoch": 0.036175846069153535,
      "grad_norm": 0.5716639757156372,
      "learning_rate": 0.0002993429377812056,
      "loss": 3.8589,
      "step": 2760
    },
    {
      "epoch": 0.0363069179752012,
      "grad_norm": 0.5484517812728882,
      "learning_rate": 0.0002993371119194337,
      "loss": 3.8259,
      "step": 2770
    },
    {
      "epoch": 0.036437989881248856,
      "grad_norm": 0.649082362651825,
      "learning_rate": 0.00029933126040113686,
      "loss": 3.8595,
      "step": 2780
    },
    {
      "epoch": 0.03656906178729651,
      "grad_norm": 0.5224399566650391,
      "learning_rate": 0.0002993253832273204,
      "loss": 3.8707,
      "step": 2790
    },
    {
      "epoch": 0.03670013369334417,
      "grad_norm": 0.469137579202652,
      "learning_rate": 0.0002993194803989941,
      "loss": 3.8745,
      "step": 2800
    },
    {
      "epoch": 0.03683120559939183,
      "grad_norm": 0.5466957688331604,
      "learning_rate": 0.00029931355191717196,
      "loss": 3.8551,
      "step": 2810
    },
    {
      "epoch": 0.036962277505439484,
      "grad_norm": 0.5486308932304382,
      "learning_rate": 0.0002993075977828726,
      "loss": 3.8241,
      "step": 2820
    },
    {
      "epoch": 0.03709334941148714,
      "grad_norm": 0.5748544931411743,
      "learning_rate": 0.00029930161799711895,
      "loss": 3.8059,
      "step": 2830
    },
    {
      "epoch": 0.0372244213175348,
      "grad_norm": 0.5815683603286743,
      "learning_rate": 0.00029929561256093827,
      "loss": 3.8804,
      "step": 2840
    },
    {
      "epoch": 0.037355493223582455,
      "grad_norm": 0.5376504063606262,
      "learning_rate": 0.00029928958147536244,
      "loss": 3.8246,
      "step": 2850
    },
    {
      "epoch": 0.03748656512963011,
      "grad_norm": 0.5153408646583557,
      "learning_rate": 0.0002992835247414275,
      "loss": 3.865,
      "step": 2860
    },
    {
      "epoch": 0.037617637035677776,
      "grad_norm": 0.655957818031311,
      "learning_rate": 0.0002992774423601741,
      "loss": 3.8933,
      "step": 2870
    },
    {
      "epoch": 0.03774870894172543,
      "grad_norm": 0.521975576877594,
      "learning_rate": 0.00029927133433264717,
      "loss": 3.8837,
      "step": 2880
    },
    {
      "epoch": 0.03787978084777309,
      "grad_norm": 0.5097159743309021,
      "learning_rate": 0.00029926520065989604,
      "loss": 3.8121,
      "step": 2890
    },
    {
      "epoch": 0.03801085275382075,
      "grad_norm": 0.5221400856971741,
      "learning_rate": 0.00029925904134297463,
      "loss": 3.82,
      "step": 2900
    },
    {
      "epoch": 0.038141924659868404,
      "grad_norm": 0.6116294860839844,
      "learning_rate": 0.000299252856382941,
      "loss": 3.8599,
      "step": 2910
    },
    {
      "epoch": 0.03827299656591606,
      "grad_norm": 0.542933464050293,
      "learning_rate": 0.00029924664578085774,
      "loss": 3.7914,
      "step": 2920
    },
    {
      "epoch": 0.03840406847196372,
      "grad_norm": 0.5418369770050049,
      "learning_rate": 0.0002992404095377919,
      "loss": 3.8463,
      "step": 2930
    },
    {
      "epoch": 0.038535140378011375,
      "grad_norm": 0.5361821055412292,
      "learning_rate": 0.0002992341476548149,
      "loss": 3.8227,
      "step": 2940
    },
    {
      "epoch": 0.03866621228405903,
      "grad_norm": 0.5375306010246277,
      "learning_rate": 0.0002992278601330025,
      "loss": 3.8327,
      "step": 2950
    },
    {
      "epoch": 0.038797284190106696,
      "grad_norm": 0.5102008581161499,
      "learning_rate": 0.00029922154697343493,
      "loss": 3.8135,
      "step": 2960
    },
    {
      "epoch": 0.03892835609615435,
      "grad_norm": 0.5300752520561218,
      "learning_rate": 0.00029921520817719686,
      "loss": 3.8217,
      "step": 2970
    },
    {
      "epoch": 0.03905942800220201,
      "grad_norm": 0.5674940347671509,
      "learning_rate": 0.0002992088437453772,
      "loss": 3.8803,
      "step": 2980
    },
    {
      "epoch": 0.03919049990824967,
      "grad_norm": 0.5605335235595703,
      "learning_rate": 0.0002992024536790695,
      "loss": 3.8171,
      "step": 2990
    },
    {
      "epoch": 0.039321571814297324,
      "grad_norm": 0.5499944090843201,
      "learning_rate": 0.0002991960379793715,
      "loss": 3.8523,
      "step": 3000
    },
    {
      "epoch": 0.039321571814297324,
      "eval_accuracy": 0.33861241446725315,
      "eval_loss": 3.819074869155884,
      "eval_runtime": 59.7894,
      "eval_samples_per_second": 33.451,
      "eval_steps_per_second": 0.535,
      "step": 3000
    },
    {
      "epoch": 0.03945264372034498,
      "grad_norm": 0.5730367302894592,
      "learning_rate": 0.00029918959664738547,
      "loss": 3.825,
      "step": 3010
    },
    {
      "epoch": 0.03958371562639264,
      "grad_norm": 0.5929002165794373,
      "learning_rate": 0.0002991831296842181,
      "loss": 3.7867,
      "step": 3020
    },
    {
      "epoch": 0.039714787532440295,
      "grad_norm": 0.6106822490692139,
      "learning_rate": 0.00029917663709098035,
      "loss": 3.8175,
      "step": 3030
    },
    {
      "epoch": 0.03984585943848795,
      "grad_norm": 0.6717346906661987,
      "learning_rate": 0.00029917011886878767,
      "loss": 3.7984,
      "step": 3040
    },
    {
      "epoch": 0.039976931344535616,
      "grad_norm": 0.5712985992431641,
      "learning_rate": 0.00029916357501875995,
      "loss": 3.8066,
      "step": 3050
    },
    {
      "epoch": 0.04010800325058327,
      "grad_norm": 0.5822254419326782,
      "learning_rate": 0.0002991570055420214,
      "loss": 3.8167,
      "step": 3060
    },
    {
      "epoch": 0.04023907515663093,
      "grad_norm": 0.5246325731277466,
      "learning_rate": 0.0002991504104397007,
      "loss": 3.7758,
      "step": 3070
    },
    {
      "epoch": 0.04037014706267859,
      "grad_norm": 0.573517382144928,
      "learning_rate": 0.00029914378971293093,
      "loss": 3.8076,
      "step": 3080
    },
    {
      "epoch": 0.040501218968726244,
      "grad_norm": 0.5353308320045471,
      "learning_rate": 0.0002991371433628495,
      "loss": 3.8271,
      "step": 3090
    },
    {
      "epoch": 0.0406322908747739,
      "grad_norm": 0.49891361594200134,
      "learning_rate": 0.00029913047139059837,
      "loss": 3.827,
      "step": 3100
    },
    {
      "epoch": 0.04076336278082156,
      "grad_norm": 0.497209757566452,
      "learning_rate": 0.0002991237737973236,
      "loss": 3.83,
      "step": 3110
    },
    {
      "epoch": 0.040894434686869215,
      "grad_norm": 0.6368744969367981,
      "learning_rate": 0.0002991170505841761,
      "loss": 3.8088,
      "step": 3120
    },
    {
      "epoch": 0.04102550659291687,
      "grad_norm": 0.5513643622398376,
      "learning_rate": 0.00029911030175231075,
      "loss": 3.7738,
      "step": 3130
    },
    {
      "epoch": 0.04115657849896453,
      "grad_norm": 0.6297922134399414,
      "learning_rate": 0.0002991035273028871,
      "loss": 3.7988,
      "step": 3140
    },
    {
      "epoch": 0.04128765040501219,
      "grad_norm": 0.7055107355117798,
      "learning_rate": 0.000299096727237069,
      "loss": 3.8424,
      "step": 3150
    },
    {
      "epoch": 0.04141872231105985,
      "grad_norm": 0.5582100749015808,
      "learning_rate": 0.00029908990155602476,
      "loss": 3.7901,
      "step": 3160
    },
    {
      "epoch": 0.04154979421710751,
      "grad_norm": 0.5210603475570679,
      "learning_rate": 0.000299083050260927,
      "loss": 3.7995,
      "step": 3170
    },
    {
      "epoch": 0.041680866123155164,
      "grad_norm": 0.4963757395744324,
      "learning_rate": 0.0002990761733529528,
      "loss": 3.7979,
      "step": 3180
    },
    {
      "epoch": 0.04181193802920282,
      "grad_norm": 0.584047257900238,
      "learning_rate": 0.00029906927083328365,
      "loss": 3.8096,
      "step": 3190
    },
    {
      "epoch": 0.04194300993525048,
      "grad_norm": 0.506237268447876,
      "learning_rate": 0.0002990623427031054,
      "loss": 3.7883,
      "step": 3200
    },
    {
      "epoch": 0.042074081841298135,
      "grad_norm": 0.5230671167373657,
      "learning_rate": 0.00029905538896360834,
      "loss": 3.8137,
      "step": 3210
    },
    {
      "epoch": 0.04220515374734579,
      "grad_norm": 0.5437023639678955,
      "learning_rate": 0.0002990484096159872,
      "loss": 3.8095,
      "step": 3220
    },
    {
      "epoch": 0.04233622565339345,
      "grad_norm": 0.6183328032493591,
      "learning_rate": 0.00029904140466144086,
      "loss": 3.769,
      "step": 3230
    },
    {
      "epoch": 0.04246729755944111,
      "grad_norm": 0.6065847873687744,
      "learning_rate": 0.000299034374101173,
      "loss": 3.7871,
      "step": 3240
    },
    {
      "epoch": 0.04259836946548877,
      "grad_norm": 0.5624334216117859,
      "learning_rate": 0.0002990273179363914,
      "loss": 3.7784,
      "step": 3250
    },
    {
      "epoch": 0.04272944137153643,
      "grad_norm": 0.5746332406997681,
      "learning_rate": 0.00029902023616830826,
      "loss": 3.7937,
      "step": 3260
    },
    {
      "epoch": 0.042860513277584084,
      "grad_norm": 0.6270084977149963,
      "learning_rate": 0.0002990131287981404,
      "loss": 3.842,
      "step": 3270
    },
    {
      "epoch": 0.04299158518363174,
      "grad_norm": 0.5226020216941833,
      "learning_rate": 0.0002990059958271088,
      "loss": 3.7564,
      "step": 3280
    },
    {
      "epoch": 0.0431226570896794,
      "grad_norm": 0.5167303085327148,
      "learning_rate": 0.0002989988372564389,
      "loss": 3.7772,
      "step": 3290
    },
    {
      "epoch": 0.043253728995727055,
      "grad_norm": 0.839910626411438,
      "learning_rate": 0.00029899165308736064,
      "loss": 3.695,
      "step": 3300
    },
    {
      "epoch": 0.04338480090177471,
      "grad_norm": 0.5034649968147278,
      "learning_rate": 0.0002989844433211082,
      "loss": 3.7686,
      "step": 3310
    },
    {
      "epoch": 0.04351587280782237,
      "grad_norm": 0.6279553771018982,
      "learning_rate": 0.00029897720795892023,
      "loss": 3.8075,
      "step": 3320
    },
    {
      "epoch": 0.04364694471387003,
      "grad_norm": 0.5339944958686829,
      "learning_rate": 0.0002989699470020399,
      "loss": 3.7631,
      "step": 3330
    },
    {
      "epoch": 0.04377801661991769,
      "grad_norm": 0.47597163915634155,
      "learning_rate": 0.00029896266045171453,
      "loss": 3.7656,
      "step": 3340
    },
    {
      "epoch": 0.04390908852596535,
      "grad_norm": 0.46320733428001404,
      "learning_rate": 0.0002989553483091961,
      "loss": 3.7752,
      "step": 3350
    },
    {
      "epoch": 0.044040160432013004,
      "grad_norm": 0.5094658136367798,
      "learning_rate": 0.0002989480105757407,
      "loss": 3.7641,
      "step": 3360
    },
    {
      "epoch": 0.04417123233806066,
      "grad_norm": 0.5029414296150208,
      "learning_rate": 0.0002989406472526091,
      "loss": 3.765,
      "step": 3370
    },
    {
      "epoch": 0.04430230424410832,
      "grad_norm": 0.6000546216964722,
      "learning_rate": 0.0002989332583410663,
      "loss": 3.78,
      "step": 3380
    },
    {
      "epoch": 0.044433376150155975,
      "grad_norm": 0.5638530850410461,
      "learning_rate": 0.0002989258438423817,
      "loss": 3.8163,
      "step": 3390
    },
    {
      "epoch": 0.04456444805620363,
      "grad_norm": 0.537567675113678,
      "learning_rate": 0.00029891840375782916,
      "loss": 3.7469,
      "step": 3400
    },
    {
      "epoch": 0.04469551996225129,
      "grad_norm": 0.5641475319862366,
      "learning_rate": 0.00029891093808868697,
      "loss": 3.7414,
      "step": 3410
    },
    {
      "epoch": 0.044826591868298946,
      "grad_norm": 0.559258759021759,
      "learning_rate": 0.00029890344683623766,
      "loss": 3.7958,
      "step": 3420
    },
    {
      "epoch": 0.04495766377434661,
      "grad_norm": 0.512962281703949,
      "learning_rate": 0.0002988959300017683,
      "loss": 3.77,
      "step": 3430
    },
    {
      "epoch": 0.04508873568039427,
      "grad_norm": 0.5854800939559937,
      "learning_rate": 0.0002988883875865702,
      "loss": 3.7907,
      "step": 3440
    },
    {
      "epoch": 0.045219807586441924,
      "grad_norm": 0.48237890005111694,
      "learning_rate": 0.0002988808195919393,
      "loss": 3.7683,
      "step": 3450
    },
    {
      "epoch": 0.04535087949248958,
      "grad_norm": 0.4636048972606659,
      "learning_rate": 0.00029887322601917577,
      "loss": 3.7812,
      "step": 3460
    },
    {
      "epoch": 0.04548195139853724,
      "grad_norm": 0.5302407145500183,
      "learning_rate": 0.0002988656068695841,
      "loss": 3.7467,
      "step": 3470
    },
    {
      "epoch": 0.045613023304584895,
      "grad_norm": 0.5752298831939697,
      "learning_rate": 0.0002988579621444735,
      "loss": 3.7947,
      "step": 3480
    },
    {
      "epoch": 0.04574409521063255,
      "grad_norm": 0.5156975984573364,
      "learning_rate": 0.0002988502918451571,
      "loss": 3.8086,
      "step": 3490
    },
    {
      "epoch": 0.04587516711668021,
      "grad_norm": 0.5584268569946289,
      "learning_rate": 0.0002988425959729529,
      "loss": 3.7478,
      "step": 3500
    },
    {
      "epoch": 0.04587516711668021,
      "eval_accuracy": 0.34467644183773216,
      "eval_loss": 3.758277416229248,
      "eval_runtime": 60.4918,
      "eval_samples_per_second": 33.062,
      "eval_steps_per_second": 0.529,
      "step": 3500
    },
    {
      "epoch": 0.046006239022727866,
      "grad_norm": 0.6193779706954956,
      "learning_rate": 0.00029883487452918295,
      "loss": 3.7491,
      "step": 3510
    },
    {
      "epoch": 0.04613731092877553,
      "grad_norm": 0.5633601546287537,
      "learning_rate": 0.00029882712751517376,
      "loss": 3.7605,
      "step": 3520
    },
    {
      "epoch": 0.04626838283482319,
      "grad_norm": 0.510349452495575,
      "learning_rate": 0.00029881935493225646,
      "loss": 3.7714,
      "step": 3530
    },
    {
      "epoch": 0.046399454740870844,
      "grad_norm": 0.4890499711036682,
      "learning_rate": 0.00029881155678176626,
      "loss": 3.7572,
      "step": 3540
    },
    {
      "epoch": 0.0465305266469185,
      "grad_norm": 0.5592243671417236,
      "learning_rate": 0.000298803733065043,
      "loss": 3.7651,
      "step": 3550
    },
    {
      "epoch": 0.04666159855296616,
      "grad_norm": 0.8634549379348755,
      "learning_rate": 0.0002987958837834307,
      "loss": 3.7482,
      "step": 3560
    },
    {
      "epoch": 0.046792670459013815,
      "grad_norm": 0.5141875147819519,
      "learning_rate": 0.000298788008938278,
      "loss": 3.7712,
      "step": 3570
    },
    {
      "epoch": 0.04692374236506147,
      "grad_norm": 0.5150090456008911,
      "learning_rate": 0.0002987801085309377,
      "loss": 3.7558,
      "step": 3580
    },
    {
      "epoch": 0.04705481427110913,
      "grad_norm": 0.6132341623306274,
      "learning_rate": 0.0002987721825627673,
      "loss": 3.7321,
      "step": 3590
    },
    {
      "epoch": 0.047185886177156786,
      "grad_norm": 0.5223958492279053,
      "learning_rate": 0.00029876423103512834,
      "loss": 3.7524,
      "step": 3600
    },
    {
      "epoch": 0.04731695808320444,
      "grad_norm": 0.5199711918830872,
      "learning_rate": 0.00029875625394938703,
      "loss": 3.7479,
      "step": 3610
    },
    {
      "epoch": 0.047448029989252107,
      "grad_norm": 0.46919021010398865,
      "learning_rate": 0.0002987482513069137,
      "loss": 3.7473,
      "step": 3620
    },
    {
      "epoch": 0.047579101895299764,
      "grad_norm": 0.5263029336929321,
      "learning_rate": 0.0002987402231090833,
      "loss": 3.7812,
      "step": 3630
    },
    {
      "epoch": 0.04771017380134742,
      "grad_norm": 0.5948681235313416,
      "learning_rate": 0.0002987321693572752,
      "loss": 3.7531,
      "step": 3640
    },
    {
      "epoch": 0.04784124570739508,
      "grad_norm": 0.48202773928642273,
      "learning_rate": 0.0002987240900528729,
      "loss": 3.7612,
      "step": 3650
    },
    {
      "epoch": 0.047972317613442735,
      "grad_norm": 0.5119797587394714,
      "learning_rate": 0.00029871598519726455,
      "loss": 3.7459,
      "step": 3660
    },
    {
      "epoch": 0.04810338951949039,
      "grad_norm": 0.4955970346927643,
      "learning_rate": 0.00029870785479184256,
      "loss": 3.7262,
      "step": 3670
    },
    {
      "epoch": 0.04823446142553805,
      "grad_norm": 0.49634048342704773,
      "learning_rate": 0.0002986996988380037,
      "loss": 3.7482,
      "step": 3680
    },
    {
      "epoch": 0.048365533331585706,
      "grad_norm": 0.6470732688903809,
      "learning_rate": 0.00029869151733714927,
      "loss": 3.7374,
      "step": 3690
    },
    {
      "epoch": 0.04849660523763336,
      "grad_norm": 0.5342593193054199,
      "learning_rate": 0.0002986833102906848,
      "loss": 3.6892,
      "step": 3700
    },
    {
      "epoch": 0.048627677143681027,
      "grad_norm": 0.45518314838409424,
      "learning_rate": 0.0002986750777000204,
      "loss": 3.7458,
      "step": 3710
    },
    {
      "epoch": 0.048758749049728684,
      "grad_norm": 0.4850049316883087,
      "learning_rate": 0.0002986668195665703,
      "loss": 3.7677,
      "step": 3720
    },
    {
      "epoch": 0.04888982095577634,
      "grad_norm": 0.5251049995422363,
      "learning_rate": 0.0002986585358917533,
      "loss": 3.7261,
      "step": 3730
    },
    {
      "epoch": 0.049020892861824,
      "grad_norm": 0.5176697373390198,
      "learning_rate": 0.0002986502266769927,
      "loss": 3.7386,
      "step": 3740
    },
    {
      "epoch": 0.049151964767871655,
      "grad_norm": 0.5143198370933533,
      "learning_rate": 0.00029864189192371597,
      "loss": 3.7698,
      "step": 3750
    },
    {
      "epoch": 0.04928303667391931,
      "grad_norm": 0.5064288973808289,
      "learning_rate": 0.00029863353163335493,
      "loss": 3.7079,
      "step": 3760
    },
    {
      "epoch": 0.04941410857996697,
      "grad_norm": 0.5063779950141907,
      "learning_rate": 0.000298625145807346,
      "loss": 3.6958,
      "step": 3770
    },
    {
      "epoch": 0.049545180486014626,
      "grad_norm": 0.5340002775192261,
      "learning_rate": 0.0002986167344471299,
      "loss": 3.7302,
      "step": 3780
    },
    {
      "epoch": 0.04967625239206228,
      "grad_norm": 0.5392060279846191,
      "learning_rate": 0.0002986082975541517,
      "loss": 3.7358,
      "step": 3790
    },
    {
      "epoch": 0.049807324298109946,
      "grad_norm": 0.5258914232254028,
      "learning_rate": 0.00029859983512986086,
      "loss": 3.7642,
      "step": 3800
    },
    {
      "epoch": 0.049938396204157603,
      "grad_norm": 0.520696759223938,
      "learning_rate": 0.0002985913471757113,
      "loss": 3.786,
      "step": 3810
    },
    {
      "epoch": 0.05006946811020526,
      "grad_norm": 0.4884369373321533,
      "learning_rate": 0.00029858283369316125,
      "loss": 3.7259,
      "step": 3820
    },
    {
      "epoch": 0.05020054001625292,
      "grad_norm": 0.4740183353424072,
      "learning_rate": 0.00029857429468367326,
      "loss": 3.72,
      "step": 3830
    },
    {
      "epoch": 0.050331611922300575,
      "grad_norm": 0.5521665215492249,
      "learning_rate": 0.00029856573014871453,
      "loss": 3.7401,
      "step": 3840
    },
    {
      "epoch": 0.05046268382834823,
      "grad_norm": 0.5022364854812622,
      "learning_rate": 0.0002985571400897564,
      "loss": 3.7483,
      "step": 3850
    },
    {
      "epoch": 0.05059375573439589,
      "grad_norm": 0.6020211577415466,
      "learning_rate": 0.0002985485245082746,
      "loss": 3.7761,
      "step": 3860
    },
    {
      "epoch": 0.050724827640443546,
      "grad_norm": 0.5678549408912659,
      "learning_rate": 0.00029853988340574933,
      "loss": 3.694,
      "step": 3870
    },
    {
      "epoch": 0.0508558995464912,
      "grad_norm": 0.4777968227863312,
      "learning_rate": 0.00029853121678366523,
      "loss": 3.7507,
      "step": 3880
    },
    {
      "epoch": 0.05098697145253886,
      "grad_norm": 0.4679000675678253,
      "learning_rate": 0.0002985225246435112,
      "loss": 3.7459,
      "step": 3890
    },
    {
      "epoch": 0.051118043358586523,
      "grad_norm": 0.4903142750263214,
      "learning_rate": 0.00029851380698678055,
      "loss": 3.7372,
      "step": 3900
    },
    {
      "epoch": 0.05124911526463418,
      "grad_norm": 0.6592577695846558,
      "learning_rate": 0.00029850506381497106,
      "loss": 3.777,
      "step": 3910
    },
    {
      "epoch": 0.05138018717068184,
      "grad_norm": 0.5241567492485046,
      "learning_rate": 0.00029849629512958476,
      "loss": 3.7499,
      "step": 3920
    },
    {
      "epoch": 0.051511259076729494,
      "grad_norm": 0.6090673804283142,
      "learning_rate": 0.00029848750093212814,
      "loss": 3.7097,
      "step": 3930
    },
    {
      "epoch": 0.05164233098277715,
      "grad_norm": 0.5108484625816345,
      "learning_rate": 0.00029847868122411214,
      "loss": 3.721,
      "step": 3940
    },
    {
      "epoch": 0.05177340288882481,
      "grad_norm": 0.49318522214889526,
      "learning_rate": 0.000298469836007052,
      "loss": 3.7443,
      "step": 3950
    },
    {
      "epoch": 0.051904474794872466,
      "grad_norm": 0.5103313326835632,
      "learning_rate": 0.00029846096528246725,
      "loss": 3.6962,
      "step": 3960
    },
    {
      "epoch": 0.05203554670092012,
      "grad_norm": 0.48358339071273804,
      "learning_rate": 0.000298452069051882,
      "loss": 3.7329,
      "step": 3970
    },
    {
      "epoch": 0.05216661860696778,
      "grad_norm": 0.5443662405014038,
      "learning_rate": 0.00029844314731682455,
      "loss": 3.6911,
      "step": 3980
    },
    {
      "epoch": 0.052297690513015443,
      "grad_norm": 0.4827999174594879,
      "learning_rate": 0.00029843420007882784,
      "loss": 3.7062,
      "step": 3990
    },
    {
      "epoch": 0.0524287624190631,
      "grad_norm": 0.5258951187133789,
      "learning_rate": 0.0002984252273394289,
      "loss": 3.7228,
      "step": 4000
    },
    {
      "epoch": 0.0524287624190631,
      "eval_accuracy": 0.3487653958944282,
      "eval_loss": 3.716583013534546,
      "eval_runtime": 59.9241,
      "eval_samples_per_second": 33.376,
      "eval_steps_per_second": 0.534,
      "step": 4000
    },
    {
      "epoch": 0.05255983432511076,
      "grad_norm": 0.4705118238925934,
      "learning_rate": 0.0002984162291001693,
      "loss": 3.6867,
      "step": 4010
    },
    {
      "epoch": 0.052690906231158414,
      "grad_norm": 0.5500466823577881,
      "learning_rate": 0.000298407205362595,
      "loss": 3.7258,
      "step": 4020
    },
    {
      "epoch": 0.05282197813720607,
      "grad_norm": 0.5881951451301575,
      "learning_rate": 0.0002983981561282563,
      "loss": 3.7065,
      "step": 4030
    },
    {
      "epoch": 0.05295305004325373,
      "grad_norm": 0.4666280150413513,
      "learning_rate": 0.00029838908139870774,
      "loss": 3.7101,
      "step": 4040
    },
    {
      "epoch": 0.053084121949301386,
      "grad_norm": 0.5772592425346375,
      "learning_rate": 0.0002983799811755086,
      "loss": 3.7088,
      "step": 4050
    },
    {
      "epoch": 0.05321519385534904,
      "grad_norm": 0.4957033097743988,
      "learning_rate": 0.0002983708554602222,
      "loss": 3.7328,
      "step": 4060
    },
    {
      "epoch": 0.0533462657613967,
      "grad_norm": 0.5589922070503235,
      "learning_rate": 0.00029836170425441637,
      "loss": 3.7586,
      "step": 4070
    },
    {
      "epoch": 0.053477337667444357,
      "grad_norm": 0.5463706851005554,
      "learning_rate": 0.0002983525275596633,
      "loss": 3.7097,
      "step": 4080
    },
    {
      "epoch": 0.05360840957349202,
      "grad_norm": 0.5326753854751587,
      "learning_rate": 0.00029834332537753965,
      "loss": 3.706,
      "step": 4090
    },
    {
      "epoch": 0.05373948147953968,
      "grad_norm": 0.4960501194000244,
      "learning_rate": 0.0002983340977096263,
      "loss": 3.6979,
      "step": 4100
    },
    {
      "epoch": 0.053870553385587334,
      "grad_norm": 0.5951488018035889,
      "learning_rate": 0.00029832484455750866,
      "loss": 3.715,
      "step": 4110
    },
    {
      "epoch": 0.05400162529163499,
      "grad_norm": 0.5008718967437744,
      "learning_rate": 0.00029831556592277637,
      "loss": 3.7594,
      "step": 4120
    },
    {
      "epoch": 0.05413269719768265,
      "grad_norm": 0.5062334537506104,
      "learning_rate": 0.00029830626180702357,
      "loss": 3.7237,
      "step": 4130
    },
    {
      "epoch": 0.054263769103730305,
      "grad_norm": 0.6344945430755615,
      "learning_rate": 0.0002982969322118487,
      "loss": 3.7494,
      "step": 4140
    },
    {
      "epoch": 0.05439484100977796,
      "grad_norm": 0.5724450349807739,
      "learning_rate": 0.00029828757713885466,
      "loss": 3.73,
      "step": 4150
    },
    {
      "epoch": 0.05452591291582562,
      "grad_norm": 0.5690779089927673,
      "learning_rate": 0.00029827819658964866,
      "loss": 3.7591,
      "step": 4160
    },
    {
      "epoch": 0.054656984821873277,
      "grad_norm": 0.5222092866897583,
      "learning_rate": 0.00029826879056584225,
      "loss": 3.6926,
      "step": 4170
    },
    {
      "epoch": 0.05478805672792094,
      "grad_norm": 0.5230311155319214,
      "learning_rate": 0.0002982593590690515,
      "loss": 3.6878,
      "step": 4180
    },
    {
      "epoch": 0.0549191286339686,
      "grad_norm": 0.5027438998222351,
      "learning_rate": 0.0002982499021008967,
      "loss": 3.7216,
      "step": 4190
    },
    {
      "epoch": 0.055050200540016254,
      "grad_norm": 0.46176785230636597,
      "learning_rate": 0.0002982404196630026,
      "loss": 3.6811,
      "step": 4200
    },
    {
      "epoch": 0.05518127244606391,
      "grad_norm": 0.5563604831695557,
      "learning_rate": 0.0002982309117569983,
      "loss": 3.6898,
      "step": 4210
    },
    {
      "epoch": 0.05531234435211157,
      "grad_norm": 0.6255161762237549,
      "learning_rate": 0.00029822137838451734,
      "loss": 3.7305,
      "step": 4220
    },
    {
      "epoch": 0.055443416258159225,
      "grad_norm": 0.6251128315925598,
      "learning_rate": 0.0002982118195471975,
      "loss": 3.7163,
      "step": 4230
    },
    {
      "epoch": 0.05557448816420688,
      "grad_norm": 0.4929603636264801,
      "learning_rate": 0.0002982022352466811,
      "loss": 3.7285,
      "step": 4240
    },
    {
      "epoch": 0.05570556007025454,
      "grad_norm": 0.5028738975524902,
      "learning_rate": 0.00029819262548461464,
      "loss": 3.7075,
      "step": 4250
    },
    {
      "epoch": 0.055836631976302196,
      "grad_norm": 0.4753704071044922,
      "learning_rate": 0.0002981829902626492,
      "loss": 3.668,
      "step": 4260
    },
    {
      "epoch": 0.05596770388234986,
      "grad_norm": 0.5016584992408752,
      "learning_rate": 0.00029817332958244005,
      "loss": 3.6895,
      "step": 4270
    },
    {
      "epoch": 0.05609877578839752,
      "grad_norm": 0.5594503283500671,
      "learning_rate": 0.000298163643445647,
      "loss": 3.7771,
      "step": 4280
    },
    {
      "epoch": 0.056229847694445174,
      "grad_norm": 0.5747581124305725,
      "learning_rate": 0.0002981539318539342,
      "loss": 3.7538,
      "step": 4290
    },
    {
      "epoch": 0.05636091960049283,
      "grad_norm": 0.46820971369743347,
      "learning_rate": 0.00029814419480897,
      "loss": 3.7655,
      "step": 4300
    },
    {
      "epoch": 0.05649199150654049,
      "grad_norm": 0.5073003768920898,
      "learning_rate": 0.0002981344323124273,
      "loss": 3.6906,
      "step": 4310
    },
    {
      "epoch": 0.056623063412588145,
      "grad_norm": 0.5288839340209961,
      "learning_rate": 0.0002981246443659833,
      "loss": 3.7187,
      "step": 4320
    },
    {
      "epoch": 0.0567541353186358,
      "grad_norm": 0.49454227089881897,
      "learning_rate": 0.0002981148309713197,
      "loss": 3.6741,
      "step": 4330
    },
    {
      "epoch": 0.05688520722468346,
      "grad_norm": 0.5309221148490906,
      "learning_rate": 0.00029810499213012235,
      "loss": 3.7016,
      "step": 4340
    },
    {
      "epoch": 0.057016279130731116,
      "grad_norm": 0.46260079741477966,
      "learning_rate": 0.0002980951278440816,
      "loss": 3.6915,
      "step": 4350
    },
    {
      "epoch": 0.057147351036778773,
      "grad_norm": 0.5022294521331787,
      "learning_rate": 0.0002980852381148923,
      "loss": 3.6973,
      "step": 4360
    },
    {
      "epoch": 0.05727842294282644,
      "grad_norm": 0.4898497760295868,
      "learning_rate": 0.00029807532294425334,
      "loss": 3.6788,
      "step": 4370
    },
    {
      "epoch": 0.057409494848874094,
      "grad_norm": 0.4465279281139374,
      "learning_rate": 0.00029806538233386826,
      "loss": 3.6907,
      "step": 4380
    },
    {
      "epoch": 0.05754056675492175,
      "grad_norm": 0.5374187231063843,
      "learning_rate": 0.00029805541628544495,
      "loss": 3.7522,
      "step": 4390
    },
    {
      "epoch": 0.05767163866096941,
      "grad_norm": 0.49404239654541016,
      "learning_rate": 0.00029804542480069547,
      "loss": 3.7174,
      "step": 4400
    },
    {
      "epoch": 0.057802710567017065,
      "grad_norm": 0.9680317044258118,
      "learning_rate": 0.00029803540788133644,
      "loss": 3.7941,
      "step": 4410
    },
    {
      "epoch": 0.05793378247306472,
      "grad_norm": 0.5890807509422302,
      "learning_rate": 0.00029802536552908887,
      "loss": 3.7073,
      "step": 4420
    },
    {
      "epoch": 0.05806485437911238,
      "grad_norm": 0.46957340836524963,
      "learning_rate": 0.000298015297745678,
      "loss": 3.6926,
      "step": 4430
    },
    {
      "epoch": 0.058195926285160036,
      "grad_norm": 0.5658926963806152,
      "learning_rate": 0.0002980052045328335,
      "loss": 3.7309,
      "step": 4440
    },
    {
      "epoch": 0.05832699819120769,
      "grad_norm": 0.5464749336242676,
      "learning_rate": 0.0002979950858922894,
      "loss": 3.7156,
      "step": 4450
    },
    {
      "epoch": 0.05845807009725536,
      "grad_norm": 0.5582205653190613,
      "learning_rate": 0.00029798494182578414,
      "loss": 3.676,
      "step": 4460
    },
    {
      "epoch": 0.058589142003303014,
      "grad_norm": 0.5002586245536804,
      "learning_rate": 0.0002979747723350605,
      "loss": 3.6928,
      "step": 4470
    },
    {
      "epoch": 0.05872021390935067,
      "grad_norm": 0.5564867854118347,
      "learning_rate": 0.0002979645774218656,
      "loss": 3.697,
      "step": 4480
    },
    {
      "epoch": 0.05885128581539833,
      "grad_norm": 0.4805472791194916,
      "learning_rate": 0.00029795435708795097,
      "loss": 3.7336,
      "step": 4490
    },
    {
      "epoch": 0.058982357721445985,
      "grad_norm": 0.43925702571868896,
      "learning_rate": 0.0002979441113350725,
      "loss": 3.7105,
      "step": 4500
    },
    {
      "epoch": 0.058982357721445985,
      "eval_accuracy": 0.3522585532746823,
      "eval_loss": 3.6766772270202637,
      "eval_runtime": 60.8588,
      "eval_samples_per_second": 32.863,
      "eval_steps_per_second": 0.526,
      "step": 4500
    },
    {
      "epoch": 0.05911342962749364,
      "grad_norm": 0.562069296836853,
      "learning_rate": 0.0002979338401649903,
      "loss": 3.6817,
      "step": 4510
    },
    {
      "epoch": 0.0592445015335413,
      "grad_norm": 0.5216554999351501,
      "learning_rate": 0.00029792354357946926,
      "loss": 3.669,
      "step": 4520
    },
    {
      "epoch": 0.059375573439588956,
      "grad_norm": 0.43828126788139343,
      "learning_rate": 0.0002979132215802781,
      "loss": 3.664,
      "step": 4530
    },
    {
      "epoch": 0.05950664534563661,
      "grad_norm": 0.48035728931427,
      "learning_rate": 0.00029790287416919044,
      "loss": 3.6798,
      "step": 4540
    },
    {
      "epoch": 0.05963771725168428,
      "grad_norm": 0.468122273683548,
      "learning_rate": 0.0002978925013479837,
      "loss": 3.6486,
      "step": 4550
    },
    {
      "epoch": 0.059768789157731934,
      "grad_norm": 0.47445327043533325,
      "learning_rate": 0.00029788210311844013,
      "loss": 3.6623,
      "step": 4560
    },
    {
      "epoch": 0.05989986106377959,
      "grad_norm": 0.5480607151985168,
      "learning_rate": 0.00029787167948234616,
      "loss": 3.743,
      "step": 4570
    },
    {
      "epoch": 0.06003093296982725,
      "grad_norm": 0.5036835074424744,
      "learning_rate": 0.00029786123044149256,
      "loss": 3.6775,
      "step": 4580
    },
    {
      "epoch": 0.060162004875874905,
      "grad_norm": 0.4812488853931427,
      "learning_rate": 0.00029785075599767446,
      "loss": 3.6848,
      "step": 4590
    },
    {
      "epoch": 0.06029307678192256,
      "grad_norm": 0.49971309304237366,
      "learning_rate": 0.0002978402561526915,
      "loss": 3.6618,
      "step": 4600
    },
    {
      "epoch": 0.06042414868797022,
      "grad_norm": 0.5052714943885803,
      "learning_rate": 0.00029782973090834757,
      "loss": 3.6919,
      "step": 4610
    },
    {
      "epoch": 0.060555220594017876,
      "grad_norm": 0.5250628590583801,
      "learning_rate": 0.0002978191802664509,
      "loss": 3.6791,
      "step": 4620
    },
    {
      "epoch": 0.06068629250006553,
      "grad_norm": 0.46273139119148254,
      "learning_rate": 0.0002978086042288141,
      "loss": 3.6619,
      "step": 4630
    },
    {
      "epoch": 0.06081736440611319,
      "grad_norm": 0.5283713936805725,
      "learning_rate": 0.00029779800279725423,
      "loss": 3.6603,
      "step": 4640
    },
    {
      "epoch": 0.060948436312160854,
      "grad_norm": 0.4948991537094116,
      "learning_rate": 0.0002977873759735926,
      "loss": 3.6829,
      "step": 4650
    },
    {
      "epoch": 0.06107950821820851,
      "grad_norm": 0.5752778649330139,
      "learning_rate": 0.0002977767237596549,
      "loss": 3.6883,
      "step": 4660
    },
    {
      "epoch": 0.06121058012425617,
      "grad_norm": 0.4794826805591583,
      "learning_rate": 0.00029776604615727124,
      "loss": 3.6909,
      "step": 4670
    },
    {
      "epoch": 0.061341652030303825,
      "grad_norm": 0.4672747850418091,
      "learning_rate": 0.00029775534316827605,
      "loss": 3.6969,
      "step": 4680
    },
    {
      "epoch": 0.06147272393635148,
      "grad_norm": 0.5124154686927795,
      "learning_rate": 0.0002977446147945082,
      "loss": 3.6528,
      "step": 4690
    },
    {
      "epoch": 0.06160379584239914,
      "grad_norm": 0.5305249094963074,
      "learning_rate": 0.0002977338610378107,
      "loss": 3.6645,
      "step": 4700
    },
    {
      "epoch": 0.061734867748446796,
      "grad_norm": 0.5009079575538635,
      "learning_rate": 0.00029772308190003125,
      "loss": 3.6289,
      "step": 4710
    },
    {
      "epoch": 0.06186593965449445,
      "grad_norm": 0.5182138681411743,
      "learning_rate": 0.00029771227738302166,
      "loss": 3.6542,
      "step": 4720
    },
    {
      "epoch": 0.06199701156054211,
      "grad_norm": 0.5290394425392151,
      "learning_rate": 0.0002977014474886381,
      "loss": 3.7444,
      "step": 4730
    },
    {
      "epoch": 0.062128083466589774,
      "grad_norm": 0.4978898763656616,
      "learning_rate": 0.00029769059221874136,
      "loss": 3.6356,
      "step": 4740
    },
    {
      "epoch": 0.06225915537263743,
      "grad_norm": 0.48885300755500793,
      "learning_rate": 0.0002976797115751962,
      "loss": 3.6583,
      "step": 4750
    },
    {
      "epoch": 0.06239022727868509,
      "grad_norm": 0.4629542827606201,
      "learning_rate": 0.0002976688055598721,
      "loss": 3.6524,
      "step": 4760
    },
    {
      "epoch": 0.06252129918473275,
      "grad_norm": 0.5347307920455933,
      "learning_rate": 0.0002976578741746426,
      "loss": 3.7007,
      "step": 4770
    },
    {
      "epoch": 0.0626523710907804,
      "grad_norm": 0.45359355211257935,
      "learning_rate": 0.0002976469174213859,
      "loss": 3.679,
      "step": 4780
    },
    {
      "epoch": 0.06278344299682806,
      "grad_norm": 0.4718404710292816,
      "learning_rate": 0.00029763593530198433,
      "loss": 3.6412,
      "step": 4790
    },
    {
      "epoch": 0.06291451490287572,
      "grad_norm": 0.5092851519584656,
      "learning_rate": 0.0002976249278183246,
      "loss": 3.6917,
      "step": 4800
    },
    {
      "epoch": 0.06304558680892337,
      "grad_norm": 0.5336654782295227,
      "learning_rate": 0.00029761389497229783,
      "loss": 3.6306,
      "step": 4810
    },
    {
      "epoch": 0.06317665871497104,
      "grad_norm": 0.6475569009780884,
      "learning_rate": 0.0002976028367657996,
      "loss": 3.7414,
      "step": 4820
    },
    {
      "epoch": 0.06330773062101869,
      "grad_norm": 0.5099215507507324,
      "learning_rate": 0.0002975917532007297,
      "loss": 3.6154,
      "step": 4830
    },
    {
      "epoch": 0.06343880252706635,
      "grad_norm": 0.4920685589313507,
      "learning_rate": 0.0002975806442789922,
      "loss": 3.6765,
      "step": 4840
    },
    {
      "epoch": 0.063569874433114,
      "grad_norm": 0.4824768006801605,
      "learning_rate": 0.0002975695100024958,
      "loss": 3.6525,
      "step": 4850
    },
    {
      "epoch": 0.06370094633916167,
      "grad_norm": 0.48799267411231995,
      "learning_rate": 0.0002975583503731533,
      "loss": 3.675,
      "step": 4860
    },
    {
      "epoch": 0.06383201824520932,
      "grad_norm": 0.48716628551483154,
      "learning_rate": 0.000297547165392882,
      "loss": 3.6708,
      "step": 4870
    },
    {
      "epoch": 0.06396309015125698,
      "grad_norm": 0.4660966992378235,
      "learning_rate": 0.0002975359550636035,
      "loss": 3.6386,
      "step": 4880
    },
    {
      "epoch": 0.06409416205730464,
      "grad_norm": 0.49145275354385376,
      "learning_rate": 0.00029752471938724375,
      "loss": 3.7652,
      "step": 4890
    },
    {
      "epoch": 0.0642252339633523,
      "grad_norm": 0.6009672284126282,
      "learning_rate": 0.00029751345836573313,
      "loss": 3.6802,
      "step": 4900
    },
    {
      "epoch": 0.06435630586939996,
      "grad_norm": 0.5045648813247681,
      "learning_rate": 0.00029750217200100616,
      "loss": 3.6672,
      "step": 4910
    },
    {
      "epoch": 0.06448737777544761,
      "grad_norm": 0.4908842444419861,
      "learning_rate": 0.00029749086029500204,
      "loss": 3.7113,
      "step": 4920
    },
    {
      "epoch": 0.06461844968149527,
      "grad_norm": 0.4836691915988922,
      "learning_rate": 0.0002974795232496641,
      "loss": 3.6645,
      "step": 4930
    },
    {
      "epoch": 0.06474952158754292,
      "grad_norm": 0.4789673686027527,
      "learning_rate": 0.00029746816086694,
      "loss": 3.6694,
      "step": 4940
    },
    {
      "epoch": 0.06488059349359059,
      "grad_norm": 0.4985313415527344,
      "learning_rate": 0.00029745677314878195,
      "loss": 3.7072,
      "step": 4950
    },
    {
      "epoch": 0.06501166539963824,
      "grad_norm": 0.5141927599906921,
      "learning_rate": 0.0002974453600971463,
      "loss": 3.7004,
      "step": 4960
    },
    {
      "epoch": 0.0651427373056859,
      "grad_norm": 0.5227395296096802,
      "learning_rate": 0.00029743392171399386,
      "loss": 3.6746,
      "step": 4970
    },
    {
      "epoch": 0.06527380921173356,
      "grad_norm": 0.48610764741897583,
      "learning_rate": 0.0002974224580012898,
      "loss": 3.6574,
      "step": 4980
    },
    {
      "epoch": 0.06540488111778121,
      "grad_norm": 0.48257651925086975,
      "learning_rate": 0.0002974109689610036,
      "loss": 3.6516,
      "step": 4990
    },
    {
      "epoch": 0.06553595302382888,
      "grad_norm": 0.5324851870536804,
      "learning_rate": 0.00029739945459510917,
      "loss": 3.6727,
      "step": 5000
    },
    {
      "epoch": 0.06553595302382888,
      "eval_accuracy": 0.3551026392961877,
      "eval_loss": 3.6477560997009277,
      "eval_runtime": 57.4778,
      "eval_samples_per_second": 34.796,
      "eval_steps_per_second": 0.557,
      "step": 5000
    },
    {
      "epoch": 0.06566702492987653,
      "grad_norm": 0.5052070021629333,
      "learning_rate": 0.00029738791490558465,
      "loss": 3.6874,
      "step": 5010
    },
    {
      "epoch": 0.06579809683592419,
      "grad_norm": 0.46584945917129517,
      "learning_rate": 0.00029737634989441256,
      "loss": 3.6993,
      "step": 5020
    },
    {
      "epoch": 0.06592916874197184,
      "grad_norm": 0.48740869760513306,
      "learning_rate": 0.00029736475956357983,
      "loss": 3.6943,
      "step": 5030
    },
    {
      "epoch": 0.0660602406480195,
      "grad_norm": 0.48296982049942017,
      "learning_rate": 0.00029735314391507776,
      "loss": 3.6633,
      "step": 5040
    },
    {
      "epoch": 0.06619131255406716,
      "grad_norm": 0.5165860056877136,
      "learning_rate": 0.00029734150295090186,
      "loss": 3.6674,
      "step": 5050
    },
    {
      "epoch": 0.06632238446011482,
      "grad_norm": 0.5146704316139221,
      "learning_rate": 0.0002973298366730521,
      "loss": 3.6604,
      "step": 5060
    },
    {
      "epoch": 0.06645345636616248,
      "grad_norm": 0.5135734677314758,
      "learning_rate": 0.0002973181450835329,
      "loss": 3.6692,
      "step": 5070
    },
    {
      "epoch": 0.06658452827221013,
      "grad_norm": 0.6608737707138062,
      "learning_rate": 0.0002973064281843528,
      "loss": 3.7411,
      "step": 5080
    },
    {
      "epoch": 0.0667156001782578,
      "grad_norm": 0.4700607359409332,
      "learning_rate": 0.0002972946859775247,
      "loss": 3.6496,
      "step": 5090
    },
    {
      "epoch": 0.06684667208430545,
      "grad_norm": 0.5591515302658081,
      "learning_rate": 0.00029728291846506616,
      "loss": 3.6373,
      "step": 5100
    },
    {
      "epoch": 0.06697774399035311,
      "grad_norm": 0.5835526585578918,
      "learning_rate": 0.0002972711256489987,
      "loss": 3.6583,
      "step": 5110
    },
    {
      "epoch": 0.06710881589640076,
      "grad_norm": 0.5377927422523499,
      "learning_rate": 0.00029725930753134844,
      "loss": 3.646,
      "step": 5120
    },
    {
      "epoch": 0.06723988780244843,
      "grad_norm": 0.45906636118888855,
      "learning_rate": 0.00029724746411414576,
      "loss": 3.6665,
      "step": 5130
    },
    {
      "epoch": 0.06737095970849608,
      "grad_norm": 0.6140027046203613,
      "learning_rate": 0.0002972355953994253,
      "loss": 3.7123,
      "step": 5140
    },
    {
      "epoch": 0.06750203161454374,
      "grad_norm": 0.5080171227455139,
      "learning_rate": 0.0002972237013892263,
      "loss": 3.6576,
      "step": 5150
    },
    {
      "epoch": 0.0676331035205914,
      "grad_norm": 0.4888867139816284,
      "learning_rate": 0.0002972117820855921,
      "loss": 3.6347,
      "step": 5160
    },
    {
      "epoch": 0.06776417542663905,
      "grad_norm": 0.48631858825683594,
      "learning_rate": 0.00029719983749057044,
      "loss": 3.6539,
      "step": 5170
    },
    {
      "epoch": 0.06789524733268672,
      "grad_norm": 0.5233666300773621,
      "learning_rate": 0.00029718786760621346,
      "loss": 3.6541,
      "step": 5180
    },
    {
      "epoch": 0.06802631923873437,
      "grad_norm": 0.486524760723114,
      "learning_rate": 0.00029717587243457765,
      "loss": 3.6401,
      "step": 5190
    },
    {
      "epoch": 0.06815739114478203,
      "grad_norm": 0.49529728293418884,
      "learning_rate": 0.00029716385197772375,
      "loss": 3.6385,
      "step": 5200
    },
    {
      "epoch": 0.06828846305082968,
      "grad_norm": 0.514244794845581,
      "learning_rate": 0.00029715180623771694,
      "loss": 3.6686,
      "step": 5210
    },
    {
      "epoch": 0.06841953495687735,
      "grad_norm": 0.5268776416778564,
      "learning_rate": 0.0002971397352166267,
      "loss": 3.6533,
      "step": 5220
    },
    {
      "epoch": 0.068550606862925,
      "grad_norm": 0.5166194438934326,
      "learning_rate": 0.00029712763891652695,
      "loss": 3.6255,
      "step": 5230
    },
    {
      "epoch": 0.06868167876897266,
      "grad_norm": 0.5018957853317261,
      "learning_rate": 0.0002971155173394957,
      "loss": 3.6492,
      "step": 5240
    },
    {
      "epoch": 0.06881275067502031,
      "grad_norm": 0.4930972456932068,
      "learning_rate": 0.00029710337048761563,
      "loss": 3.6345,
      "step": 5250
    },
    {
      "epoch": 0.06894382258106797,
      "grad_norm": 0.5227683186531067,
      "learning_rate": 0.0002970911983629735,
      "loss": 3.6705,
      "step": 5260
    },
    {
      "epoch": 0.06907489448711564,
      "grad_norm": 0.48323237895965576,
      "learning_rate": 0.0002970790009676606,
      "loss": 3.6558,
      "step": 5270
    },
    {
      "epoch": 0.06920596639316329,
      "grad_norm": 0.5217201113700867,
      "learning_rate": 0.00029706677830377236,
      "loss": 3.6167,
      "step": 5280
    },
    {
      "epoch": 0.06933703829921095,
      "grad_norm": 0.5185927152633667,
      "learning_rate": 0.0002970545303734087,
      "loss": 3.656,
      "step": 5290
    },
    {
      "epoch": 0.0694681102052586,
      "grad_norm": 0.44609251618385315,
      "learning_rate": 0.000297042257178674,
      "loss": 3.6509,
      "step": 5300
    },
    {
      "epoch": 0.06959918211130627,
      "grad_norm": 0.5058116316795349,
      "learning_rate": 0.0002970299587216766,
      "loss": 3.6385,
      "step": 5310
    },
    {
      "epoch": 0.06973025401735392,
      "grad_norm": 0.5857456922531128,
      "learning_rate": 0.0002970176350045296,
      "loss": 3.658,
      "step": 5320
    },
    {
      "epoch": 0.06986132592340158,
      "grad_norm": 0.49739158153533936,
      "learning_rate": 0.00029700528602935014,
      "loss": 3.6331,
      "step": 5330
    },
    {
      "epoch": 0.06999239782944923,
      "grad_norm": 0.5257708430290222,
      "learning_rate": 0.0002969929117982598,
      "loss": 3.6689,
      "step": 5340
    },
    {
      "epoch": 0.0701234697354969,
      "grad_norm": 0.635816216468811,
      "learning_rate": 0.0002969805123133846,
      "loss": 3.6006,
      "step": 5350
    },
    {
      "epoch": 0.07025454164154456,
      "grad_norm": 0.517143189907074,
      "learning_rate": 0.0002969680875768547,
      "loss": 3.6675,
      "step": 5360
    },
    {
      "epoch": 0.07038561354759221,
      "grad_norm": 0.485242635011673,
      "learning_rate": 0.0002969556375908048,
      "loss": 3.6456,
      "step": 5370
    },
    {
      "epoch": 0.07051668545363987,
      "grad_norm": 0.43611156940460205,
      "learning_rate": 0.0002969431623573738,
      "loss": 3.6308,
      "step": 5380
    },
    {
      "epoch": 0.07064775735968752,
      "grad_norm": 0.480699747800827,
      "learning_rate": 0.0002969306618787049,
      "loss": 3.6529,
      "step": 5390
    },
    {
      "epoch": 0.07077882926573519,
      "grad_norm": 0.536634087562561,
      "learning_rate": 0.0002969181361569458,
      "loss": 3.6242,
      "step": 5400
    },
    {
      "epoch": 0.07090990117178284,
      "grad_norm": 0.561854898929596,
      "learning_rate": 0.0002969055851942485,
      "loss": 3.6433,
      "step": 5410
    },
    {
      "epoch": 0.0710409730778305,
      "grad_norm": 0.5024386048316956,
      "learning_rate": 0.0002968930089927692,
      "loss": 3.667,
      "step": 5420
    },
    {
      "epoch": 0.07117204498387815,
      "grad_norm": 0.47561752796173096,
      "learning_rate": 0.0002968804075546686,
      "loss": 3.6381,
      "step": 5430
    },
    {
      "epoch": 0.07130311688992581,
      "grad_norm": 0.48047977685928345,
      "learning_rate": 0.00029686778088211155,
      "loss": 3.6134,
      "step": 5440
    },
    {
      "epoch": 0.07143418879597348,
      "grad_norm": 0.5246568918228149,
      "learning_rate": 0.0002968551289772674,
      "loss": 3.6295,
      "step": 5450
    },
    {
      "epoch": 0.07156526070202113,
      "grad_norm": 0.44749024510383606,
      "learning_rate": 0.0002968424518423099,
      "loss": 3.6633,
      "step": 5460
    },
    {
      "epoch": 0.07169633260806879,
      "grad_norm": 0.5438427925109863,
      "learning_rate": 0.0002968297494794168,
      "loss": 3.622,
      "step": 5470
    },
    {
      "epoch": 0.07182740451411644,
      "grad_norm": 0.46368175745010376,
      "learning_rate": 0.00029681702189077056,
      "loss": 3.6406,
      "step": 5480
    },
    {
      "epoch": 0.0719584764201641,
      "grad_norm": 0.4888322651386261,
      "learning_rate": 0.0002968042690785578,
      "loss": 3.7165,
      "step": 5490
    },
    {
      "epoch": 0.07208954832621176,
      "grad_norm": 0.4643484950065613,
      "learning_rate": 0.0002967914910449694,
      "loss": 3.6376,
      "step": 5500
    },
    {
      "epoch": 0.07208954832621176,
      "eval_accuracy": 0.3579242424242424,
      "eval_loss": 3.617338180541992,
      "eval_runtime": 58.6235,
      "eval_samples_per_second": 34.116,
      "eval_steps_per_second": 0.546,
      "step": 5500
    },
    {
      "epoch": 0.07222062023225942,
      "grad_norm": 0.5033567547798157,
      "learning_rate": 0.0002967786877922008,
      "loss": 3.6349,
      "step": 5510
    },
    {
      "epoch": 0.07235169213830707,
      "grad_norm": 0.48570674657821655,
      "learning_rate": 0.0002967658593224515,
      "loss": 3.6114,
      "step": 5520
    },
    {
      "epoch": 0.07248276404435473,
      "grad_norm": 0.5471339821815491,
      "learning_rate": 0.00029675300563792553,
      "loss": 3.6315,
      "step": 5530
    },
    {
      "epoch": 0.0726138359504024,
      "grad_norm": 0.517301082611084,
      "learning_rate": 0.0002967401267408312,
      "loss": 3.6175,
      "step": 5540
    },
    {
      "epoch": 0.07274490785645005,
      "grad_norm": 0.4708433449268341,
      "learning_rate": 0.00029672722263338107,
      "loss": 3.6615,
      "step": 5550
    },
    {
      "epoch": 0.07287597976249771,
      "grad_norm": 0.4930337369441986,
      "learning_rate": 0.00029671429331779215,
      "loss": 3.6275,
      "step": 5560
    },
    {
      "epoch": 0.07300705166854536,
      "grad_norm": 0.47179871797561646,
      "learning_rate": 0.00029670133879628575,
      "loss": 3.5884,
      "step": 5570
    },
    {
      "epoch": 0.07313812357459303,
      "grad_norm": 0.5165817141532898,
      "learning_rate": 0.0002966883590710875,
      "loss": 3.6039,
      "step": 5580
    },
    {
      "epoch": 0.07326919548064068,
      "grad_norm": 0.4755318760871887,
      "learning_rate": 0.0002966753541444273,
      "loss": 3.6264,
      "step": 5590
    },
    {
      "epoch": 0.07340026738668834,
      "grad_norm": 0.4987680912017822,
      "learning_rate": 0.00029666232401853943,
      "loss": 3.6492,
      "step": 5600
    },
    {
      "epoch": 0.07353133929273599,
      "grad_norm": 0.4880028963088989,
      "learning_rate": 0.0002966492686956626,
      "loss": 3.619,
      "step": 5610
    },
    {
      "epoch": 0.07366241119878365,
      "grad_norm": 0.4575103521347046,
      "learning_rate": 0.0002966361881780396,
      "loss": 3.6273,
      "step": 5620
    },
    {
      "epoch": 0.07379348310483132,
      "grad_norm": 0.5023475885391235,
      "learning_rate": 0.0002966230824679178,
      "loss": 3.6423,
      "step": 5630
    },
    {
      "epoch": 0.07392455501087897,
      "grad_norm": 0.5324191451072693,
      "learning_rate": 0.00029660995156754875,
      "loss": 3.61,
      "step": 5640
    },
    {
      "epoch": 0.07405562691692663,
      "grad_norm": 0.5010618567466736,
      "learning_rate": 0.0002965967954791884,
      "loss": 3.6221,
      "step": 5650
    },
    {
      "epoch": 0.07418669882297428,
      "grad_norm": 0.4758966863155365,
      "learning_rate": 0.0002965836142050971,
      "loss": 3.6339,
      "step": 5660
    },
    {
      "epoch": 0.07431777072902195,
      "grad_norm": 0.5429836511611938,
      "learning_rate": 0.0002965704077475392,
      "loss": 3.6158,
      "step": 5670
    },
    {
      "epoch": 0.0744488426350696,
      "grad_norm": 0.47053462266921997,
      "learning_rate": 0.00029655717610878383,
      "loss": 3.6357,
      "step": 5680
    },
    {
      "epoch": 0.07457991454111726,
      "grad_norm": 0.49886468052864075,
      "learning_rate": 0.0002965439192911041,
      "loss": 3.5996,
      "step": 5690
    },
    {
      "epoch": 0.07471098644716491,
      "grad_norm": 0.4686560332775116,
      "learning_rate": 0.00029653063729677756,
      "loss": 3.6218,
      "step": 5700
    },
    {
      "epoch": 0.07484205835321257,
      "grad_norm": 0.4660986065864563,
      "learning_rate": 0.00029651733012808616,
      "loss": 3.6113,
      "step": 5710
    },
    {
      "epoch": 0.07497313025926022,
      "grad_norm": 0.46515005826950073,
      "learning_rate": 0.0002965039977873161,
      "loss": 3.5852,
      "step": 5720
    },
    {
      "epoch": 0.07510420216530789,
      "grad_norm": 0.5323407649993896,
      "learning_rate": 0.00029649064027675783,
      "loss": 3.6281,
      "step": 5730
    },
    {
      "epoch": 0.07523527407135555,
      "grad_norm": 0.4844474792480469,
      "learning_rate": 0.00029647725759870627,
      "loss": 3.6191,
      "step": 5740
    },
    {
      "epoch": 0.0753663459774032,
      "grad_norm": 0.5009616613388062,
      "learning_rate": 0.0002964638497554606,
      "loss": 3.5831,
      "step": 5750
    },
    {
      "epoch": 0.07549741788345087,
      "grad_norm": 0.5001603364944458,
      "learning_rate": 0.00029645041674932436,
      "loss": 3.6572,
      "step": 5760
    },
    {
      "epoch": 0.07562848978949852,
      "grad_norm": 0.5076395273208618,
      "learning_rate": 0.0002964369585826053,
      "loss": 3.6371,
      "step": 5770
    },
    {
      "epoch": 0.07575956169554618,
      "grad_norm": 0.4922044277191162,
      "learning_rate": 0.00029642347525761563,
      "loss": 3.6136,
      "step": 5780
    },
    {
      "epoch": 0.07589063360159383,
      "grad_norm": 0.48166152834892273,
      "learning_rate": 0.00029640996677667177,
      "loss": 3.6764,
      "step": 5790
    },
    {
      "epoch": 0.0760217055076415,
      "grad_norm": 0.4808330833911896,
      "learning_rate": 0.0002963964331420946,
      "loss": 3.6406,
      "step": 5800
    },
    {
      "epoch": 0.07615277741368914,
      "grad_norm": 0.4927902817726135,
      "learning_rate": 0.0002963828743562091,
      "loss": 3.6103,
      "step": 5810
    },
    {
      "epoch": 0.07628384931973681,
      "grad_norm": 0.47091490030288696,
      "learning_rate": 0.00029636929042134484,
      "loss": 3.5437,
      "step": 5820
    },
    {
      "epoch": 0.07641492122578447,
      "grad_norm": 0.5401109457015991,
      "learning_rate": 0.0002963556813398355,
      "loss": 3.6095,
      "step": 5830
    },
    {
      "epoch": 0.07654599313183212,
      "grad_norm": 0.6588603258132935,
      "learning_rate": 0.00029634204711401923,
      "loss": 3.6749,
      "step": 5840
    },
    {
      "epoch": 0.07667706503787979,
      "grad_norm": 0.5430197715759277,
      "learning_rate": 0.00029632838774623833,
      "loss": 3.6395,
      "step": 5850
    },
    {
      "epoch": 0.07680813694392744,
      "grad_norm": 0.47560471296310425,
      "learning_rate": 0.00029631470323883955,
      "loss": 3.6697,
      "step": 5860
    },
    {
      "epoch": 0.0769392088499751,
      "grad_norm": 0.5048189163208008,
      "learning_rate": 0.000296300993594174,
      "loss": 3.6027,
      "step": 5870
    },
    {
      "epoch": 0.07707028075602275,
      "grad_norm": 0.471696674823761,
      "learning_rate": 0.00029628725881459693,
      "loss": 3.5997,
      "step": 5880
    },
    {
      "epoch": 0.07720135266207041,
      "grad_norm": 0.4908413887023926,
      "learning_rate": 0.000296273498902468,
      "loss": 3.6414,
      "step": 5890
    },
    {
      "epoch": 0.07733242456811806,
      "grad_norm": 0.46346020698547363,
      "learning_rate": 0.0002962597138601514,
      "loss": 3.597,
      "step": 5900
    },
    {
      "epoch": 0.07746349647416573,
      "grad_norm": 0.4762059152126312,
      "learning_rate": 0.0002962459036900152,
      "loss": 3.5985,
      "step": 5910
    },
    {
      "epoch": 0.07759456838021339,
      "grad_norm": 0.509049117565155,
      "learning_rate": 0.0002962320683944321,
      "loss": 3.6464,
      "step": 5920
    },
    {
      "epoch": 0.07772564028626104,
      "grad_norm": 0.48113760352134705,
      "learning_rate": 0.0002962182079757791,
      "loss": 3.6072,
      "step": 5930
    },
    {
      "epoch": 0.0778567121923087,
      "grad_norm": 0.5319749712944031,
      "learning_rate": 0.0002962043224364374,
      "loss": 3.6365,
      "step": 5940
    },
    {
      "epoch": 0.07798778409835636,
      "grad_norm": 0.4811880886554718,
      "learning_rate": 0.00029619041177879265,
      "loss": 3.5996,
      "step": 5950
    },
    {
      "epoch": 0.07811885600440402,
      "grad_norm": 0.5448269844055176,
      "learning_rate": 0.0002961764760052347,
      "loss": 3.6289,
      "step": 5960
    },
    {
      "epoch": 0.07824992791045167,
      "grad_norm": 0.5355154871940613,
      "learning_rate": 0.00029616251511815764,
      "loss": 3.6001,
      "step": 5970
    },
    {
      "epoch": 0.07838099981649933,
      "grad_norm": 0.5143796801567078,
      "learning_rate": 0.0002961485291199601,
      "loss": 3.6674,
      "step": 5980
    },
    {
      "epoch": 0.07851207172254698,
      "grad_norm": 0.44844329357147217,
      "learning_rate": 0.00029613451801304493,
      "loss": 3.6245,
      "step": 5990
    },
    {
      "epoch": 0.07864314362859465,
      "grad_norm": 0.5102342367172241,
      "learning_rate": 0.00029612048179981926,
      "loss": 3.5868,
      "step": 6000
    },
    {
      "epoch": 0.07864314362859465,
      "eval_accuracy": 0.3598147605083089,
      "eval_loss": 3.597665786743164,
      "eval_runtime": 62.0575,
      "eval_samples_per_second": 32.228,
      "eval_steps_per_second": 0.516,
      "step": 6000
    },
    {
      "epoch": 0.07877421553464231,
      "grad_norm": 0.4596799612045288,
      "learning_rate": 0.00029610642048269453,
      "loss": 3.6254,
      "step": 6010
    },
    {
      "epoch": 0.07890528744068996,
      "grad_norm": 0.47384923696517944,
      "learning_rate": 0.00029609233406408653,
      "loss": 3.6126,
      "step": 6020
    },
    {
      "epoch": 0.07903635934673763,
      "grad_norm": 0.5375943183898926,
      "learning_rate": 0.0002960782225464153,
      "loss": 3.6433,
      "step": 6030
    },
    {
      "epoch": 0.07916743125278528,
      "grad_norm": 0.4941160976886749,
      "learning_rate": 0.00029606408593210525,
      "loss": 3.5984,
      "step": 6040
    },
    {
      "epoch": 0.07929850315883294,
      "grad_norm": 0.513177752494812,
      "learning_rate": 0.0002960499242235852,
      "loss": 3.5971,
      "step": 6050
    },
    {
      "epoch": 0.07942957506488059,
      "grad_norm": 0.4601130783557892,
      "learning_rate": 0.000296035737423288,
      "loss": 3.5837,
      "step": 6060
    },
    {
      "epoch": 0.07956064697092825,
      "grad_norm": 0.4826773703098297,
      "learning_rate": 0.0002960215255336511,
      "loss": 3.6205,
      "step": 6070
    },
    {
      "epoch": 0.0796917188769759,
      "grad_norm": 0.48354971408843994,
      "learning_rate": 0.00029600728855711607,
      "loss": 3.6058,
      "step": 6080
    },
    {
      "epoch": 0.07982279078302357,
      "grad_norm": 0.5530339479446411,
      "learning_rate": 0.00029599302649612895,
      "loss": 3.577,
      "step": 6090
    },
    {
      "epoch": 0.07995386268907123,
      "grad_norm": 0.5434287190437317,
      "learning_rate": 0.00029597873935313993,
      "loss": 3.6517,
      "step": 6100
    },
    {
      "epoch": 0.08008493459511888,
      "grad_norm": 0.8914490342140198,
      "learning_rate": 0.0002959644271306035,
      "loss": 3.5986,
      "step": 6110
    },
    {
      "epoch": 0.08021600650116655,
      "grad_norm": 0.5456758141517639,
      "learning_rate": 0.00029595008983097877,
      "loss": 3.6027,
      "step": 6120
    },
    {
      "epoch": 0.0803470784072142,
      "grad_norm": 0.4839215874671936,
      "learning_rate": 0.0002959357274567287,
      "loss": 3.595,
      "step": 6130
    },
    {
      "epoch": 0.08047815031326186,
      "grad_norm": 0.5667473673820496,
      "learning_rate": 0.00029592134001032094,
      "loss": 3.5895,
      "step": 6140
    },
    {
      "epoch": 0.08060922221930951,
      "grad_norm": 0.5036835074424744,
      "learning_rate": 0.00029590692749422717,
      "loss": 3.61,
      "step": 6150
    },
    {
      "epoch": 0.08074029412535717,
      "grad_norm": 0.5362805724143982,
      "learning_rate": 0.0002958924899109236,
      "loss": 3.5802,
      "step": 6160
    },
    {
      "epoch": 0.08087136603140482,
      "grad_norm": 0.4782995879650116,
      "learning_rate": 0.00029587802726289064,
      "loss": 3.6181,
      "step": 6170
    },
    {
      "epoch": 0.08100243793745249,
      "grad_norm": 0.46912720799446106,
      "learning_rate": 0.0002958635395526129,
      "loss": 3.6466,
      "step": 6180
    },
    {
      "epoch": 0.08113350984350014,
      "grad_norm": 0.5034590363502502,
      "learning_rate": 0.0002958490267825795,
      "loss": 3.6329,
      "step": 6190
    },
    {
      "epoch": 0.0812645817495478,
      "grad_norm": 0.4898591935634613,
      "learning_rate": 0.0002958344889552838,
      "loss": 3.6044,
      "step": 6200
    },
    {
      "epoch": 0.08139565365559547,
      "grad_norm": 0.49792569875717163,
      "learning_rate": 0.0002958199260732234,
      "loss": 3.6333,
      "step": 6210
    },
    {
      "epoch": 0.08152672556164312,
      "grad_norm": 0.5021080374717712,
      "learning_rate": 0.00029580533813890024,
      "loss": 3.5974,
      "step": 6220
    },
    {
      "epoch": 0.08165779746769078,
      "grad_norm": 0.5469182133674622,
      "learning_rate": 0.00029579072515482056,
      "loss": 3.5965,
      "step": 6230
    },
    {
      "epoch": 0.08178886937373843,
      "grad_norm": 0.5059979557991028,
      "learning_rate": 0.0002957760871234949,
      "loss": 3.6119,
      "step": 6240
    },
    {
      "epoch": 0.0819199412797861,
      "grad_norm": 0.49016863107681274,
      "learning_rate": 0.00029576142404743813,
      "loss": 3.5939,
      "step": 6250
    },
    {
      "epoch": 0.08205101318583374,
      "grad_norm": 0.4785931706428528,
      "learning_rate": 0.00029574673592916944,
      "loss": 3.5804,
      "step": 6260
    },
    {
      "epoch": 0.08218208509188141,
      "grad_norm": 0.48176640272140503,
      "learning_rate": 0.00029573202277121226,
      "loss": 3.6031,
      "step": 6270
    },
    {
      "epoch": 0.08231315699792906,
      "grad_norm": 0.5348442792892456,
      "learning_rate": 0.0002957172845760943,
      "loss": 3.5713,
      "step": 6280
    },
    {
      "epoch": 0.08244422890397672,
      "grad_norm": 0.47209852933883667,
      "learning_rate": 0.0002957025213463478,
      "loss": 3.598,
      "step": 6290
    },
    {
      "epoch": 0.08257530081002439,
      "grad_norm": 0.4687283933162689,
      "learning_rate": 0.0002956877330845089,
      "loss": 3.5578,
      "step": 6300
    },
    {
      "epoch": 0.08270637271607204,
      "grad_norm": 0.4971533715724945,
      "learning_rate": 0.0002956729197931183,
      "loss": 3.6386,
      "step": 6310
    },
    {
      "epoch": 0.0828374446221197,
      "grad_norm": 0.4881878197193146,
      "learning_rate": 0.00029565808147472114,
      "loss": 3.593,
      "step": 6320
    },
    {
      "epoch": 0.08296851652816735,
      "grad_norm": 0.5033032298088074,
      "learning_rate": 0.0002956432181318665,
      "loss": 3.5918,
      "step": 6330
    },
    {
      "epoch": 0.08309958843421501,
      "grad_norm": 0.4330627918243408,
      "learning_rate": 0.0002956283297671081,
      "loss": 3.5864,
      "step": 6340
    },
    {
      "epoch": 0.08323066034026266,
      "grad_norm": 0.4908876419067383,
      "learning_rate": 0.0002956134163830036,
      "loss": 3.5721,
      "step": 6350
    },
    {
      "epoch": 0.08336173224631033,
      "grad_norm": 0.4548991024494171,
      "learning_rate": 0.0002955984779821153,
      "loss": 3.6188,
      "step": 6360
    },
    {
      "epoch": 0.08349280415235798,
      "grad_norm": 0.5071555972099304,
      "learning_rate": 0.0002955835145670097,
      "loss": 3.5869,
      "step": 6370
    },
    {
      "epoch": 0.08362387605840564,
      "grad_norm": 0.488614559173584,
      "learning_rate": 0.0002955685261402574,
      "loss": 3.5946,
      "step": 6380
    },
    {
      "epoch": 0.0837549479644533,
      "grad_norm": 0.48327845335006714,
      "learning_rate": 0.0002955535127044336,
      "loss": 3.5836,
      "step": 6390
    },
    {
      "epoch": 0.08388601987050096,
      "grad_norm": 0.6187971830368042,
      "learning_rate": 0.00029553847426211756,
      "loss": 3.6075,
      "step": 6400
    },
    {
      "epoch": 0.08401709177654862,
      "grad_norm": 0.5842393636703491,
      "learning_rate": 0.000295523410815893,
      "loss": 3.6403,
      "step": 6410
    },
    {
      "epoch": 0.08414816368259627,
      "grad_norm": 0.475318044424057,
      "learning_rate": 0.0002955083223683477,
      "loss": 3.5948,
      "step": 6420
    },
    {
      "epoch": 0.08427923558864393,
      "grad_norm": 0.5049416422843933,
      "learning_rate": 0.0002954932089220741,
      "loss": 3.5718,
      "step": 6430
    },
    {
      "epoch": 0.08441030749469158,
      "grad_norm": 0.438046932220459,
      "learning_rate": 0.0002954780704796687,
      "loss": 3.5789,
      "step": 6440
    },
    {
      "epoch": 0.08454137940073925,
      "grad_norm": 0.4366346299648285,
      "learning_rate": 0.0002954629070437322,
      "loss": 3.5555,
      "step": 6450
    },
    {
      "epoch": 0.0846724513067869,
      "grad_norm": 0.5139963626861572,
      "learning_rate": 0.0002954477186168698,
      "loss": 3.5847,
      "step": 6460
    },
    {
      "epoch": 0.08480352321283456,
      "grad_norm": 0.4718990623950958,
      "learning_rate": 0.00029543250520169085,
      "loss": 3.6012,
      "step": 6470
    },
    {
      "epoch": 0.08493459511888223,
      "grad_norm": 0.5247958898544312,
      "learning_rate": 0.00029541726680080917,
      "loss": 3.6666,
      "step": 6480
    },
    {
      "epoch": 0.08506566702492988,
      "grad_norm": 0.5552542805671692,
      "learning_rate": 0.0002954020034168427,
      "loss": 3.5825,
      "step": 6490
    },
    {
      "epoch": 0.08519673893097754,
      "grad_norm": 0.47670796513557434,
      "learning_rate": 0.00029538671505241373,
      "loss": 3.6041,
      "step": 6500
    },
    {
      "epoch": 0.08519673893097754,
      "eval_accuracy": 0.36129130009775173,
      "eval_loss": 3.5803511142730713,
      "eval_runtime": 60.797,
      "eval_samples_per_second": 32.896,
      "eval_steps_per_second": 0.526,
      "step": 6500
    },
    {
      "epoch": 0.08532781083702519,
      "grad_norm": 0.46810227632522583,
      "learning_rate": 0.0002953714017101488,
      "loss": 3.5733,
      "step": 6510
    },
    {
      "epoch": 0.08545888274307285,
      "grad_norm": 0.45893773436546326,
      "learning_rate": 0.0002953560633926789,
      "loss": 3.5315,
      "step": 6520
    },
    {
      "epoch": 0.0855899546491205,
      "grad_norm": 0.45480263233184814,
      "learning_rate": 0.00029534070010263906,
      "loss": 3.5776,
      "step": 6530
    },
    {
      "epoch": 0.08572102655516817,
      "grad_norm": 0.4754258394241333,
      "learning_rate": 0.0002953253118426688,
      "loss": 3.5857,
      "step": 6540
    },
    {
      "epoch": 0.08585209846121582,
      "grad_norm": 0.4808047413825989,
      "learning_rate": 0.0002953098986154119,
      "loss": 3.6145,
      "step": 6550
    },
    {
      "epoch": 0.08598317036726348,
      "grad_norm": 0.4620581865310669,
      "learning_rate": 0.0002952944604235163,
      "loss": 3.5703,
      "step": 6560
    },
    {
      "epoch": 0.08611424227331115,
      "grad_norm": 0.5404554009437561,
      "learning_rate": 0.00029527899726963444,
      "loss": 3.5888,
      "step": 6570
    },
    {
      "epoch": 0.0862453141793588,
      "grad_norm": 0.4914781153202057,
      "learning_rate": 0.00029526350915642284,
      "loss": 3.6239,
      "step": 6580
    },
    {
      "epoch": 0.08637638608540646,
      "grad_norm": 0.4948068857192993,
      "learning_rate": 0.00029524799608654246,
      "loss": 3.5759,
      "step": 6590
    },
    {
      "epoch": 0.08650745799145411,
      "grad_norm": 0.4793195426464081,
      "learning_rate": 0.0002952324580626584,
      "loss": 3.5855,
      "step": 6600
    },
    {
      "epoch": 0.08663852989750177,
      "grad_norm": 0.5116032958030701,
      "learning_rate": 0.00029521689508744026,
      "loss": 3.6058,
      "step": 6610
    },
    {
      "epoch": 0.08676960180354942,
      "grad_norm": 0.5280539989471436,
      "learning_rate": 0.0002952013071635617,
      "loss": 3.5574,
      "step": 6620
    },
    {
      "epoch": 0.08690067370959709,
      "grad_norm": 0.5247043371200562,
      "learning_rate": 0.0002951856942937008,
      "loss": 3.5841,
      "step": 6630
    },
    {
      "epoch": 0.08703174561564474,
      "grad_norm": 0.5255647301673889,
      "learning_rate": 0.00029517005648053995,
      "loss": 3.5602,
      "step": 6640
    },
    {
      "epoch": 0.0871628175216924,
      "grad_norm": 0.46197646856307983,
      "learning_rate": 0.0002951543937267657,
      "loss": 3.6182,
      "step": 6650
    },
    {
      "epoch": 0.08729388942774007,
      "grad_norm": 0.48573994636535645,
      "learning_rate": 0.00029513870603506897,
      "loss": 3.5896,
      "step": 6660
    },
    {
      "epoch": 0.08742496133378772,
      "grad_norm": 0.49037501215934753,
      "learning_rate": 0.00029512299340814495,
      "loss": 3.6131,
      "step": 6670
    },
    {
      "epoch": 0.08755603323983538,
      "grad_norm": 0.4632072448730469,
      "learning_rate": 0.0002951072558486931,
      "loss": 3.6018,
      "step": 6680
    },
    {
      "epoch": 0.08768710514588303,
      "grad_norm": 0.47344645857810974,
      "learning_rate": 0.00029509149335941716,
      "loss": 3.5462,
      "step": 6690
    },
    {
      "epoch": 0.0878181770519307,
      "grad_norm": 0.47739177942276,
      "learning_rate": 0.00029507570594302523,
      "loss": 3.5418,
      "step": 6700
    },
    {
      "epoch": 0.08794924895797834,
      "grad_norm": 0.48223811388015747,
      "learning_rate": 0.00029505989360222967,
      "loss": 3.5923,
      "step": 6710
    },
    {
      "epoch": 0.08808032086402601,
      "grad_norm": 0.46866801381111145,
      "learning_rate": 0.0002950440563397469,
      "loss": 3.5822,
      "step": 6720
    },
    {
      "epoch": 0.08821139277007366,
      "grad_norm": 0.46003222465515137,
      "learning_rate": 0.00029502819415829797,
      "loss": 3.5526,
      "step": 6730
    },
    {
      "epoch": 0.08834246467612132,
      "grad_norm": 0.4742317497730255,
      "learning_rate": 0.000295012307060608,
      "loss": 3.6064,
      "step": 6740
    },
    {
      "epoch": 0.08847353658216897,
      "grad_norm": 0.5065898895263672,
      "learning_rate": 0.00029499639504940644,
      "loss": 3.6302,
      "step": 6750
    },
    {
      "epoch": 0.08860460848821664,
      "grad_norm": 0.7090433239936829,
      "learning_rate": 0.00029498045812742697,
      "loss": 3.6591,
      "step": 6760
    },
    {
      "epoch": 0.0887356803942643,
      "grad_norm": 0.4633511006832123,
      "learning_rate": 0.0002949644962974077,
      "loss": 3.609,
      "step": 6770
    },
    {
      "epoch": 0.08886675230031195,
      "grad_norm": 0.48159927129745483,
      "learning_rate": 0.0002949485095620908,
      "loss": 3.5617,
      "step": 6780
    },
    {
      "epoch": 0.08899782420635961,
      "grad_norm": 0.5400639772415161,
      "learning_rate": 0.0002949324979242229,
      "loss": 3.5979,
      "step": 6790
    },
    {
      "epoch": 0.08912889611240726,
      "grad_norm": 0.4782787561416626,
      "learning_rate": 0.0002949164613865548,
      "loss": 3.6082,
      "step": 6800
    },
    {
      "epoch": 0.08925996801845493,
      "grad_norm": 0.4989713430404663,
      "learning_rate": 0.0002949003999518416,
      "loss": 3.6175,
      "step": 6810
    },
    {
      "epoch": 0.08939103992450258,
      "grad_norm": 0.47920992970466614,
      "learning_rate": 0.00029488431362284284,
      "loss": 3.5905,
      "step": 6820
    },
    {
      "epoch": 0.08952211183055024,
      "grad_norm": 0.49157336354255676,
      "learning_rate": 0.0002948682024023221,
      "loss": 3.6266,
      "step": 6830
    },
    {
      "epoch": 0.08965318373659789,
      "grad_norm": 0.49132001399993896,
      "learning_rate": 0.00029485206629304735,
      "loss": 3.6072,
      "step": 6840
    },
    {
      "epoch": 0.08978425564264556,
      "grad_norm": 0.46873581409454346,
      "learning_rate": 0.0002948359052977908,
      "loss": 3.5741,
      "step": 6850
    },
    {
      "epoch": 0.08991532754869322,
      "grad_norm": 0.47375303506851196,
      "learning_rate": 0.0002948197194193289,
      "loss": 3.6192,
      "step": 6860
    },
    {
      "epoch": 0.09004639945474087,
      "grad_norm": 0.47037723660469055,
      "learning_rate": 0.00029480350866044255,
      "loss": 3.4951,
      "step": 6870
    },
    {
      "epoch": 0.09017747136078853,
      "grad_norm": 0.46293631196022034,
      "learning_rate": 0.00029478727302391665,
      "loss": 3.5775,
      "step": 6880
    },
    {
      "epoch": 0.09030854326683618,
      "grad_norm": 0.4646480977535248,
      "learning_rate": 0.0002947710125125407,
      "loss": 3.5841,
      "step": 6890
    },
    {
      "epoch": 0.09043961517288385,
      "grad_norm": 0.478644996881485,
      "learning_rate": 0.0002947547271291082,
      "loss": 3.5721,
      "step": 6900
    },
    {
      "epoch": 0.0905706870789315,
      "grad_norm": 0.46336767077445984,
      "learning_rate": 0.000294738416876417,
      "loss": 3.5704,
      "step": 6910
    },
    {
      "epoch": 0.09070175898497916,
      "grad_norm": 0.6859182715415955,
      "learning_rate": 0.00029472208175726933,
      "loss": 3.6357,
      "step": 6920
    },
    {
      "epoch": 0.09083283089102681,
      "grad_norm": 0.4686596095561981,
      "learning_rate": 0.0002947057217744716,
      "loss": 3.5719,
      "step": 6930
    },
    {
      "epoch": 0.09096390279707448,
      "grad_norm": 0.4931657016277313,
      "learning_rate": 0.0002946893369308344,
      "loss": 3.5498,
      "step": 6940
    },
    {
      "epoch": 0.09109497470312214,
      "grad_norm": 0.4427664279937744,
      "learning_rate": 0.00029467292722917273,
      "loss": 3.5498,
      "step": 6950
    },
    {
      "epoch": 0.09122604660916979,
      "grad_norm": 0.4552489221096039,
      "learning_rate": 0.0002946564926723059,
      "loss": 3.5616,
      "step": 6960
    },
    {
      "epoch": 0.09135711851521745,
      "grad_norm": 0.442079097032547,
      "learning_rate": 0.00029464003326305733,
      "loss": 3.5964,
      "step": 6970
    },
    {
      "epoch": 0.0914881904212651,
      "grad_norm": 0.6060337424278259,
      "learning_rate": 0.00029462354900425486,
      "loss": 3.6194,
      "step": 6980
    },
    {
      "epoch": 0.09161926232731277,
      "grad_norm": 0.49838823080062866,
      "learning_rate": 0.00029460703989873043,
      "loss": 3.5856,
      "step": 6990
    },
    {
      "epoch": 0.09175033423336042,
      "grad_norm": 0.48732805252075195,
      "learning_rate": 0.00029459050594932043,
      "loss": 3.5936,
      "step": 7000
    },
    {
      "epoch": 0.09175033423336042,
      "eval_accuracy": 0.36401026392961877,
      "eval_loss": 3.560344696044922,
      "eval_runtime": 60.1149,
      "eval_samples_per_second": 33.27,
      "eval_steps_per_second": 0.532,
      "step": 7000
    },
    {
      "epoch": 0.09188140613940808,
      "grad_norm": 0.43881261348724365,
      "learning_rate": 0.0002945739471588654,
      "loss": 3.5713,
      "step": 7010
    },
    {
      "epoch": 0.09201247804545573,
      "grad_norm": 0.43614882230758667,
      "learning_rate": 0.00029455736353021025,
      "loss": 3.5778,
      "step": 7020
    },
    {
      "epoch": 0.0921435499515034,
      "grad_norm": 0.4552080035209656,
      "learning_rate": 0.00029454075506620404,
      "loss": 3.5531,
      "step": 7030
    },
    {
      "epoch": 0.09227462185755106,
      "grad_norm": 0.4761829674243927,
      "learning_rate": 0.0002945241217697001,
      "loss": 3.5604,
      "step": 7040
    },
    {
      "epoch": 0.09240569376359871,
      "grad_norm": 0.46336135268211365,
      "learning_rate": 0.0002945074636435562,
      "loss": 3.5628,
      "step": 7050
    },
    {
      "epoch": 0.09253676566964637,
      "grad_norm": 0.4259665310382843,
      "learning_rate": 0.00029449078069063416,
      "loss": 3.5787,
      "step": 7060
    },
    {
      "epoch": 0.09266783757569402,
      "grad_norm": 0.5578000545501709,
      "learning_rate": 0.0002944740729138002,
      "loss": 3.5903,
      "step": 7070
    },
    {
      "epoch": 0.09279890948174169,
      "grad_norm": 0.4593062698841095,
      "learning_rate": 0.0002944573403159247,
      "loss": 3.5929,
      "step": 7080
    },
    {
      "epoch": 0.09292998138778934,
      "grad_norm": 0.4525621831417084,
      "learning_rate": 0.0002944405828998824,
      "loss": 3.5851,
      "step": 7090
    },
    {
      "epoch": 0.093061053293837,
      "grad_norm": 0.46507716178894043,
      "learning_rate": 0.00029442380066855233,
      "loss": 3.5415,
      "step": 7100
    },
    {
      "epoch": 0.09319212519988465,
      "grad_norm": 0.5256420373916626,
      "learning_rate": 0.00029440699362481766,
      "loss": 3.5509,
      "step": 7110
    },
    {
      "epoch": 0.09332319710593232,
      "grad_norm": 0.5470519661903381,
      "learning_rate": 0.0002943901617715659,
      "loss": 3.55,
      "step": 7120
    },
    {
      "epoch": 0.09345426901197998,
      "grad_norm": 0.6189631223678589,
      "learning_rate": 0.0002943733051116888,
      "loss": 3.5695,
      "step": 7130
    },
    {
      "epoch": 0.09358534091802763,
      "grad_norm": 0.5014888644218445,
      "learning_rate": 0.00029435642364808234,
      "loss": 3.5517,
      "step": 7140
    },
    {
      "epoch": 0.09371641282407529,
      "grad_norm": 0.4354492127895355,
      "learning_rate": 0.0002943395173836469,
      "loss": 3.5816,
      "step": 7150
    },
    {
      "epoch": 0.09384748473012294,
      "grad_norm": 0.44574570655822754,
      "learning_rate": 0.00029432258632128695,
      "loss": 3.5582,
      "step": 7160
    },
    {
      "epoch": 0.09397855663617061,
      "grad_norm": 0.4585030674934387,
      "learning_rate": 0.0002943056304639113,
      "loss": 3.5786,
      "step": 7170
    },
    {
      "epoch": 0.09410962854221826,
      "grad_norm": 0.47453171014785767,
      "learning_rate": 0.00029428864981443303,
      "loss": 3.5703,
      "step": 7180
    },
    {
      "epoch": 0.09424070044826592,
      "grad_norm": 0.4796355366706848,
      "learning_rate": 0.0002942716443757695,
      "loss": 3.5586,
      "step": 7190
    },
    {
      "epoch": 0.09437177235431357,
      "grad_norm": 0.45695897936820984,
      "learning_rate": 0.00029425461415084225,
      "loss": 3.5855,
      "step": 7200
    },
    {
      "epoch": 0.09450284426036124,
      "grad_norm": 0.4512489438056946,
      "learning_rate": 0.0002942375591425771,
      "loss": 3.556,
      "step": 7210
    },
    {
      "epoch": 0.09463391616640889,
      "grad_norm": 0.5440936088562012,
      "learning_rate": 0.00029422047935390416,
      "loss": 3.5431,
      "step": 7220
    },
    {
      "epoch": 0.09476498807245655,
      "grad_norm": 0.48467716574668884,
      "learning_rate": 0.0002942033747877578,
      "loss": 3.5598,
      "step": 7230
    },
    {
      "epoch": 0.09489605997850421,
      "grad_norm": 0.490763396024704,
      "learning_rate": 0.0002941862454470766,
      "loss": 3.5649,
      "step": 7240
    },
    {
      "epoch": 0.09502713188455186,
      "grad_norm": 0.5141284465789795,
      "learning_rate": 0.0002941690913348035,
      "loss": 3.5035,
      "step": 7250
    },
    {
      "epoch": 0.09515820379059953,
      "grad_norm": 0.5250519514083862,
      "learning_rate": 0.0002941519124538855,
      "loss": 3.5512,
      "step": 7260
    },
    {
      "epoch": 0.09528927569664718,
      "grad_norm": 0.49070683121681213,
      "learning_rate": 0.00029413470880727404,
      "loss": 3.5083,
      "step": 7270
    },
    {
      "epoch": 0.09542034760269484,
      "grad_norm": 0.48466047644615173,
      "learning_rate": 0.0002941174803979248,
      "loss": 3.5536,
      "step": 7280
    },
    {
      "epoch": 0.09555141950874249,
      "grad_norm": 0.44810783863067627,
      "learning_rate": 0.0002941002272287977,
      "loss": 3.5153,
      "step": 7290
    },
    {
      "epoch": 0.09568249141479016,
      "grad_norm": 0.4396667182445526,
      "learning_rate": 0.0002940829493028568,
      "loss": 3.5212,
      "step": 7300
    },
    {
      "epoch": 0.0958135633208378,
      "grad_norm": 0.488755464553833,
      "learning_rate": 0.0002940656466230704,
      "loss": 3.5599,
      "step": 7310
    },
    {
      "epoch": 0.09594463522688547,
      "grad_norm": 0.45529067516326904,
      "learning_rate": 0.00029404831919241137,
      "loss": 3.552,
      "step": 7320
    },
    {
      "epoch": 0.09607570713293313,
      "grad_norm": 0.45739734172821045,
      "learning_rate": 0.00029403096701385644,
      "loss": 3.524,
      "step": 7330
    },
    {
      "epoch": 0.09620677903898078,
      "grad_norm": 0.4366675615310669,
      "learning_rate": 0.0002940135900903868,
      "loss": 3.5244,
      "step": 7340
    },
    {
      "epoch": 0.09633785094502845,
      "grad_norm": 0.4646051824092865,
      "learning_rate": 0.00029399618842498794,
      "loss": 3.506,
      "step": 7350
    },
    {
      "epoch": 0.0964689228510761,
      "grad_norm": 0.43333420157432556,
      "learning_rate": 0.0002939787620206494,
      "loss": 3.5719,
      "step": 7360
    },
    {
      "epoch": 0.09659999475712376,
      "grad_norm": 0.4883551597595215,
      "learning_rate": 0.00029396131088036515,
      "loss": 3.5641,
      "step": 7370
    },
    {
      "epoch": 0.09673106666317141,
      "grad_norm": 0.4403475821018219,
      "learning_rate": 0.0002939438350071333,
      "loss": 3.5258,
      "step": 7380
    },
    {
      "epoch": 0.09686213856921908,
      "grad_norm": 0.4596627652645111,
      "learning_rate": 0.00029392633440395637,
      "loss": 3.5269,
      "step": 7390
    },
    {
      "epoch": 0.09699321047526673,
      "grad_norm": 0.47244781255722046,
      "learning_rate": 0.0002939088090738408,
      "loss": 3.5622,
      "step": 7400
    },
    {
      "epoch": 0.09712428238131439,
      "grad_norm": 0.4625895917415619,
      "learning_rate": 0.00029389125901979764,
      "loss": 3.5422,
      "step": 7410
    },
    {
      "epoch": 0.09725535428736205,
      "grad_norm": 0.47102269530296326,
      "learning_rate": 0.00029387368424484204,
      "loss": 3.5236,
      "step": 7420
    },
    {
      "epoch": 0.0973864261934097,
      "grad_norm": 0.4610661566257477,
      "learning_rate": 0.0002938560847519933,
      "loss": 3.4814,
      "step": 7430
    },
    {
      "epoch": 0.09751749809945737,
      "grad_norm": 0.476470947265625,
      "learning_rate": 0.0002938384605442752,
      "loss": 3.5808,
      "step": 7440
    },
    {
      "epoch": 0.09764857000550502,
      "grad_norm": 0.498090922832489,
      "learning_rate": 0.0002938208116247155,
      "loss": 3.5364,
      "step": 7450
    },
    {
      "epoch": 0.09777964191155268,
      "grad_norm": 0.455168753862381,
      "learning_rate": 0.00029380313799634637,
      "loss": 3.5205,
      "step": 7460
    },
    {
      "epoch": 0.09791071381760033,
      "grad_norm": 0.5717991590499878,
      "learning_rate": 0.0002937854396622042,
      "loss": 3.5475,
      "step": 7470
    },
    {
      "epoch": 0.098041785723648,
      "grad_norm": 0.5302655696868896,
      "learning_rate": 0.00029376771662532967,
      "loss": 3.5738,
      "step": 7480
    },
    {
      "epoch": 0.09817285762969565,
      "grad_norm": 0.46155864000320435,
      "learning_rate": 0.00029374996888876755,
      "loss": 3.5143,
      "step": 7490
    },
    {
      "epoch": 0.09830392953574331,
      "grad_norm": 0.4753029942512512,
      "learning_rate": 0.000293732196455567,
      "loss": 3.5633,
      "step": 7500
    },
    {
      "epoch": 0.09830392953574331,
      "eval_accuracy": 0.36511827956989246,
      "eval_loss": 3.5411956310272217,
      "eval_runtime": 58.321,
      "eval_samples_per_second": 34.293,
      "eval_steps_per_second": 0.549,
      "step": 7500
    },
    {
      "epoch": 0.09843500144179097,
      "grad_norm": 0.5458186268806458,
      "learning_rate": 0.00029371439932878135,
      "loss": 3.5422,
      "step": 7510
    },
    {
      "epoch": 0.09856607334783862,
      "grad_norm": 0.507274329662323,
      "learning_rate": 0.00029369657751146815,
      "loss": 3.5538,
      "step": 7520
    },
    {
      "epoch": 0.09869714525388629,
      "grad_norm": 0.5154192447662354,
      "learning_rate": 0.0002936787310066894,
      "loss": 3.5235,
      "step": 7530
    },
    {
      "epoch": 0.09882821715993394,
      "grad_norm": 0.4681265950202942,
      "learning_rate": 0.000293660859817511,
      "loss": 3.4941,
      "step": 7540
    },
    {
      "epoch": 0.0989592890659816,
      "grad_norm": 0.4517514407634735,
      "learning_rate": 0.0002936429639470033,
      "loss": 3.5172,
      "step": 7550
    },
    {
      "epoch": 0.09909036097202925,
      "grad_norm": 0.4764167368412018,
      "learning_rate": 0.000293625043398241,
      "loss": 3.5241,
      "step": 7560
    },
    {
      "epoch": 0.09922143287807692,
      "grad_norm": 0.4595801830291748,
      "learning_rate": 0.0002936070981743027,
      "loss": 3.4947,
      "step": 7570
    },
    {
      "epoch": 0.09935250478412457,
      "grad_norm": 0.449593186378479,
      "learning_rate": 0.0002935891282782716,
      "loss": 3.462,
      "step": 7580
    },
    {
      "epoch": 0.09948357669017223,
      "grad_norm": 0.47141632437705994,
      "learning_rate": 0.0002935711337132349,
      "loss": 3.5301,
      "step": 7590
    },
    {
      "epoch": 0.09961464859621989,
      "grad_norm": 0.431039422750473,
      "learning_rate": 0.00029355311448228417,
      "loss": 3.5215,
      "step": 7600
    },
    {
      "epoch": 0.09974572050226754,
      "grad_norm": 0.4559817314147949,
      "learning_rate": 0.00029353507058851506,
      "loss": 3.5492,
      "step": 7610
    },
    {
      "epoch": 0.09987679240831521,
      "grad_norm": 0.5659443736076355,
      "learning_rate": 0.00029351700203502767,
      "loss": 3.5193,
      "step": 7620
    },
    {
      "epoch": 0.10000786431436286,
      "grad_norm": 0.4569407105445862,
      "learning_rate": 0.0002934989088249261,
      "loss": 3.5582,
      "step": 7630
    },
    {
      "epoch": 0.10013893622041052,
      "grad_norm": 0.43534642457962036,
      "learning_rate": 0.000293480790961319,
      "loss": 3.52,
      "step": 7640
    },
    {
      "epoch": 0.10027000812645817,
      "grad_norm": 0.5535159707069397,
      "learning_rate": 0.00029346264844731893,
      "loss": 3.5071,
      "step": 7650
    },
    {
      "epoch": 0.10040108003250584,
      "grad_norm": 0.4479416310787201,
      "learning_rate": 0.00029344448128604283,
      "loss": 3.5117,
      "step": 7660
    },
    {
      "epoch": 0.10053215193855349,
      "grad_norm": 0.43781858682632446,
      "learning_rate": 0.0002934262894806119,
      "loss": 3.5399,
      "step": 7670
    },
    {
      "epoch": 0.10066322384460115,
      "grad_norm": 0.49844059348106384,
      "learning_rate": 0.00029340807303415153,
      "loss": 3.5649,
      "step": 7680
    },
    {
      "epoch": 0.1007942957506488,
      "grad_norm": 0.46877720952033997,
      "learning_rate": 0.0002933898319497914,
      "loss": 3.5462,
      "step": 7690
    },
    {
      "epoch": 0.10092536765669646,
      "grad_norm": 0.4481637477874756,
      "learning_rate": 0.0002933715662306653,
      "loss": 3.5202,
      "step": 7700
    },
    {
      "epoch": 0.10105643956274413,
      "grad_norm": 0.5598857402801514,
      "learning_rate": 0.00029335327587991136,
      "loss": 3.5352,
      "step": 7710
    },
    {
      "epoch": 0.10118751146879178,
      "grad_norm": 0.540005624294281,
      "learning_rate": 0.000293334960900672,
      "loss": 3.5646,
      "step": 7720
    },
    {
      "epoch": 0.10131858337483944,
      "grad_norm": 0.4770313501358032,
      "learning_rate": 0.0002933166212960937,
      "loss": 3.5449,
      "step": 7730
    },
    {
      "epoch": 0.10144965528088709,
      "grad_norm": 0.4288584887981415,
      "learning_rate": 0.00029329825706932714,
      "loss": 3.5574,
      "step": 7740
    },
    {
      "epoch": 0.10158072718693476,
      "grad_norm": 0.4816032350063324,
      "learning_rate": 0.0002932798682235276,
      "loss": 3.5358,
      "step": 7750
    },
    {
      "epoch": 0.1017117990929824,
      "grad_norm": 0.4582579433917999,
      "learning_rate": 0.0002932614547618541,
      "loss": 3.5067,
      "step": 7760
    },
    {
      "epoch": 0.10184287099903007,
      "grad_norm": 0.4778504967689514,
      "learning_rate": 0.0002932430166874703,
      "loss": 3.53,
      "step": 7770
    },
    {
      "epoch": 0.10197394290507772,
      "grad_norm": 0.47503966093063354,
      "learning_rate": 0.00029322455400354375,
      "loss": 3.54,
      "step": 7780
    },
    {
      "epoch": 0.10210501481112538,
      "grad_norm": 0.48486900329589844,
      "learning_rate": 0.00029320606671324654,
      "loss": 3.5303,
      "step": 7790
    },
    {
      "epoch": 0.10223608671717305,
      "grad_norm": 0.4400011897087097,
      "learning_rate": 0.00029318755481975474,
      "loss": 3.5235,
      "step": 7800
    },
    {
      "epoch": 0.1023671586232207,
      "grad_norm": 0.44769418239593506,
      "learning_rate": 0.00029316901832624875,
      "loss": 3.5339,
      "step": 7810
    },
    {
      "epoch": 0.10249823052926836,
      "grad_norm": 0.47030365467071533,
      "learning_rate": 0.0002931504572359133,
      "loss": 3.5271,
      "step": 7820
    },
    {
      "epoch": 0.10262930243531601,
      "grad_norm": 0.4971461892127991,
      "learning_rate": 0.0002931318715519371,
      "loss": 3.4976,
      "step": 7830
    },
    {
      "epoch": 0.10276037434136367,
      "grad_norm": 0.6969183087348938,
      "learning_rate": 0.00029311326127751325,
      "loss": 3.5682,
      "step": 7840
    },
    {
      "epoch": 0.10289144624741133,
      "grad_norm": 0.46795719861984253,
      "learning_rate": 0.0002930946264158391,
      "loss": 3.477,
      "step": 7850
    },
    {
      "epoch": 0.10302251815345899,
      "grad_norm": 0.4499169886112213,
      "learning_rate": 0.00029307596697011613,
      "loss": 3.5127,
      "step": 7860
    },
    {
      "epoch": 0.10315359005950664,
      "grad_norm": 0.46941062808036804,
      "learning_rate": 0.0002930572829435501,
      "loss": 3.5192,
      "step": 7870
    },
    {
      "epoch": 0.1032846619655543,
      "grad_norm": 0.5304393768310547,
      "learning_rate": 0.000293038574339351,
      "loss": 3.511,
      "step": 7880
    },
    {
      "epoch": 0.10341573387160197,
      "grad_norm": 0.4679034948348999,
      "learning_rate": 0.0002930198411607329,
      "loss": 3.5195,
      "step": 7890
    },
    {
      "epoch": 0.10354680577764962,
      "grad_norm": 0.44153067469596863,
      "learning_rate": 0.0002930010834109144,
      "loss": 3.5363,
      "step": 7900
    },
    {
      "epoch": 0.10367787768369728,
      "grad_norm": 0.5041334629058838,
      "learning_rate": 0.000292982301093118,
      "loss": 3.5727,
      "step": 7910
    },
    {
      "epoch": 0.10380894958974493,
      "grad_norm": 0.46002861857414246,
      "learning_rate": 0.00029296349421057056,
      "loss": 3.4813,
      "step": 7920
    },
    {
      "epoch": 0.1039400214957926,
      "grad_norm": 0.5418364405632019,
      "learning_rate": 0.00029294466276650325,
      "loss": 3.544,
      "step": 7930
    },
    {
      "epoch": 0.10407109340184025,
      "grad_norm": 0.5058448314666748,
      "learning_rate": 0.00029292580676415123,
      "loss": 3.5432,
      "step": 7940
    },
    {
      "epoch": 0.10420216530788791,
      "grad_norm": 0.4582832455635071,
      "learning_rate": 0.0002929069262067541,
      "loss": 3.4978,
      "step": 7950
    },
    {
      "epoch": 0.10433323721393556,
      "grad_norm": 0.4767458140850067,
      "learning_rate": 0.0002928880210975556,
      "loss": 3.4747,
      "step": 7960
    },
    {
      "epoch": 0.10446430911998322,
      "grad_norm": 0.4693649113178253,
      "learning_rate": 0.00029286909143980355,
      "loss": 3.5175,
      "step": 7970
    },
    {
      "epoch": 0.10459538102603089,
      "grad_norm": 0.4615069031715393,
      "learning_rate": 0.00029285013723675033,
      "loss": 3.512,
      "step": 7980
    },
    {
      "epoch": 0.10472645293207854,
      "grad_norm": 0.6384862661361694,
      "learning_rate": 0.00029283115849165216,
      "loss": 3.4915,
      "step": 7990
    },
    {
      "epoch": 0.1048575248381262,
      "grad_norm": 0.44873589277267456,
      "learning_rate": 0.0002928121552077697,
      "loss": 3.5418,
      "step": 8000
    },
    {
      "epoch": 0.1048575248381262,
      "eval_accuracy": 0.3662678396871945,
      "eval_loss": 3.5298736095428467,
      "eval_runtime": 57.482,
      "eval_samples_per_second": 34.793,
      "eval_steps_per_second": 0.557,
      "step": 8000
    },
    {
      "epoch": 0.10498859674417385,
      "grad_norm": 0.4701840579509735,
      "learning_rate": 0.0002927931273883677,
      "loss": 3.5307,
      "step": 8010
    },
    {
      "epoch": 0.10511966865022151,
      "grad_norm": 0.4925515949726105,
      "learning_rate": 0.0002927740750367153,
      "loss": 3.536,
      "step": 8020
    },
    {
      "epoch": 0.10525074055626917,
      "grad_norm": 0.4784519374370575,
      "learning_rate": 0.0002927549981560857,
      "loss": 3.5176,
      "step": 8030
    },
    {
      "epoch": 0.10538181246231683,
      "grad_norm": 0.4505259394645691,
      "learning_rate": 0.00029273589674975634,
      "loss": 3.5424,
      "step": 8040
    },
    {
      "epoch": 0.10551288436836448,
      "grad_norm": 0.5178182125091553,
      "learning_rate": 0.0002927167708210089,
      "loss": 3.4844,
      "step": 8050
    },
    {
      "epoch": 0.10564395627441214,
      "grad_norm": 0.6316171884536743,
      "learning_rate": 0.0002926976203731293,
      "loss": 3.5357,
      "step": 8060
    },
    {
      "epoch": 0.1057750281804598,
      "grad_norm": 0.4602052867412567,
      "learning_rate": 0.0002926784454094076,
      "loss": 3.5122,
      "step": 8070
    },
    {
      "epoch": 0.10590610008650746,
      "grad_norm": 0.47283709049224854,
      "learning_rate": 0.0002926592459331382,
      "loss": 3.6148,
      "step": 8080
    },
    {
      "epoch": 0.10603717199255512,
      "grad_norm": 0.48447901010513306,
      "learning_rate": 0.00029264002194761944,
      "loss": 3.5297,
      "step": 8090
    },
    {
      "epoch": 0.10616824389860277,
      "grad_norm": 0.4643007218837738,
      "learning_rate": 0.00029262077345615416,
      "loss": 3.497,
      "step": 8100
    },
    {
      "epoch": 0.10629931580465043,
      "grad_norm": 0.4748530685901642,
      "learning_rate": 0.0002926015004620494,
      "loss": 3.48,
      "step": 8110
    },
    {
      "epoch": 0.10643038771069809,
      "grad_norm": 0.4466969966888428,
      "learning_rate": 0.00029258220296861614,
      "loss": 3.5158,
      "step": 8120
    },
    {
      "epoch": 0.10656145961674575,
      "grad_norm": 0.45363420248031616,
      "learning_rate": 0.00029256288097916983,
      "loss": 3.522,
      "step": 8130
    },
    {
      "epoch": 0.1066925315227934,
      "grad_norm": 0.4423820674419403,
      "learning_rate": 0.00029254353449703006,
      "loss": 3.5195,
      "step": 8140
    },
    {
      "epoch": 0.10682360342884106,
      "grad_norm": 0.46928536891937256,
      "learning_rate": 0.0002925241635255206,
      "loss": 3.4754,
      "step": 8150
    },
    {
      "epoch": 0.10695467533488871,
      "grad_norm": 0.4678310453891754,
      "learning_rate": 0.0002925047680679694,
      "loss": 3.5884,
      "step": 8160
    },
    {
      "epoch": 0.10708574724093638,
      "grad_norm": 0.49261412024497986,
      "learning_rate": 0.0002924853481277086,
      "loss": 3.4893,
      "step": 8170
    },
    {
      "epoch": 0.10721681914698404,
      "grad_norm": 0.5520643591880798,
      "learning_rate": 0.00029246590370807474,
      "loss": 3.5572,
      "step": 8180
    },
    {
      "epoch": 0.10734789105303169,
      "grad_norm": 0.568009078502655,
      "learning_rate": 0.0002924464348124083,
      "loss": 3.4881,
      "step": 8190
    },
    {
      "epoch": 0.10747896295907935,
      "grad_norm": 0.5396344065666199,
      "learning_rate": 0.0002924269414440542,
      "loss": 3.5097,
      "step": 8200
    },
    {
      "epoch": 0.107610034865127,
      "grad_norm": 0.44680163264274597,
      "learning_rate": 0.00029240742360636137,
      "loss": 3.5081,
      "step": 8210
    },
    {
      "epoch": 0.10774110677117467,
      "grad_norm": 0.48444864153862,
      "learning_rate": 0.0002923878813026831,
      "loss": 3.4902,
      "step": 8220
    },
    {
      "epoch": 0.10787217867722232,
      "grad_norm": 0.4674491286277771,
      "learning_rate": 0.00029236831453637675,
      "loss": 3.5473,
      "step": 8230
    },
    {
      "epoch": 0.10800325058326998,
      "grad_norm": 0.5855780243873596,
      "learning_rate": 0.00029234872331080394,
      "loss": 3.51,
      "step": 8240
    },
    {
      "epoch": 0.10813432248931763,
      "grad_norm": 0.4637541174888611,
      "learning_rate": 0.00029232910762933044,
      "loss": 3.5064,
      "step": 8250
    },
    {
      "epoch": 0.1082653943953653,
      "grad_norm": 0.5322461724281311,
      "learning_rate": 0.00029230946749532645,
      "loss": 3.5415,
      "step": 8260
    },
    {
      "epoch": 0.10839646630141296,
      "grad_norm": 0.5352371335029602,
      "learning_rate": 0.0002922898029121661,
      "loss": 3.5075,
      "step": 8270
    },
    {
      "epoch": 0.10852753820746061,
      "grad_norm": 0.4658830761909485,
      "learning_rate": 0.0002922701138832278,
      "loss": 3.5473,
      "step": 8280
    },
    {
      "epoch": 0.10865861011350827,
      "grad_norm": 0.48253655433654785,
      "learning_rate": 0.0002922504004118942,
      "loss": 3.5267,
      "step": 8290
    },
    {
      "epoch": 0.10878968201955592,
      "grad_norm": 0.46246853470802307,
      "learning_rate": 0.00029223066250155214,
      "loss": 3.515,
      "step": 8300
    },
    {
      "epoch": 0.10892075392560359,
      "grad_norm": 0.44831106066703796,
      "learning_rate": 0.00029221090015559264,
      "loss": 3.5303,
      "step": 8310
    },
    {
      "epoch": 0.10905182583165124,
      "grad_norm": 0.47510993480682373,
      "learning_rate": 0.0002921911133774109,
      "loss": 3.5065,
      "step": 8320
    },
    {
      "epoch": 0.1091828977376989,
      "grad_norm": 0.5800129771232605,
      "learning_rate": 0.00029217130217040635,
      "loss": 3.5934,
      "step": 8330
    },
    {
      "epoch": 0.10931396964374655,
      "grad_norm": 0.5854555368423462,
      "learning_rate": 0.0002921514665379827,
      "loss": 3.5187,
      "step": 8340
    },
    {
      "epoch": 0.10944504154979422,
      "grad_norm": 0.47814205288887024,
      "learning_rate": 0.0002921316064835476,
      "loss": 3.4998,
      "step": 8350
    },
    {
      "epoch": 0.10957611345584188,
      "grad_norm": 0.4468466639518738,
      "learning_rate": 0.00029211172201051323,
      "loss": 3.527,
      "step": 8360
    },
    {
      "epoch": 0.10970718536188953,
      "grad_norm": 0.5606037974357605,
      "learning_rate": 0.00029209181312229564,
      "loss": 3.4783,
      "step": 8370
    },
    {
      "epoch": 0.1098382572679372,
      "grad_norm": 0.516117513179779,
      "learning_rate": 0.0002920718798223154,
      "loss": 3.5061,
      "step": 8380
    },
    {
      "epoch": 0.10996932917398484,
      "grad_norm": 0.45420533418655396,
      "learning_rate": 0.00029205192211399696,
      "loss": 3.5119,
      "step": 8390
    },
    {
      "epoch": 0.11010040108003251,
      "grad_norm": 0.48712828755378723,
      "learning_rate": 0.00029203194000076917,
      "loss": 3.5043,
      "step": 8400
    },
    {
      "epoch": 0.11023147298608016,
      "grad_norm": 0.4711209535598755,
      "learning_rate": 0.000292011933486065,
      "loss": 3.472,
      "step": 8410
    },
    {
      "epoch": 0.11036254489212782,
      "grad_norm": 0.4305958151817322,
      "learning_rate": 0.0002919919025733216,
      "loss": 3.5023,
      "step": 8420
    },
    {
      "epoch": 0.11049361679817547,
      "grad_norm": 0.4968533515930176,
      "learning_rate": 0.00029197184726598044,
      "loss": 3.496,
      "step": 8430
    },
    {
      "epoch": 0.11062468870422314,
      "grad_norm": 0.47156596183776855,
      "learning_rate": 0.000291951767567487,
      "loss": 3.5145,
      "step": 8440
    },
    {
      "epoch": 0.1107557606102708,
      "grad_norm": 0.4705654978752136,
      "learning_rate": 0.00029193166348129095,
      "loss": 3.5044,
      "step": 8450
    },
    {
      "epoch": 0.11088683251631845,
      "grad_norm": 0.49704211950302124,
      "learning_rate": 0.00029191153501084634,
      "loss": 3.521,
      "step": 8460
    },
    {
      "epoch": 0.11101790442236611,
      "grad_norm": 0.4435335695743561,
      "learning_rate": 0.0002918913821596113,
      "loss": 3.5182,
      "step": 8470
    },
    {
      "epoch": 0.11114897632841376,
      "grad_norm": 0.4872967302799225,
      "learning_rate": 0.00029187120493104807,
      "loss": 3.5169,
      "step": 8480
    },
    {
      "epoch": 0.11128004823446143,
      "grad_norm": 0.5348435640335083,
      "learning_rate": 0.0002918510033286232,
      "loss": 3.4882,
      "step": 8490
    },
    {
      "epoch": 0.11141112014050908,
      "grad_norm": 0.520009458065033,
      "learning_rate": 0.0002918307773558074,
      "loss": 3.6253,
      "step": 8500
    },
    {
      "epoch": 0.11141112014050908,
      "eval_accuracy": 0.367697458455523,
      "eval_loss": 3.5144271850585938,
      "eval_runtime": 59.1827,
      "eval_samples_per_second": 33.794,
      "eval_steps_per_second": 0.541,
      "step": 8500
    },
    {
      "epoch": 0.11154219204655674,
      "grad_norm": 0.45221784710884094,
      "learning_rate": 0.0002918105270160755,
      "loss": 3.4948,
      "step": 8510
    },
    {
      "epoch": 0.11167326395260439,
      "grad_norm": 0.9741812944412231,
      "learning_rate": 0.00029179025231290657,
      "loss": 3.4811,
      "step": 8520
    },
    {
      "epoch": 0.11180433585865206,
      "grad_norm": 0.4499918520450592,
      "learning_rate": 0.0002917699532497839,
      "loss": 3.5252,
      "step": 8530
    },
    {
      "epoch": 0.11193540776469972,
      "grad_norm": 0.5078301429748535,
      "learning_rate": 0.0002917496298301949,
      "loss": 3.5002,
      "step": 8540
    },
    {
      "epoch": 0.11206647967074737,
      "grad_norm": 0.5004239082336426,
      "learning_rate": 0.00029172928205763117,
      "loss": 3.5181,
      "step": 8550
    },
    {
      "epoch": 0.11219755157679503,
      "grad_norm": 0.4823358356952667,
      "learning_rate": 0.0002917089099355885,
      "loss": 3.5276,
      "step": 8560
    },
    {
      "epoch": 0.11232862348284268,
      "grad_norm": 0.479449599981308,
      "learning_rate": 0.000291688513467567,
      "loss": 3.521,
      "step": 8570
    },
    {
      "epoch": 0.11245969538889035,
      "grad_norm": 0.45943817496299744,
      "learning_rate": 0.0002916680926570707,
      "loss": 3.4874,
      "step": 8580
    },
    {
      "epoch": 0.112590767294938,
      "grad_norm": 0.5079334378242493,
      "learning_rate": 0.00029164764750760794,
      "loss": 3.5352,
      "step": 8590
    },
    {
      "epoch": 0.11272183920098566,
      "grad_norm": 0.4453621506690979,
      "learning_rate": 0.00029162717802269135,
      "loss": 3.5319,
      "step": 8600
    },
    {
      "epoch": 0.11285291110703331,
      "grad_norm": 0.4498366415500641,
      "learning_rate": 0.00029160668420583763,
      "loss": 3.5018,
      "step": 8610
    },
    {
      "epoch": 0.11298398301308098,
      "grad_norm": 0.49653908610343933,
      "learning_rate": 0.0002915861660605676,
      "loss": 3.5869,
      "step": 8620
    },
    {
      "epoch": 0.11311505491912864,
      "grad_norm": 0.48615044355392456,
      "learning_rate": 0.0002915656235904064,
      "loss": 3.52,
      "step": 8630
    },
    {
      "epoch": 0.11324612682517629,
      "grad_norm": 0.4671148955821991,
      "learning_rate": 0.0002915450567988833,
      "loss": 3.4758,
      "step": 8640
    },
    {
      "epoch": 0.11337719873122395,
      "grad_norm": 0.5058296918869019,
      "learning_rate": 0.0002915244656895317,
      "loss": 3.4601,
      "step": 8650
    },
    {
      "epoch": 0.1135082706372716,
      "grad_norm": 0.4719547629356384,
      "learning_rate": 0.0002915038502658891,
      "loss": 3.5325,
      "step": 8660
    },
    {
      "epoch": 0.11363934254331927,
      "grad_norm": 0.6268478631973267,
      "learning_rate": 0.00029148321053149744,
      "loss": 3.5487,
      "step": 8670
    },
    {
      "epoch": 0.11377041444936692,
      "grad_norm": 1.1889880895614624,
      "learning_rate": 0.0002914625464899027,
      "loss": 3.5187,
      "step": 8680
    },
    {
      "epoch": 0.11390148635541458,
      "grad_norm": 0.49866530299186707,
      "learning_rate": 0.00029144185814465487,
      "loss": 3.5116,
      "step": 8690
    },
    {
      "epoch": 0.11403255826146223,
      "grad_norm": 0.49476513266563416,
      "learning_rate": 0.0002914211454993084,
      "loss": 3.526,
      "step": 8700
    },
    {
      "epoch": 0.1141636301675099,
      "grad_norm": 0.4714028537273407,
      "learning_rate": 0.0002914004085574217,
      "loss": 3.4961,
      "step": 8710
    },
    {
      "epoch": 0.11429470207355755,
      "grad_norm": 0.5784475207328796,
      "learning_rate": 0.00029137964732255746,
      "loss": 3.5972,
      "step": 8720
    },
    {
      "epoch": 0.11442577397960521,
      "grad_norm": 0.47380244731903076,
      "learning_rate": 0.0002913588617982825,
      "loss": 3.5166,
      "step": 8730
    },
    {
      "epoch": 0.11455684588565287,
      "grad_norm": 0.42808353900909424,
      "learning_rate": 0.0002913380519881679,
      "loss": 3.5026,
      "step": 8740
    },
    {
      "epoch": 0.11468791779170052,
      "grad_norm": 0.43910810351371765,
      "learning_rate": 0.00029131721789578874,
      "loss": 3.499,
      "step": 8750
    },
    {
      "epoch": 0.11481898969774819,
      "grad_norm": 0.4526660442352295,
      "learning_rate": 0.0002912963595247244,
      "loss": 3.5013,
      "step": 8760
    },
    {
      "epoch": 0.11495006160379584,
      "grad_norm": 0.6203173398971558,
      "learning_rate": 0.00029127547687855847,
      "loss": 3.5681,
      "step": 8770
    },
    {
      "epoch": 0.1150811335098435,
      "grad_norm": 0.4945323169231415,
      "learning_rate": 0.0002912545699608786,
      "loss": 3.5111,
      "step": 8780
    },
    {
      "epoch": 0.11521220541589115,
      "grad_norm": 0.49398332834243774,
      "learning_rate": 0.00029123363877527666,
      "loss": 3.5125,
      "step": 8790
    },
    {
      "epoch": 0.11534327732193882,
      "grad_norm": 0.5326728820800781,
      "learning_rate": 0.0002912126833253487,
      "loss": 3.5413,
      "step": 8800
    },
    {
      "epoch": 0.11547434922798647,
      "grad_norm": 0.7283675074577332,
      "learning_rate": 0.0002911917036146948,
      "loss": 3.5464,
      "step": 8810
    },
    {
      "epoch": 0.11560542113403413,
      "grad_norm": 0.4735192656517029,
      "learning_rate": 0.00029117069964691955,
      "loss": 3.4845,
      "step": 8820
    },
    {
      "epoch": 0.1157364930400818,
      "grad_norm": 0.43747878074645996,
      "learning_rate": 0.0002911496714256314,
      "loss": 3.5,
      "step": 8830
    },
    {
      "epoch": 0.11586756494612944,
      "grad_norm": 0.4925345480442047,
      "learning_rate": 0.00029112861895444296,
      "loss": 3.5018,
      "step": 8840
    },
    {
      "epoch": 0.11599863685217711,
      "grad_norm": 0.445500910282135,
      "learning_rate": 0.0002911075422369712,
      "loss": 3.5437,
      "step": 8850
    },
    {
      "epoch": 0.11612970875822476,
      "grad_norm": 0.4846246540546417,
      "learning_rate": 0.00029108644127683714,
      "loss": 3.5004,
      "step": 8860
    },
    {
      "epoch": 0.11626078066427242,
      "grad_norm": 0.5170527100563049,
      "learning_rate": 0.00029106531607766596,
      "loss": 3.4909,
      "step": 8870
    },
    {
      "epoch": 0.11639185257032007,
      "grad_norm": 0.4501773416996002,
      "learning_rate": 0.00029104416664308704,
      "loss": 3.5,
      "step": 8880
    },
    {
      "epoch": 0.11652292447636774,
      "grad_norm": 0.5002649426460266,
      "learning_rate": 0.0002910229929767339,
      "loss": 3.5583,
      "step": 8890
    },
    {
      "epoch": 0.11665399638241539,
      "grad_norm": 0.5204038023948669,
      "learning_rate": 0.00029100179508224433,
      "loss": 3.505,
      "step": 8900
    },
    {
      "epoch": 0.11678506828846305,
      "grad_norm": 0.4864109754562378,
      "learning_rate": 0.00029098057296326004,
      "loss": 3.5271,
      "step": 8910
    },
    {
      "epoch": 0.11691614019451071,
      "grad_norm": 0.515388011932373,
      "learning_rate": 0.0002909593266234271,
      "loss": 3.505,
      "step": 8920
    },
    {
      "epoch": 0.11704721210055836,
      "grad_norm": 0.4775398373603821,
      "learning_rate": 0.00029093805606639575,
      "loss": 3.5156,
      "step": 8930
    },
    {
      "epoch": 0.11717828400660603,
      "grad_norm": 0.462382435798645,
      "learning_rate": 0.0002909167612958203,
      "loss": 3.5034,
      "step": 8940
    },
    {
      "epoch": 0.11730935591265368,
      "grad_norm": 0.44869083166122437,
      "learning_rate": 0.0002908954423153592,
      "loss": 3.5453,
      "step": 8950
    },
    {
      "epoch": 0.11744042781870134,
      "grad_norm": 0.47136276960372925,
      "learning_rate": 0.00029087409912867515,
      "loss": 3.4759,
      "step": 8960
    },
    {
      "epoch": 0.11757149972474899,
      "grad_norm": 0.4949701726436615,
      "learning_rate": 0.00029085273173943497,
      "loss": 3.4989,
      "step": 8970
    },
    {
      "epoch": 0.11770257163079666,
      "grad_norm": 0.6363611817359924,
      "learning_rate": 0.0002908313401513096,
      "loss": 3.5066,
      "step": 8980
    },
    {
      "epoch": 0.1178336435368443,
      "grad_norm": 0.48528364300727844,
      "learning_rate": 0.0002908099243679742,
      "loss": 3.5054,
      "step": 8990
    },
    {
      "epoch": 0.11796471544289197,
      "grad_norm": 0.49855613708496094,
      "learning_rate": 0.0002907884843931081,
      "loss": 3.4898,
      "step": 9000
    },
    {
      "epoch": 0.11796471544289197,
      "eval_accuracy": 0.36874389051808404,
      "eval_loss": 3.504366636276245,
      "eval_runtime": 56.0692,
      "eval_samples_per_second": 35.67,
      "eval_steps_per_second": 0.571,
      "step": 9000
    },
    {
      "epoch": 0.11809578734893963,
      "grad_norm": 0.5045238733291626,
      "learning_rate": 0.0002907670202303946,
      "loss": 3.52,
      "step": 9010
    },
    {
      "epoch": 0.11822685925498728,
      "grad_norm": 0.49751725792884827,
      "learning_rate": 0.0002907455318835215,
      "loss": 3.5517,
      "step": 9020
    },
    {
      "epoch": 0.11835793116103495,
      "grad_norm": 0.5183396339416504,
      "learning_rate": 0.00029072401935618047,
      "loss": 3.4864,
      "step": 9030
    },
    {
      "epoch": 0.1184890030670826,
      "grad_norm": 0.5227517485618591,
      "learning_rate": 0.0002907024826520674,
      "loss": 3.5074,
      "step": 9040
    },
    {
      "epoch": 0.11862007497313026,
      "grad_norm": 0.45498913526535034,
      "learning_rate": 0.00029068092177488236,
      "loss": 3.5089,
      "step": 9050
    },
    {
      "epoch": 0.11875114687917791,
      "grad_norm": 0.46289879083633423,
      "learning_rate": 0.0002906593367283296,
      "loss": 3.5091,
      "step": 9060
    },
    {
      "epoch": 0.11888221878522558,
      "grad_norm": 0.5254165530204773,
      "learning_rate": 0.00029063772751611745,
      "loss": 3.5093,
      "step": 9070
    },
    {
      "epoch": 0.11901329069127323,
      "grad_norm": 0.4961933195590973,
      "learning_rate": 0.00029061609414195844,
      "loss": 3.4877,
      "step": 9080
    },
    {
      "epoch": 0.11914436259732089,
      "grad_norm": 0.4799545109272003,
      "learning_rate": 0.0002905944366095693,
      "loss": 3.5252,
      "step": 9090
    },
    {
      "epoch": 0.11927543450336855,
      "grad_norm": 0.49481505155563354,
      "learning_rate": 0.0002905727549226707,
      "loss": 3.5153,
      "step": 9100
    },
    {
      "epoch": 0.1194065064094162,
      "grad_norm": 0.4765486717224121,
      "learning_rate": 0.0002905510490849878,
      "loss": 3.486,
      "step": 9110
    },
    {
      "epoch": 0.11953757831546387,
      "grad_norm": 0.5067988634109497,
      "learning_rate": 0.0002905293191002497,
      "loss": 3.5121,
      "step": 9120
    },
    {
      "epoch": 0.11966865022151152,
      "grad_norm": 0.43447136878967285,
      "learning_rate": 0.00029050756497218954,
      "loss": 3.5027,
      "step": 9130
    },
    {
      "epoch": 0.11979972212755918,
      "grad_norm": 0.44820019602775574,
      "learning_rate": 0.0002904857867045448,
      "loss": 3.5095,
      "step": 9140
    },
    {
      "epoch": 0.11993079403360683,
      "grad_norm": 0.44943761825561523,
      "learning_rate": 0.0002904639843010571,
      "loss": 3.492,
      "step": 9150
    },
    {
      "epoch": 0.1200618659396545,
      "grad_norm": 0.5042577981948853,
      "learning_rate": 0.0002904421577654721,
      "loss": 3.4885,
      "step": 9160
    },
    {
      "epoch": 0.12019293784570215,
      "grad_norm": 0.4754837453365326,
      "learning_rate": 0.0002904203071015397,
      "loss": 3.5267,
      "step": 9170
    },
    {
      "epoch": 0.12032400975174981,
      "grad_norm": 0.4582735300064087,
      "learning_rate": 0.0002903984323130138,
      "loss": 3.4596,
      "step": 9180
    },
    {
      "epoch": 0.12045508165779746,
      "grad_norm": 0.4480108618736267,
      "learning_rate": 0.00029037653340365264,
      "loss": 3.5177,
      "step": 9190
    },
    {
      "epoch": 0.12058615356384512,
      "grad_norm": 0.4538060128688812,
      "learning_rate": 0.0002903546103772185,
      "loss": 3.4773,
      "step": 9200
    },
    {
      "epoch": 0.12071722546989279,
      "grad_norm": 0.4471026062965393,
      "learning_rate": 0.0002903326632374779,
      "loss": 3.5581,
      "step": 9210
    },
    {
      "epoch": 0.12084829737594044,
      "grad_norm": 0.45538967847824097,
      "learning_rate": 0.0002903106919882012,
      "loss": 3.5018,
      "step": 9220
    },
    {
      "epoch": 0.1209793692819881,
      "grad_norm": 0.45529353618621826,
      "learning_rate": 0.0002902886966331633,
      "loss": 3.493,
      "step": 9230
    },
    {
      "epoch": 0.12111044118803575,
      "grad_norm": 0.9161607027053833,
      "learning_rate": 0.00029026667717614303,
      "loss": 3.5246,
      "step": 9240
    },
    {
      "epoch": 0.12124151309408342,
      "grad_norm": 0.49503448605537415,
      "learning_rate": 0.00029024463362092336,
      "loss": 3.5357,
      "step": 9250
    },
    {
      "epoch": 0.12137258500013107,
      "grad_norm": 0.5013350248336792,
      "learning_rate": 0.00029022256597129145,
      "loss": 3.5617,
      "step": 9260
    },
    {
      "epoch": 0.12150365690617873,
      "grad_norm": 0.50674968957901,
      "learning_rate": 0.00029020047423103856,
      "loss": 3.4601,
      "step": 9270
    },
    {
      "epoch": 0.12163472881222638,
      "grad_norm": 0.5524303913116455,
      "learning_rate": 0.00029017835840396014,
      "loss": 3.4894,
      "step": 9280
    },
    {
      "epoch": 0.12176580071827404,
      "grad_norm": 0.5000428557395935,
      "learning_rate": 0.0002901562184938558,
      "loss": 3.4641,
      "step": 9290
    },
    {
      "epoch": 0.12189687262432171,
      "grad_norm": 0.45507073402404785,
      "learning_rate": 0.0002901340545045291,
      "loss": 3.4676,
      "step": 9300
    },
    {
      "epoch": 0.12202794453036936,
      "grad_norm": 0.43615269660949707,
      "learning_rate": 0.00029011186643978804,
      "loss": 3.4979,
      "step": 9310
    },
    {
      "epoch": 0.12215901643641702,
      "grad_norm": 0.4739636182785034,
      "learning_rate": 0.0002900896543034444,
      "loss": 3.4911,
      "step": 9320
    },
    {
      "epoch": 0.12229008834246467,
      "grad_norm": 0.47106558084487915,
      "learning_rate": 0.0002900674180993145,
      "loss": 3.475,
      "step": 9330
    },
    {
      "epoch": 0.12242116024851234,
      "grad_norm": 0.4603668749332428,
      "learning_rate": 0.0002900451578312184,
      "loss": 3.5401,
      "step": 9340
    },
    {
      "epoch": 0.12255223215455999,
      "grad_norm": 0.45501139760017395,
      "learning_rate": 0.0002900228735029806,
      "loss": 3.4497,
      "step": 9350
    },
    {
      "epoch": 0.12268330406060765,
      "grad_norm": 0.4499923884868622,
      "learning_rate": 0.0002900005651184295,
      "loss": 3.4611,
      "step": 9360
    },
    {
      "epoch": 0.1228143759666553,
      "grad_norm": 0.4462062120437622,
      "learning_rate": 0.00028997823268139784,
      "loss": 3.512,
      "step": 9370
    },
    {
      "epoch": 0.12294544787270296,
      "grad_norm": 0.45755869150161743,
      "learning_rate": 0.0002899558761957223,
      "loss": 3.4664,
      "step": 9380
    },
    {
      "epoch": 0.12307651977875063,
      "grad_norm": 0.5310529470443726,
      "learning_rate": 0.0002899334956652438,
      "loss": 3.5023,
      "step": 9390
    },
    {
      "epoch": 0.12320759168479828,
      "grad_norm": 0.46713870763778687,
      "learning_rate": 0.00028991109109380753,
      "loss": 3.4883,
      "step": 9400
    },
    {
      "epoch": 0.12333866359084594,
      "grad_norm": 0.4533595144748688,
      "learning_rate": 0.0002898886624852625,
      "loss": 3.4577,
      "step": 9410
    },
    {
      "epoch": 0.12346973549689359,
      "grad_norm": 0.8377737402915955,
      "learning_rate": 0.000289866209843462,
      "loss": 3.5512,
      "step": 9420
    },
    {
      "epoch": 0.12360080740294126,
      "grad_norm": 0.4522849917411804,
      "learning_rate": 0.0002898437331722635,
      "loss": 3.454,
      "step": 9430
    },
    {
      "epoch": 0.1237318793089889,
      "grad_norm": 0.53324294090271,
      "learning_rate": 0.0002898212324755286,
      "loss": 3.5197,
      "step": 9440
    },
    {
      "epoch": 0.12386295121503657,
      "grad_norm": 0.4963279664516449,
      "learning_rate": 0.0002897987077571229,
      "loss": 3.4942,
      "step": 9450
    },
    {
      "epoch": 0.12399402312108422,
      "grad_norm": 0.5305104851722717,
      "learning_rate": 0.0002897761590209162,
      "loss": 3.4696,
      "step": 9460
    },
    {
      "epoch": 0.12412509502713188,
      "grad_norm": 0.4586170017719269,
      "learning_rate": 0.0002897535862707826,
      "loss": 3.496,
      "step": 9470
    },
    {
      "epoch": 0.12425616693317955,
      "grad_norm": 0.45205289125442505,
      "learning_rate": 0.00028973098951059996,
      "loss": 3.5128,
      "step": 9480
    },
    {
      "epoch": 0.1243872388392272,
      "grad_norm": 0.4941801428794861,
      "learning_rate": 0.0002897083687442506,
      "loss": 3.5171,
      "step": 9490
    },
    {
      "epoch": 0.12451831074527486,
      "grad_norm": 0.5905492901802063,
      "learning_rate": 0.0002896857239756207,
      "loss": 3.5816,
      "step": 9500
    },
    {
      "epoch": 0.12451831074527486,
      "eval_accuracy": 0.36890224828934504,
      "eval_loss": 3.50140643119812,
      "eval_runtime": 57.0906,
      "eval_samples_per_second": 35.032,
      "eval_steps_per_second": 0.561,
      "step": 9500
    },
    {
      "epoch": 0.12464938265132251,
      "grad_norm": 0.5543625354766846,
      "learning_rate": 0.0002896630552086008,
      "loss": 3.5092,
      "step": 9510
    },
    {
      "epoch": 0.12478045455737018,
      "grad_norm": 0.49343088269233704,
      "learning_rate": 0.00028964036244708545,
      "loss": 3.5269,
      "step": 9520
    },
    {
      "epoch": 0.12491152646341783,
      "grad_norm": 0.481702595949173,
      "learning_rate": 0.0002896176456949733,
      "loss": 3.4797,
      "step": 9530
    },
    {
      "epoch": 0.1250425983694655,
      "grad_norm": 0.44378378987312317,
      "learning_rate": 0.0002895949049561671,
      "loss": 3.5101,
      "step": 9540
    },
    {
      "epoch": 0.12517367027551315,
      "grad_norm": 0.45204079151153564,
      "learning_rate": 0.00028957214023457383,
      "loss": 3.473,
      "step": 9550
    },
    {
      "epoch": 0.1253047421815608,
      "grad_norm": 0.4531859755516052,
      "learning_rate": 0.00028954935153410456,
      "loss": 3.4877,
      "step": 9560
    },
    {
      "epoch": 0.12543581408760845,
      "grad_norm": 0.46899449825286865,
      "learning_rate": 0.0002895265388586744,
      "loss": 3.4669,
      "step": 9570
    },
    {
      "epoch": 0.12556688599365612,
      "grad_norm": 0.48416972160339355,
      "learning_rate": 0.0002895037022122026,
      "loss": 3.4843,
      "step": 9580
    },
    {
      "epoch": 0.12569795789970378,
      "grad_norm": 0.5221899151802063,
      "learning_rate": 0.0002894808415986127,
      "loss": 3.5322,
      "step": 9590
    },
    {
      "epoch": 0.12582902980575145,
      "grad_norm": 0.5146622061729431,
      "learning_rate": 0.000289457957021832,
      "loss": 3.486,
      "step": 9600
    },
    {
      "epoch": 0.12596010171179908,
      "grad_norm": 0.45903176069259644,
      "learning_rate": 0.00028943504848579236,
      "loss": 3.4827,
      "step": 9610
    },
    {
      "epoch": 0.12609117361784675,
      "grad_norm": 0.5619129538536072,
      "learning_rate": 0.0002894121159944294,
      "loss": 3.5272,
      "step": 9620
    },
    {
      "epoch": 0.1262222455238944,
      "grad_norm": 0.48688507080078125,
      "learning_rate": 0.000289389159551683,
      "loss": 3.4894,
      "step": 9630
    },
    {
      "epoch": 0.12635331742994207,
      "grad_norm": 0.5024608373641968,
      "learning_rate": 0.00028936617916149716,
      "loss": 3.4702,
      "step": 9640
    },
    {
      "epoch": 0.1264843893359897,
      "grad_norm": 0.485359787940979,
      "learning_rate": 0.00028934317482781993,
      "loss": 3.5179,
      "step": 9650
    },
    {
      "epoch": 0.12661546124203737,
      "grad_norm": 0.44778189063072205,
      "learning_rate": 0.00028932014655460355,
      "loss": 3.4802,
      "step": 9660
    },
    {
      "epoch": 0.12674653314808504,
      "grad_norm": 0.5721172094345093,
      "learning_rate": 0.0002892970943458044,
      "loss": 3.4688,
      "step": 9670
    },
    {
      "epoch": 0.1268776050541327,
      "grad_norm": 0.5213291645050049,
      "learning_rate": 0.00028927401820538274,
      "loss": 3.4726,
      "step": 9680
    },
    {
      "epoch": 0.12700867696018037,
      "grad_norm": 0.5082741975784302,
      "learning_rate": 0.00028925091813730334,
      "loss": 3.4871,
      "step": 9690
    },
    {
      "epoch": 0.127139748866228,
      "grad_norm": 0.5057971477508545,
      "learning_rate": 0.0002892277941455347,
      "loss": 3.4934,
      "step": 9700
    },
    {
      "epoch": 0.12727082077227567,
      "grad_norm": 0.45714524388313293,
      "learning_rate": 0.0002892046462340496,
      "loss": 3.5154,
      "step": 9710
    },
    {
      "epoch": 0.12740189267832333,
      "grad_norm": 0.4501226842403412,
      "learning_rate": 0.0002891814744068249,
      "loss": 3.4871,
      "step": 9720
    },
    {
      "epoch": 0.127532964584371,
      "grad_norm": 0.5218084454536438,
      "learning_rate": 0.0002891582786678417,
      "loss": 3.5375,
      "step": 9730
    },
    {
      "epoch": 0.12766403649041863,
      "grad_norm": 0.4602382779121399,
      "learning_rate": 0.00028913505902108505,
      "loss": 3.472,
      "step": 9740
    },
    {
      "epoch": 0.1277951083964663,
      "grad_norm": 0.47582870721817017,
      "learning_rate": 0.000289111815470544,
      "loss": 3.4769,
      "step": 9750
    },
    {
      "epoch": 0.12792618030251396,
      "grad_norm": 0.49297818541526794,
      "learning_rate": 0.00028908854802021206,
      "loss": 3.4706,
      "step": 9760
    },
    {
      "epoch": 0.12805725220856162,
      "grad_norm": 0.4634343683719635,
      "learning_rate": 0.0002890652566740865,
      "loss": 3.476,
      "step": 9770
    },
    {
      "epoch": 0.1281883241146093,
      "grad_norm": 0.5093823671340942,
      "learning_rate": 0.00028904194143616896,
      "loss": 3.5159,
      "step": 9780
    },
    {
      "epoch": 0.12831939602065692,
      "grad_norm": 0.6196965575218201,
      "learning_rate": 0.00028901860231046495,
      "loss": 3.5342,
      "step": 9790
    },
    {
      "epoch": 0.1284504679267046,
      "grad_norm": 0.4709146022796631,
      "learning_rate": 0.0002889952393009842,
      "loss": 3.5328,
      "step": 9800
    },
    {
      "epoch": 0.12858153983275225,
      "grad_norm": 0.45561230182647705,
      "learning_rate": 0.0002889718524117406,
      "loss": 3.4865,
      "step": 9810
    },
    {
      "epoch": 0.12871261173879991,
      "grad_norm": 0.5158346891403198,
      "learning_rate": 0.0002889484416467521,
      "loss": 3.5253,
      "step": 9820
    },
    {
      "epoch": 0.12884368364484755,
      "grad_norm": 0.5189052820205688,
      "learning_rate": 0.00028892500701004065,
      "loss": 3.501,
      "step": 9830
    },
    {
      "epoch": 0.12897475555089521,
      "grad_norm": 0.6089820265769958,
      "learning_rate": 0.0002889015485056324,
      "loss": 3.4052,
      "step": 9840
    },
    {
      "epoch": 0.12910582745694288,
      "grad_norm": 0.501613974571228,
      "learning_rate": 0.0002888780661375577,
      "loss": 3.52,
      "step": 9850
    },
    {
      "epoch": 0.12923689936299054,
      "grad_norm": 0.49026432633399963,
      "learning_rate": 0.0002888545599098507,
      "loss": 3.4665,
      "step": 9860
    },
    {
      "epoch": 0.1293679712690382,
      "grad_norm": 0.4951883554458618,
      "learning_rate": 0.00028883102982655,
      "loss": 3.4567,
      "step": 9870
    },
    {
      "epoch": 0.12949904317508584,
      "grad_norm": 0.5906792283058167,
      "learning_rate": 0.000288807475891698,
      "loss": 3.4975,
      "step": 9880
    },
    {
      "epoch": 0.1296301150811335,
      "grad_norm": 0.48943591117858887,
      "learning_rate": 0.0002887838981093414,
      "loss": 3.5039,
      "step": 9890
    },
    {
      "epoch": 0.12976118698718117,
      "grad_norm": 0.48656216263771057,
      "learning_rate": 0.0002887602964835309,
      "loss": 3.4557,
      "step": 9900
    },
    {
      "epoch": 0.12989225889322883,
      "grad_norm": 0.46317243576049805,
      "learning_rate": 0.0002887366710183214,
      "loss": 3.4069,
      "step": 9910
    },
    {
      "epoch": 0.13002333079927647,
      "grad_norm": 0.5517776608467102,
      "learning_rate": 0.0002887130217177718,
      "loss": 3.4366,
      "step": 9920
    },
    {
      "epoch": 0.13015440270532413,
      "grad_norm": 0.5146608948707581,
      "learning_rate": 0.000288689348585945,
      "loss": 3.4669,
      "step": 9930
    },
    {
      "epoch": 0.1302854746113718,
      "grad_norm": 0.46838003396987915,
      "learning_rate": 0.00028866565162690826,
      "loss": 3.4813,
      "step": 9940
    },
    {
      "epoch": 0.13041654651741946,
      "grad_norm": 0.47433462738990784,
      "learning_rate": 0.00028864193084473264,
      "loss": 3.4765,
      "step": 9950
    },
    {
      "epoch": 0.13054761842346713,
      "grad_norm": 0.4596497714519501,
      "learning_rate": 0.0002886181862434936,
      "loss": 3.4834,
      "step": 9960
    },
    {
      "epoch": 0.13067869032951476,
      "grad_norm": 0.6324459910392761,
      "learning_rate": 0.0002885944178272703,
      "loss": 3.4732,
      "step": 9970
    },
    {
      "epoch": 0.13080976223556243,
      "grad_norm": 0.49346238374710083,
      "learning_rate": 0.00028857062560014643,
      "loss": 3.4612,
      "step": 9980
    },
    {
      "epoch": 0.1309408341416101,
      "grad_norm": 0.4459896385669708,
      "learning_rate": 0.0002885468095662095,
      "loss": 3.4895,
      "step": 9990
    },
    {
      "epoch": 0.13107190604765775,
      "grad_norm": 0.5026695728302002,
      "learning_rate": 0.0002885229697295511,
      "loss": 3.4952,
      "step": 10000
    },
    {
      "epoch": 0.13107190604765775,
      "eval_accuracy": 0.37182209188660803,
      "eval_loss": 3.479508876800537,
      "eval_runtime": 59.681,
      "eval_samples_per_second": 33.511,
      "eval_steps_per_second": 0.536,
      "step": 10000
    },
    {
      "epoch": 0.1312029779537054,
      "grad_norm": 0.542897641658783,
      "learning_rate": 0.0002884991060942671,
      "loss": 3.4807,
      "step": 10010
    },
    {
      "epoch": 0.13133404985975305,
      "grad_norm": 0.47724440693855286,
      "learning_rate": 0.0002884752186644572,
      "loss": 3.4733,
      "step": 10020
    },
    {
      "epoch": 0.13146512176580072,
      "grad_norm": 0.47636890411376953,
      "learning_rate": 0.00028845130744422547,
      "loss": 3.4925,
      "step": 10030
    },
    {
      "epoch": 0.13159619367184838,
      "grad_norm": 0.4408959448337555,
      "learning_rate": 0.00028842737243767977,
      "loss": 3.5224,
      "step": 10040
    },
    {
      "epoch": 0.13172726557789605,
      "grad_norm": 0.46676045656204224,
      "learning_rate": 0.00028840341364893234,
      "loss": 3.4638,
      "step": 10050
    },
    {
      "epoch": 0.13185833748394368,
      "grad_norm": 0.48547276854515076,
      "learning_rate": 0.0002883794310820993,
      "loss": 3.4692,
      "step": 10060
    },
    {
      "epoch": 0.13198940938999135,
      "grad_norm": 0.4853609502315521,
      "learning_rate": 0.0002883554247413009,
      "loss": 3.4892,
      "step": 10070
    },
    {
      "epoch": 0.132120481296039,
      "grad_norm": 0.4766978919506073,
      "learning_rate": 0.0002883313946306615,
      "loss": 3.4192,
      "step": 10080
    },
    {
      "epoch": 0.13225155320208667,
      "grad_norm": 0.47026965022087097,
      "learning_rate": 0.0002883073407543096,
      "loss": 3.4906,
      "step": 10090
    },
    {
      "epoch": 0.1323826251081343,
      "grad_norm": 0.49355441331863403,
      "learning_rate": 0.00028828326311637766,
      "loss": 3.4666,
      "step": 10100
    },
    {
      "epoch": 0.13251369701418197,
      "grad_norm": 0.5487005114555359,
      "learning_rate": 0.0002882591617210023,
      "loss": 3.5077,
      "step": 10110
    },
    {
      "epoch": 0.13264476892022964,
      "grad_norm": 0.5189139246940613,
      "learning_rate": 0.0002882350365723242,
      "loss": 3.5323,
      "step": 10120
    },
    {
      "epoch": 0.1327758408262773,
      "grad_norm": 0.5090906023979187,
      "learning_rate": 0.0002882108876744882,
      "loss": 3.5082,
      "step": 10130
    },
    {
      "epoch": 0.13290691273232497,
      "grad_norm": 0.47560951113700867,
      "learning_rate": 0.00028818671503164295,
      "loss": 3.5138,
      "step": 10140
    },
    {
      "epoch": 0.1330379846383726,
      "grad_norm": 0.4869980812072754,
      "learning_rate": 0.0002881625186479416,
      "loss": 3.4801,
      "step": 10150
    },
    {
      "epoch": 0.13316905654442027,
      "grad_norm": 0.49297258257865906,
      "learning_rate": 0.0002881382985275411,
      "loss": 3.4601,
      "step": 10160
    },
    {
      "epoch": 0.13330012845046793,
      "grad_norm": 0.4600953757762909,
      "learning_rate": 0.0002881140546746025,
      "loss": 3.4914,
      "step": 10170
    },
    {
      "epoch": 0.1334312003565156,
      "grad_norm": 0.48152002692222595,
      "learning_rate": 0.0002880897870932909,
      "loss": 3.4391,
      "step": 10180
    },
    {
      "epoch": 0.13356227226256323,
      "grad_norm": 0.4700385332107544,
      "learning_rate": 0.0002880654957877756,
      "loss": 3.4843,
      "step": 10190
    },
    {
      "epoch": 0.1336933441686109,
      "grad_norm": 0.46671298146247864,
      "learning_rate": 0.0002880411807622299,
      "loss": 3.4708,
      "step": 10200
    },
    {
      "epoch": 0.13382441607465856,
      "grad_norm": 0.47941797971725464,
      "learning_rate": 0.0002880168420208312,
      "loss": 3.4545,
      "step": 10210
    },
    {
      "epoch": 0.13395548798070622,
      "grad_norm": 0.46498405933380127,
      "learning_rate": 0.000287992479567761,
      "loss": 3.4798,
      "step": 10220
    },
    {
      "epoch": 0.1340865598867539,
      "grad_norm": 0.5226965546607971,
      "learning_rate": 0.00028796809340720476,
      "loss": 3.4821,
      "step": 10230
    },
    {
      "epoch": 0.13421763179280152,
      "grad_norm": 0.45690470933914185,
      "learning_rate": 0.0002879436835433521,
      "loss": 3.4954,
      "step": 10240
    },
    {
      "epoch": 0.1343487036988492,
      "grad_norm": 0.5046117305755615,
      "learning_rate": 0.00028791924998039677,
      "loss": 3.4379,
      "step": 10250
    },
    {
      "epoch": 0.13447977560489685,
      "grad_norm": 0.49336785078048706,
      "learning_rate": 0.0002878947927225365,
      "loss": 3.5118,
      "step": 10260
    },
    {
      "epoch": 0.13461084751094451,
      "grad_norm": 0.4532913267612457,
      "learning_rate": 0.00028787031177397303,
      "loss": 3.4546,
      "step": 10270
    },
    {
      "epoch": 0.13474191941699215,
      "grad_norm": 0.42878860235214233,
      "learning_rate": 0.00028784580713891235,
      "loss": 3.4697,
      "step": 10280
    },
    {
      "epoch": 0.13487299132303981,
      "grad_norm": 0.5561381578445435,
      "learning_rate": 0.00028782127882156437,
      "loss": 3.5145,
      "step": 10290
    },
    {
      "epoch": 0.13500406322908748,
      "grad_norm": 0.46265798807144165,
      "learning_rate": 0.00028779672682614316,
      "loss": 3.4684,
      "step": 10300
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 0.5181275010108948,
      "learning_rate": 0.0002877721511568668,
      "loss": 3.4341,
      "step": 10310
    },
    {
      "epoch": 0.1352662070411828,
      "grad_norm": 0.4461798369884491,
      "learning_rate": 0.00028774755181795747,
      "loss": 3.4982,
      "step": 10320
    },
    {
      "epoch": 0.13539727894723044,
      "grad_norm": 0.5217160582542419,
      "learning_rate": 0.0002877229288136414,
      "loss": 3.497,
      "step": 10330
    },
    {
      "epoch": 0.1355283508532781,
      "grad_norm": 0.4260563552379608,
      "learning_rate": 0.00028769828214814883,
      "loss": 3.4698,
      "step": 10340
    },
    {
      "epoch": 0.13565942275932577,
      "grad_norm": 0.44918951392173767,
      "learning_rate": 0.00028767361182571423,
      "loss": 3.432,
      "step": 10350
    },
    {
      "epoch": 0.13579049466537343,
      "grad_norm": 0.4920472204685211,
      "learning_rate": 0.0002876489178505759,
      "loss": 3.4681,
      "step": 10360
    },
    {
      "epoch": 0.13592156657142107,
      "grad_norm": 0.4882318079471588,
      "learning_rate": 0.0002876242002269765,
      "loss": 3.4789,
      "step": 10370
    },
    {
      "epoch": 0.13605263847746873,
      "grad_norm": 0.5243890285491943,
      "learning_rate": 0.0002875994589591625,
      "loss": 3.4802,
      "step": 10380
    },
    {
      "epoch": 0.1361837103835164,
      "grad_norm": 0.5118129849433899,
      "learning_rate": 0.0002875746940513845,
      "loss": 3.4974,
      "step": 10390
    },
    {
      "epoch": 0.13631478228956406,
      "grad_norm": 0.4777592420578003,
      "learning_rate": 0.0002875499055078971,
      "loss": 3.4544,
      "step": 10400
    },
    {
      "epoch": 0.1364458541956117,
      "grad_norm": 0.4404730498790741,
      "learning_rate": 0.0002875250933329592,
      "loss": 3.486,
      "step": 10410
    },
    {
      "epoch": 0.13657692610165936,
      "grad_norm": 0.4383951425552368,
      "learning_rate": 0.00028750025753083357,
      "loss": 3.446,
      "step": 10420
    },
    {
      "epoch": 0.13670799800770703,
      "grad_norm": 0.5083426833152771,
      "learning_rate": 0.0002874753981057869,
      "loss": 3.4809,
      "step": 10430
    },
    {
      "epoch": 0.1368390699137547,
      "grad_norm": 0.4439893364906311,
      "learning_rate": 0.00028745051506209033,
      "loss": 3.5053,
      "step": 10440
    },
    {
      "epoch": 0.13697014181980235,
      "grad_norm": 0.5202581286430359,
      "learning_rate": 0.00028742560840401874,
      "loss": 3.4364,
      "step": 10450
    },
    {
      "epoch": 0.13710121372585,
      "grad_norm": 0.487450510263443,
      "learning_rate": 0.0002874006781358512,
      "loss": 3.5002,
      "step": 10460
    },
    {
      "epoch": 0.13723228563189765,
      "grad_norm": 0.4697534143924713,
      "learning_rate": 0.00028737572426187067,
      "loss": 3.4833,
      "step": 10470
    },
    {
      "epoch": 0.13736335753794532,
      "grad_norm": 0.5115428566932678,
      "learning_rate": 0.00028735074678636443,
      "loss": 3.507,
      "step": 10480
    },
    {
      "epoch": 0.13749442944399298,
      "grad_norm": 0.5819740295410156,
      "learning_rate": 0.00028732574571362367,
      "loss": 3.5597,
      "step": 10490
    },
    {
      "epoch": 0.13762550135004062,
      "grad_norm": 0.4898456931114197,
      "learning_rate": 0.0002873007210479435,
      "loss": 3.5101,
      "step": 10500
    },
    {
      "epoch": 0.13762550135004062,
      "eval_accuracy": 0.372539100684262,
      "eval_loss": 3.4707252979278564,
      "eval_runtime": 56.9876,
      "eval_samples_per_second": 35.095,
      "eval_steps_per_second": 0.562,
      "step": 10500
    },
    {
      "epoch": 0.13775657325608828,
      "grad_norm": 0.44379517436027527,
      "learning_rate": 0.0002872756727936234,
      "loss": 3.4375,
      "step": 10510
    },
    {
      "epoch": 0.13788764516213595,
      "grad_norm": 0.4594014286994934,
      "learning_rate": 0.0002872506009549667,
      "loss": 3.5189,
      "step": 10520
    },
    {
      "epoch": 0.1380187170681836,
      "grad_norm": 0.4826613962650299,
      "learning_rate": 0.00028722550553628075,
      "loss": 3.4517,
      "step": 10530
    },
    {
      "epoch": 0.13814978897423127,
      "grad_norm": 0.469264954328537,
      "learning_rate": 0.000287200386541877,
      "loss": 3.5392,
      "step": 10540
    },
    {
      "epoch": 0.1382808608802789,
      "grad_norm": 0.4690016806125641,
      "learning_rate": 0.00028717524397607104,
      "loss": 3.473,
      "step": 10550
    },
    {
      "epoch": 0.13841193278632657,
      "grad_norm": 0.4696325957775116,
      "learning_rate": 0.0002871500778431824,
      "loss": 3.5064,
      "step": 10560
    },
    {
      "epoch": 0.13854300469237424,
      "grad_norm": 0.4364059567451477,
      "learning_rate": 0.0002871248881475346,
      "loss": 3.4771,
      "step": 10570
    },
    {
      "epoch": 0.1386740765984219,
      "grad_norm": 0.5359991788864136,
      "learning_rate": 0.00028709967489345544,
      "loss": 3.4899,
      "step": 10580
    },
    {
      "epoch": 0.13880514850446954,
      "grad_norm": 0.46471062302589417,
      "learning_rate": 0.00028707443808527656,
      "loss": 3.4901,
      "step": 10590
    },
    {
      "epoch": 0.1389362204105172,
      "grad_norm": 0.49836158752441406,
      "learning_rate": 0.0002870491777273337,
      "loss": 3.4474,
      "step": 10600
    },
    {
      "epoch": 0.13906729231656487,
      "grad_norm": 0.4775247871875763,
      "learning_rate": 0.0002870238938239667,
      "loss": 3.478,
      "step": 10610
    },
    {
      "epoch": 0.13919836422261253,
      "grad_norm": 0.4349680542945862,
      "learning_rate": 0.00028699858637951933,
      "loss": 3.465,
      "step": 10620
    },
    {
      "epoch": 0.1393294361286602,
      "grad_norm": 0.4677911102771759,
      "learning_rate": 0.00028697325539833957,
      "loss": 3.4818,
      "step": 10630
    },
    {
      "epoch": 0.13946050803470783,
      "grad_norm": 0.47013041377067566,
      "learning_rate": 0.0002869479008847793,
      "loss": 3.4427,
      "step": 10640
    },
    {
      "epoch": 0.1395915799407555,
      "grad_norm": 0.4582670331001282,
      "learning_rate": 0.0002869225228431945,
      "loss": 3.4091,
      "step": 10650
    },
    {
      "epoch": 0.13972265184680316,
      "grad_norm": 0.4598885476589203,
      "learning_rate": 0.00028689712127794514,
      "loss": 3.4623,
      "step": 10660
    },
    {
      "epoch": 0.13985372375285082,
      "grad_norm": 0.4532322287559509,
      "learning_rate": 0.0002868716961933954,
      "loss": 3.489,
      "step": 10670
    },
    {
      "epoch": 0.13998479565889846,
      "grad_norm": 0.46565166115760803,
      "learning_rate": 0.0002868462475939133,
      "loss": 3.4657,
      "step": 10680
    },
    {
      "epoch": 0.14011586756494612,
      "grad_norm": 0.46127745509147644,
      "learning_rate": 0.000286820775483871,
      "loss": 3.4607,
      "step": 10690
    },
    {
      "epoch": 0.1402469394709938,
      "grad_norm": 0.45777079463005066,
      "learning_rate": 0.0002867952798676446,
      "loss": 3.4545,
      "step": 10700
    },
    {
      "epoch": 0.14037801137704145,
      "grad_norm": 0.6560996174812317,
      "learning_rate": 0.00028676976074961443,
      "loss": 3.5527,
      "step": 10710
    },
    {
      "epoch": 0.14050908328308911,
      "grad_norm": 0.4788513779640198,
      "learning_rate": 0.0002867442181341647,
      "loss": 3.4837,
      "step": 10720
    },
    {
      "epoch": 0.14064015518913675,
      "grad_norm": 0.4573829472064972,
      "learning_rate": 0.00028671865202568365,
      "loss": 3.5107,
      "step": 10730
    },
    {
      "epoch": 0.14077122709518441,
      "grad_norm": 0.4475380480289459,
      "learning_rate": 0.0002866930624285637,
      "loss": 3.4704,
      "step": 10740
    },
    {
      "epoch": 0.14090229900123208,
      "grad_norm": 0.48022469878196716,
      "learning_rate": 0.0002866674493472012,
      "loss": 3.4593,
      "step": 10750
    },
    {
      "epoch": 0.14103337090727974,
      "grad_norm": 0.48298606276512146,
      "learning_rate": 0.0002866418127859965,
      "loss": 3.4361,
      "step": 10760
    },
    {
      "epoch": 0.14116444281332738,
      "grad_norm": 0.49679988622665405,
      "learning_rate": 0.00028661615274935397,
      "loss": 3.4811,
      "step": 10770
    },
    {
      "epoch": 0.14129551471937504,
      "grad_norm": 0.5553948879241943,
      "learning_rate": 0.0002865904692416822,
      "loss": 3.4679,
      "step": 10780
    },
    {
      "epoch": 0.1414265866254227,
      "grad_norm": 0.4999101161956787,
      "learning_rate": 0.0002865647622673937,
      "loss": 3.4851,
      "step": 10790
    },
    {
      "epoch": 0.14155765853147037,
      "grad_norm": 0.5213942527770996,
      "learning_rate": 0.00028653903183090487,
      "loss": 3.4994,
      "step": 10800
    },
    {
      "epoch": 0.14168873043751803,
      "grad_norm": 0.5320174694061279,
      "learning_rate": 0.00028651327793663634,
      "loss": 3.4908,
      "step": 10810
    },
    {
      "epoch": 0.14181980234356567,
      "grad_norm": 0.48200204968452454,
      "learning_rate": 0.00028648750058901277,
      "loss": 3.4965,
      "step": 10820
    },
    {
      "epoch": 0.14195087424961333,
      "grad_norm": 0.5126617550849915,
      "learning_rate": 0.0002864616997924626,
      "loss": 3.4656,
      "step": 10830
    },
    {
      "epoch": 0.142081946155661,
      "grad_norm": 0.534378170967102,
      "learning_rate": 0.0002864358755514186,
      "loss": 3.4787,
      "step": 10840
    },
    {
      "epoch": 0.14221301806170866,
      "grad_norm": 0.5089064836502075,
      "learning_rate": 0.0002864100278703175,
      "loss": 3.4635,
      "step": 10850
    },
    {
      "epoch": 0.1423440899677563,
      "grad_norm": 0.5454674363136292,
      "learning_rate": 0.0002863841567535999,
      "loss": 3.49,
      "step": 10860
    },
    {
      "epoch": 0.14247516187380396,
      "grad_norm": 0.4899287819862366,
      "learning_rate": 0.00028635826220571055,
      "loss": 3.4722,
      "step": 10870
    },
    {
      "epoch": 0.14260623377985163,
      "grad_norm": 0.46908432245254517,
      "learning_rate": 0.0002863323442310982,
      "loss": 3.4202,
      "step": 10880
    },
    {
      "epoch": 0.1427373056858993,
      "grad_norm": 0.544380247592926,
      "learning_rate": 0.00028630640283421564,
      "loss": 3.4769,
      "step": 10890
    },
    {
      "epoch": 0.14286837759194695,
      "grad_norm": 0.4295310974121094,
      "learning_rate": 0.0002862804380195197,
      "loss": 3.4427,
      "step": 10900
    },
    {
      "epoch": 0.1429994494979946,
      "grad_norm": 0.5383688807487488,
      "learning_rate": 0.00028625444979147116,
      "loss": 3.4452,
      "step": 10910
    },
    {
      "epoch": 0.14313052140404225,
      "grad_norm": 0.5293220281600952,
      "learning_rate": 0.0002862284381545349,
      "loss": 3.4586,
      "step": 10920
    },
    {
      "epoch": 0.14326159331008992,
      "grad_norm": 0.48412325978279114,
      "learning_rate": 0.0002862024031131798,
      "loss": 3.4738,
      "step": 10930
    },
    {
      "epoch": 0.14339266521613758,
      "grad_norm": 0.5150004029273987,
      "learning_rate": 0.0002861763446718788,
      "loss": 3.422,
      "step": 10940
    },
    {
      "epoch": 0.14352373712218522,
      "grad_norm": 0.4714004695415497,
      "learning_rate": 0.0002861502628351086,
      "loss": 3.5168,
      "step": 10950
    },
    {
      "epoch": 0.14365480902823288,
      "grad_norm": 0.5016045570373535,
      "learning_rate": 0.00028612415760735036,
      "loss": 3.5183,
      "step": 10960
    },
    {
      "epoch": 0.14378588093428055,
      "grad_norm": 0.5568922162055969,
      "learning_rate": 0.00028609802899308895,
      "loss": 3.4697,
      "step": 10970
    },
    {
      "epoch": 0.1439169528403282,
      "grad_norm": 0.45118725299835205,
      "learning_rate": 0.00028607187699681327,
      "loss": 3.4296,
      "step": 10980
    },
    {
      "epoch": 0.14404802474637587,
      "grad_norm": 0.5002226233482361,
      "learning_rate": 0.00028604570162301643,
      "loss": 3.4779,
      "step": 10990
    },
    {
      "epoch": 0.1441790966524235,
      "grad_norm": 0.4887906610965729,
      "learning_rate": 0.00028601950287619537,
      "loss": 3.4427,
      "step": 11000
    },
    {
      "epoch": 0.1441790966524235,
      "eval_accuracy": 0.3736564027370479,
      "eval_loss": 3.457526445388794,
      "eval_runtime": 60.8378,
      "eval_samples_per_second": 32.874,
      "eval_steps_per_second": 0.526,
      "step": 11000
    },
    {
      "epoch": 0.14431016855847117,
      "grad_norm": 0.4549509584903717,
      "learning_rate": 0.00028599328076085105,
      "loss": 3.4518,
      "step": 11010
    },
    {
      "epoch": 0.14444124046451884,
      "grad_norm": 0.5148573517799377,
      "learning_rate": 0.0002859670352814886,
      "loss": 3.4529,
      "step": 11020
    },
    {
      "epoch": 0.1445723123705665,
      "grad_norm": 0.47731995582580566,
      "learning_rate": 0.00028594076644261704,
      "loss": 3.4355,
      "step": 11030
    },
    {
      "epoch": 0.14470338427661414,
      "grad_norm": 0.44726118445396423,
      "learning_rate": 0.00028591447424874933,
      "loss": 3.4319,
      "step": 11040
    },
    {
      "epoch": 0.1448344561826618,
      "grad_norm": 0.4602685570716858,
      "learning_rate": 0.0002858881587044027,
      "loss": 3.4587,
      "step": 11050
    },
    {
      "epoch": 0.14496552808870947,
      "grad_norm": 0.4795949459075928,
      "learning_rate": 0.00028586181981409816,
      "loss": 3.4722,
      "step": 11060
    },
    {
      "epoch": 0.14509659999475713,
      "grad_norm": 0.46392470598220825,
      "learning_rate": 0.00028583545758236075,
      "loss": 3.458,
      "step": 11070
    },
    {
      "epoch": 0.1452276719008048,
      "grad_norm": 0.48206326365470886,
      "learning_rate": 0.00028580907201371963,
      "loss": 3.4798,
      "step": 11080
    },
    {
      "epoch": 0.14535874380685243,
      "grad_norm": 0.4673754572868347,
      "learning_rate": 0.0002857826631127079,
      "loss": 3.4547,
      "step": 11090
    },
    {
      "epoch": 0.1454898157129001,
      "grad_norm": 0.4722890853881836,
      "learning_rate": 0.00028575623088386267,
      "loss": 3.481,
      "step": 11100
    },
    {
      "epoch": 0.14562088761894776,
      "grad_norm": 0.4599805176258087,
      "learning_rate": 0.0002857297753317251,
      "loss": 3.4458,
      "step": 11110
    },
    {
      "epoch": 0.14575195952499542,
      "grad_norm": 0.482412725687027,
      "learning_rate": 0.0002857032964608403,
      "loss": 3.4436,
      "step": 11120
    },
    {
      "epoch": 0.14588303143104306,
      "grad_norm": 0.534850537776947,
      "learning_rate": 0.0002856767942757575,
      "loss": 3.4541,
      "step": 11130
    },
    {
      "epoch": 0.14601410333709072,
      "grad_norm": 0.48207658529281616,
      "learning_rate": 0.0002856502687810297,
      "loss": 3.4724,
      "step": 11140
    },
    {
      "epoch": 0.1461451752431384,
      "grad_norm": 0.4739237129688263,
      "learning_rate": 0.0002856237199812141,
      "loss": 3.4751,
      "step": 11150
    },
    {
      "epoch": 0.14627624714918605,
      "grad_norm": 0.47671934962272644,
      "learning_rate": 0.00028559714788087195,
      "loss": 3.4522,
      "step": 11160
    },
    {
      "epoch": 0.14640731905523371,
      "grad_norm": 0.48334044218063354,
      "learning_rate": 0.0002855705524845683,
      "loss": 3.4493,
      "step": 11170
    },
    {
      "epoch": 0.14653839096128135,
      "grad_norm": 0.6427947282791138,
      "learning_rate": 0.0002855439337968724,
      "loss": 3.4512,
      "step": 11180
    },
    {
      "epoch": 0.14666946286732901,
      "grad_norm": 0.5009649991989136,
      "learning_rate": 0.0002855172918223573,
      "loss": 3.4595,
      "step": 11190
    },
    {
      "epoch": 0.14680053477337668,
      "grad_norm": 0.45900410413742065,
      "learning_rate": 0.00028549062656560024,
      "loss": 3.4661,
      "step": 11200
    },
    {
      "epoch": 0.14693160667942434,
      "grad_norm": 0.5070620775222778,
      "learning_rate": 0.00028546393803118237,
      "loss": 3.4461,
      "step": 11210
    },
    {
      "epoch": 0.14706267858547198,
      "grad_norm": 0.48640307784080505,
      "learning_rate": 0.00028543722622368885,
      "loss": 3.4591,
      "step": 11220
    },
    {
      "epoch": 0.14719375049151964,
      "grad_norm": 0.5324881076812744,
      "learning_rate": 0.0002854104911477089,
      "loss": 3.4238,
      "step": 11230
    },
    {
      "epoch": 0.1473248223975673,
      "grad_norm": 0.43761202692985535,
      "learning_rate": 0.00028538373280783563,
      "loss": 3.4866,
      "step": 11240
    },
    {
      "epoch": 0.14745589430361497,
      "grad_norm": 0.537097156047821,
      "learning_rate": 0.00028535695120866615,
      "loss": 3.4624,
      "step": 11250
    },
    {
      "epoch": 0.14758696620966263,
      "grad_norm": 0.4827416241168976,
      "learning_rate": 0.0002853301463548017,
      "loss": 3.4096,
      "step": 11260
    },
    {
      "epoch": 0.14771803811571027,
      "grad_norm": 0.49781015515327454,
      "learning_rate": 0.0002853033182508474,
      "loss": 3.4423,
      "step": 11270
    },
    {
      "epoch": 0.14784911002175793,
      "grad_norm": 0.5164894461631775,
      "learning_rate": 0.0002852764669014123,
      "loss": 3.5159,
      "step": 11280
    },
    {
      "epoch": 0.1479801819278056,
      "grad_norm": 0.4543289840221405,
      "learning_rate": 0.0002852495923111096,
      "loss": 3.455,
      "step": 11290
    },
    {
      "epoch": 0.14811125383385326,
      "grad_norm": 0.5365880727767944,
      "learning_rate": 0.0002852226944845565,
      "loss": 3.4687,
      "step": 11300
    },
    {
      "epoch": 0.1482423257399009,
      "grad_norm": 0.455738365650177,
      "learning_rate": 0.0002851957734263741,
      "loss": 3.4649,
      "step": 11310
    },
    {
      "epoch": 0.14837339764594856,
      "grad_norm": 0.495217502117157,
      "learning_rate": 0.0002851688291411874,
      "loss": 3.4737,
      "step": 11320
    },
    {
      "epoch": 0.14850446955199623,
      "grad_norm": 0.454673707485199,
      "learning_rate": 0.00028514186163362555,
      "loss": 3.4225,
      "step": 11330
    },
    {
      "epoch": 0.1486355414580439,
      "grad_norm": 0.46012046933174133,
      "learning_rate": 0.0002851148709083217,
      "loss": 3.4579,
      "step": 11340
    },
    {
      "epoch": 0.14876661336409155,
      "grad_norm": 0.661840558052063,
      "learning_rate": 0.0002850878569699129,
      "loss": 3.4554,
      "step": 11350
    },
    {
      "epoch": 0.1488976852701392,
      "grad_norm": 0.4678546190261841,
      "learning_rate": 0.0002850608198230402,
      "loss": 3.5092,
      "step": 11360
    },
    {
      "epoch": 0.14902875717618685,
      "grad_norm": 0.5044662356376648,
      "learning_rate": 0.0002850337594723487,
      "loss": 3.4545,
      "step": 11370
    },
    {
      "epoch": 0.14915982908223452,
      "grad_norm": 0.43815484642982483,
      "learning_rate": 0.0002850066759224874,
      "loss": 3.4489,
      "step": 11380
    },
    {
      "epoch": 0.14929090098828218,
      "grad_norm": 0.46379292011260986,
      "learning_rate": 0.00028497956917810936,
      "loss": 3.4691,
      "step": 11390
    },
    {
      "epoch": 0.14942197289432982,
      "grad_norm": 0.4638902246952057,
      "learning_rate": 0.0002849524392438715,
      "loss": 3.5029,
      "step": 11400
    },
    {
      "epoch": 0.14955304480037748,
      "grad_norm": 0.48007163405418396,
      "learning_rate": 0.00028492528612443496,
      "loss": 3.4265,
      "step": 11410
    },
    {
      "epoch": 0.14968411670642515,
      "grad_norm": 0.564456582069397,
      "learning_rate": 0.00028489810982446464,
      "loss": 3.4715,
      "step": 11420
    },
    {
      "epoch": 0.1498151886124728,
      "grad_norm": 0.47137540578842163,
      "learning_rate": 0.00028487091034862943,
      "loss": 3.4983,
      "step": 11430
    },
    {
      "epoch": 0.14994626051852045,
      "grad_norm": 0.5990513563156128,
      "learning_rate": 0.0002848436877016025,
      "loss": 3.417,
      "step": 11440
    },
    {
      "epoch": 0.1500773324245681,
      "grad_norm": 0.5240691900253296,
      "learning_rate": 0.0002848164418880605,
      "loss": 3.4729,
      "step": 11450
    },
    {
      "epoch": 0.15020840433061577,
      "grad_norm": 0.5165992379188538,
      "learning_rate": 0.0002847891729126845,
      "loss": 3.4476,
      "step": 11460
    },
    {
      "epoch": 0.15033947623666344,
      "grad_norm": 0.4929254353046417,
      "learning_rate": 0.00028476188078015944,
      "loss": 3.481,
      "step": 11470
    },
    {
      "epoch": 0.1504705481427111,
      "grad_norm": 0.49835100769996643,
      "learning_rate": 0.00028473456549517405,
      "loss": 3.5156,
      "step": 11480
    },
    {
      "epoch": 0.15060162004875874,
      "grad_norm": 0.4938746392726898,
      "learning_rate": 0.00028470722706242115,
      "loss": 3.4502,
      "step": 11490
    },
    {
      "epoch": 0.1507326919548064,
      "grad_norm": 0.4674617350101471,
      "learning_rate": 0.00028467986548659773,
      "loss": 3.4453,
      "step": 11500
    },
    {
      "epoch": 0.1507326919548064,
      "eval_accuracy": 0.37421945259042033,
      "eval_loss": 3.4517383575439453,
      "eval_runtime": 60.6656,
      "eval_samples_per_second": 32.968,
      "eval_steps_per_second": 0.527,
      "step": 11500
    },
    {
      "epoch": 0.15086376386085407,
      "grad_norm": 0.49305862188339233,
      "learning_rate": 0.0002846524807724044,
      "loss": 3.4421,
      "step": 11510
    },
    {
      "epoch": 0.15099483576690173,
      "grad_norm": 0.44207972288131714,
      "learning_rate": 0.00028462507292454606,
      "loss": 3.453,
      "step": 11520
    },
    {
      "epoch": 0.15112590767294937,
      "grad_norm": 0.46912840008735657,
      "learning_rate": 0.00028459764194773134,
      "loss": 3.4694,
      "step": 11530
    },
    {
      "epoch": 0.15125697957899703,
      "grad_norm": 0.4455344080924988,
      "learning_rate": 0.0002845701878466731,
      "loss": 3.4517,
      "step": 11540
    },
    {
      "epoch": 0.1513880514850447,
      "grad_norm": 0.4816332459449768,
      "learning_rate": 0.00028454271062608786,
      "loss": 3.4605,
      "step": 11550
    },
    {
      "epoch": 0.15151912339109236,
      "grad_norm": 0.45779168605804443,
      "learning_rate": 0.0002845152102906964,
      "loss": 3.4353,
      "step": 11560
    },
    {
      "epoch": 0.15165019529714002,
      "grad_norm": 0.48170971870422363,
      "learning_rate": 0.00028448768684522337,
      "loss": 3.4608,
      "step": 11570
    },
    {
      "epoch": 0.15178126720318766,
      "grad_norm": 0.4514457583427429,
      "learning_rate": 0.00028446014029439724,
      "loss": 3.4576,
      "step": 11580
    },
    {
      "epoch": 0.15191233910923532,
      "grad_norm": 0.5135019421577454,
      "learning_rate": 0.00028443257064295074,
      "loss": 3.4571,
      "step": 11590
    },
    {
      "epoch": 0.152043411015283,
      "grad_norm": 0.4359944462776184,
      "learning_rate": 0.0002844049778956203,
      "loss": 3.4281,
      "step": 11600
    },
    {
      "epoch": 0.15217448292133065,
      "grad_norm": 0.5181604623794556,
      "learning_rate": 0.0002843773620571465,
      "loss": 3.5211,
      "step": 11610
    },
    {
      "epoch": 0.1523055548273783,
      "grad_norm": 0.4558055102825165,
      "learning_rate": 0.00028434972313227377,
      "loss": 3.5045,
      "step": 11620
    },
    {
      "epoch": 0.15243662673342595,
      "grad_norm": 0.4575844705104828,
      "learning_rate": 0.00028432206112575056,
      "loss": 3.4452,
      "step": 11630
    },
    {
      "epoch": 0.15256769863947361,
      "grad_norm": 0.4772001802921295,
      "learning_rate": 0.00028429437604232937,
      "loss": 3.478,
      "step": 11640
    },
    {
      "epoch": 0.15269877054552128,
      "grad_norm": 0.4719308316707611,
      "learning_rate": 0.0002842666678867664,
      "loss": 3.4507,
      "step": 11650
    },
    {
      "epoch": 0.15282984245156894,
      "grad_norm": 0.6219125390052795,
      "learning_rate": 0.00028423893666382205,
      "loss": 3.487,
      "step": 11660
    },
    {
      "epoch": 0.15296091435761658,
      "grad_norm": 0.432285875082016,
      "learning_rate": 0.0002842111823782607,
      "loss": 3.4387,
      "step": 11670
    },
    {
      "epoch": 0.15309198626366424,
      "grad_norm": 0.5074673891067505,
      "learning_rate": 0.0002841834050348506,
      "loss": 3.4571,
      "step": 11680
    },
    {
      "epoch": 0.1532230581697119,
      "grad_norm": 0.7799618244171143,
      "learning_rate": 0.00028415560463836386,
      "loss": 3.5287,
      "step": 11690
    },
    {
      "epoch": 0.15335413007575957,
      "grad_norm": 0.5166928172111511,
      "learning_rate": 0.00028412778119357674,
      "loss": 3.4774,
      "step": 11700
    },
    {
      "epoch": 0.1534852019818072,
      "grad_norm": 0.48976626992225647,
      "learning_rate": 0.00028409993470526946,
      "loss": 3.4091,
      "step": 11710
    },
    {
      "epoch": 0.15361627388785487,
      "grad_norm": 0.45122426748275757,
      "learning_rate": 0.00028407206517822595,
      "loss": 3.4851,
      "step": 11720
    },
    {
      "epoch": 0.15374734579390253,
      "grad_norm": 0.5581915974617004,
      "learning_rate": 0.0002840441726172344,
      "loss": 3.4857,
      "step": 11730
    },
    {
      "epoch": 0.1538784176999502,
      "grad_norm": 0.4775165021419525,
      "learning_rate": 0.00028401625702708684,
      "loss": 3.468,
      "step": 11740
    },
    {
      "epoch": 0.15400948960599786,
      "grad_norm": 0.5234624743461609,
      "learning_rate": 0.0002839883184125792,
      "loss": 3.4315,
      "step": 11750
    },
    {
      "epoch": 0.1541405615120455,
      "grad_norm": 0.47337770462036133,
      "learning_rate": 0.00028396035677851134,
      "loss": 3.4486,
      "step": 11760
    },
    {
      "epoch": 0.15427163341809316,
      "grad_norm": 0.4782180190086365,
      "learning_rate": 0.00028393237212968723,
      "loss": 3.4353,
      "step": 11770
    },
    {
      "epoch": 0.15440270532414083,
      "grad_norm": 0.4975164532661438,
      "learning_rate": 0.00028390436447091474,
      "loss": 3.4551,
      "step": 11780
    },
    {
      "epoch": 0.1545337772301885,
      "grad_norm": 0.5054335594177246,
      "learning_rate": 0.0002838763338070056,
      "loss": 3.4394,
      "step": 11790
    },
    {
      "epoch": 0.15466484913623613,
      "grad_norm": 0.43333151936531067,
      "learning_rate": 0.0002838482801427756,
      "loss": 3.4609,
      "step": 11800
    },
    {
      "epoch": 0.1547959210422838,
      "grad_norm": 0.45587974786758423,
      "learning_rate": 0.00028382020348304433,
      "loss": 3.4731,
      "step": 11810
    },
    {
      "epoch": 0.15492699294833145,
      "grad_norm": 0.4675145447254181,
      "learning_rate": 0.0002837921038326355,
      "loss": 3.4858,
      "step": 11820
    },
    {
      "epoch": 0.15505806485437912,
      "grad_norm": 0.4495203197002411,
      "learning_rate": 0.00028376398119637683,
      "loss": 3.443,
      "step": 11830
    },
    {
      "epoch": 0.15518913676042678,
      "grad_norm": 0.4804924428462982,
      "learning_rate": 0.0002837358355790997,
      "loss": 3.475,
      "step": 11840
    },
    {
      "epoch": 0.15532020866647442,
      "grad_norm": 0.528252363204956,
      "learning_rate": 0.0002837076669856397,
      "loss": 3.4667,
      "step": 11850
    },
    {
      "epoch": 0.15545128057252208,
      "grad_norm": 0.4776580333709717,
      "learning_rate": 0.0002836794754208362,
      "loss": 3.543,
      "step": 11860
    },
    {
      "epoch": 0.15558235247856975,
      "grad_norm": 0.46633362770080566,
      "learning_rate": 0.00028365126088953263,
      "loss": 3.4873,
      "step": 11870
    },
    {
      "epoch": 0.1557134243846174,
      "grad_norm": 0.48364025354385376,
      "learning_rate": 0.0002836230233965763,
      "loss": 3.4753,
      "step": 11880
    },
    {
      "epoch": 0.15584449629066505,
      "grad_norm": 0.4588400721549988,
      "learning_rate": 0.00028359476294681854,
      "loss": 3.4552,
      "step": 11890
    },
    {
      "epoch": 0.1559755681967127,
      "grad_norm": 0.444388210773468,
      "learning_rate": 0.0002835664795451145,
      "loss": 3.4745,
      "step": 11900
    },
    {
      "epoch": 0.15610664010276037,
      "grad_norm": 0.46250733733177185,
      "learning_rate": 0.0002835381731963234,
      "loss": 3.4183,
      "step": 11910
    },
    {
      "epoch": 0.15623771200880804,
      "grad_norm": 0.5620143413543701,
      "learning_rate": 0.0002835098439053083,
      "loss": 3.4487,
      "step": 11920
    },
    {
      "epoch": 0.1563687839148557,
      "grad_norm": 0.47562694549560547,
      "learning_rate": 0.00028348149167693634,
      "loss": 3.4505,
      "step": 11930
    },
    {
      "epoch": 0.15649985582090334,
      "grad_norm": 0.4566475749015808,
      "learning_rate": 0.0002834531165160784,
      "loss": 3.4335,
      "step": 11940
    },
    {
      "epoch": 0.156630927726951,
      "grad_norm": 0.4564806818962097,
      "learning_rate": 0.0002834247184276095,
      "loss": 3.4387,
      "step": 11950
    },
    {
      "epoch": 0.15676199963299867,
      "grad_norm": 0.45230308175086975,
      "learning_rate": 0.0002833962974164085,
      "loss": 3.4462,
      "step": 11960
    },
    {
      "epoch": 0.15689307153904633,
      "grad_norm": 0.532983124256134,
      "learning_rate": 0.0002833678534873581,
      "loss": 3.4872,
      "step": 11970
    },
    {
      "epoch": 0.15702414344509397,
      "grad_norm": 0.4635123014450073,
      "learning_rate": 0.00028333938664534515,
      "loss": 3.4783,
      "step": 11980
    },
    {
      "epoch": 0.15715521535114163,
      "grad_norm": 0.5696286559104919,
      "learning_rate": 0.00028331089689526037,
      "loss": 3.4411,
      "step": 11990
    },
    {
      "epoch": 0.1572862872571893,
      "grad_norm": 0.49350249767303467,
      "learning_rate": 0.0002832823842419982,
      "loss": 3.4727,
      "step": 12000
    },
    {
      "epoch": 0.1572862872571893,
      "eval_accuracy": 0.3744755620723363,
      "eval_loss": 3.448422431945801,
      "eval_runtime": 63.4292,
      "eval_samples_per_second": 31.531,
      "eval_steps_per_second": 0.504,
      "step": 12000
    },
    {
      "epoch": 0.15741735916323696,
      "grad_norm": 0.7076990008354187,
      "learning_rate": 0.00028325384869045734,
      "loss": 3.427,
      "step": 12010
    },
    {
      "epoch": 0.15754843106928462,
      "grad_norm": 0.5164821743965149,
      "learning_rate": 0.0002832252902455403,
      "loss": 3.4495,
      "step": 12020
    },
    {
      "epoch": 0.15767950297533226,
      "grad_norm": 0.514133095741272,
      "learning_rate": 0.00028319670891215336,
      "loss": 3.4307,
      "step": 12030
    },
    {
      "epoch": 0.15781057488137992,
      "grad_norm": 0.585483193397522,
      "learning_rate": 0.000283168104695207,
      "loss": 3.4255,
      "step": 12040
    },
    {
      "epoch": 0.1579416467874276,
      "grad_norm": 0.44780194759368896,
      "learning_rate": 0.00028313947759961547,
      "loss": 3.4214,
      "step": 12050
    },
    {
      "epoch": 0.15807271869347525,
      "grad_norm": 0.46449315547943115,
      "learning_rate": 0.0002831108276302969,
      "loss": 3.4482,
      "step": 12060
    },
    {
      "epoch": 0.1582037905995229,
      "grad_norm": 0.5253653526306152,
      "learning_rate": 0.0002830821547921736,
      "loss": 3.4892,
      "step": 12070
    },
    {
      "epoch": 0.15833486250557055,
      "grad_norm": 0.48446130752563477,
      "learning_rate": 0.0002830534590901715,
      "loss": 3.4957,
      "step": 12080
    },
    {
      "epoch": 0.15846593441161821,
      "grad_norm": 0.4688984751701355,
      "learning_rate": 0.0002830247405292206,
      "loss": 3.4668,
      "step": 12090
    },
    {
      "epoch": 0.15859700631766588,
      "grad_norm": 0.5757489800453186,
      "learning_rate": 0.0002829959991142549,
      "loss": 3.4337,
      "step": 12100
    },
    {
      "epoch": 0.15872807822371354,
      "grad_norm": 0.467786580324173,
      "learning_rate": 0.0002829672348502122,
      "loss": 3.4431,
      "step": 12110
    },
    {
      "epoch": 0.15885915012976118,
      "grad_norm": 0.5425522923469543,
      "learning_rate": 0.00028293844774203436,
      "loss": 3.4775,
      "step": 12120
    },
    {
      "epoch": 0.15899022203580884,
      "grad_norm": 0.9176479578018188,
      "learning_rate": 0.000282909637794667,
      "loss": 3.4935,
      "step": 12130
    },
    {
      "epoch": 0.1591212939418565,
      "grad_norm": 0.48447319865226746,
      "learning_rate": 0.00028288080501305974,
      "loss": 3.4529,
      "step": 12140
    },
    {
      "epoch": 0.15925236584790417,
      "grad_norm": 0.5930560827255249,
      "learning_rate": 0.0002828519494021662,
      "loss": 3.4385,
      "step": 12150
    },
    {
      "epoch": 0.1593834377539518,
      "grad_norm": 0.4580259621143341,
      "learning_rate": 0.00028282307096694383,
      "loss": 3.4508,
      "step": 12160
    },
    {
      "epoch": 0.15951450965999947,
      "grad_norm": 0.6035920977592468,
      "learning_rate": 0.000282794169712354,
      "loss": 3.4558,
      "step": 12170
    },
    {
      "epoch": 0.15964558156604713,
      "grad_norm": 0.5065329670906067,
      "learning_rate": 0.000282765245643362,
      "loss": 3.4523,
      "step": 12180
    },
    {
      "epoch": 0.1597766534720948,
      "grad_norm": 0.5191739201545715,
      "learning_rate": 0.0002827362987649371,
      "loss": 3.4502,
      "step": 12190
    },
    {
      "epoch": 0.15990772537814246,
      "grad_norm": 0.48273390531539917,
      "learning_rate": 0.00028270732908205256,
      "loss": 3.4531,
      "step": 12200
    },
    {
      "epoch": 0.1600387972841901,
      "grad_norm": 0.45366406440734863,
      "learning_rate": 0.00028267833659968526,
      "loss": 3.4153,
      "step": 12210
    },
    {
      "epoch": 0.16016986919023776,
      "grad_norm": 0.4847581386566162,
      "learning_rate": 0.0002826493213228163,
      "loss": 3.4411,
      "step": 12220
    },
    {
      "epoch": 0.16030094109628543,
      "grad_norm": 0.5094118714332581,
      "learning_rate": 0.00028262028325643055,
      "loss": 3.588,
      "step": 12230
    },
    {
      "epoch": 0.1604320130023331,
      "grad_norm": 0.496362566947937,
      "learning_rate": 0.00028259122240551686,
      "loss": 3.4443,
      "step": 12240
    },
    {
      "epoch": 0.16056308490838073,
      "grad_norm": 0.4638541638851166,
      "learning_rate": 0.00028256213877506786,
      "loss": 3.5058,
      "step": 12250
    },
    {
      "epoch": 0.1606941568144284,
      "grad_norm": 0.5724948644638062,
      "learning_rate": 0.0002825330323700803,
      "loss": 3.4386,
      "step": 12260
    },
    {
      "epoch": 0.16082522872047605,
      "grad_norm": 0.8594478368759155,
      "learning_rate": 0.0002825039031955548,
      "loss": 3.4571,
      "step": 12270
    },
    {
      "epoch": 0.16095630062652372,
      "grad_norm": 0.48069873452186584,
      "learning_rate": 0.0002824747512564957,
      "loss": 3.4791,
      "step": 12280
    },
    {
      "epoch": 0.16108737253257138,
      "grad_norm": 0.46263888478279114,
      "learning_rate": 0.0002824455765579115,
      "loss": 3.4341,
      "step": 12290
    },
    {
      "epoch": 0.16121844443861902,
      "grad_norm": 0.47283491492271423,
      "learning_rate": 0.00028241637910481434,
      "loss": 3.4388,
      "step": 12300
    },
    {
      "epoch": 0.16134951634466668,
      "grad_norm": 0.44265952706336975,
      "learning_rate": 0.00028238715890222053,
      "loss": 3.4505,
      "step": 12310
    },
    {
      "epoch": 0.16148058825071435,
      "grad_norm": 0.4213998019695282,
      "learning_rate": 0.00028235791595515017,
      "loss": 3.4622,
      "step": 12320
    },
    {
      "epoch": 0.161611660156762,
      "grad_norm": 0.4900745153427124,
      "learning_rate": 0.00028232865026862723,
      "loss": 3.4376,
      "step": 12330
    },
    {
      "epoch": 0.16174273206280965,
      "grad_norm": 0.44946059584617615,
      "learning_rate": 0.00028229936184767976,
      "loss": 3.4879,
      "step": 12340
    },
    {
      "epoch": 0.1618738039688573,
      "grad_norm": 0.4458317756652832,
      "learning_rate": 0.00028227005069733944,
      "loss": 3.4345,
      "step": 12350
    },
    {
      "epoch": 0.16200487587490497,
      "grad_norm": 0.5682699680328369,
      "learning_rate": 0.00028224071682264205,
      "loss": 3.3999,
      "step": 12360
    },
    {
      "epoch": 0.16213594778095264,
      "grad_norm": 0.48975199460983276,
      "learning_rate": 0.0002822113602286273,
      "loss": 3.4245,
      "step": 12370
    },
    {
      "epoch": 0.16226701968700027,
      "grad_norm": 0.46664556860923767,
      "learning_rate": 0.00028218198092033866,
      "loss": 3.448,
      "step": 12380
    },
    {
      "epoch": 0.16239809159304794,
      "grad_norm": 0.4667070508003235,
      "learning_rate": 0.00028215257890282366,
      "loss": 3.428,
      "step": 12390
    },
    {
      "epoch": 0.1625291634990956,
      "grad_norm": 0.44727087020874023,
      "learning_rate": 0.0002821231541811335,
      "loss": 3.4334,
      "step": 12400
    },
    {
      "epoch": 0.16266023540514327,
      "grad_norm": 0.47045594453811646,
      "learning_rate": 0.0002820937067603236,
      "loss": 3.4417,
      "step": 12410
    },
    {
      "epoch": 0.16279130731119093,
      "grad_norm": 0.4491787254810333,
      "learning_rate": 0.00028206423664545296,
      "loss": 3.4971,
      "step": 12420
    },
    {
      "epoch": 0.16292237921723857,
      "grad_norm": 0.477226197719574,
      "learning_rate": 0.00028203474384158464,
      "loss": 3.4361,
      "step": 12430
    },
    {
      "epoch": 0.16305345112328623,
      "grad_norm": 0.491279661655426,
      "learning_rate": 0.00028200522835378574,
      "loss": 3.4506,
      "step": 12440
    },
    {
      "epoch": 0.1631845230293339,
      "grad_norm": 0.42322397232055664,
      "learning_rate": 0.0002819756901871269,
      "loss": 3.4618,
      "step": 12450
    },
    {
      "epoch": 0.16331559493538156,
      "grad_norm": 0.5434552431106567,
      "learning_rate": 0.00028194612934668304,
      "loss": 3.4791,
      "step": 12460
    },
    {
      "epoch": 0.1634466668414292,
      "grad_norm": 0.4583127498626709,
      "learning_rate": 0.0002819165458375327,
      "loss": 3.4479,
      "step": 12470
    },
    {
      "epoch": 0.16357773874747686,
      "grad_norm": 0.4839136004447937,
      "learning_rate": 0.00028188693966475835,
      "loss": 3.4589,
      "step": 12480
    },
    {
      "epoch": 0.16370881065352452,
      "grad_norm": 0.4750353693962097,
      "learning_rate": 0.0002818573108334465,
      "loss": 3.4451,
      "step": 12490
    },
    {
      "epoch": 0.1638398825595722,
      "grad_norm": 0.4929468333721161,
      "learning_rate": 0.0002818276593486874,
      "loss": 3.4677,
      "step": 12500
    },
    {
      "epoch": 0.1638398825595722,
      "eval_accuracy": 0.3751622678396872,
      "eval_loss": 3.441032886505127,
      "eval_runtime": 61.4262,
      "eval_samples_per_second": 32.559,
      "eval_steps_per_second": 0.521,
      "step": 12500
    },
    {
      "epoch": 0.16397095446561985,
      "grad_norm": 0.4697575271129608,
      "learning_rate": 0.0002817979852155753,
      "loss": 3.4215,
      "step": 12510
    },
    {
      "epoch": 0.1641020263716675,
      "grad_norm": 0.4972652196884155,
      "learning_rate": 0.00028176828843920845,
      "loss": 3.44,
      "step": 12520
    },
    {
      "epoch": 0.16423309827771515,
      "grad_norm": 0.592870831489563,
      "learning_rate": 0.00028173856902468855,
      "loss": 3.4488,
      "step": 12530
    },
    {
      "epoch": 0.16436417018376281,
      "grad_norm": 0.4708905816078186,
      "learning_rate": 0.0002817088269771216,
      "loss": 3.4357,
      "step": 12540
    },
    {
      "epoch": 0.16449524208981048,
      "grad_norm": 0.4656590223312378,
      "learning_rate": 0.0002816790623016174,
      "loss": 3.4499,
      "step": 12550
    },
    {
      "epoch": 0.16462631399585811,
      "grad_norm": 0.48974642157554626,
      "learning_rate": 0.0002816492750032896,
      "loss": 3.4247,
      "step": 12560
    },
    {
      "epoch": 0.16475738590190578,
      "grad_norm": 0.5321942567825317,
      "learning_rate": 0.00028161946508725564,
      "loss": 3.4493,
      "step": 12570
    },
    {
      "epoch": 0.16488845780795344,
      "grad_norm": 0.4596943259239197,
      "learning_rate": 0.0002815896325586371,
      "loss": 3.4652,
      "step": 12580
    },
    {
      "epoch": 0.1650195297140011,
      "grad_norm": 0.5785807371139526,
      "learning_rate": 0.00028155977742255913,
      "loss": 3.4563,
      "step": 12590
    },
    {
      "epoch": 0.16515060162004877,
      "grad_norm": 0.45556116104125977,
      "learning_rate": 0.00028152989968415105,
      "loss": 3.4439,
      "step": 12600
    },
    {
      "epoch": 0.1652816735260964,
      "grad_norm": 0.45564770698547363,
      "learning_rate": 0.0002814999993485459,
      "loss": 3.4303,
      "step": 12610
    },
    {
      "epoch": 0.16541274543214407,
      "grad_norm": 0.5735878944396973,
      "learning_rate": 0.00028147007642088066,
      "loss": 3.4776,
      "step": 12620
    },
    {
      "epoch": 0.16554381733819173,
      "grad_norm": 0.4601587653160095,
      "learning_rate": 0.0002814401309062961,
      "loss": 3.4243,
      "step": 12630
    },
    {
      "epoch": 0.1656748892442394,
      "grad_norm": 0.4892955422401428,
      "learning_rate": 0.000281410162809937,
      "loss": 3.4428,
      "step": 12640
    },
    {
      "epoch": 0.16580596115028703,
      "grad_norm": 0.8085596561431885,
      "learning_rate": 0.0002813801721369519,
      "loss": 3.4271,
      "step": 12650
    },
    {
      "epoch": 0.1659370330563347,
      "grad_norm": 0.48498865962028503,
      "learning_rate": 0.0002813501588924934,
      "loss": 3.4602,
      "step": 12660
    },
    {
      "epoch": 0.16606810496238236,
      "grad_norm": 0.4812212586402893,
      "learning_rate": 0.0002813201230817178,
      "loss": 3.4236,
      "step": 12670
    },
    {
      "epoch": 0.16619917686843003,
      "grad_norm": 0.4619809091091156,
      "learning_rate": 0.00028129006470978523,
      "loss": 3.4345,
      "step": 12680
    },
    {
      "epoch": 0.1663302487744777,
      "grad_norm": 0.5014129877090454,
      "learning_rate": 0.0002812599837818599,
      "loss": 3.449,
      "step": 12690
    },
    {
      "epoch": 0.16646132068052533,
      "grad_norm": 0.49772366881370544,
      "learning_rate": 0.0002812298803031098,
      "loss": 3.4368,
      "step": 12700
    },
    {
      "epoch": 0.166592392586573,
      "grad_norm": 0.46379750967025757,
      "learning_rate": 0.0002811997542787068,
      "loss": 3.4885,
      "step": 12710
    },
    {
      "epoch": 0.16672346449262065,
      "grad_norm": 0.6322084069252014,
      "learning_rate": 0.0002811696057138266,
      "loss": 3.4667,
      "step": 12720
    },
    {
      "epoch": 0.16685453639866832,
      "grad_norm": 0.45088011026382446,
      "learning_rate": 0.0002811394346136488,
      "loss": 3.4614,
      "step": 12730
    },
    {
      "epoch": 0.16698560830471595,
      "grad_norm": 0.5803069472312927,
      "learning_rate": 0.00028110924098335684,
      "loss": 3.4432,
      "step": 12740
    },
    {
      "epoch": 0.16711668021076362,
      "grad_norm": 0.5069870948791504,
      "learning_rate": 0.0002810790248281382,
      "loss": 3.4453,
      "step": 12750
    },
    {
      "epoch": 0.16724775211681128,
      "grad_norm": 0.5113685131072998,
      "learning_rate": 0.000281048786153184,
      "loss": 3.4341,
      "step": 12760
    },
    {
      "epoch": 0.16737882402285895,
      "grad_norm": 0.4828410744667053,
      "learning_rate": 0.0002810185249636893,
      "loss": 3.4399,
      "step": 12770
    },
    {
      "epoch": 0.1675098959289066,
      "grad_norm": 0.4671396017074585,
      "learning_rate": 0.0002809882412648532,
      "loss": 3.4533,
      "step": 12780
    },
    {
      "epoch": 0.16764096783495425,
      "grad_norm": 0.5339388847351074,
      "learning_rate": 0.0002809579350618783,
      "loss": 3.4333,
      "step": 12790
    },
    {
      "epoch": 0.1677720397410019,
      "grad_norm": 0.48319700360298157,
      "learning_rate": 0.0002809276063599715,
      "loss": 3.4455,
      "step": 12800
    },
    {
      "epoch": 0.16790311164704957,
      "grad_norm": 0.4890466630458832,
      "learning_rate": 0.0002808972551643432,
      "loss": 3.401,
      "step": 12810
    },
    {
      "epoch": 0.16803418355309724,
      "grad_norm": 0.4491477608680725,
      "learning_rate": 0.0002808668814802079,
      "loss": 3.457,
      "step": 12820
    },
    {
      "epoch": 0.16816525545914487,
      "grad_norm": 0.47704818844795227,
      "learning_rate": 0.0002808364853127839,
      "loss": 3.3966,
      "step": 12830
    },
    {
      "epoch": 0.16829632736519254,
      "grad_norm": 0.46327733993530273,
      "learning_rate": 0.0002808060666672933,
      "loss": 3.4784,
      "step": 12840
    },
    {
      "epoch": 0.1684273992712402,
      "grad_norm": 0.48739147186279297,
      "learning_rate": 0.0002807756255489621,
      "loss": 3.4415,
      "step": 12850
    },
    {
      "epoch": 0.16855847117728787,
      "grad_norm": 0.4859282076358795,
      "learning_rate": 0.0002807451619630202,
      "loss": 3.4633,
      "step": 12860
    },
    {
      "epoch": 0.16868954308333553,
      "grad_norm": 0.42382189631462097,
      "learning_rate": 0.0002807146759147013,
      "loss": 3.4624,
      "step": 12870
    },
    {
      "epoch": 0.16882061498938317,
      "grad_norm": 0.4561120867729187,
      "learning_rate": 0.000280684167409243,
      "loss": 3.4661,
      "step": 12880
    },
    {
      "epoch": 0.16895168689543083,
      "grad_norm": 0.47691744565963745,
      "learning_rate": 0.00028065363645188676,
      "loss": 3.4472,
      "step": 12890
    },
    {
      "epoch": 0.1690827588014785,
      "grad_norm": 0.5770523548126221,
      "learning_rate": 0.00028062308304787785,
      "loss": 3.4754,
      "step": 12900
    },
    {
      "epoch": 0.16921383070752616,
      "grad_norm": 0.5005218982696533,
      "learning_rate": 0.00028059250720246544,
      "loss": 3.4228,
      "step": 12910
    },
    {
      "epoch": 0.1693449026135738,
      "grad_norm": 0.47962647676467896,
      "learning_rate": 0.00028056190892090254,
      "loss": 3.4688,
      "step": 12920
    },
    {
      "epoch": 0.16947597451962146,
      "grad_norm": 0.4576781392097473,
      "learning_rate": 0.0002805312882084461,
      "loss": 3.4075,
      "step": 12930
    },
    {
      "epoch": 0.16960704642566912,
      "grad_norm": 0.5028262734413147,
      "learning_rate": 0.0002805006450703567,
      "loss": 3.4436,
      "step": 12940
    },
    {
      "epoch": 0.1697381183317168,
      "grad_norm": 0.46324482560157776,
      "learning_rate": 0.000280469979511899,
      "loss": 3.4583,
      "step": 12950
    },
    {
      "epoch": 0.16986919023776445,
      "grad_norm": 0.42851150035858154,
      "learning_rate": 0.0002804392915383414,
      "loss": 3.4515,
      "step": 12960
    },
    {
      "epoch": 0.1700002621438121,
      "grad_norm": 0.5224888920783997,
      "learning_rate": 0.0002804085811549562,
      "loss": 3.4239,
      "step": 12970
    },
    {
      "epoch": 0.17013133404985975,
      "grad_norm": 0.47697246074676514,
      "learning_rate": 0.00028037784836701954,
      "loss": 3.4538,
      "step": 12980
    },
    {
      "epoch": 0.17026240595590741,
      "grad_norm": 0.4745752811431885,
      "learning_rate": 0.0002803470931798113,
      "loss": 3.4665,
      "step": 12990
    },
    {
      "epoch": 0.17039347786195508,
      "grad_norm": 0.4403124153614044,
      "learning_rate": 0.00028031631559861546,
      "loss": 3.4091,
      "step": 13000
    },
    {
      "epoch": 0.17039347786195508,
      "eval_accuracy": 0.37676197458455524,
      "eval_loss": 3.4280595779418945,
      "eval_runtime": 61.1234,
      "eval_samples_per_second": 32.721,
      "eval_steps_per_second": 0.524,
      "step": 13000
    },
    {
      "epoch": 0.17052454976800271,
      "grad_norm": 0.46125295758247375,
      "learning_rate": 0.0002802855156287196,
      "loss": 3.4219,
      "step": 13010
    },
    {
      "epoch": 0.17065562167405038,
      "grad_norm": 0.4351152181625366,
      "learning_rate": 0.00028025469327541523,
      "loss": 3.4399,
      "step": 13020
    },
    {
      "epoch": 0.17078669358009804,
      "grad_norm": 0.6186047792434692,
      "learning_rate": 0.0002802238485439977,
      "loss": 3.5254,
      "step": 13030
    },
    {
      "epoch": 0.1709177654861457,
      "grad_norm": 0.522502064704895,
      "learning_rate": 0.00028019298143976633,
      "loss": 3.4626,
      "step": 13040
    },
    {
      "epoch": 0.17104883739219337,
      "grad_norm": 0.478019118309021,
      "learning_rate": 0.00028016209196802405,
      "loss": 3.4582,
      "step": 13050
    },
    {
      "epoch": 0.171179909298241,
      "grad_norm": 0.5090868473052979,
      "learning_rate": 0.0002801311801340778,
      "loss": 3.4494,
      "step": 13060
    },
    {
      "epoch": 0.17131098120428867,
      "grad_norm": 0.4480034112930298,
      "learning_rate": 0.00028010024594323835,
      "loss": 3.4607,
      "step": 13070
    },
    {
      "epoch": 0.17144205311033633,
      "grad_norm": 0.900874137878418,
      "learning_rate": 0.0002800692894008203,
      "loss": 3.4143,
      "step": 13080
    },
    {
      "epoch": 0.171573125016384,
      "grad_norm": 0.5846623182296753,
      "learning_rate": 0.00028003831051214196,
      "loss": 3.4112,
      "step": 13090
    },
    {
      "epoch": 0.17170419692243163,
      "grad_norm": 0.4678378403186798,
      "learning_rate": 0.0002800073092825256,
      "loss": 3.4448,
      "step": 13100
    },
    {
      "epoch": 0.1718352688284793,
      "grad_norm": 0.423756867647171,
      "learning_rate": 0.00027997628571729745,
      "loss": 3.4025,
      "step": 13110
    },
    {
      "epoch": 0.17196634073452696,
      "grad_norm": 0.457555890083313,
      "learning_rate": 0.0002799452398217873,
      "loss": 3.4641,
      "step": 13120
    },
    {
      "epoch": 0.17209741264057463,
      "grad_norm": 0.7598799467086792,
      "learning_rate": 0.00027991417160132904,
      "loss": 3.4898,
      "step": 13130
    },
    {
      "epoch": 0.1722284845466223,
      "grad_norm": 0.4833213686943054,
      "learning_rate": 0.00027988308106126017,
      "loss": 3.413,
      "step": 13140
    },
    {
      "epoch": 0.17235955645266993,
      "grad_norm": 0.45229774713516235,
      "learning_rate": 0.00027985196820692215,
      "loss": 3.4299,
      "step": 13150
    },
    {
      "epoch": 0.1724906283587176,
      "grad_norm": 0.46664178371429443,
      "learning_rate": 0.0002798208330436603,
      "loss": 3.4508,
      "step": 13160
    },
    {
      "epoch": 0.17262170026476525,
      "grad_norm": 0.44830113649368286,
      "learning_rate": 0.00027978967557682377,
      "loss": 3.3941,
      "step": 13170
    },
    {
      "epoch": 0.17275277217081292,
      "grad_norm": 0.4428168833255768,
      "learning_rate": 0.00027975849581176536,
      "loss": 3.4241,
      "step": 13180
    },
    {
      "epoch": 0.17288384407686055,
      "grad_norm": 0.49060672521591187,
      "learning_rate": 0.0002797272937538419,
      "loss": 3.4431,
      "step": 13190
    },
    {
      "epoch": 0.17301491598290822,
      "grad_norm": 0.4401482939720154,
      "learning_rate": 0.000279696069408414,
      "loss": 3.413,
      "step": 13200
    },
    {
      "epoch": 0.17314598788895588,
      "grad_norm": 0.45251721143722534,
      "learning_rate": 0.0002796648227808461,
      "loss": 3.4184,
      "step": 13210
    },
    {
      "epoch": 0.17327705979500355,
      "grad_norm": 0.4663998484611511,
      "learning_rate": 0.00027963355387650645,
      "loss": 3.4368,
      "step": 13220
    },
    {
      "epoch": 0.1734081317010512,
      "grad_norm": 0.4752306640148163,
      "learning_rate": 0.00027960226270076717,
      "loss": 3.4283,
      "step": 13230
    },
    {
      "epoch": 0.17353920360709885,
      "grad_norm": 0.4840189516544342,
      "learning_rate": 0.000279570949259004,
      "loss": 3.3978,
      "step": 13240
    },
    {
      "epoch": 0.1736702755131465,
      "grad_norm": 0.5018368363380432,
      "learning_rate": 0.00027953961355659695,
      "loss": 3.4412,
      "step": 13250
    },
    {
      "epoch": 0.17380134741919417,
      "grad_norm": 0.45143285393714905,
      "learning_rate": 0.0002795082555989293,
      "loss": 3.4164,
      "step": 13260
    },
    {
      "epoch": 0.17393241932524184,
      "grad_norm": 0.4826856553554535,
      "learning_rate": 0.00027947687539138863,
      "loss": 3.4462,
      "step": 13270
    },
    {
      "epoch": 0.17406349123128947,
      "grad_norm": 0.4343440532684326,
      "learning_rate": 0.00027944547293936613,
      "loss": 3.4808,
      "step": 13280
    },
    {
      "epoch": 0.17419456313733714,
      "grad_norm": 0.5318564772605896,
      "learning_rate": 0.00027941404824825675,
      "loss": 3.435,
      "step": 13290
    },
    {
      "epoch": 0.1743256350433848,
      "grad_norm": 0.5284613370895386,
      "learning_rate": 0.0002793826013234593,
      "loss": 3.4437,
      "step": 13300
    },
    {
      "epoch": 0.17445670694943247,
      "grad_norm": 0.5311353206634521,
      "learning_rate": 0.0002793511321703766,
      "loss": 3.4235,
      "step": 13310
    },
    {
      "epoch": 0.17458777885548013,
      "grad_norm": 0.5232999324798584,
      "learning_rate": 0.00027931964079441497,
      "loss": 3.463,
      "step": 13320
    },
    {
      "epoch": 0.17471885076152777,
      "grad_norm": 0.5391507744789124,
      "learning_rate": 0.0002792881272009848,
      "loss": 3.4568,
      "step": 13330
    },
    {
      "epoch": 0.17484992266757543,
      "grad_norm": 0.48401257395744324,
      "learning_rate": 0.00027925659139550025,
      "loss": 3.4111,
      "step": 13340
    },
    {
      "epoch": 0.1749809945736231,
      "grad_norm": 0.5197456479072571,
      "learning_rate": 0.0002792250333833792,
      "loss": 3.4713,
      "step": 13350
    },
    {
      "epoch": 0.17511206647967076,
      "grad_norm": 0.4938254654407501,
      "learning_rate": 0.0002791934531700434,
      "loss": 3.4343,
      "step": 13360
    },
    {
      "epoch": 0.1752431383857184,
      "grad_norm": 0.4608302712440491,
      "learning_rate": 0.00027916185076091847,
      "loss": 3.4331,
      "step": 13370
    },
    {
      "epoch": 0.17537421029176606,
      "grad_norm": 0.6390073895454407,
      "learning_rate": 0.0002791302261614337,
      "loss": 3.4549,
      "step": 13380
    },
    {
      "epoch": 0.17550528219781372,
      "grad_norm": 0.46695172786712646,
      "learning_rate": 0.00027909857937702235,
      "loss": 3.4378,
      "step": 13390
    },
    {
      "epoch": 0.1756363541038614,
      "grad_norm": 0.45665082335472107,
      "learning_rate": 0.0002790669104131214,
      "loss": 3.4637,
      "step": 13400
    },
    {
      "epoch": 0.17576742600990902,
      "grad_norm": 0.437671035528183,
      "learning_rate": 0.00027903521927517176,
      "loss": 3.477,
      "step": 13410
    },
    {
      "epoch": 0.1758984979159567,
      "grad_norm": 0.5254935622215271,
      "learning_rate": 0.0002790035059686179,
      "loss": 3.4375,
      "step": 13420
    },
    {
      "epoch": 0.17602956982200435,
      "grad_norm": 0.503588855266571,
      "learning_rate": 0.00027897177049890836,
      "loss": 3.4484,
      "step": 13430
    },
    {
      "epoch": 0.17616064172805201,
      "grad_norm": 0.46020036935806274,
      "learning_rate": 0.0002789400128714953,
      "loss": 3.4297,
      "step": 13440
    },
    {
      "epoch": 0.17629171363409968,
      "grad_norm": 0.5382055640220642,
      "learning_rate": 0.00027890823309183486,
      "loss": 3.4048,
      "step": 13450
    },
    {
      "epoch": 0.17642278554014731,
      "grad_norm": 0.5180526375770569,
      "learning_rate": 0.0002788764311653868,
      "loss": 3.4538,
      "step": 13460
    },
    {
      "epoch": 0.17655385744619498,
      "grad_norm": 0.5745498538017273,
      "learning_rate": 0.00027884460709761484,
      "loss": 3.4795,
      "step": 13470
    },
    {
      "epoch": 0.17668492935224264,
      "grad_norm": 0.5094558000564575,
      "learning_rate": 0.00027881276089398646,
      "loss": 3.4626,
      "step": 13480
    },
    {
      "epoch": 0.1768160012582903,
      "grad_norm": 0.5203001499176025,
      "learning_rate": 0.00027878089255997283,
      "loss": 3.4214,
      "step": 13490
    },
    {
      "epoch": 0.17694707316433794,
      "grad_norm": 0.4974258542060852,
      "learning_rate": 0.0002787490021010491,
      "loss": 3.4212,
      "step": 13500
    },
    {
      "epoch": 0.17694707316433794,
      "eval_accuracy": 0.3774687194525904,
      "eval_loss": 3.4224255084991455,
      "eval_runtime": 61.8687,
      "eval_samples_per_second": 32.327,
      "eval_steps_per_second": 0.517,
      "step": 13500
    },
    {
      "epoch": 0.1770781450703856,
      "grad_norm": 0.4879973828792572,
      "learning_rate": 0.0002787170895226941,
      "loss": 3.4293,
      "step": 13510
    },
    {
      "epoch": 0.17720921697643327,
      "grad_norm": 0.48761534690856934,
      "learning_rate": 0.0002786851548303905,
      "loss": 3.4443,
      "step": 13520
    },
    {
      "epoch": 0.17734028888248093,
      "grad_norm": 0.4671953320503235,
      "learning_rate": 0.00027865319802962483,
      "loss": 3.489,
      "step": 13530
    },
    {
      "epoch": 0.1774713607885286,
      "grad_norm": 0.4607062041759491,
      "learning_rate": 0.0002786212191258873,
      "loss": 3.4487,
      "step": 13540
    },
    {
      "epoch": 0.17760243269457623,
      "grad_norm": 0.45916101336479187,
      "learning_rate": 0.000278589218124672,
      "loss": 3.4239,
      "step": 13550
    },
    {
      "epoch": 0.1777335046006239,
      "grad_norm": 0.4751562774181366,
      "learning_rate": 0.0002785571950314768,
      "loss": 3.4505,
      "step": 13560
    },
    {
      "epoch": 0.17786457650667156,
      "grad_norm": 0.48061585426330566,
      "learning_rate": 0.00027852514985180323,
      "loss": 3.4294,
      "step": 13570
    },
    {
      "epoch": 0.17799564841271923,
      "grad_norm": 0.9624294638633728,
      "learning_rate": 0.00027849308259115683,
      "loss": 3.4647,
      "step": 13580
    },
    {
      "epoch": 0.17812672031876686,
      "grad_norm": 0.5418807864189148,
      "learning_rate": 0.0002784609932550469,
      "loss": 3.406,
      "step": 13590
    },
    {
      "epoch": 0.17825779222481453,
      "grad_norm": 0.5914803743362427,
      "learning_rate": 0.00027842888184898636,
      "loss": 3.4695,
      "step": 13600
    },
    {
      "epoch": 0.1783888641308622,
      "grad_norm": 0.6168108582496643,
      "learning_rate": 0.0002783967483784922,
      "loss": 3.4013,
      "step": 13610
    },
    {
      "epoch": 0.17851993603690985,
      "grad_norm": 0.4759767949581146,
      "learning_rate": 0.0002783645928490849,
      "loss": 3.43,
      "step": 13620
    },
    {
      "epoch": 0.17865100794295752,
      "grad_norm": 0.44222530722618103,
      "learning_rate": 0.00027833241526628886,
      "loss": 3.4469,
      "step": 13630
    },
    {
      "epoch": 0.17878207984900515,
      "grad_norm": 0.5154469013214111,
      "learning_rate": 0.0002783002156356324,
      "loss": 3.4622,
      "step": 13640
    },
    {
      "epoch": 0.17891315175505282,
      "grad_norm": 0.4683372974395752,
      "learning_rate": 0.00027826799396264733,
      "loss": 3.4046,
      "step": 13650
    },
    {
      "epoch": 0.17904422366110048,
      "grad_norm": 0.4847538471221924,
      "learning_rate": 0.00027823575025286964,
      "loss": 3.4834,
      "step": 13660
    },
    {
      "epoch": 0.17917529556714815,
      "grad_norm": 0.4713755249977112,
      "learning_rate": 0.00027820348451183867,
      "loss": 3.458,
      "step": 13670
    },
    {
      "epoch": 0.17930636747319578,
      "grad_norm": 0.5047947764396667,
      "learning_rate": 0.00027817119674509793,
      "loss": 3.412,
      "step": 13680
    },
    {
      "epoch": 0.17943743937924345,
      "grad_norm": 0.4931299388408661,
      "learning_rate": 0.00027813888695819443,
      "loss": 3.4228,
      "step": 13690
    },
    {
      "epoch": 0.1795685112852911,
      "grad_norm": 0.47706812620162964,
      "learning_rate": 0.00027810655515667917,
      "loss": 3.4156,
      "step": 13700
    },
    {
      "epoch": 0.17969958319133877,
      "grad_norm": 0.5775094032287598,
      "learning_rate": 0.00027807420134610673,
      "loss": 3.3819,
      "step": 13710
    },
    {
      "epoch": 0.17983065509738644,
      "grad_norm": 0.4836021363735199,
      "learning_rate": 0.0002780418255320357,
      "loss": 3.4348,
      "step": 13720
    },
    {
      "epoch": 0.17996172700343407,
      "grad_norm": 0.4540124833583832,
      "learning_rate": 0.0002780094277200283,
      "loss": 3.4552,
      "step": 13730
    },
    {
      "epoch": 0.18009279890948174,
      "grad_norm": 0.45619773864746094,
      "learning_rate": 0.00027797700791565056,
      "loss": 3.4309,
      "step": 13740
    },
    {
      "epoch": 0.1802238708155294,
      "grad_norm": 0.5055285692214966,
      "learning_rate": 0.00027794456612447225,
      "loss": 3.423,
      "step": 13750
    },
    {
      "epoch": 0.18035494272157707,
      "grad_norm": 0.4827156662940979,
      "learning_rate": 0.000277912102352067,
      "loss": 3.4032,
      "step": 13760
    },
    {
      "epoch": 0.1804860146276247,
      "grad_norm": 0.5969164967536926,
      "learning_rate": 0.0002778796166040122,
      "loss": 3.4777,
      "step": 13770
    },
    {
      "epoch": 0.18061708653367237,
      "grad_norm": 0.5232666730880737,
      "learning_rate": 0.00027784710888588887,
      "loss": 3.5064,
      "step": 13780
    },
    {
      "epoch": 0.18074815843972003,
      "grad_norm": 0.45465904474258423,
      "learning_rate": 0.00027781457920328205,
      "loss": 3.3831,
      "step": 13790
    },
    {
      "epoch": 0.1808792303457677,
      "grad_norm": 0.47327592968940735,
      "learning_rate": 0.0002777820275617804,
      "loss": 3.3936,
      "step": 13800
    },
    {
      "epoch": 0.18101030225181536,
      "grad_norm": 0.5152847170829773,
      "learning_rate": 0.00027774945396697636,
      "loss": 3.4698,
      "step": 13810
    },
    {
      "epoch": 0.181141374157863,
      "grad_norm": 0.646497368812561,
      "learning_rate": 0.0002777168584244662,
      "loss": 3.539,
      "step": 13820
    },
    {
      "epoch": 0.18127244606391066,
      "grad_norm": 0.5064975023269653,
      "learning_rate": 0.0002776842409398499,
      "loss": 3.4444,
      "step": 13830
    },
    {
      "epoch": 0.18140351796995832,
      "grad_norm": 0.45902004837989807,
      "learning_rate": 0.00027765160151873116,
      "loss": 3.4342,
      "step": 13840
    },
    {
      "epoch": 0.181534589876006,
      "grad_norm": 0.507665753364563,
      "learning_rate": 0.00027761894016671763,
      "loss": 3.4519,
      "step": 13850
    },
    {
      "epoch": 0.18166566178205362,
      "grad_norm": 0.49543237686157227,
      "learning_rate": 0.00027758625688942056,
      "loss": 3.4495,
      "step": 13860
    },
    {
      "epoch": 0.1817967336881013,
      "grad_norm": 0.4499450922012329,
      "learning_rate": 0.0002775535516924551,
      "loss": 3.4778,
      "step": 13870
    },
    {
      "epoch": 0.18192780559414895,
      "grad_norm": 0.49522531032562256,
      "learning_rate": 0.00027752082458144,
      "loss": 3.4299,
      "step": 13880
    },
    {
      "epoch": 0.18205887750019661,
      "grad_norm": 0.49619540572166443,
      "learning_rate": 0.0002774880755619979,
      "loss": 3.4561,
      "step": 13890
    },
    {
      "epoch": 0.18218994940624428,
      "grad_norm": 0.4431461691856384,
      "learning_rate": 0.00027745530463975523,
      "loss": 3.4401,
      "step": 13900
    },
    {
      "epoch": 0.18232102131229191,
      "grad_norm": 0.4803423285484314,
      "learning_rate": 0.000277422511820342,
      "loss": 3.372,
      "step": 13910
    },
    {
      "epoch": 0.18245209321833958,
      "grad_norm": 0.5019609928131104,
      "learning_rate": 0.00027738969710939225,
      "loss": 3.4153,
      "step": 13920
    },
    {
      "epoch": 0.18258316512438724,
      "grad_norm": 0.5056695342063904,
      "learning_rate": 0.0002773568605125435,
      "loss": 3.4941,
      "step": 13930
    },
    {
      "epoch": 0.1827142370304349,
      "grad_norm": 0.5480809807777405,
      "learning_rate": 0.00027732400203543727,
      "loss": 3.4246,
      "step": 13940
    },
    {
      "epoch": 0.18284530893648254,
      "grad_norm": 0.4850194454193115,
      "learning_rate": 0.0002772911216837187,
      "loss": 3.4142,
      "step": 13950
    },
    {
      "epoch": 0.1829763808425302,
      "grad_norm": 0.4969313442707062,
      "learning_rate": 0.0002772582194630367,
      "loss": 3.4911,
      "step": 13960
    },
    {
      "epoch": 0.18310745274857787,
      "grad_norm": 0.5064730048179626,
      "learning_rate": 0.00027722529537904406,
      "loss": 3.4184,
      "step": 13970
    },
    {
      "epoch": 0.18323852465462553,
      "grad_norm": 0.4935951232910156,
      "learning_rate": 0.0002771923494373971,
      "loss": 3.5414,
      "step": 13980
    },
    {
      "epoch": 0.1833695965606732,
      "grad_norm": 0.5500324368476868,
      "learning_rate": 0.00027715938164375606,
      "loss": 3.4907,
      "step": 13990
    },
    {
      "epoch": 0.18350066846672083,
      "grad_norm": 0.444823682308197,
      "learning_rate": 0.00027712639200378494,
      "loss": 3.4353,
      "step": 14000
    },
    {
      "epoch": 0.18350066846672083,
      "eval_accuracy": 0.3776867057673509,
      "eval_loss": 3.41703462600708,
      "eval_runtime": 62.1723,
      "eval_samples_per_second": 32.169,
      "eval_steps_per_second": 0.515,
      "step": 14000
    },
    {
      "epoch": 0.1836317403727685,
      "grad_norm": 0.4800029695034027,
      "learning_rate": 0.0002770933805231514,
      "loss": 3.4663,
      "step": 14010
    },
    {
      "epoch": 0.18376281227881616,
      "grad_norm": 0.5142637491226196,
      "learning_rate": 0.000277060347207527,
      "loss": 3.3631,
      "step": 14020
    },
    {
      "epoch": 0.18389388418486383,
      "grad_norm": 0.5115611553192139,
      "learning_rate": 0.0002770272920625868,
      "loss": 3.4308,
      "step": 14030
    },
    {
      "epoch": 0.18402495609091146,
      "grad_norm": 0.4980021119117737,
      "learning_rate": 0.00027699421509400985,
      "loss": 3.4255,
      "step": 14040
    },
    {
      "epoch": 0.18415602799695913,
      "grad_norm": 0.45525217056274414,
      "learning_rate": 0.0002769611163074789,
      "loss": 3.3977,
      "step": 14050
    },
    {
      "epoch": 0.1842870999030068,
      "grad_norm": 0.4756547808647156,
      "learning_rate": 0.00027692799570868033,
      "loss": 3.4332,
      "step": 14060
    },
    {
      "epoch": 0.18441817180905445,
      "grad_norm": 0.4435168504714966,
      "learning_rate": 0.00027689485330330445,
      "loss": 3.4408,
      "step": 14070
    },
    {
      "epoch": 0.18454924371510212,
      "grad_norm": 0.49747538566589355,
      "learning_rate": 0.0002768616890970451,
      "loss": 3.402,
      "step": 14080
    },
    {
      "epoch": 0.18468031562114975,
      "grad_norm": 0.4744519889354706,
      "learning_rate": 0.0002768285030956001,
      "loss": 3.4426,
      "step": 14090
    },
    {
      "epoch": 0.18481138752719742,
      "grad_norm": 0.7319813370704651,
      "learning_rate": 0.0002767952953046707,
      "loss": 3.4486,
      "step": 14100
    },
    {
      "epoch": 0.18494245943324508,
      "grad_norm": 0.4884132444858551,
      "learning_rate": 0.0002767620657299624,
      "loss": 3.3838,
      "step": 14110
    },
    {
      "epoch": 0.18507353133929275,
      "grad_norm": 0.47535035014152527,
      "learning_rate": 0.00027672881437718386,
      "loss": 3.3997,
      "step": 14120
    },
    {
      "epoch": 0.18520460324534038,
      "grad_norm": 0.4394412040710449,
      "learning_rate": 0.0002766955412520479,
      "loss": 3.4188,
      "step": 14130
    },
    {
      "epoch": 0.18533567515138805,
      "grad_norm": 0.5982511639595032,
      "learning_rate": 0.00027666224636027084,
      "loss": 3.4885,
      "step": 14140
    },
    {
      "epoch": 0.1854667470574357,
      "grad_norm": 0.5425320267677307,
      "learning_rate": 0.00027662892970757287,
      "loss": 3.4318,
      "step": 14150
    },
    {
      "epoch": 0.18559781896348337,
      "grad_norm": 0.5134817957878113,
      "learning_rate": 0.0002765955912996779,
      "loss": 3.4038,
      "step": 14160
    },
    {
      "epoch": 0.18572889086953104,
      "grad_norm": 0.5584385991096497,
      "learning_rate": 0.0002765622311423136,
      "loss": 3.4026,
      "step": 14170
    },
    {
      "epoch": 0.18585996277557867,
      "grad_norm": 0.49094322323799133,
      "learning_rate": 0.0002765288492412112,
      "loss": 3.4532,
      "step": 14180
    },
    {
      "epoch": 0.18599103468162634,
      "grad_norm": 0.4672970473766327,
      "learning_rate": 0.00027649544560210597,
      "loss": 3.4321,
      "step": 14190
    },
    {
      "epoch": 0.186122106587674,
      "grad_norm": 0.4445798099040985,
      "learning_rate": 0.0002764620202307366,
      "loss": 3.4267,
      "step": 14200
    },
    {
      "epoch": 0.18625317849372167,
      "grad_norm": 0.44376811385154724,
      "learning_rate": 0.00027642857313284577,
      "loss": 3.4634,
      "step": 14210
    },
    {
      "epoch": 0.1863842503997693,
      "grad_norm": 0.5125358700752258,
      "learning_rate": 0.0002763951043141797,
      "loss": 3.3695,
      "step": 14220
    },
    {
      "epoch": 0.18651532230581697,
      "grad_norm": 0.4908403158187866,
      "learning_rate": 0.0002763616137804885,
      "loss": 3.4586,
      "step": 14230
    },
    {
      "epoch": 0.18664639421186463,
      "grad_norm": 0.5442969799041748,
      "learning_rate": 0.0002763281015375259,
      "loss": 3.4225,
      "step": 14240
    },
    {
      "epoch": 0.1867774661179123,
      "grad_norm": 0.5176172852516174,
      "learning_rate": 0.00027629456759104936,
      "loss": 3.4067,
      "step": 14250
    },
    {
      "epoch": 0.18690853802395996,
      "grad_norm": 0.4638959765434265,
      "learning_rate": 0.0002762610119468202,
      "loss": 3.3803,
      "step": 14260
    },
    {
      "epoch": 0.1870396099300076,
      "grad_norm": 0.4630998373031616,
      "learning_rate": 0.00027622743461060324,
      "loss": 3.4628,
      "step": 14270
    },
    {
      "epoch": 0.18717068183605526,
      "grad_norm": 0.4978286623954773,
      "learning_rate": 0.0002761938355881672,
      "loss": 3.4297,
      "step": 14280
    },
    {
      "epoch": 0.18730175374210292,
      "grad_norm": 0.44554421305656433,
      "learning_rate": 0.0002761602148852846,
      "loss": 3.4325,
      "step": 14290
    },
    {
      "epoch": 0.18743282564815059,
      "grad_norm": 0.48802322149276733,
      "learning_rate": 0.00027612657250773145,
      "loss": 3.414,
      "step": 14300
    },
    {
      "epoch": 0.18756389755419822,
      "grad_norm": 0.48546743392944336,
      "learning_rate": 0.0002760929084612876,
      "loss": 3.4122,
      "step": 14310
    },
    {
      "epoch": 0.1876949694602459,
      "grad_norm": 0.5193252563476562,
      "learning_rate": 0.0002760592227517367,
      "loss": 3.4235,
      "step": 14320
    },
    {
      "epoch": 0.18782604136629355,
      "grad_norm": 0.44588741660118103,
      "learning_rate": 0.00027602551538486597,
      "loss": 3.4279,
      "step": 14330
    },
    {
      "epoch": 0.18795711327234121,
      "grad_norm": 0.45645618438720703,
      "learning_rate": 0.0002759917863664665,
      "loss": 3.417,
      "step": 14340
    },
    {
      "epoch": 0.18808818517838885,
      "grad_norm": 0.48375260829925537,
      "learning_rate": 0.000275958035702333,
      "loss": 3.4272,
      "step": 14350
    },
    {
      "epoch": 0.18821925708443651,
      "grad_norm": 0.4768548607826233,
      "learning_rate": 0.00027592426339826385,
      "loss": 3.406,
      "step": 14360
    },
    {
      "epoch": 0.18835032899048418,
      "grad_norm": 0.44193676114082336,
      "learning_rate": 0.00027589046946006136,
      "loss": 3.4009,
      "step": 14370
    },
    {
      "epoch": 0.18848140089653184,
      "grad_norm": 0.46580153703689575,
      "learning_rate": 0.00027585665389353133,
      "loss": 3.3828,
      "step": 14380
    },
    {
      "epoch": 0.1886124728025795,
      "grad_norm": 0.49439114332199097,
      "learning_rate": 0.00027582281670448334,
      "loss": 3.4195,
      "step": 14390
    },
    {
      "epoch": 0.18874354470862714,
      "grad_norm": 0.4502621591091156,
      "learning_rate": 0.00027578895789873083,
      "loss": 3.3898,
      "step": 14400
    },
    {
      "epoch": 0.1888746166146748,
      "grad_norm": 0.4472695589065552,
      "learning_rate": 0.00027575507748209076,
      "loss": 3.3686,
      "step": 14410
    },
    {
      "epoch": 0.18900568852072247,
      "grad_norm": 0.5233445167541504,
      "learning_rate": 0.0002757211754603839,
      "loss": 3.408,
      "step": 14420
    },
    {
      "epoch": 0.18913676042677013,
      "grad_norm": 0.48751455545425415,
      "learning_rate": 0.00027568725183943466,
      "loss": 3.3935,
      "step": 14430
    },
    {
      "epoch": 0.18926783233281777,
      "grad_norm": 0.499214231967926,
      "learning_rate": 0.00027565330662507125,
      "loss": 3.4041,
      "step": 14440
    },
    {
      "epoch": 0.18939890423886543,
      "grad_norm": 0.4653128683567047,
      "learning_rate": 0.00027561933982312554,
      "loss": 3.4231,
      "step": 14450
    },
    {
      "epoch": 0.1895299761449131,
      "grad_norm": 0.46365872025489807,
      "learning_rate": 0.0002755853514394332,
      "loss": 3.3864,
      "step": 14460
    },
    {
      "epoch": 0.18966104805096076,
      "grad_norm": 0.4903241693973541,
      "learning_rate": 0.00027555134147983336,
      "loss": 3.4113,
      "step": 14470
    },
    {
      "epoch": 0.18979211995700843,
      "grad_norm": 0.5719674825668335,
      "learning_rate": 0.0002755173099501691,
      "loss": 3.4458,
      "step": 14480
    },
    {
      "epoch": 0.18992319186305606,
      "grad_norm": 0.4818357229232788,
      "learning_rate": 0.00027548325685628725,
      "loss": 3.4136,
      "step": 14490
    },
    {
      "epoch": 0.19005426376910373,
      "grad_norm": 0.6275627017021179,
      "learning_rate": 0.0002754491822040381,
      "loss": 3.4187,
      "step": 14500
    },
    {
      "epoch": 0.19005426376910373,
      "eval_accuracy": 0.37800293255131967,
      "eval_loss": 3.417149305343628,
      "eval_runtime": 57.467,
      "eval_samples_per_second": 34.803,
      "eval_steps_per_second": 0.557,
      "step": 14500
    },
    {
      "epoch": 0.1901853356751514,
      "grad_norm": 0.4966403841972351,
      "learning_rate": 0.0002754150859992757,
      "loss": 3.4032,
      "step": 14510
    },
    {
      "epoch": 0.19031640758119905,
      "grad_norm": 0.6149167418479919,
      "learning_rate": 0.00027538096824785807,
      "loss": 3.4258,
      "step": 14520
    },
    {
      "epoch": 0.1904474794872467,
      "grad_norm": 0.4607338011264801,
      "learning_rate": 0.0002753468289556466,
      "loss": 3.4159,
      "step": 14530
    },
    {
      "epoch": 0.19057855139329435,
      "grad_norm": 0.4628077447414398,
      "learning_rate": 0.00027531266812850656,
      "loss": 3.4555,
      "step": 14540
    },
    {
      "epoch": 0.19070962329934202,
      "grad_norm": 0.4515349566936493,
      "learning_rate": 0.0002752784857723068,
      "loss": 3.4053,
      "step": 14550
    },
    {
      "epoch": 0.19084069520538968,
      "grad_norm": 0.46982747316360474,
      "learning_rate": 0.0002752442818929201,
      "loss": 3.4249,
      "step": 14560
    },
    {
      "epoch": 0.19097176711143735,
      "grad_norm": 0.4639160931110382,
      "learning_rate": 0.00027521005649622254,
      "loss": 3.3928,
      "step": 14570
    },
    {
      "epoch": 0.19110283901748498,
      "grad_norm": 0.5255354046821594,
      "learning_rate": 0.0002751758095880944,
      "loss": 3.358,
      "step": 14580
    },
    {
      "epoch": 0.19123391092353265,
      "grad_norm": 0.5032148361206055,
      "learning_rate": 0.00027514154117441915,
      "loss": 3.4195,
      "step": 14590
    },
    {
      "epoch": 0.1913649828295803,
      "grad_norm": 0.507500171661377,
      "learning_rate": 0.0002751072512610844,
      "loss": 3.436,
      "step": 14600
    },
    {
      "epoch": 0.19149605473562797,
      "grad_norm": 0.48513293266296387,
      "learning_rate": 0.0002750729398539811,
      "loss": 3.4203,
      "step": 14610
    },
    {
      "epoch": 0.1916271266416756,
      "grad_norm": 0.4644482135772705,
      "learning_rate": 0.0002750386069590042,
      "loss": 3.4048,
      "step": 14620
    },
    {
      "epoch": 0.19175819854772327,
      "grad_norm": 0.4690842628479004,
      "learning_rate": 0.000275004252582052,
      "loss": 3.4494,
      "step": 14630
    },
    {
      "epoch": 0.19188927045377094,
      "grad_norm": 0.4733312427997589,
      "learning_rate": 0.0002749698767290268,
      "loss": 3.4064,
      "step": 14640
    },
    {
      "epoch": 0.1920203423598186,
      "grad_norm": 0.4718853235244751,
      "learning_rate": 0.00027493547940583444,
      "loss": 3.4082,
      "step": 14650
    },
    {
      "epoch": 0.19215141426586627,
      "grad_norm": 0.507323145866394,
      "learning_rate": 0.0002749010606183845,
      "loss": 3.4612,
      "step": 14660
    },
    {
      "epoch": 0.1922824861719139,
      "grad_norm": 0.4553130269050598,
      "learning_rate": 0.0002748666203725902,
      "loss": 3.4052,
      "step": 14670
    },
    {
      "epoch": 0.19241355807796157,
      "grad_norm": 0.9733585715293884,
      "learning_rate": 0.00027483215867436843,
      "loss": 3.4629,
      "step": 14680
    },
    {
      "epoch": 0.19254462998400923,
      "grad_norm": 0.4549850821495056,
      "learning_rate": 0.00027479767552963984,
      "loss": 3.4402,
      "step": 14690
    },
    {
      "epoch": 0.1926757018900569,
      "grad_norm": 0.5259763598442078,
      "learning_rate": 0.00027476317094432865,
      "loss": 3.4031,
      "step": 14700
    },
    {
      "epoch": 0.19280677379610453,
      "grad_norm": 0.4788714647293091,
      "learning_rate": 0.000274728644924363,
      "loss": 3.414,
      "step": 14710
    },
    {
      "epoch": 0.1929378457021522,
      "grad_norm": 0.4303590655326843,
      "learning_rate": 0.0002746940974756745,
      "loss": 3.4145,
      "step": 14720
    },
    {
      "epoch": 0.19306891760819986,
      "grad_norm": 0.4757668673992157,
      "learning_rate": 0.0002746595286041984,
      "loss": 3.3729,
      "step": 14730
    },
    {
      "epoch": 0.19319998951424752,
      "grad_norm": 0.46910107135772705,
      "learning_rate": 0.00027462493831587384,
      "loss": 3.3951,
      "step": 14740
    },
    {
      "epoch": 0.19333106142029519,
      "grad_norm": 0.5050699710845947,
      "learning_rate": 0.00027459032661664347,
      "loss": 3.3661,
      "step": 14750
    },
    {
      "epoch": 0.19346213332634282,
      "grad_norm": 0.47570332884788513,
      "learning_rate": 0.0002745556935124537,
      "loss": 3.3687,
      "step": 14760
    },
    {
      "epoch": 0.1935932052323905,
      "grad_norm": 0.49932152032852173,
      "learning_rate": 0.0002745210390092545,
      "loss": 3.4041,
      "step": 14770
    },
    {
      "epoch": 0.19372427713843815,
      "grad_norm": 0.5171705484390259,
      "learning_rate": 0.00027448636311299974,
      "loss": 3.4008,
      "step": 14780
    },
    {
      "epoch": 0.19385534904448581,
      "grad_norm": 0.45184949040412903,
      "learning_rate": 0.00027445166582964683,
      "loss": 3.4104,
      "step": 14790
    },
    {
      "epoch": 0.19398642095053345,
      "grad_norm": 0.5198200941085815,
      "learning_rate": 0.00027441694716515673,
      "loss": 3.4581,
      "step": 14800
    },
    {
      "epoch": 0.19411749285658111,
      "grad_norm": 0.49282753467559814,
      "learning_rate": 0.0002743822071254943,
      "loss": 3.3992,
      "step": 14810
    },
    {
      "epoch": 0.19424856476262878,
      "grad_norm": 0.5191429257392883,
      "learning_rate": 0.000274347445716628,
      "loss": 3.4264,
      "step": 14820
    },
    {
      "epoch": 0.19437963666867644,
      "grad_norm": 0.5468852519989014,
      "learning_rate": 0.0002743126629445298,
      "loss": 3.3845,
      "step": 14830
    },
    {
      "epoch": 0.1945107085747241,
      "grad_norm": 0.5020343661308289,
      "learning_rate": 0.0002742778588151756,
      "loss": 3.3852,
      "step": 14840
    },
    {
      "epoch": 0.19464178048077174,
      "grad_norm": 0.4812702238559723,
      "learning_rate": 0.0002742430333345449,
      "loss": 3.4496,
      "step": 14850
    },
    {
      "epoch": 0.1947728523868194,
      "grad_norm": 0.48560580611228943,
      "learning_rate": 0.00027420818650862063,
      "loss": 3.3885,
      "step": 14860
    },
    {
      "epoch": 0.19490392429286707,
      "grad_norm": 0.5241788029670715,
      "learning_rate": 0.0002741733183433897,
      "loss": 3.379,
      "step": 14870
    },
    {
      "epoch": 0.19503499619891473,
      "grad_norm": 0.5088839530944824,
      "learning_rate": 0.00027413842884484245,
      "loss": 3.3852,
      "step": 14880
    },
    {
      "epoch": 0.19516606810496237,
      "grad_norm": 0.5261641144752502,
      "learning_rate": 0.00027410351801897314,
      "loss": 3.4436,
      "step": 14890
    },
    {
      "epoch": 0.19529714001101003,
      "grad_norm": 0.4630206525325775,
      "learning_rate": 0.00027406858587177945,
      "loss": 3.4211,
      "step": 14900
    },
    {
      "epoch": 0.1954282119170577,
      "grad_norm": 0.46623384952545166,
      "learning_rate": 0.00027403363240926284,
      "loss": 3.3959,
      "step": 14910
    },
    {
      "epoch": 0.19555928382310536,
      "grad_norm": 0.5889549255371094,
      "learning_rate": 0.00027399865763742836,
      "loss": 3.411,
      "step": 14920
    },
    {
      "epoch": 0.19569035572915303,
      "grad_norm": 0.551916778087616,
      "learning_rate": 0.00027396366156228483,
      "loss": 3.4219,
      "step": 14930
    },
    {
      "epoch": 0.19582142763520066,
      "grad_norm": 0.47522786259651184,
      "learning_rate": 0.0002739286441898447,
      "loss": 3.4157,
      "step": 14940
    },
    {
      "epoch": 0.19595249954124833,
      "grad_norm": 0.5226032137870789,
      "learning_rate": 0.00027389360552612397,
      "loss": 3.3909,
      "step": 14950
    },
    {
      "epoch": 0.196083571447296,
      "grad_norm": 0.5070422291755676,
      "learning_rate": 0.0002738585455771424,
      "loss": 3.412,
      "step": 14960
    },
    {
      "epoch": 0.19621464335334365,
      "grad_norm": 0.5165011286735535,
      "learning_rate": 0.0002738234643489235,
      "loss": 3.3884,
      "step": 14970
    },
    {
      "epoch": 0.1963457152593913,
      "grad_norm": 0.8619230389595032,
      "learning_rate": 0.0002737883618474941,
      "loss": 3.4324,
      "step": 14980
    },
    {
      "epoch": 0.19647678716543895,
      "grad_norm": 0.46162718534469604,
      "learning_rate": 0.0002737532380788851,
      "loss": 3.3963,
      "step": 14990
    },
    {
      "epoch": 0.19660785907148662,
      "grad_norm": 0.5595199465751648,
      "learning_rate": 0.0002737180930491308,
      "loss": 3.3533,
      "step": 15000
    },
    {
      "epoch": 0.19660785907148662,
      "eval_accuracy": 0.3786510263929619,
      "eval_loss": 3.4083900451660156,
      "eval_runtime": 58.7258,
      "eval_samples_per_second": 34.057,
      "eval_steps_per_second": 0.545,
      "step": 15000
    },
    {
      "epoch": 0.19673893097753428,
      "grad_norm": 0.5293156504631042,
      "learning_rate": 0.00027368292676426914,
      "loss": 3.3765,
      "step": 15010
    },
    {
      "epoch": 0.19687000288358195,
      "grad_norm": 0.5147075653076172,
      "learning_rate": 0.0002736477392303419,
      "loss": 3.4125,
      "step": 15020
    },
    {
      "epoch": 0.19700107478962958,
      "grad_norm": 0.45647475123405457,
      "learning_rate": 0.0002736125304533943,
      "loss": 3.4018,
      "step": 15030
    },
    {
      "epoch": 0.19713214669567725,
      "grad_norm": 0.5064095854759216,
      "learning_rate": 0.0002735773004394754,
      "loss": 3.4241,
      "step": 15040
    },
    {
      "epoch": 0.1972632186017249,
      "grad_norm": 0.4744664132595062,
      "learning_rate": 0.0002735420491946377,
      "loss": 3.4255,
      "step": 15050
    },
    {
      "epoch": 0.19739429050777257,
      "grad_norm": 0.4954212009906769,
      "learning_rate": 0.0002735067767249376,
      "loss": 3.4187,
      "step": 15060
    },
    {
      "epoch": 0.1975253624138202,
      "grad_norm": 0.5786910057067871,
      "learning_rate": 0.0002734714830364348,
      "loss": 3.4156,
      "step": 15070
    },
    {
      "epoch": 0.19765643431986787,
      "grad_norm": 0.5505225658416748,
      "learning_rate": 0.00027343616813519305,
      "loss": 3.4252,
      "step": 15080
    },
    {
      "epoch": 0.19778750622591554,
      "grad_norm": 0.4847007095813751,
      "learning_rate": 0.0002734008320272795,
      "loss": 3.3617,
      "step": 15090
    },
    {
      "epoch": 0.1979185781319632,
      "grad_norm": 0.5412439107894897,
      "learning_rate": 0.0002733654747187649,
      "loss": 3.4095,
      "step": 15100
    },
    {
      "epoch": 0.19804965003801087,
      "grad_norm": 0.5054135918617249,
      "learning_rate": 0.0002733300962157238,
      "loss": 3.3772,
      "step": 15110
    },
    {
      "epoch": 0.1981807219440585,
      "grad_norm": 0.4472280740737915,
      "learning_rate": 0.00027329469652423444,
      "loss": 3.4057,
      "step": 15120
    },
    {
      "epoch": 0.19831179385010617,
      "grad_norm": 0.46590396761894226,
      "learning_rate": 0.00027325927565037837,
      "loss": 3.407,
      "step": 15130
    },
    {
      "epoch": 0.19844286575615383,
      "grad_norm": 0.8282036781311035,
      "learning_rate": 0.00027322383360024104,
      "loss": 3.4397,
      "step": 15140
    },
    {
      "epoch": 0.1985739376622015,
      "grad_norm": 0.4920446276664734,
      "learning_rate": 0.00027318837037991164,
      "loss": 3.4628,
      "step": 15150
    },
    {
      "epoch": 0.19870500956824913,
      "grad_norm": 0.49810758233070374,
      "learning_rate": 0.0002731528859954827,
      "loss": 3.3805,
      "step": 15160
    },
    {
      "epoch": 0.1988360814742968,
      "grad_norm": 0.4850003123283386,
      "learning_rate": 0.0002731173804530506,
      "loss": 3.3887,
      "step": 15170
    },
    {
      "epoch": 0.19896715338034446,
      "grad_norm": 0.5460561513900757,
      "learning_rate": 0.0002730818537587152,
      "loss": 3.4237,
      "step": 15180
    },
    {
      "epoch": 0.19909822528639212,
      "grad_norm": 0.5468739867210388,
      "learning_rate": 0.0002730463059185803,
      "loss": 3.4224,
      "step": 15190
    },
    {
      "epoch": 0.19922929719243979,
      "grad_norm": 0.46255752444267273,
      "learning_rate": 0.0002730107369387528,
      "loss": 3.3904,
      "step": 15200
    },
    {
      "epoch": 0.19936036909848742,
      "grad_norm": 0.4776484966278076,
      "learning_rate": 0.00027297514682534383,
      "loss": 3.4195,
      "step": 15210
    },
    {
      "epoch": 0.19949144100453509,
      "grad_norm": 0.49248263239860535,
      "learning_rate": 0.0002729395355844677,
      "loss": 3.3942,
      "step": 15220
    },
    {
      "epoch": 0.19962251291058275,
      "grad_norm": 0.4770093262195587,
      "learning_rate": 0.0002729039032222426,
      "loss": 3.4095,
      "step": 15230
    },
    {
      "epoch": 0.19975358481663041,
      "grad_norm": 0.47482526302337646,
      "learning_rate": 0.0002728682497447903,
      "loss": 3.4062,
      "step": 15240
    },
    {
      "epoch": 0.19988465672267805,
      "grad_norm": 0.5152221918106079,
      "learning_rate": 0.000272832575158236,
      "loss": 3.444,
      "step": 15250
    },
    {
      "epoch": 0.20001572862872571,
      "grad_norm": 0.4705754518508911,
      "learning_rate": 0.00027279687946870886,
      "loss": 3.4385,
      "step": 15260
    },
    {
      "epoch": 0.20014680053477338,
      "grad_norm": 0.477688729763031,
      "learning_rate": 0.00027276116268234145,
      "loss": 3.4121,
      "step": 15270
    },
    {
      "epoch": 0.20027787244082104,
      "grad_norm": 0.4842638671398163,
      "learning_rate": 0.00027272542480527,
      "loss": 3.3725,
      "step": 15280
    },
    {
      "epoch": 0.2004089443468687,
      "grad_norm": 0.5012391209602356,
      "learning_rate": 0.00027268966584363436,
      "loss": 3.3723,
      "step": 15290
    },
    {
      "epoch": 0.20054001625291634,
      "grad_norm": 0.5003477334976196,
      "learning_rate": 0.00027265388580357804,
      "loss": 3.3668,
      "step": 15300
    },
    {
      "epoch": 0.200671088158964,
      "grad_norm": 0.6785241365432739,
      "learning_rate": 0.0002726180846912481,
      "loss": 3.4575,
      "step": 15310
    },
    {
      "epoch": 0.20080216006501167,
      "grad_norm": 0.485784649848938,
      "learning_rate": 0.00027258226251279536,
      "loss": 3.4338,
      "step": 15320
    },
    {
      "epoch": 0.20093323197105933,
      "grad_norm": 0.5904166102409363,
      "learning_rate": 0.00027254641927437415,
      "loss": 3.3999,
      "step": 15330
    },
    {
      "epoch": 0.20106430387710697,
      "grad_norm": 0.48969829082489014,
      "learning_rate": 0.00027251055498214233,
      "loss": 3.4449,
      "step": 15340
    },
    {
      "epoch": 0.20119537578315463,
      "grad_norm": 0.49957844614982605,
      "learning_rate": 0.0002724746696422616,
      "loss": 3.353,
      "step": 15350
    },
    {
      "epoch": 0.2013264476892023,
      "grad_norm": 0.5234807133674622,
      "learning_rate": 0.0002724387632608971,
      "loss": 3.4701,
      "step": 15360
    },
    {
      "epoch": 0.20145751959524996,
      "grad_norm": 0.640742301940918,
      "learning_rate": 0.00027240283584421766,
      "loss": 3.4618,
      "step": 15370
    },
    {
      "epoch": 0.2015885915012976,
      "grad_norm": 0.5477535128593445,
      "learning_rate": 0.0002723668873983958,
      "loss": 3.4289,
      "step": 15380
    },
    {
      "epoch": 0.20171966340734526,
      "grad_norm": 0.48281335830688477,
      "learning_rate": 0.00027233091792960737,
      "loss": 3.3355,
      "step": 15390
    },
    {
      "epoch": 0.20185073531339293,
      "grad_norm": 0.5090018510818481,
      "learning_rate": 0.0002722949274440322,
      "loss": 3.42,
      "step": 15400
    },
    {
      "epoch": 0.2019818072194406,
      "grad_norm": 0.45641839504241943,
      "learning_rate": 0.00027225891594785337,
      "loss": 3.4224,
      "step": 15410
    },
    {
      "epoch": 0.20211287912548825,
      "grad_norm": 0.4841395914554596,
      "learning_rate": 0.0002722228834472579,
      "loss": 3.4309,
      "step": 15420
    },
    {
      "epoch": 0.2022439510315359,
      "grad_norm": 0.454244464635849,
      "learning_rate": 0.0002721868299484363,
      "loss": 3.404,
      "step": 15430
    },
    {
      "epoch": 0.20237502293758355,
      "grad_norm": 0.4650370478630066,
      "learning_rate": 0.0002721507554575826,
      "loss": 3.4193,
      "step": 15440
    },
    {
      "epoch": 0.20250609484363122,
      "grad_norm": 0.4992330074310303,
      "learning_rate": 0.0002721146599808945,
      "loss": 3.4089,
      "step": 15450
    },
    {
      "epoch": 0.20263716674967888,
      "grad_norm": 0.5103819370269775,
      "learning_rate": 0.0002720785435245732,
      "loss": 3.4369,
      "step": 15460
    },
    {
      "epoch": 0.20276823865572652,
      "grad_norm": 0.5182070732116699,
      "learning_rate": 0.0002720424060948238,
      "loss": 3.4335,
      "step": 15470
    },
    {
      "epoch": 0.20289931056177418,
      "grad_norm": 0.4663074314594269,
      "learning_rate": 0.00027200624769785467,
      "loss": 3.394,
      "step": 15480
    },
    {
      "epoch": 0.20303038246782185,
      "grad_norm": 0.5089011192321777,
      "learning_rate": 0.00027197006833987793,
      "loss": 3.4036,
      "step": 15490
    },
    {
      "epoch": 0.2031614543738695,
      "grad_norm": 0.4646686911582947,
      "learning_rate": 0.00027193386802710937,
      "loss": 3.3841,
      "step": 15500
    },
    {
      "epoch": 0.2031614543738695,
      "eval_accuracy": 0.3801373411534702,
      "eval_loss": 3.398216962814331,
      "eval_runtime": 58.7775,
      "eval_samples_per_second": 34.027,
      "eval_steps_per_second": 0.544,
      "step": 15500
    },
    {
      "epoch": 0.20329252627991717,
      "grad_norm": 0.48524144291877747,
      "learning_rate": 0.0002718976467657683,
      "loss": 3.3851,
      "step": 15510
    },
    {
      "epoch": 0.2034235981859648,
      "grad_norm": 0.4884251654148102,
      "learning_rate": 0.0002718614045620775,
      "loss": 3.3823,
      "step": 15520
    },
    {
      "epoch": 0.20355467009201247,
      "grad_norm": 0.5551469922065735,
      "learning_rate": 0.0002718251414222636,
      "loss": 3.3581,
      "step": 15530
    },
    {
      "epoch": 0.20368574199806014,
      "grad_norm": 0.4551747739315033,
      "learning_rate": 0.0002717888573525567,
      "loss": 3.3864,
      "step": 15540
    },
    {
      "epoch": 0.2038168139041078,
      "grad_norm": 0.5031114220619202,
      "learning_rate": 0.00027175255235919045,
      "loss": 3.4049,
      "step": 15550
    },
    {
      "epoch": 0.20394788581015544,
      "grad_norm": 0.4581761062145233,
      "learning_rate": 0.00027171622644840216,
      "loss": 3.3859,
      "step": 15560
    },
    {
      "epoch": 0.2040789577162031,
      "grad_norm": 0.7916029095649719,
      "learning_rate": 0.00027167987962643277,
      "loss": 3.4355,
      "step": 15570
    },
    {
      "epoch": 0.20421002962225077,
      "grad_norm": 0.5009720325469971,
      "learning_rate": 0.00027164351189952674,
      "loss": 3.3696,
      "step": 15580
    },
    {
      "epoch": 0.20434110152829843,
      "grad_norm": 0.5387327075004578,
      "learning_rate": 0.0002716071232739321,
      "loss": 3.4045,
      "step": 15590
    },
    {
      "epoch": 0.2044721734343461,
      "grad_norm": 0.5183151960372925,
      "learning_rate": 0.00027157071375590063,
      "loss": 3.4194,
      "step": 15600
    },
    {
      "epoch": 0.20460324534039373,
      "grad_norm": 0.47220396995544434,
      "learning_rate": 0.00027153428335168747,
      "loss": 3.3999,
      "step": 15610
    },
    {
      "epoch": 0.2047343172464414,
      "grad_norm": 0.7329874634742737,
      "learning_rate": 0.0002714978320675514,
      "loss": 3.3769,
      "step": 15620
    },
    {
      "epoch": 0.20486538915248906,
      "grad_norm": 0.5052440762519836,
      "learning_rate": 0.00027146135990975507,
      "loss": 3.395,
      "step": 15630
    },
    {
      "epoch": 0.20499646105853672,
      "grad_norm": 0.506623387336731,
      "learning_rate": 0.00027142486688456433,
      "loss": 3.4356,
      "step": 15640
    },
    {
      "epoch": 0.20512753296458436,
      "grad_norm": 0.4946392774581909,
      "learning_rate": 0.00027138835299824886,
      "loss": 3.3691,
      "step": 15650
    },
    {
      "epoch": 0.20525860487063202,
      "grad_norm": 0.6985766887664795,
      "learning_rate": 0.0002713518182570818,
      "loss": 3.4513,
      "step": 15660
    },
    {
      "epoch": 0.20538967677667969,
      "grad_norm": 0.5396666526794434,
      "learning_rate": 0.0002713152626673399,
      "loss": 3.4048,
      "step": 15670
    },
    {
      "epoch": 0.20552074868272735,
      "grad_norm": 0.49957504868507385,
      "learning_rate": 0.00027127868623530357,
      "loss": 3.4662,
      "step": 15680
    },
    {
      "epoch": 0.20565182058877501,
      "grad_norm": 0.4544884264469147,
      "learning_rate": 0.00027124208896725673,
      "loss": 3.4043,
      "step": 15690
    },
    {
      "epoch": 0.20578289249482265,
      "grad_norm": 0.4907263219356537,
      "learning_rate": 0.0002712054708694869,
      "loss": 3.404,
      "step": 15700
    },
    {
      "epoch": 0.20591396440087031,
      "grad_norm": 0.5310547351837158,
      "learning_rate": 0.0002711688319482851,
      "loss": 3.4209,
      "step": 15710
    },
    {
      "epoch": 0.20604503630691798,
      "grad_norm": 0.539485514163971,
      "learning_rate": 0.000271132172209946,
      "loss": 3.4492,
      "step": 15720
    },
    {
      "epoch": 0.20617610821296564,
      "grad_norm": 0.49285343289375305,
      "learning_rate": 0.000271095491660768,
      "loss": 3.3726,
      "step": 15730
    },
    {
      "epoch": 0.20630718011901328,
      "grad_norm": 0.5236189961433411,
      "learning_rate": 0.0002710587903070527,
      "loss": 3.4089,
      "step": 15740
    },
    {
      "epoch": 0.20643825202506094,
      "grad_norm": 0.514589250087738,
      "learning_rate": 0.00027102206815510563,
      "loss": 3.4077,
      "step": 15750
    },
    {
      "epoch": 0.2065693239311086,
      "grad_norm": 0.4693169593811035,
      "learning_rate": 0.00027098532521123577,
      "loss": 3.408,
      "step": 15760
    },
    {
      "epoch": 0.20670039583715627,
      "grad_norm": 0.49514663219451904,
      "learning_rate": 0.00027094856148175553,
      "loss": 3.3955,
      "step": 15770
    },
    {
      "epoch": 0.20683146774320393,
      "grad_norm": 0.5016627311706543,
      "learning_rate": 0.00027091177697298116,
      "loss": 3.3633,
      "step": 15780
    },
    {
      "epoch": 0.20696253964925157,
      "grad_norm": 0.46157896518707275,
      "learning_rate": 0.00027087497169123225,
      "loss": 3.4188,
      "step": 15790
    },
    {
      "epoch": 0.20709361155529923,
      "grad_norm": 0.4769839644432068,
      "learning_rate": 0.00027083814564283213,
      "loss": 3.4198,
      "step": 15800
    },
    {
      "epoch": 0.2072246834613469,
      "grad_norm": 0.5373361706733704,
      "learning_rate": 0.0002708012988341075,
      "loss": 3.4776,
      "step": 15810
    },
    {
      "epoch": 0.20735575536739456,
      "grad_norm": 0.5045587420463562,
      "learning_rate": 0.0002707644312713888,
      "loss": 3.3963,
      "step": 15820
    },
    {
      "epoch": 0.2074868272734422,
      "grad_norm": 0.4794834852218628,
      "learning_rate": 0.0002707275429610101,
      "loss": 3.4071,
      "step": 15830
    },
    {
      "epoch": 0.20761789917948986,
      "grad_norm": 0.8172100782394409,
      "learning_rate": 0.0002706906339093087,
      "loss": 3.3906,
      "step": 15840
    },
    {
      "epoch": 0.20774897108553753,
      "grad_norm": 0.4557173252105713,
      "learning_rate": 0.0002706537041226258,
      "loss": 3.3912,
      "step": 15850
    },
    {
      "epoch": 0.2078800429915852,
      "grad_norm": 0.49864286184310913,
      "learning_rate": 0.000270616753607306,
      "loss": 3.3933,
      "step": 15860
    },
    {
      "epoch": 0.20801111489763285,
      "grad_norm": 0.46710920333862305,
      "learning_rate": 0.0002705797823696975,
      "loss": 3.4217,
      "step": 15870
    },
    {
      "epoch": 0.2081421868036805,
      "grad_norm": 0.5150077939033508,
      "learning_rate": 0.00027054279041615215,
      "loss": 3.4291,
      "step": 15880
    },
    {
      "epoch": 0.20827325870972815,
      "grad_norm": 0.497726708650589,
      "learning_rate": 0.0002705057777530251,
      "loss": 3.4486,
      "step": 15890
    },
    {
      "epoch": 0.20840433061577582,
      "grad_norm": 0.7510843873023987,
      "learning_rate": 0.0002704687443866754,
      "loss": 3.4358,
      "step": 15900
    },
    {
      "epoch": 0.20853540252182348,
      "grad_norm": 0.5004276633262634,
      "learning_rate": 0.00027043169032346534,
      "loss": 3.4119,
      "step": 15910
    },
    {
      "epoch": 0.20866647442787112,
      "grad_norm": 0.5384702682495117,
      "learning_rate": 0.00027039461556976104,
      "loss": 3.4035,
      "step": 15920
    },
    {
      "epoch": 0.20879754633391878,
      "grad_norm": 0.5392207503318787,
      "learning_rate": 0.0002703575201319319,
      "loss": 3.4343,
      "step": 15930
    },
    {
      "epoch": 0.20892861823996645,
      "grad_norm": 0.45744046568870544,
      "learning_rate": 0.00027032040401635114,
      "loss": 3.4069,
      "step": 15940
    },
    {
      "epoch": 0.2090596901460141,
      "grad_norm": 0.49621301889419556,
      "learning_rate": 0.00027028326722939536,
      "loss": 3.4054,
      "step": 15950
    },
    {
      "epoch": 0.20919076205206177,
      "grad_norm": 0.7787452340126038,
      "learning_rate": 0.00027024610977744476,
      "loss": 3.4429,
      "step": 15960
    },
    {
      "epoch": 0.2093218339581094,
      "grad_norm": 0.8236386179924011,
      "learning_rate": 0.0002702089316668831,
      "loss": 3.4734,
      "step": 15970
    },
    {
      "epoch": 0.20945290586415707,
      "grad_norm": 0.5062335133552551,
      "learning_rate": 0.0002701717329040977,
      "loss": 3.3808,
      "step": 15980
    },
    {
      "epoch": 0.20958397777020474,
      "grad_norm": 0.49789005517959595,
      "learning_rate": 0.00027013451349547937,
      "loss": 3.4037,
      "step": 15990
    },
    {
      "epoch": 0.2097150496762524,
      "grad_norm": 0.4680810868740082,
      "learning_rate": 0.0002700972734474225,
      "loss": 3.3842,
      "step": 16000
    },
    {
      "epoch": 0.2097150496762524,
      "eval_accuracy": 0.3799149560117302,
      "eval_loss": 3.398092746734619,
      "eval_runtime": 57.6962,
      "eval_samples_per_second": 34.664,
      "eval_steps_per_second": 0.555,
      "step": 16000
    },
    {
      "epoch": 0.20984612158230004,
      "grad_norm": 0.4790894389152527,
      "learning_rate": 0.00027006001276632513,
      "loss": 3.3945,
      "step": 16010
    },
    {
      "epoch": 0.2099771934883477,
      "grad_norm": 0.48230820894241333,
      "learning_rate": 0.0002700227314585886,
      "loss": 3.3946,
      "step": 16020
    },
    {
      "epoch": 0.21010826539439537,
      "grad_norm": 0.5224131345748901,
      "learning_rate": 0.00026998542953061803,
      "loss": 3.4728,
      "step": 16030
    },
    {
      "epoch": 0.21023933730044303,
      "grad_norm": 0.49105599522590637,
      "learning_rate": 0.000269948106988822,
      "loss": 3.4629,
      "step": 16040
    },
    {
      "epoch": 0.2103704092064907,
      "grad_norm": 0.5059372186660767,
      "learning_rate": 0.0002699107638396126,
      "loss": 3.4044,
      "step": 16050
    },
    {
      "epoch": 0.21050148111253833,
      "grad_norm": 0.5101101994514465,
      "learning_rate": 0.0002698734000894055,
      "loss": 3.3487,
      "step": 16060
    },
    {
      "epoch": 0.210632553018586,
      "grad_norm": 0.7494931221008301,
      "learning_rate": 0.0002698360157446199,
      "loss": 3.3808,
      "step": 16070
    },
    {
      "epoch": 0.21076362492463366,
      "grad_norm": 0.4944099187850952,
      "learning_rate": 0.00026979861081167844,
      "loss": 3.4006,
      "step": 16080
    },
    {
      "epoch": 0.21089469683068132,
      "grad_norm": 0.49640953540802,
      "learning_rate": 0.00026976118529700754,
      "loss": 3.3768,
      "step": 16090
    },
    {
      "epoch": 0.21102576873672896,
      "grad_norm": 0.5246527194976807,
      "learning_rate": 0.0002697237392070369,
      "loss": 3.4529,
      "step": 16100
    },
    {
      "epoch": 0.21115684064277662,
      "grad_norm": 0.5306735038757324,
      "learning_rate": 0.0002696862725481999,
      "loss": 3.3914,
      "step": 16110
    },
    {
      "epoch": 0.21128791254882429,
      "grad_norm": 0.5840965509414673,
      "learning_rate": 0.0002696487853269334,
      "loss": 3.4386,
      "step": 16120
    },
    {
      "epoch": 0.21141898445487195,
      "grad_norm": 0.48307162523269653,
      "learning_rate": 0.00026961127754967785,
      "loss": 3.343,
      "step": 16130
    },
    {
      "epoch": 0.2115500563609196,
      "grad_norm": 0.47478798031806946,
      "learning_rate": 0.00026957374922287715,
      "loss": 3.389,
      "step": 16140
    },
    {
      "epoch": 0.21168112826696725,
      "grad_norm": 0.5520452260971069,
      "learning_rate": 0.00026953620035297874,
      "loss": 3.3588,
      "step": 16150
    },
    {
      "epoch": 0.21181220017301491,
      "grad_norm": 0.5338848829269409,
      "learning_rate": 0.0002694986309464337,
      "loss": 3.4211,
      "step": 16160
    },
    {
      "epoch": 0.21194327207906258,
      "grad_norm": 0.5740928649902344,
      "learning_rate": 0.0002694610410096965,
      "loss": 3.425,
      "step": 16170
    },
    {
      "epoch": 0.21207434398511024,
      "grad_norm": 0.4831336438655853,
      "learning_rate": 0.00026942343054922515,
      "loss": 3.434,
      "step": 16180
    },
    {
      "epoch": 0.21220541589115788,
      "grad_norm": 0.48146599531173706,
      "learning_rate": 0.0002693857995714813,
      "loss": 3.4053,
      "step": 16190
    },
    {
      "epoch": 0.21233648779720554,
      "grad_norm": 0.483006089925766,
      "learning_rate": 0.00026934814808293007,
      "loss": 3.4174,
      "step": 16200
    },
    {
      "epoch": 0.2124675597032532,
      "grad_norm": 0.4695867598056793,
      "learning_rate": 0.00026931047609004004,
      "loss": 3.4086,
      "step": 16210
    },
    {
      "epoch": 0.21259863160930087,
      "grad_norm": 0.4822619557380676,
      "learning_rate": 0.00026927278359928337,
      "loss": 3.4147,
      "step": 16220
    },
    {
      "epoch": 0.21272970351534853,
      "grad_norm": 0.4940435588359833,
      "learning_rate": 0.0002692350706171358,
      "loss": 3.3986,
      "step": 16230
    },
    {
      "epoch": 0.21286077542139617,
      "grad_norm": 0.4866895079612732,
      "learning_rate": 0.0002691973371500764,
      "loss": 3.3688,
      "step": 16240
    },
    {
      "epoch": 0.21299184732744383,
      "grad_norm": 0.4760785698890686,
      "learning_rate": 0.000269159583204588,
      "loss": 3.4021,
      "step": 16250
    },
    {
      "epoch": 0.2131229192334915,
      "grad_norm": 0.4633254110813141,
      "learning_rate": 0.0002691218087871567,
      "loss": 3.4662,
      "step": 16260
    },
    {
      "epoch": 0.21325399113953916,
      "grad_norm": 0.5595280528068542,
      "learning_rate": 0.00026908401390427243,
      "loss": 3.4803,
      "step": 16270
    },
    {
      "epoch": 0.2133850630455868,
      "grad_norm": 0.9013370871543884,
      "learning_rate": 0.00026904619856242827,
      "loss": 3.4155,
      "step": 16280
    },
    {
      "epoch": 0.21351613495163446,
      "grad_norm": 0.5218617916107178,
      "learning_rate": 0.0002690083627681211,
      "loss": 3.362,
      "step": 16290
    },
    {
      "epoch": 0.21364720685768213,
      "grad_norm": 0.48159685730934143,
      "learning_rate": 0.00026897050652785123,
      "loss": 3.3826,
      "step": 16300
    },
    {
      "epoch": 0.2137782787637298,
      "grad_norm": 0.5270137190818787,
      "learning_rate": 0.0002689326298481224,
      "loss": 3.4076,
      "step": 16310
    },
    {
      "epoch": 0.21390935066977743,
      "grad_norm": 0.5090985894203186,
      "learning_rate": 0.00026889473273544203,
      "loss": 3.4183,
      "step": 16320
    },
    {
      "epoch": 0.2140404225758251,
      "grad_norm": 0.647125244140625,
      "learning_rate": 0.0002688568151963208,
      "loss": 3.3951,
      "step": 16330
    },
    {
      "epoch": 0.21417149448187275,
      "grad_norm": 0.5214628577232361,
      "learning_rate": 0.00026881887723727315,
      "loss": 3.4031,
      "step": 16340
    },
    {
      "epoch": 0.21430256638792042,
      "grad_norm": 0.5041720271110535,
      "learning_rate": 0.00026878091886481685,
      "loss": 3.407,
      "step": 16350
    },
    {
      "epoch": 0.21443363829396808,
      "grad_norm": 0.42975032329559326,
      "learning_rate": 0.00026874294008547337,
      "loss": 3.4151,
      "step": 16360
    },
    {
      "epoch": 0.21456471020001572,
      "grad_norm": 0.46399039030075073,
      "learning_rate": 0.0002687049409057675,
      "loss": 3.3904,
      "step": 16370
    },
    {
      "epoch": 0.21469578210606338,
      "grad_norm": 0.5741137266159058,
      "learning_rate": 0.00026866692133222756,
      "loss": 3.4165,
      "step": 16380
    },
    {
      "epoch": 0.21482685401211105,
      "grad_norm": 0.499250590801239,
      "learning_rate": 0.00026862888137138547,
      "loss": 3.4787,
      "step": 16390
    },
    {
      "epoch": 0.2149579259181587,
      "grad_norm": 0.5438000559806824,
      "learning_rate": 0.0002685908210297766,
      "loss": 3.3536,
      "step": 16400
    },
    {
      "epoch": 0.21508899782420635,
      "grad_norm": 0.5131533741950989,
      "learning_rate": 0.0002685527403139398,
      "loss": 3.4516,
      "step": 16410
    },
    {
      "epoch": 0.215220069730254,
      "grad_norm": 0.4960278570652008,
      "learning_rate": 0.0002685146392304175,
      "loss": 3.3776,
      "step": 16420
    },
    {
      "epoch": 0.21535114163630167,
      "grad_norm": 0.6120615601539612,
      "learning_rate": 0.00026847651778575543,
      "loss": 3.4406,
      "step": 16430
    },
    {
      "epoch": 0.21548221354234934,
      "grad_norm": 0.47769850492477417,
      "learning_rate": 0.00026843837598650313,
      "loss": 3.4178,
      "step": 16440
    },
    {
      "epoch": 0.215613285448397,
      "grad_norm": 0.5606678128242493,
      "learning_rate": 0.00026840021383921334,
      "loss": 3.4329,
      "step": 16450
    },
    {
      "epoch": 0.21574435735444464,
      "grad_norm": 0.4822315275669098,
      "learning_rate": 0.0002683620313504425,
      "loss": 3.4115,
      "step": 16460
    },
    {
      "epoch": 0.2158754292604923,
      "grad_norm": 0.5292935371398926,
      "learning_rate": 0.0002683238285267504,
      "loss": 3.4472,
      "step": 16470
    },
    {
      "epoch": 0.21600650116653997,
      "grad_norm": 0.503404438495636,
      "learning_rate": 0.00026828560537470045,
      "loss": 3.4398,
      "step": 16480
    },
    {
      "epoch": 0.21613757307258763,
      "grad_norm": 0.481086403131485,
      "learning_rate": 0.0002682473619008595,
      "loss": 3.403,
      "step": 16490
    },
    {
      "epoch": 0.21626864497863527,
      "grad_norm": 0.5071761012077332,
      "learning_rate": 0.00026820909811179776,
      "loss": 3.3662,
      "step": 16500
    },
    {
      "epoch": 0.21626864497863527,
      "eval_accuracy": 0.38066373411534704,
      "eval_loss": 3.3885767459869385,
      "eval_runtime": 56.0998,
      "eval_samples_per_second": 35.651,
      "eval_steps_per_second": 0.57,
      "step": 16500
    },
    {
      "epoch": 0.21639971688468293,
      "grad_norm": 0.47781041264533997,
      "learning_rate": 0.00026817081401408924,
      "loss": 3.4016,
      "step": 16510
    },
    {
      "epoch": 0.2165307887907306,
      "grad_norm": 0.4860817492008209,
      "learning_rate": 0.00026813250961431106,
      "loss": 3.4665,
      "step": 16520
    },
    {
      "epoch": 0.21666186069677826,
      "grad_norm": 0.5249307155609131,
      "learning_rate": 0.0002680941849190442,
      "loss": 3.4431,
      "step": 16530
    },
    {
      "epoch": 0.21679293260282592,
      "grad_norm": 0.5495378375053406,
      "learning_rate": 0.0002680558399348728,
      "loss": 3.4441,
      "step": 16540
    },
    {
      "epoch": 0.21692400450887356,
      "grad_norm": 0.47377732396125793,
      "learning_rate": 0.00026801747466838474,
      "loss": 3.391,
      "step": 16550
    },
    {
      "epoch": 0.21705507641492122,
      "grad_norm": 0.4828034043312073,
      "learning_rate": 0.0002679790891261712,
      "loss": 3.4124,
      "step": 16560
    },
    {
      "epoch": 0.21718614832096889,
      "grad_norm": 0.5329654812812805,
      "learning_rate": 0.00026794068331482696,
      "loss": 3.3743,
      "step": 16570
    },
    {
      "epoch": 0.21731722022701655,
      "grad_norm": 0.4866692125797272,
      "learning_rate": 0.00026790225724095026,
      "loss": 3.3968,
      "step": 16580
    },
    {
      "epoch": 0.21744829213306419,
      "grad_norm": 0.4528242349624634,
      "learning_rate": 0.0002678638109111428,
      "loss": 3.397,
      "step": 16590
    },
    {
      "epoch": 0.21757936403911185,
      "grad_norm": 0.521208643913269,
      "learning_rate": 0.00026782534433200976,
      "loss": 3.3993,
      "step": 16600
    },
    {
      "epoch": 0.21771043594515951,
      "grad_norm": 0.5434349775314331,
      "learning_rate": 0.0002677868575101597,
      "loss": 3.4584,
      "step": 16610
    },
    {
      "epoch": 0.21784150785120718,
      "grad_norm": 0.49142688512802124,
      "learning_rate": 0.00026774835045220496,
      "loss": 3.3952,
      "step": 16620
    },
    {
      "epoch": 0.21797257975725484,
      "grad_norm": 0.4871850311756134,
      "learning_rate": 0.000267709823164761,
      "loss": 3.3869,
      "step": 16630
    },
    {
      "epoch": 0.21810365166330248,
      "grad_norm": 0.4870157539844513,
      "learning_rate": 0.000267671275654447,
      "loss": 3.376,
      "step": 16640
    },
    {
      "epoch": 0.21823472356935014,
      "grad_norm": 0.47191616892814636,
      "learning_rate": 0.0002676327079278855,
      "loss": 3.4133,
      "step": 16650
    },
    {
      "epoch": 0.2183657954753978,
      "grad_norm": 2.497199296951294,
      "learning_rate": 0.0002675941199917025,
      "loss": 3.3803,
      "step": 16660
    },
    {
      "epoch": 0.21849686738144547,
      "grad_norm": 0.5264095067977905,
      "learning_rate": 0.0002675555118525276,
      "loss": 3.4002,
      "step": 16670
    },
    {
      "epoch": 0.2186279392874931,
      "grad_norm": 0.4669231176376343,
      "learning_rate": 0.0002675168835169937,
      "loss": 3.3922,
      "step": 16680
    },
    {
      "epoch": 0.21875901119354077,
      "grad_norm": 0.5854026079177856,
      "learning_rate": 0.0002674782349917373,
      "loss": 3.3831,
      "step": 16690
    },
    {
      "epoch": 0.21889008309958843,
      "grad_norm": 0.5240497589111328,
      "learning_rate": 0.0002674395662833983,
      "loss": 3.3908,
      "step": 16700
    },
    {
      "epoch": 0.2190211550056361,
      "grad_norm": 0.5014669895172119,
      "learning_rate": 0.00026740087739862006,
      "loss": 3.4195,
      "step": 16710
    },
    {
      "epoch": 0.21915222691168376,
      "grad_norm": 0.4932710826396942,
      "learning_rate": 0.00026736216834404947,
      "loss": 3.4121,
      "step": 16720
    },
    {
      "epoch": 0.2192832988177314,
      "grad_norm": 0.4622075855731964,
      "learning_rate": 0.00026732343912633694,
      "loss": 3.376,
      "step": 16730
    },
    {
      "epoch": 0.21941437072377906,
      "grad_norm": 0.5236517786979675,
      "learning_rate": 0.0002672846897521361,
      "loss": 3.4383,
      "step": 16740
    },
    {
      "epoch": 0.21954544262982673,
      "grad_norm": 0.5652025938034058,
      "learning_rate": 0.0002672459202281043,
      "loss": 3.5106,
      "step": 16750
    },
    {
      "epoch": 0.2196765145358744,
      "grad_norm": 0.46889278292655945,
      "learning_rate": 0.00026720713056090225,
      "loss": 3.4028,
      "step": 16760
    },
    {
      "epoch": 0.21980758644192203,
      "grad_norm": 0.5682184100151062,
      "learning_rate": 0.00026716832075719405,
      "loss": 3.332,
      "step": 16770
    },
    {
      "epoch": 0.2199386583479697,
      "grad_norm": 0.500281810760498,
      "learning_rate": 0.00026712949082364736,
      "loss": 3.429,
      "step": 16780
    },
    {
      "epoch": 0.22006973025401735,
      "grad_norm": 0.5777886509895325,
      "learning_rate": 0.0002670906407669333,
      "loss": 3.4471,
      "step": 16790
    },
    {
      "epoch": 0.22020080216006502,
      "grad_norm": 0.5012169480323792,
      "learning_rate": 0.00026705177059372644,
      "loss": 3.3898,
      "step": 16800
    },
    {
      "epoch": 0.22033187406611268,
      "grad_norm": 0.5232200026512146,
      "learning_rate": 0.0002670128803107047,
      "loss": 3.359,
      "step": 16810
    },
    {
      "epoch": 0.22046294597216032,
      "grad_norm": 0.5235132575035095,
      "learning_rate": 0.00026697396992454963,
      "loss": 3.4381,
      "step": 16820
    },
    {
      "epoch": 0.22059401787820798,
      "grad_norm": 0.46975916624069214,
      "learning_rate": 0.000266935039441946,
      "loss": 3.4152,
      "step": 16830
    },
    {
      "epoch": 0.22072508978425565,
      "grad_norm": 0.49765726923942566,
      "learning_rate": 0.00026689608886958235,
      "loss": 3.3686,
      "step": 16840
    },
    {
      "epoch": 0.2208561616903033,
      "grad_norm": 0.4771338105201721,
      "learning_rate": 0.0002668571182141504,
      "loss": 3.3952,
      "step": 16850
    },
    {
      "epoch": 0.22098723359635095,
      "grad_norm": 0.6371210813522339,
      "learning_rate": 0.00026681812748234536,
      "loss": 3.4805,
      "step": 16860
    },
    {
      "epoch": 0.2211183055023986,
      "grad_norm": 0.48688623309135437,
      "learning_rate": 0.00026677911668086606,
      "loss": 3.4167,
      "step": 16870
    },
    {
      "epoch": 0.22124937740844627,
      "grad_norm": 0.5382996201515198,
      "learning_rate": 0.0002667400858164146,
      "loss": 3.372,
      "step": 16880
    },
    {
      "epoch": 0.22138044931449394,
      "grad_norm": 0.49394339323043823,
      "learning_rate": 0.00026670103489569657,
      "loss": 3.3829,
      "step": 16890
    },
    {
      "epoch": 0.2215115212205416,
      "grad_norm": 0.4759417772293091,
      "learning_rate": 0.00026666196392542105,
      "loss": 3.4361,
      "step": 16900
    },
    {
      "epoch": 0.22164259312658924,
      "grad_norm": 0.5409964919090271,
      "learning_rate": 0.0002666228729123006,
      "loss": 3.3687,
      "step": 16910
    },
    {
      "epoch": 0.2217736650326369,
      "grad_norm": 0.5456223487854004,
      "learning_rate": 0.0002665837618630511,
      "loss": 3.3696,
      "step": 16920
    },
    {
      "epoch": 0.22190473693868457,
      "grad_norm": 0.4861427843570709,
      "learning_rate": 0.0002665446307843919,
      "loss": 3.3887,
      "step": 16930
    },
    {
      "epoch": 0.22203580884473223,
      "grad_norm": 0.5568569898605347,
      "learning_rate": 0.00026650547968304594,
      "loss": 3.4374,
      "step": 16940
    },
    {
      "epoch": 0.22216688075077987,
      "grad_norm": 0.4905017614364624,
      "learning_rate": 0.00026646630856573935,
      "loss": 3.4062,
      "step": 16950
    },
    {
      "epoch": 0.22229795265682753,
      "grad_norm": 0.45244428515434265,
      "learning_rate": 0.0002664271174392019,
      "loss": 3.3939,
      "step": 16960
    },
    {
      "epoch": 0.2224290245628752,
      "grad_norm": 0.48985737562179565,
      "learning_rate": 0.00026638790631016676,
      "loss": 3.4237,
      "step": 16970
    },
    {
      "epoch": 0.22256009646892286,
      "grad_norm": 0.5541404485702515,
      "learning_rate": 0.00026634867518537047,
      "loss": 3.3386,
      "step": 16980
    },
    {
      "epoch": 0.22269116837497052,
      "grad_norm": 0.594668984413147,
      "learning_rate": 0.0002663094240715531,
      "loss": 3.3583,
      "step": 16990
    },
    {
      "epoch": 0.22282224028101816,
      "grad_norm": 0.6855695247650146,
      "learning_rate": 0.000266270152975458,
      "loss": 3.3758,
      "step": 17000
    },
    {
      "epoch": 0.22282224028101816,
      "eval_accuracy": 0.3801622678396872,
      "eval_loss": 3.3947768211364746,
      "eval_runtime": 56.228,
      "eval_samples_per_second": 35.569,
      "eval_steps_per_second": 0.569,
      "step": 17000
    },
    {
      "epoch": 0.22295331218706582,
      "grad_norm": 0.4556143581867218,
      "learning_rate": 0.00026623086190383216,
      "loss": 3.4151,
      "step": 17010
    },
    {
      "epoch": 0.22308438409311349,
      "grad_norm": 0.5514528751373291,
      "learning_rate": 0.00026619155086342577,
      "loss": 3.4415,
      "step": 17020
    },
    {
      "epoch": 0.22321545599916115,
      "grad_norm": 0.44785982370376587,
      "learning_rate": 0.0002661522198609927,
      "loss": 3.38,
      "step": 17030
    },
    {
      "epoch": 0.22334652790520879,
      "grad_norm": 0.5807173848152161,
      "learning_rate": 0.0002661128689032901,
      "loss": 3.405,
      "step": 17040
    },
    {
      "epoch": 0.22347759981125645,
      "grad_norm": 0.729180634021759,
      "learning_rate": 0.0002660734979970785,
      "loss": 3.4221,
      "step": 17050
    },
    {
      "epoch": 0.2236086717173041,
      "grad_norm": 0.49227374792099,
      "learning_rate": 0.000266034107149122,
      "loss": 3.4085,
      "step": 17060
    },
    {
      "epoch": 0.22373974362335178,
      "grad_norm": 0.5061467289924622,
      "learning_rate": 0.000265994696366188,
      "loss": 3.4207,
      "step": 17070
    },
    {
      "epoch": 0.22387081552939944,
      "grad_norm": 0.5126314759254456,
      "learning_rate": 0.0002659552656550474,
      "loss": 3.4411,
      "step": 17080
    },
    {
      "epoch": 0.22400188743544708,
      "grad_norm": 0.4785802364349365,
      "learning_rate": 0.00026591581502247456,
      "loss": 3.432,
      "step": 17090
    },
    {
      "epoch": 0.22413295934149474,
      "grad_norm": 0.5242347717285156,
      "learning_rate": 0.0002658763444752471,
      "loss": 3.4183,
      "step": 17100
    },
    {
      "epoch": 0.2242640312475424,
      "grad_norm": 0.5275418758392334,
      "learning_rate": 0.00026583685402014625,
      "loss": 3.441,
      "step": 17110
    },
    {
      "epoch": 0.22439510315359007,
      "grad_norm": 0.5344134569168091,
      "learning_rate": 0.0002657973436639566,
      "loss": 3.4579,
      "step": 17120
    },
    {
      "epoch": 0.2245261750596377,
      "grad_norm": 0.5219182372093201,
      "learning_rate": 0.00026575781341346607,
      "loss": 3.385,
      "step": 17130
    },
    {
      "epoch": 0.22465724696568537,
      "grad_norm": 0.48795098066329956,
      "learning_rate": 0.00026571826327546606,
      "loss": 3.3983,
      "step": 17140
    },
    {
      "epoch": 0.22478831887173303,
      "grad_norm": 0.5252710580825806,
      "learning_rate": 0.0002656786932567514,
      "loss": 3.3763,
      "step": 17150
    },
    {
      "epoch": 0.2249193907777807,
      "grad_norm": 0.48478052020072937,
      "learning_rate": 0.0002656391033641204,
      "loss": 3.3829,
      "step": 17160
    },
    {
      "epoch": 0.22505046268382836,
      "grad_norm": 0.4631688594818115,
      "learning_rate": 0.00026559949360437465,
      "loss": 3.4068,
      "step": 17170
    },
    {
      "epoch": 0.225181534589876,
      "grad_norm": 0.4956544041633606,
      "learning_rate": 0.00026555986398431927,
      "loss": 3.3948,
      "step": 17180
    },
    {
      "epoch": 0.22531260649592366,
      "grad_norm": 0.47481346130371094,
      "learning_rate": 0.00026552021451076266,
      "loss": 3.3873,
      "step": 17190
    },
    {
      "epoch": 0.22544367840197133,
      "grad_norm": 0.469371497631073,
      "learning_rate": 0.0002654805451905167,
      "loss": 3.3988,
      "step": 17200
    },
    {
      "epoch": 0.225574750308019,
      "grad_norm": 0.5332944393157959,
      "learning_rate": 0.0002654408560303968,
      "loss": 3.4691,
      "step": 17210
    },
    {
      "epoch": 0.22570582221406663,
      "grad_norm": 0.5205812454223633,
      "learning_rate": 0.00026540114703722155,
      "loss": 3.3764,
      "step": 17220
    },
    {
      "epoch": 0.2258368941201143,
      "grad_norm": 0.5361905694007874,
      "learning_rate": 0.0002653614182178131,
      "loss": 3.4148,
      "step": 17230
    },
    {
      "epoch": 0.22596796602616195,
      "grad_norm": 0.5105152130126953,
      "learning_rate": 0.000265321669578997,
      "loss": 3.3807,
      "step": 17240
    },
    {
      "epoch": 0.22609903793220962,
      "grad_norm": 0.4876307547092438,
      "learning_rate": 0.0002652819011276022,
      "loss": 3.4038,
      "step": 17250
    },
    {
      "epoch": 0.22623010983825728,
      "grad_norm": 0.4757630228996277,
      "learning_rate": 0.0002652421128704609,
      "loss": 3.3972,
      "step": 17260
    },
    {
      "epoch": 0.22636118174430492,
      "grad_norm": 0.4577835500240326,
      "learning_rate": 0.000265202304814409,
      "loss": 3.3619,
      "step": 17270
    },
    {
      "epoch": 0.22649225365035258,
      "grad_norm": 0.5464780330657959,
      "learning_rate": 0.00026516247696628547,
      "loss": 3.411,
      "step": 17280
    },
    {
      "epoch": 0.22662332555640025,
      "grad_norm": 0.5261518359184265,
      "learning_rate": 0.00026512262933293295,
      "loss": 3.3831,
      "step": 17290
    },
    {
      "epoch": 0.2267543974624479,
      "grad_norm": 0.5163315534591675,
      "learning_rate": 0.0002650827619211974,
      "loss": 3.413,
      "step": 17300
    },
    {
      "epoch": 0.22688546936849555,
      "grad_norm": 0.46411749720573425,
      "learning_rate": 0.000265042874737928,
      "loss": 3.3772,
      "step": 17310
    },
    {
      "epoch": 0.2270165412745432,
      "grad_norm": 0.6696628928184509,
      "learning_rate": 0.00026500296778997763,
      "loss": 3.4095,
      "step": 17320
    },
    {
      "epoch": 0.22714761318059087,
      "grad_norm": 0.4646836221218109,
      "learning_rate": 0.00026496304108420234,
      "loss": 3.4462,
      "step": 17330
    },
    {
      "epoch": 0.22727868508663854,
      "grad_norm": 0.49629825353622437,
      "learning_rate": 0.00026492309462746165,
      "loss": 3.4107,
      "step": 17340
    },
    {
      "epoch": 0.22740975699268617,
      "grad_norm": 0.4764888286590576,
      "learning_rate": 0.0002648831284266185,
      "loss": 3.365,
      "step": 17350
    },
    {
      "epoch": 0.22754082889873384,
      "grad_norm": 0.4344521164894104,
      "learning_rate": 0.00026484314248853915,
      "loss": 3.3619,
      "step": 17360
    },
    {
      "epoch": 0.2276719008047815,
      "grad_norm": 0.4751974642276764,
      "learning_rate": 0.0002648031368200934,
      "loss": 3.3961,
      "step": 17370
    },
    {
      "epoch": 0.22780297271082917,
      "grad_norm": 0.49443814158439636,
      "learning_rate": 0.0002647631114281542,
      "loss": 3.4113,
      "step": 17380
    },
    {
      "epoch": 0.22793404461687683,
      "grad_norm": 0.5711185336112976,
      "learning_rate": 0.00026472306631959805,
      "loss": 3.3569,
      "step": 17390
    },
    {
      "epoch": 0.22806511652292447,
      "grad_norm": 0.4946565628051758,
      "learning_rate": 0.00026468300150130487,
      "loss": 3.3694,
      "step": 17400
    },
    {
      "epoch": 0.22819618842897213,
      "grad_norm": 0.5110031366348267,
      "learning_rate": 0.00026464291698015787,
      "loss": 3.4144,
      "step": 17410
    },
    {
      "epoch": 0.2283272603350198,
      "grad_norm": 0.5114426612854004,
      "learning_rate": 0.0002646028127630436,
      "loss": 3.3822,
      "step": 17420
    },
    {
      "epoch": 0.22845833224106746,
      "grad_norm": 0.5218921303749084,
      "learning_rate": 0.0002645626888568523,
      "loss": 3.4439,
      "step": 17430
    },
    {
      "epoch": 0.2285894041471151,
      "grad_norm": 0.6223999261856079,
      "learning_rate": 0.0002645225452684771,
      "loss": 3.3837,
      "step": 17440
    },
    {
      "epoch": 0.22872047605316276,
      "grad_norm": 0.4877152144908905,
      "learning_rate": 0.000264482382004815,
      "loss": 3.4235,
      "step": 17450
    },
    {
      "epoch": 0.22885154795921042,
      "grad_norm": 0.46455150842666626,
      "learning_rate": 0.00026444219907276596,
      "loss": 3.4017,
      "step": 17460
    },
    {
      "epoch": 0.22898261986525809,
      "grad_norm": 0.5336849689483643,
      "learning_rate": 0.0002644019964792337,
      "loss": 3.4226,
      "step": 17470
    },
    {
      "epoch": 0.22911369177130575,
      "grad_norm": 0.5674910545349121,
      "learning_rate": 0.00026436177423112497,
      "loss": 3.4585,
      "step": 17480
    },
    {
      "epoch": 0.22924476367735339,
      "grad_norm": 0.5229107737541199,
      "learning_rate": 0.00026432153233535026,
      "loss": 3.3604,
      "step": 17490
    },
    {
      "epoch": 0.22937583558340105,
      "grad_norm": 0.5497515797615051,
      "learning_rate": 0.00026428127079882304,
      "loss": 3.4416,
      "step": 17500
    },
    {
      "epoch": 0.22937583558340105,
      "eval_accuracy": 0.3809115347018573,
      "eval_loss": 3.387723207473755,
      "eval_runtime": 58.5014,
      "eval_samples_per_second": 34.187,
      "eval_steps_per_second": 0.547,
      "step": 17500
    },
    {
      "epoch": 0.2295069074894487,
      "grad_norm": 0.48802050948143005,
      "learning_rate": 0.00026424098962846046,
      "loss": 3.3911,
      "step": 17510
    },
    {
      "epoch": 0.22963797939549638,
      "grad_norm": 0.5731046795845032,
      "learning_rate": 0.0002642006888311829,
      "loss": 3.3897,
      "step": 17520
    },
    {
      "epoch": 0.22976905130154401,
      "grad_norm": 0.457011878490448,
      "learning_rate": 0.0002641603684139141,
      "loss": 3.4318,
      "step": 17530
    },
    {
      "epoch": 0.22990012320759168,
      "grad_norm": 0.8286596536636353,
      "learning_rate": 0.00026412002838358136,
      "loss": 3.3713,
      "step": 17540
    },
    {
      "epoch": 0.23003119511363934,
      "grad_norm": 0.760231614112854,
      "learning_rate": 0.0002640796687471151,
      "loss": 3.3936,
      "step": 17550
    },
    {
      "epoch": 0.230162267019687,
      "grad_norm": 0.6058096885681152,
      "learning_rate": 0.0002640392895114493,
      "loss": 3.4133,
      "step": 17560
    },
    {
      "epoch": 0.23029333892573467,
      "grad_norm": 0.5323788523674011,
      "learning_rate": 0.00026399889068352107,
      "loss": 3.3966,
      "step": 17570
    },
    {
      "epoch": 0.2304244108317823,
      "grad_norm": 0.5030145049095154,
      "learning_rate": 0.0002639584722702712,
      "loss": 3.4084,
      "step": 17580
    },
    {
      "epoch": 0.23055548273782997,
      "grad_norm": 0.5138178467750549,
      "learning_rate": 0.00026391803427864354,
      "loss": 3.4586,
      "step": 17590
    },
    {
      "epoch": 0.23068655464387763,
      "grad_norm": 0.5164147019386292,
      "learning_rate": 0.00026387757671558553,
      "loss": 3.3893,
      "step": 17600
    },
    {
      "epoch": 0.2308176265499253,
      "grad_norm": 0.5105056762695312,
      "learning_rate": 0.00026383709958804793,
      "loss": 3.448,
      "step": 17610
    },
    {
      "epoch": 0.23094869845597293,
      "grad_norm": 0.49030616879463196,
      "learning_rate": 0.0002637966029029848,
      "loss": 3.3714,
      "step": 17620
    },
    {
      "epoch": 0.2310797703620206,
      "grad_norm": 0.45526233315467834,
      "learning_rate": 0.00026375608666735355,
      "loss": 3.3565,
      "step": 17630
    },
    {
      "epoch": 0.23121084226806826,
      "grad_norm": 0.5430101156234741,
      "learning_rate": 0.00026371555088811494,
      "loss": 3.3659,
      "step": 17640
    },
    {
      "epoch": 0.23134191417411593,
      "grad_norm": 0.5124925971031189,
      "learning_rate": 0.0002636749955722332,
      "loss": 3.3638,
      "step": 17650
    },
    {
      "epoch": 0.2314729860801636,
      "grad_norm": 0.48366084694862366,
      "learning_rate": 0.00026363442072667583,
      "loss": 3.3961,
      "step": 17660
    },
    {
      "epoch": 0.23160405798621123,
      "grad_norm": 0.5245600938796997,
      "learning_rate": 0.00026359382635841367,
      "loss": 3.5081,
      "step": 17670
    },
    {
      "epoch": 0.2317351298922589,
      "grad_norm": 0.4796691834926605,
      "learning_rate": 0.000263553212474421,
      "loss": 3.4048,
      "step": 17680
    },
    {
      "epoch": 0.23186620179830655,
      "grad_norm": 0.4758095443248749,
      "learning_rate": 0.00026351257908167533,
      "loss": 3.3914,
      "step": 17690
    },
    {
      "epoch": 0.23199727370435422,
      "grad_norm": 0.5025807023048401,
      "learning_rate": 0.0002634719261871576,
      "loss": 3.3682,
      "step": 17700
    },
    {
      "epoch": 0.23212834561040185,
      "grad_norm": 0.463545024394989,
      "learning_rate": 0.00026343125379785215,
      "loss": 3.3526,
      "step": 17710
    },
    {
      "epoch": 0.23225941751644952,
      "grad_norm": 0.5044845342636108,
      "learning_rate": 0.00026339056192074654,
      "loss": 3.3717,
      "step": 17720
    },
    {
      "epoch": 0.23239048942249718,
      "grad_norm": 0.47991499304771423,
      "learning_rate": 0.00026334985056283177,
      "loss": 3.3904,
      "step": 17730
    },
    {
      "epoch": 0.23252156132854485,
      "grad_norm": 0.4708031117916107,
      "learning_rate": 0.0002633091197311022,
      "loss": 3.4025,
      "step": 17740
    },
    {
      "epoch": 0.2326526332345925,
      "grad_norm": 0.48496532440185547,
      "learning_rate": 0.00026326836943255544,
      "loss": 3.3623,
      "step": 17750
    },
    {
      "epoch": 0.23278370514064015,
      "grad_norm": 0.5322645902633667,
      "learning_rate": 0.00026322759967419246,
      "loss": 3.3449,
      "step": 17760
    },
    {
      "epoch": 0.2329147770466878,
      "grad_norm": 0.49165719747543335,
      "learning_rate": 0.00026318681046301777,
      "loss": 3.4149,
      "step": 17770
    },
    {
      "epoch": 0.23304584895273547,
      "grad_norm": 0.4983372986316681,
      "learning_rate": 0.00026314600180603895,
      "loss": 3.4188,
      "step": 17780
    },
    {
      "epoch": 0.23317692085878314,
      "grad_norm": 0.49107345938682556,
      "learning_rate": 0.00026310517371026706,
      "loss": 3.3693,
      "step": 17790
    },
    {
      "epoch": 0.23330799276483077,
      "grad_norm": 0.5023448467254639,
      "learning_rate": 0.0002630643261827165,
      "loss": 3.3876,
      "step": 17800
    },
    {
      "epoch": 0.23343906467087844,
      "grad_norm": 0.47192686796188354,
      "learning_rate": 0.000263023459230405,
      "loss": 3.3335,
      "step": 17810
    },
    {
      "epoch": 0.2335701365769261,
      "grad_norm": 0.49780160188674927,
      "learning_rate": 0.0002629825728603535,
      "loss": 3.406,
      "step": 17820
    },
    {
      "epoch": 0.23370120848297377,
      "grad_norm": 0.5062437057495117,
      "learning_rate": 0.00026294166707958654,
      "loss": 3.4254,
      "step": 17830
    },
    {
      "epoch": 0.23383228038902143,
      "grad_norm": 0.4637068510055542,
      "learning_rate": 0.0002629007418951318,
      "loss": 3.3844,
      "step": 17840
    },
    {
      "epoch": 0.23396335229506907,
      "grad_norm": 0.5053417086601257,
      "learning_rate": 0.00026285979731402025,
      "loss": 3.4175,
      "step": 17850
    },
    {
      "epoch": 0.23409442420111673,
      "grad_norm": 0.6807954907417297,
      "learning_rate": 0.0002628188333432864,
      "loss": 3.3639,
      "step": 17860
    },
    {
      "epoch": 0.2342254961071644,
      "grad_norm": 0.5615761876106262,
      "learning_rate": 0.00026277784998996796,
      "loss": 3.4024,
      "step": 17870
    },
    {
      "epoch": 0.23435656801321206,
      "grad_norm": 0.49657776951789856,
      "learning_rate": 0.0002627368472611059,
      "loss": 3.4123,
      "step": 17880
    },
    {
      "epoch": 0.2344876399192597,
      "grad_norm": 0.4727565050125122,
      "learning_rate": 0.0002626958251637447,
      "loss": 3.4139,
      "step": 17890
    },
    {
      "epoch": 0.23461871182530736,
      "grad_norm": 0.47801733016967773,
      "learning_rate": 0.00026265478370493203,
      "loss": 3.4146,
      "step": 17900
    },
    {
      "epoch": 0.23474978373135502,
      "grad_norm": 0.46431678533554077,
      "learning_rate": 0.0002626137228917188,
      "loss": 3.3666,
      "step": 17910
    },
    {
      "epoch": 0.23488085563740269,
      "grad_norm": 0.7370932698249817,
      "learning_rate": 0.00026257264273115956,
      "loss": 3.4096,
      "step": 17920
    },
    {
      "epoch": 0.23501192754345035,
      "grad_norm": 0.4802114963531494,
      "learning_rate": 0.0002625315432303119,
      "loss": 3.3829,
      "step": 17930
    },
    {
      "epoch": 0.23514299944949799,
      "grad_norm": 0.45758768916130066,
      "learning_rate": 0.0002624904243962369,
      "loss": 3.3801,
      "step": 17940
    },
    {
      "epoch": 0.23527407135554565,
      "grad_norm": 0.45234042406082153,
      "learning_rate": 0.0002624492862359988,
      "loss": 3.3897,
      "step": 17950
    },
    {
      "epoch": 0.2354051432615933,
      "grad_norm": 0.4375002682209015,
      "learning_rate": 0.00026240812875666527,
      "loss": 3.3504,
      "step": 17960
    },
    {
      "epoch": 0.23553621516764098,
      "grad_norm": 0.5750641226768494,
      "learning_rate": 0.0002623669519653073,
      "loss": 3.3797,
      "step": 17970
    },
    {
      "epoch": 0.2356672870736886,
      "grad_norm": 0.4941924214363098,
      "learning_rate": 0.00026232575586899917,
      "loss": 3.3997,
      "step": 17980
    },
    {
      "epoch": 0.23579835897973628,
      "grad_norm": 0.6209459900856018,
      "learning_rate": 0.00026228454047481844,
      "loss": 3.3941,
      "step": 17990
    },
    {
      "epoch": 0.23592943088578394,
      "grad_norm": 0.4788353741168976,
      "learning_rate": 0.0002622433057898461,
      "loss": 3.3979,
      "step": 18000
    },
    {
      "epoch": 0.23592943088578394,
      "eval_accuracy": 0.3819144672531769,
      "eval_loss": 3.3799145221710205,
      "eval_runtime": 58.1225,
      "eval_samples_per_second": 34.41,
      "eval_steps_per_second": 0.551,
      "step": 18000
    },
    {
      "epoch": 0.2360605027918316,
      "grad_norm": 1.327889323234558,
      "learning_rate": 0.0002622020518211663,
      "loss": 3.4198,
      "step": 18010
    },
    {
      "epoch": 0.23619157469787927,
      "grad_norm": 4.272329807281494,
      "learning_rate": 0.0002621607785758667,
      "loss": 3.3981,
      "step": 18020
    },
    {
      "epoch": 0.2363226466039269,
      "grad_norm": 0.5897387862205505,
      "learning_rate": 0.00026211948606103805,
      "loss": 3.4245,
      "step": 18030
    },
    {
      "epoch": 0.23645371850997457,
      "grad_norm": 0.5627679228782654,
      "learning_rate": 0.00026207817428377454,
      "loss": 3.4919,
      "step": 18040
    },
    {
      "epoch": 0.23658479041602223,
      "grad_norm": 0.45954766869544983,
      "learning_rate": 0.00026203684325117375,
      "loss": 3.3899,
      "step": 18050
    },
    {
      "epoch": 0.2367158623220699,
      "grad_norm": 0.7642937898635864,
      "learning_rate": 0.00026199549297033627,
      "loss": 3.3632,
      "step": 18060
    },
    {
      "epoch": 0.23684693422811753,
      "grad_norm": 0.5129378437995911,
      "learning_rate": 0.00026195412344836634,
      "loss": 3.4092,
      "step": 18070
    },
    {
      "epoch": 0.2369780061341652,
      "grad_norm": 0.4724496901035309,
      "learning_rate": 0.00026191273469237133,
      "loss": 3.3683,
      "step": 18080
    },
    {
      "epoch": 0.23710907804021286,
      "grad_norm": 0.5164384245872498,
      "learning_rate": 0.0002618713267094619,
      "loss": 3.3788,
      "step": 18090
    },
    {
      "epoch": 0.23724014994626053,
      "grad_norm": 0.5096123814582825,
      "learning_rate": 0.0002618298995067521,
      "loss": 3.4118,
      "step": 18100
    },
    {
      "epoch": 0.2373712218523082,
      "grad_norm": 0.6319358348846436,
      "learning_rate": 0.0002617884530913592,
      "loss": 3.3861,
      "step": 18110
    },
    {
      "epoch": 0.23750229375835583,
      "grad_norm": 0.5916162729263306,
      "learning_rate": 0.0002617469874704039,
      "loss": 3.3369,
      "step": 18120
    },
    {
      "epoch": 0.2376333656644035,
      "grad_norm": 0.5185272097587585,
      "learning_rate": 0.0002617055026510099,
      "loss": 3.4323,
      "step": 18130
    },
    {
      "epoch": 0.23776443757045115,
      "grad_norm": 0.5723345875740051,
      "learning_rate": 0.00026166399864030465,
      "loss": 3.4499,
      "step": 18140
    },
    {
      "epoch": 0.23789550947649882,
      "grad_norm": 0.49790388345718384,
      "learning_rate": 0.0002616224754454185,
      "loss": 3.4287,
      "step": 18150
    },
    {
      "epoch": 0.23802658138254645,
      "grad_norm": 0.6142022609710693,
      "learning_rate": 0.0002615809330734853,
      "loss": 3.4504,
      "step": 18160
    },
    {
      "epoch": 0.23815765328859412,
      "grad_norm": 0.542665958404541,
      "learning_rate": 0.0002615393715316421,
      "loss": 3.3603,
      "step": 18170
    },
    {
      "epoch": 0.23828872519464178,
      "grad_norm": 0.49192872643470764,
      "learning_rate": 0.00026149779082702937,
      "loss": 3.3842,
      "step": 18180
    },
    {
      "epoch": 0.23841979710068945,
      "grad_norm": 0.5645651817321777,
      "learning_rate": 0.0002614561909667907,
      "loss": 3.3944,
      "step": 18190
    },
    {
      "epoch": 0.2385508690067371,
      "grad_norm": 0.46613574028015137,
      "learning_rate": 0.00026141457195807315,
      "loss": 3.4094,
      "step": 18200
    },
    {
      "epoch": 0.23868194091278475,
      "grad_norm": 0.5053684711456299,
      "learning_rate": 0.0002613729338080269,
      "loss": 3.3888,
      "step": 18210
    },
    {
      "epoch": 0.2388130128188324,
      "grad_norm": 0.5207937955856323,
      "learning_rate": 0.0002613312765238055,
      "loss": 3.3752,
      "step": 18220
    },
    {
      "epoch": 0.23894408472488007,
      "grad_norm": 0.5006843209266663,
      "learning_rate": 0.00026128960011256587,
      "loss": 3.4086,
      "step": 18230
    },
    {
      "epoch": 0.23907515663092774,
      "grad_norm": 0.49254661798477173,
      "learning_rate": 0.00026124790458146804,
      "loss": 3.3988,
      "step": 18240
    },
    {
      "epoch": 0.23920622853697537,
      "grad_norm": 0.522190272808075,
      "learning_rate": 0.0002612061899376754,
      "loss": 3.4184,
      "step": 18250
    },
    {
      "epoch": 0.23933730044302304,
      "grad_norm": 0.505527138710022,
      "learning_rate": 0.00026116445618835476,
      "loss": 3.3962,
      "step": 18260
    },
    {
      "epoch": 0.2394683723490707,
      "grad_norm": 0.4459320604801178,
      "learning_rate": 0.00026112270334067604,
      "loss": 3.363,
      "step": 18270
    },
    {
      "epoch": 0.23959944425511837,
      "grad_norm": 0.48850515484809875,
      "learning_rate": 0.0002610809314018124,
      "loss": 3.3902,
      "step": 18280
    },
    {
      "epoch": 0.239730516161166,
      "grad_norm": 0.5111159682273865,
      "learning_rate": 0.00026103914037894053,
      "loss": 3.4386,
      "step": 18290
    },
    {
      "epoch": 0.23986158806721367,
      "grad_norm": 0.4699200689792633,
      "learning_rate": 0.00026099733027924016,
      "loss": 3.4089,
      "step": 18300
    },
    {
      "epoch": 0.23999265997326133,
      "grad_norm": 0.49904584884643555,
      "learning_rate": 0.00026095550110989435,
      "loss": 3.3723,
      "step": 18310
    },
    {
      "epoch": 0.240123731879309,
      "grad_norm": 0.49701058864593506,
      "learning_rate": 0.0002609136528780895,
      "loss": 3.3827,
      "step": 18320
    },
    {
      "epoch": 0.24025480378535666,
      "grad_norm": 0.5063018202781677,
      "learning_rate": 0.00026087178559101523,
      "loss": 3.4561,
      "step": 18330
    },
    {
      "epoch": 0.2403858756914043,
      "grad_norm": 0.5363951921463013,
      "learning_rate": 0.0002608298992558645,
      "loss": 3.3811,
      "step": 18340
    },
    {
      "epoch": 0.24051694759745196,
      "grad_norm": 0.5137924551963806,
      "learning_rate": 0.0002607879938798334,
      "loss": 3.4193,
      "step": 18350
    },
    {
      "epoch": 0.24064801950349962,
      "grad_norm": 0.6772361993789673,
      "learning_rate": 0.0002607460694701216,
      "loss": 3.4356,
      "step": 18360
    },
    {
      "epoch": 0.24077909140954729,
      "grad_norm": 0.5464832186698914,
      "learning_rate": 0.0002607041260339316,
      "loss": 3.3921,
      "step": 18370
    },
    {
      "epoch": 0.24091016331559492,
      "grad_norm": 0.6228564381599426,
      "learning_rate": 0.0002606621635784695,
      "loss": 3.482,
      "step": 18380
    },
    {
      "epoch": 0.24104123522164259,
      "grad_norm": 0.5203681588172913,
      "learning_rate": 0.0002606201821109446,
      "loss": 3.3886,
      "step": 18390
    },
    {
      "epoch": 0.24117230712769025,
      "grad_norm": 0.5531490445137024,
      "learning_rate": 0.0002605781816385693,
      "loss": 3.3623,
      "step": 18400
    },
    {
      "epoch": 0.2413033790337379,
      "grad_norm": 0.6129160523414612,
      "learning_rate": 0.00026053616216855947,
      "loss": 3.3966,
      "step": 18410
    },
    {
      "epoch": 0.24143445093978558,
      "grad_norm": 0.6550762057304382,
      "learning_rate": 0.00026049412370813423,
      "loss": 3.4231,
      "step": 18420
    },
    {
      "epoch": 0.2415655228458332,
      "grad_norm": 0.4645429849624634,
      "learning_rate": 0.00026045206626451584,
      "loss": 3.395,
      "step": 18430
    },
    {
      "epoch": 0.24169659475188088,
      "grad_norm": 0.4609980583190918,
      "learning_rate": 0.0002604099898449299,
      "loss": 3.4284,
      "step": 18440
    },
    {
      "epoch": 0.24182766665792854,
      "grad_norm": 0.5266175866127014,
      "learning_rate": 0.00026036789445660525,
      "loss": 3.3876,
      "step": 18450
    },
    {
      "epoch": 0.2419587385639762,
      "grad_norm": 0.5135475993156433,
      "learning_rate": 0.00026032578010677394,
      "loss": 3.3933,
      "step": 18460
    },
    {
      "epoch": 0.24208981047002384,
      "grad_norm": 0.5176001191139221,
      "learning_rate": 0.00026028364680267147,
      "loss": 3.4338,
      "step": 18470
    },
    {
      "epoch": 0.2422208823760715,
      "grad_norm": 0.4797724783420563,
      "learning_rate": 0.0002602414945515363,
      "loss": 3.308,
      "step": 18480
    },
    {
      "epoch": 0.24235195428211917,
      "grad_norm": 0.5928671956062317,
      "learning_rate": 0.0002601993233606104,
      "loss": 3.3591,
      "step": 18490
    },
    {
      "epoch": 0.24248302618816683,
      "grad_norm": 0.5590078234672546,
      "learning_rate": 0.0002601571332371389,
      "loss": 3.3946,
      "step": 18500
    },
    {
      "epoch": 0.24248302618816683,
      "eval_accuracy": 0.3822741935483871,
      "eval_loss": 3.3758177757263184,
      "eval_runtime": 54.6966,
      "eval_samples_per_second": 36.565,
      "eval_steps_per_second": 0.585,
      "step": 18500
    },
    {
      "epoch": 0.2426140980942145,
      "grad_norm": 0.5177618861198425,
      "learning_rate": 0.00026011492418837,
      "loss": 3.4273,
      "step": 18510
    },
    {
      "epoch": 0.24274517000026213,
      "grad_norm": 0.48212575912475586,
      "learning_rate": 0.00026007269622155564,
      "loss": 3.3906,
      "step": 18520
    },
    {
      "epoch": 0.2428762419063098,
      "grad_norm": 0.5299989581108093,
      "learning_rate": 0.0002600304493439505,
      "loss": 3.365,
      "step": 18530
    },
    {
      "epoch": 0.24300731381235746,
      "grad_norm": 0.7468590140342712,
      "learning_rate": 0.0002599881835628127,
      "loss": 3.3926,
      "step": 18540
    },
    {
      "epoch": 0.24313838571840513,
      "grad_norm": 0.5878967046737671,
      "learning_rate": 0.00025994589888540364,
      "loss": 3.4076,
      "step": 18550
    },
    {
      "epoch": 0.24326945762445276,
      "grad_norm": 0.54641193151474,
      "learning_rate": 0.00025990359531898803,
      "loss": 3.3917,
      "step": 18560
    },
    {
      "epoch": 0.24340052953050043,
      "grad_norm": 0.49342408776283264,
      "learning_rate": 0.0002598612728708337,
      "loss": 3.3902,
      "step": 18570
    },
    {
      "epoch": 0.2435316014365481,
      "grad_norm": 0.5096401572227478,
      "learning_rate": 0.0002598189315482117,
      "loss": 3.3766,
      "step": 18580
    },
    {
      "epoch": 0.24366267334259575,
      "grad_norm": 0.5253897309303284,
      "learning_rate": 0.0002597765713583964,
      "loss": 3.3741,
      "step": 18590
    },
    {
      "epoch": 0.24379374524864342,
      "grad_norm": 0.47128984332084656,
      "learning_rate": 0.00025973419230866543,
      "loss": 3.362,
      "step": 18600
    },
    {
      "epoch": 0.24392481715469105,
      "grad_norm": 0.49379560351371765,
      "learning_rate": 0.00025969179440629963,
      "loss": 3.4197,
      "step": 18610
    },
    {
      "epoch": 0.24405588906073872,
      "grad_norm": 0.4624928832054138,
      "learning_rate": 0.0002596493776585831,
      "loss": 3.3699,
      "step": 18620
    },
    {
      "epoch": 0.24418696096678638,
      "grad_norm": 0.46999022364616394,
      "learning_rate": 0.000259606942072803,
      "loss": 3.3894,
      "step": 18630
    },
    {
      "epoch": 0.24431803287283405,
      "grad_norm": 0.485188752412796,
      "learning_rate": 0.0002595644876562501,
      "loss": 3.378,
      "step": 18640
    },
    {
      "epoch": 0.24444910477888168,
      "grad_norm": 0.5165528059005737,
      "learning_rate": 0.0002595220144162181,
      "loss": 3.4071,
      "step": 18650
    },
    {
      "epoch": 0.24458017668492935,
      "grad_norm": 0.46077853441238403,
      "learning_rate": 0.0002594795223600039,
      "loss": 3.4017,
      "step": 18660
    },
    {
      "epoch": 0.244711248590977,
      "grad_norm": 0.5288727879524231,
      "learning_rate": 0.00025943701149490795,
      "loss": 3.4272,
      "step": 18670
    },
    {
      "epoch": 0.24484232049702467,
      "grad_norm": 0.5579582452774048,
      "learning_rate": 0.00025939448182823354,
      "loss": 3.389,
      "step": 18680
    },
    {
      "epoch": 0.24497339240307234,
      "grad_norm": 0.5361863970756531,
      "learning_rate": 0.00025935193336728754,
      "loss": 3.3533,
      "step": 18690
    },
    {
      "epoch": 0.24510446430911997,
      "grad_norm": 0.5100134015083313,
      "learning_rate": 0.0002593093661193798,
      "loss": 3.3543,
      "step": 18700
    },
    {
      "epoch": 0.24523553621516764,
      "grad_norm": 0.5155029296875,
      "learning_rate": 0.0002592667800918235,
      "loss": 3.3888,
      "step": 18710
    },
    {
      "epoch": 0.2453666081212153,
      "grad_norm": 0.494831919670105,
      "learning_rate": 0.0002592241752919351,
      "loss": 3.3843,
      "step": 18720
    },
    {
      "epoch": 0.24549768002726297,
      "grad_norm": 0.5021178126335144,
      "learning_rate": 0.00025918155172703416,
      "loss": 3.3732,
      "step": 18730
    },
    {
      "epoch": 0.2456287519333106,
      "grad_norm": 0.504138708114624,
      "learning_rate": 0.0002591389094044435,
      "loss": 3.3575,
      "step": 18740
    },
    {
      "epoch": 0.24575982383935827,
      "grad_norm": 0.5163382291793823,
      "learning_rate": 0.0002590962483314894,
      "loss": 3.3862,
      "step": 18750
    },
    {
      "epoch": 0.24589089574540593,
      "grad_norm": 0.46512117981910706,
      "learning_rate": 0.00025905356851550076,
      "loss": 3.3667,
      "step": 18760
    },
    {
      "epoch": 0.2460219676514536,
      "grad_norm": 0.5206557512283325,
      "learning_rate": 0.0002590108699638104,
      "loss": 3.372,
      "step": 18770
    },
    {
      "epoch": 0.24615303955750126,
      "grad_norm": 0.4781223237514496,
      "learning_rate": 0.000258968152683754,
      "loss": 3.2837,
      "step": 18780
    },
    {
      "epoch": 0.2462841114635489,
      "grad_norm": 0.5260870456695557,
      "learning_rate": 0.00025892541668267043,
      "loss": 3.3634,
      "step": 18790
    },
    {
      "epoch": 0.24641518336959656,
      "grad_norm": 0.47804683446884155,
      "learning_rate": 0.00025888266196790187,
      "loss": 3.3721,
      "step": 18800
    },
    {
      "epoch": 0.24654625527564422,
      "grad_norm": 0.49942153692245483,
      "learning_rate": 0.0002588398885467937,
      "loss": 3.3406,
      "step": 18810
    },
    {
      "epoch": 0.24667732718169189,
      "grad_norm": 0.5488879084587097,
      "learning_rate": 0.00025879709642669455,
      "loss": 3.41,
      "step": 18820
    },
    {
      "epoch": 0.24680839908773952,
      "grad_norm": 0.4843831956386566,
      "learning_rate": 0.00025875428561495615,
      "loss": 3.4211,
      "step": 18830
    },
    {
      "epoch": 0.24693947099378719,
      "grad_norm": 0.4863773584365845,
      "learning_rate": 0.00025871145611893365,
      "loss": 3.3889,
      "step": 18840
    },
    {
      "epoch": 0.24707054289983485,
      "grad_norm": 0.4832908511161804,
      "learning_rate": 0.00025866860794598515,
      "loss": 3.4014,
      "step": 18850
    },
    {
      "epoch": 0.2472016148058825,
      "grad_norm": 0.46896177530288696,
      "learning_rate": 0.00025862574110347206,
      "loss": 3.3942,
      "step": 18860
    },
    {
      "epoch": 0.24733268671193018,
      "grad_norm": 0.47052544355392456,
      "learning_rate": 0.00025858285559875914,
      "loss": 3.374,
      "step": 18870
    },
    {
      "epoch": 0.2474637586179778,
      "grad_norm": 0.5489212870597839,
      "learning_rate": 0.00025853995143921416,
      "loss": 3.3656,
      "step": 18880
    },
    {
      "epoch": 0.24759483052402548,
      "grad_norm": 0.47955843806266785,
      "learning_rate": 0.00025849702863220815,
      "loss": 3.379,
      "step": 18890
    },
    {
      "epoch": 0.24772590243007314,
      "grad_norm": 0.500724732875824,
      "learning_rate": 0.00025845408718511544,
      "loss": 3.3699,
      "step": 18900
    },
    {
      "epoch": 0.2478569743361208,
      "grad_norm": 0.5118017792701721,
      "learning_rate": 0.0002584111271053135,
      "loss": 3.3938,
      "step": 18910
    },
    {
      "epoch": 0.24798804624216844,
      "grad_norm": 0.46557512879371643,
      "learning_rate": 0.00025836814840018285,
      "loss": 3.3854,
      "step": 18920
    },
    {
      "epoch": 0.2481191181482161,
      "grad_norm": 0.48239007592201233,
      "learning_rate": 0.00025832515107710753,
      "loss": 3.4083,
      "step": 18930
    },
    {
      "epoch": 0.24825019005426377,
      "grad_norm": 0.48925870656967163,
      "learning_rate": 0.00025828213514347446,
      "loss": 3.4299,
      "step": 18940
    },
    {
      "epoch": 0.24838126196031143,
      "grad_norm": 0.5083717107772827,
      "learning_rate": 0.000258239100606674,
      "loss": 3.399,
      "step": 18950
    },
    {
      "epoch": 0.2485123338663591,
      "grad_norm": 0.5038092732429504,
      "learning_rate": 0.00025819604747409945,
      "loss": 3.3806,
      "step": 18960
    },
    {
      "epoch": 0.24864340577240673,
      "grad_norm": 0.4976921081542969,
      "learning_rate": 0.0002581529757531476,
      "loss": 3.3708,
      "step": 18970
    },
    {
      "epoch": 0.2487744776784544,
      "grad_norm": 0.4652520418167114,
      "learning_rate": 0.0002581098854512183,
      "loss": 3.4195,
      "step": 18980
    },
    {
      "epoch": 0.24890554958450206,
      "grad_norm": 0.46985769271850586,
      "learning_rate": 0.0002580667765757144,
      "loss": 3.3771,
      "step": 18990
    },
    {
      "epoch": 0.24903662149054973,
      "grad_norm": 0.5051211714744568,
      "learning_rate": 0.00025802364913404233,
      "loss": 3.3987,
      "step": 19000
    },
    {
      "epoch": 0.24903662149054973,
      "eval_accuracy": 0.38285288367546433,
      "eval_loss": 3.368293046951294,
      "eval_runtime": 58.5706,
      "eval_samples_per_second": 34.147,
      "eval_steps_per_second": 0.546,
      "step": 19000
    },
    {
      "epoch": 0.24916769339659736,
      "grad_norm": 0.5215790271759033,
      "learning_rate": 0.00025798050313361137,
      "loss": 3.3712,
      "step": 19010
    },
    {
      "epoch": 0.24929876530264503,
      "grad_norm": 0.564456045627594,
      "learning_rate": 0.00025793733858183425,
      "loss": 3.3806,
      "step": 19020
    },
    {
      "epoch": 0.2494298372086927,
      "grad_norm": 0.5583539009094238,
      "learning_rate": 0.0002578941554861266,
      "loss": 3.3838,
      "step": 19030
    },
    {
      "epoch": 0.24956090911474035,
      "grad_norm": 0.6268367171287537,
      "learning_rate": 0.00025785095385390754,
      "loss": 3.4205,
      "step": 19040
    },
    {
      "epoch": 0.24969198102078802,
      "grad_norm": 0.7370564341545105,
      "learning_rate": 0.0002578077336925991,
      "loss": 3.4061,
      "step": 19050
    },
    {
      "epoch": 0.24982305292683565,
      "grad_norm": 0.5384109616279602,
      "learning_rate": 0.0002577644950096268,
      "loss": 3.382,
      "step": 19060
    },
    {
      "epoch": 0.24995412483288332,
      "grad_norm": 0.5439771413803101,
      "learning_rate": 0.0002577212378124189,
      "loss": 3.3555,
      "step": 19070
    },
    {
      "epoch": 0.250085196738931,
      "grad_norm": 0.5593909621238708,
      "learning_rate": 0.0002576779621084073,
      "loss": 3.4644,
      "step": 19080
    },
    {
      "epoch": 0.25021626864497865,
      "grad_norm": 0.5148481726646423,
      "learning_rate": 0.0002576346679050269,
      "loss": 3.3707,
      "step": 19090
    },
    {
      "epoch": 0.2503473405510263,
      "grad_norm": 0.5286930799484253,
      "learning_rate": 0.00025759135520971564,
      "loss": 3.4159,
      "step": 19100
    },
    {
      "epoch": 0.250478412457074,
      "grad_norm": 0.5317391753196716,
      "learning_rate": 0.00025754802402991485,
      "loss": 3.3591,
      "step": 19110
    },
    {
      "epoch": 0.2506094843631216,
      "grad_norm": 0.5086570382118225,
      "learning_rate": 0.00025750467437306887,
      "loss": 3.373,
      "step": 19120
    },
    {
      "epoch": 0.25074055626916925,
      "grad_norm": 0.4801681339740753,
      "learning_rate": 0.00025746130624662534,
      "loss": 3.3854,
      "step": 19130
    },
    {
      "epoch": 0.2508716281752169,
      "grad_norm": 0.4954985976219177,
      "learning_rate": 0.000257417919658035,
      "loss": 3.3755,
      "step": 19140
    },
    {
      "epoch": 0.2510027000812646,
      "grad_norm": 0.4759616255760193,
      "learning_rate": 0.0002573745146147519,
      "loss": 3.3647,
      "step": 19150
    },
    {
      "epoch": 0.25113377198731224,
      "grad_norm": 0.48420652747154236,
      "learning_rate": 0.00025733109112423296,
      "loss": 3.474,
      "step": 19160
    },
    {
      "epoch": 0.2512648438933599,
      "grad_norm": 0.499133825302124,
      "learning_rate": 0.0002572876491939386,
      "loss": 3.3929,
      "step": 19170
    },
    {
      "epoch": 0.25139591579940757,
      "grad_norm": 0.5237839818000793,
      "learning_rate": 0.0002572441888313322,
      "loss": 3.3554,
      "step": 19180
    },
    {
      "epoch": 0.25152698770545523,
      "grad_norm": 0.5094130635261536,
      "learning_rate": 0.00025720071004388037,
      "loss": 3.3683,
      "step": 19190
    },
    {
      "epoch": 0.2516580596115029,
      "grad_norm": 0.48563820123672485,
      "learning_rate": 0.00025715721283905285,
      "loss": 3.3841,
      "step": 19200
    },
    {
      "epoch": 0.2517891315175505,
      "grad_norm": 0.513314425945282,
      "learning_rate": 0.00025711369722432266,
      "loss": 3.3543,
      "step": 19210
    },
    {
      "epoch": 0.25192020342359817,
      "grad_norm": 0.5647361278533936,
      "learning_rate": 0.0002570701632071658,
      "loss": 3.3721,
      "step": 19220
    },
    {
      "epoch": 0.25205127532964583,
      "grad_norm": 0.49494507908821106,
      "learning_rate": 0.0002570266107950617,
      "loss": 3.3783,
      "step": 19230
    },
    {
      "epoch": 0.2521823472356935,
      "grad_norm": 0.6295251846313477,
      "learning_rate": 0.0002569830399954926,
      "loss": 3.3988,
      "step": 19240
    },
    {
      "epoch": 0.25231341914174116,
      "grad_norm": 0.5498766303062439,
      "learning_rate": 0.0002569394508159442,
      "loss": 3.4212,
      "step": 19250
    },
    {
      "epoch": 0.2524444910477888,
      "grad_norm": 0.49350452423095703,
      "learning_rate": 0.00025689584326390514,
      "loss": 3.3973,
      "step": 19260
    },
    {
      "epoch": 0.2525755629538365,
      "grad_norm": 0.5391107797622681,
      "learning_rate": 0.0002568522173468674,
      "loss": 3.3808,
      "step": 19270
    },
    {
      "epoch": 0.25270663485988415,
      "grad_norm": 0.5277228355407715,
      "learning_rate": 0.000256808573072326,
      "loss": 3.3741,
      "step": 19280
    },
    {
      "epoch": 0.2528377067659318,
      "grad_norm": 0.5601983666419983,
      "learning_rate": 0.0002567649104477792,
      "loss": 3.397,
      "step": 19290
    },
    {
      "epoch": 0.2529687786719794,
      "grad_norm": 0.5053668022155762,
      "learning_rate": 0.0002567212294807283,
      "loss": 3.387,
      "step": 19300
    },
    {
      "epoch": 0.2530998505780271,
      "grad_norm": 0.5177274942398071,
      "learning_rate": 0.00025667753017867784,
      "loss": 3.3641,
      "step": 19310
    },
    {
      "epoch": 0.25323092248407475,
      "grad_norm": 0.5592427253723145,
      "learning_rate": 0.00025663381254913543,
      "loss": 3.3866,
      "step": 19320
    },
    {
      "epoch": 0.2533619943901224,
      "grad_norm": 0.49000266194343567,
      "learning_rate": 0.00025659007659961186,
      "loss": 3.3628,
      "step": 19330
    },
    {
      "epoch": 0.2534930662961701,
      "grad_norm": 0.5149667859077454,
      "learning_rate": 0.00025654632233762124,
      "loss": 3.374,
      "step": 19340
    },
    {
      "epoch": 0.25362413820221774,
      "grad_norm": 0.5025848746299744,
      "learning_rate": 0.00025650254977068056,
      "loss": 3.367,
      "step": 19350
    },
    {
      "epoch": 0.2537552101082654,
      "grad_norm": 0.4912576973438263,
      "learning_rate": 0.00025645875890631,
      "loss": 3.4284,
      "step": 19360
    },
    {
      "epoch": 0.25388628201431307,
      "grad_norm": 0.5368895530700684,
      "learning_rate": 0.0002564149497520331,
      "loss": 3.3885,
      "step": 19370
    },
    {
      "epoch": 0.25401735392036073,
      "grad_norm": 0.6058098077774048,
      "learning_rate": 0.0002563711223153763,
      "loss": 3.4206,
      "step": 19380
    },
    {
      "epoch": 0.25414842582640834,
      "grad_norm": 0.5442577004432678,
      "learning_rate": 0.0002563272766038693,
      "loss": 3.3945,
      "step": 19390
    },
    {
      "epoch": 0.254279497732456,
      "grad_norm": 0.522678792476654,
      "learning_rate": 0.00025628341262504486,
      "loss": 3.3673,
      "step": 19400
    },
    {
      "epoch": 0.25441056963850367,
      "grad_norm": 0.5306452512741089,
      "learning_rate": 0.00025623953038643904,
      "loss": 3.3315,
      "step": 19410
    },
    {
      "epoch": 0.25454164154455133,
      "grad_norm": 0.4932625889778137,
      "learning_rate": 0.00025619562989559084,
      "loss": 3.3506,
      "step": 19420
    },
    {
      "epoch": 0.254672713450599,
      "grad_norm": 0.5194747447967529,
      "learning_rate": 0.00025615171116004256,
      "loss": 3.3502,
      "step": 19430
    },
    {
      "epoch": 0.25480378535664666,
      "grad_norm": 0.5342203378677368,
      "learning_rate": 0.00025610777418733945,
      "loss": 3.3773,
      "step": 19440
    },
    {
      "epoch": 0.2549348572626943,
      "grad_norm": 0.524432361125946,
      "learning_rate": 0.00025606381898503,
      "loss": 3.412,
      "step": 19450
    },
    {
      "epoch": 0.255065929168742,
      "grad_norm": 0.5470812916755676,
      "learning_rate": 0.00025601984556066604,
      "loss": 3.45,
      "step": 19460
    },
    {
      "epoch": 0.25519700107478965,
      "grad_norm": 0.5186784863471985,
      "learning_rate": 0.0002559758539218021,
      "loss": 3.3873,
      "step": 19470
    },
    {
      "epoch": 0.25532807298083726,
      "grad_norm": 0.569594144821167,
      "learning_rate": 0.0002559318440759962,
      "loss": 3.3906,
      "step": 19480
    },
    {
      "epoch": 0.2554591448868849,
      "grad_norm": 0.49763914942741394,
      "learning_rate": 0.0002558878160308092,
      "loss": 3.3831,
      "step": 19490
    },
    {
      "epoch": 0.2555902167929326,
      "grad_norm": 0.5492805242538452,
      "learning_rate": 0.0002558437697938054,
      "loss": 3.3083,
      "step": 19500
    },
    {
      "epoch": 0.2555902167929326,
      "eval_accuracy": 0.38351270772238516,
      "eval_loss": 3.3659300804138184,
      "eval_runtime": 60.2158,
      "eval_samples_per_second": 33.214,
      "eval_steps_per_second": 0.531,
      "step": 19500
    },
    {
      "epoch": 0.25572128869898025,
      "grad_norm": 0.468669593334198,
      "learning_rate": 0.00025579970537255206,
      "loss": 3.3493,
      "step": 19510
    },
    {
      "epoch": 0.2558523606050279,
      "grad_norm": 0.5602644085884094,
      "learning_rate": 0.00025575562277461945,
      "loss": 3.3673,
      "step": 19520
    },
    {
      "epoch": 0.2559834325110756,
      "grad_norm": 0.47940757870674133,
      "learning_rate": 0.00025571152200758114,
      "loss": 3.4144,
      "step": 19530
    },
    {
      "epoch": 0.25611450441712325,
      "grad_norm": 0.5363076329231262,
      "learning_rate": 0.0002556674030790138,
      "loss": 3.3628,
      "step": 19540
    },
    {
      "epoch": 0.2562455763231709,
      "grad_norm": 0.45793336629867554,
      "learning_rate": 0.00025562326599649714,
      "loss": 3.3658,
      "step": 19550
    },
    {
      "epoch": 0.2563766482292186,
      "grad_norm": 0.49759840965270996,
      "learning_rate": 0.00025557911076761397,
      "loss": 3.3541,
      "step": 19560
    },
    {
      "epoch": 0.2565077201352662,
      "grad_norm": 0.5248744487762451,
      "learning_rate": 0.00025553493739995044,
      "loss": 3.3666,
      "step": 19570
    },
    {
      "epoch": 0.25663879204131385,
      "grad_norm": 0.5986993908882141,
      "learning_rate": 0.0002554907459010955,
      "loss": 3.3594,
      "step": 19580
    },
    {
      "epoch": 0.2567698639473615,
      "grad_norm": 0.5177679061889648,
      "learning_rate": 0.00025544653627864144,
      "loss": 3.391,
      "step": 19590
    },
    {
      "epoch": 0.2569009358534092,
      "grad_norm": 0.5337527990341187,
      "learning_rate": 0.0002554023085401836,
      "loss": 3.3424,
      "step": 19600
    },
    {
      "epoch": 0.25703200775945684,
      "grad_norm": 0.4825301170349121,
      "learning_rate": 0.00025535806269332044,
      "loss": 3.3926,
      "step": 19610
    },
    {
      "epoch": 0.2571630796655045,
      "grad_norm": 0.5543971657752991,
      "learning_rate": 0.00025531379874565345,
      "loss": 3.3943,
      "step": 19620
    },
    {
      "epoch": 0.25729415157155217,
      "grad_norm": 0.4479270279407501,
      "learning_rate": 0.0002552695167047873,
      "loss": 3.424,
      "step": 19630
    },
    {
      "epoch": 0.25742522347759983,
      "grad_norm": 0.4700687527656555,
      "learning_rate": 0.0002552252165783299,
      "loss": 3.3494,
      "step": 19640
    },
    {
      "epoch": 0.2575562953836475,
      "grad_norm": 0.5026695728302002,
      "learning_rate": 0.000255180898373892,
      "loss": 3.3622,
      "step": 19650
    },
    {
      "epoch": 0.2576873672896951,
      "grad_norm": 0.5069868564605713,
      "learning_rate": 0.00025513656209908764,
      "loss": 3.341,
      "step": 19660
    },
    {
      "epoch": 0.25781843919574277,
      "grad_norm": 0.502629280090332,
      "learning_rate": 0.00025509220776153384,
      "loss": 3.3761,
      "step": 19670
    },
    {
      "epoch": 0.25794951110179043,
      "grad_norm": 0.5309900641441345,
      "learning_rate": 0.0002550478353688509,
      "loss": 3.4294,
      "step": 19680
    },
    {
      "epoch": 0.2580805830078381,
      "grad_norm": 0.47560250759124756,
      "learning_rate": 0.0002550034449286621,
      "loss": 3.3844,
      "step": 19690
    },
    {
      "epoch": 0.25821165491388576,
      "grad_norm": 0.46928149461746216,
      "learning_rate": 0.0002549590364485938,
      "loss": 3.3845,
      "step": 19700
    },
    {
      "epoch": 0.2583427268199334,
      "grad_norm": 0.5022599101066589,
      "learning_rate": 0.00025491460993627555,
      "loss": 3.3958,
      "step": 19710
    },
    {
      "epoch": 0.2584737987259811,
      "grad_norm": 0.5018314123153687,
      "learning_rate": 0.0002548701653993398,
      "loss": 3.3173,
      "step": 19720
    },
    {
      "epoch": 0.25860487063202875,
      "grad_norm": 0.7498799562454224,
      "learning_rate": 0.00025482570284542246,
      "loss": 3.3885,
      "step": 19730
    },
    {
      "epoch": 0.2587359425380764,
      "grad_norm": 0.5063115358352661,
      "learning_rate": 0.00025478122228216217,
      "loss": 3.3426,
      "step": 19740
    },
    {
      "epoch": 0.258867014444124,
      "grad_norm": 0.4781665503978729,
      "learning_rate": 0.0002547367237172009,
      "loss": 3.4264,
      "step": 19750
    },
    {
      "epoch": 0.2589980863501717,
      "grad_norm": 0.5142199993133545,
      "learning_rate": 0.0002546922071581835,
      "loss": 3.4084,
      "step": 19760
    },
    {
      "epoch": 0.25912915825621935,
      "grad_norm": 0.5027515888214111,
      "learning_rate": 0.0002546476726127582,
      "loss": 3.3759,
      "step": 19770
    },
    {
      "epoch": 0.259260230162267,
      "grad_norm": 0.5489360690116882,
      "learning_rate": 0.000254603120088576,
      "loss": 3.4007,
      "step": 19780
    },
    {
      "epoch": 0.2593913020683147,
      "grad_norm": 0.4964730739593506,
      "learning_rate": 0.00025455854959329123,
      "loss": 3.4307,
      "step": 19790
    },
    {
      "epoch": 0.25952237397436234,
      "grad_norm": 0.4799068570137024,
      "learning_rate": 0.00025451396113456126,
      "loss": 3.3818,
      "step": 19800
    },
    {
      "epoch": 0.25965344588041,
      "grad_norm": 0.49567481875419617,
      "learning_rate": 0.0002544693547200465,
      "loss": 3.3557,
      "step": 19810
    },
    {
      "epoch": 0.25978451778645767,
      "grad_norm": 0.49198588728904724,
      "learning_rate": 0.0002544247303574103,
      "loss": 3.4057,
      "step": 19820
    },
    {
      "epoch": 0.25991558969250533,
      "grad_norm": 0.5088213086128235,
      "learning_rate": 0.00025438008805431944,
      "loss": 3.3798,
      "step": 19830
    },
    {
      "epoch": 0.26004666159855294,
      "grad_norm": 0.4931320250034332,
      "learning_rate": 0.00025433542781844344,
      "loss": 3.3643,
      "step": 19840
    },
    {
      "epoch": 0.2601777335046006,
      "grad_norm": 0.5095168352127075,
      "learning_rate": 0.0002542907496574552,
      "loss": 3.318,
      "step": 19850
    },
    {
      "epoch": 0.26030880541064827,
      "grad_norm": 0.4951655864715576,
      "learning_rate": 0.0002542460535790305,
      "loss": 3.399,
      "step": 19860
    },
    {
      "epoch": 0.26043987731669593,
      "grad_norm": 0.5496547818183899,
      "learning_rate": 0.00025420133959084813,
      "loss": 3.4608,
      "step": 19870
    },
    {
      "epoch": 0.2605709492227436,
      "grad_norm": 0.4908362030982971,
      "learning_rate": 0.00025415660770059027,
      "loss": 3.3989,
      "step": 19880
    },
    {
      "epoch": 0.26070202112879126,
      "grad_norm": 0.5039595365524292,
      "learning_rate": 0.0002541118579159418,
      "loss": 3.3964,
      "step": 19890
    },
    {
      "epoch": 0.2608330930348389,
      "grad_norm": 0.4787905812263489,
      "learning_rate": 0.00025406709024459095,
      "loss": 3.4117,
      "step": 19900
    },
    {
      "epoch": 0.2609641649408866,
      "grad_norm": 0.5055891871452332,
      "learning_rate": 0.00025402230469422896,
      "loss": 3.3542,
      "step": 19910
    },
    {
      "epoch": 0.26109523684693425,
      "grad_norm": 0.5189982652664185,
      "learning_rate": 0.00025397750127255,
      "loss": 3.359,
      "step": 19920
    },
    {
      "epoch": 0.26122630875298186,
      "grad_norm": 0.5714228749275208,
      "learning_rate": 0.0002539326799872515,
      "loss": 3.3808,
      "step": 19930
    },
    {
      "epoch": 0.2613573806590295,
      "grad_norm": 0.5211257934570312,
      "learning_rate": 0.0002538878408460339,
      "loss": 3.3588,
      "step": 19940
    },
    {
      "epoch": 0.2614884525650772,
      "grad_norm": 0.48032626509666443,
      "learning_rate": 0.0002538429838566007,
      "loss": 3.4149,
      "step": 19950
    },
    {
      "epoch": 0.26161952447112485,
      "grad_norm": 0.4825182855129242,
      "learning_rate": 0.0002537981090266583,
      "loss": 3.4104,
      "step": 19960
    },
    {
      "epoch": 0.2617505963771725,
      "grad_norm": 0.515887975692749,
      "learning_rate": 0.0002537532163639165,
      "loss": 3.405,
      "step": 19970
    },
    {
      "epoch": 0.2618816682832202,
      "grad_norm": 0.50924152135849,
      "learning_rate": 0.00025370830587608786,
      "loss": 3.4044,
      "step": 19980
    },
    {
      "epoch": 0.26201274018926785,
      "grad_norm": 0.4855426251888275,
      "learning_rate": 0.0002536633775708882,
      "loss": 3.3744,
      "step": 19990
    },
    {
      "epoch": 0.2621438120953155,
      "grad_norm": 0.4671946167945862,
      "learning_rate": 0.0002536184314560363,
      "loss": 3.3883,
      "step": 20000
    },
    {
      "epoch": 0.2621438120953155,
      "eval_accuracy": 0.3844364613880743,
      "eval_loss": 3.3576464653015137,
      "eval_runtime": 60.0593,
      "eval_samples_per_second": 33.3,
      "eval_steps_per_second": 0.533,
      "step": 20000
    },
    {
      "epoch": 0.2622748840013632,
      "grad_norm": 0.47316285967826843,
      "learning_rate": 0.00025357346753925405,
      "loss": 3.3579,
      "step": 20010
    },
    {
      "epoch": 0.2624059559074108,
      "grad_norm": 0.5318315029144287,
      "learning_rate": 0.0002535284858282663,
      "loss": 3.3494,
      "step": 20020
    },
    {
      "epoch": 0.26253702781345845,
      "grad_norm": 0.5010706186294556,
      "learning_rate": 0.0002534834863308011,
      "loss": 3.3757,
      "step": 20030
    },
    {
      "epoch": 0.2626680997195061,
      "grad_norm": 0.5707141757011414,
      "learning_rate": 0.0002534384690545894,
      "loss": 3.3622,
      "step": 20040
    },
    {
      "epoch": 0.2627991716255538,
      "grad_norm": 0.5315442085266113,
      "learning_rate": 0.00025339343400736545,
      "loss": 3.3712,
      "step": 20050
    },
    {
      "epoch": 0.26293024353160144,
      "grad_norm": 0.5701140761375427,
      "learning_rate": 0.00025334838119686626,
      "loss": 3.3868,
      "step": 20060
    },
    {
      "epoch": 0.2630613154376491,
      "grad_norm": 0.4901539087295532,
      "learning_rate": 0.0002533033106308321,
      "loss": 3.3429,
      "step": 20070
    },
    {
      "epoch": 0.26319238734369677,
      "grad_norm": 0.5093066692352295,
      "learning_rate": 0.00025325822231700613,
      "loss": 3.3706,
      "step": 20080
    },
    {
      "epoch": 0.26332345924974443,
      "grad_norm": 0.46033746004104614,
      "learning_rate": 0.00025321311626313477,
      "loss": 3.3728,
      "step": 20090
    },
    {
      "epoch": 0.2634545311557921,
      "grad_norm": 0.4557162821292877,
      "learning_rate": 0.00025316799247696725,
      "loss": 3.3724,
      "step": 20100
    },
    {
      "epoch": 0.2635856030618397,
      "grad_norm": 0.5097328424453735,
      "learning_rate": 0.000253122850966256,
      "loss": 3.3456,
      "step": 20110
    },
    {
      "epoch": 0.26371667496788737,
      "grad_norm": 0.5740048885345459,
      "learning_rate": 0.0002530776917387565,
      "loss": 3.4153,
      "step": 20120
    },
    {
      "epoch": 0.26384774687393503,
      "grad_norm": 0.5523825883865356,
      "learning_rate": 0.00025303251480222715,
      "loss": 3.3674,
      "step": 20130
    },
    {
      "epoch": 0.2639788187799827,
      "grad_norm": 0.5145972967147827,
      "learning_rate": 0.00025298732016442956,
      "loss": 3.3635,
      "step": 20140
    },
    {
      "epoch": 0.26410989068603036,
      "grad_norm": 0.7871749997138977,
      "learning_rate": 0.0002529421078331282,
      "loss": 3.3519,
      "step": 20150
    },
    {
      "epoch": 0.264240962592078,
      "grad_norm": 0.479594886302948,
      "learning_rate": 0.0002528968778160907,
      "loss": 3.4041,
      "step": 20160
    },
    {
      "epoch": 0.2643720344981257,
      "grad_norm": 0.502040684223175,
      "learning_rate": 0.00025285163012108775,
      "loss": 3.3556,
      "step": 20170
    },
    {
      "epoch": 0.26450310640417335,
      "grad_norm": 0.5546539425849915,
      "learning_rate": 0.000252806364755893,
      "loss": 3.421,
      "step": 20180
    },
    {
      "epoch": 0.264634178310221,
      "grad_norm": 0.5052618384361267,
      "learning_rate": 0.00025276108172828316,
      "loss": 3.3754,
      "step": 20190
    },
    {
      "epoch": 0.2647652502162686,
      "grad_norm": 1.4250965118408203,
      "learning_rate": 0.00025271578104603794,
      "loss": 3.421,
      "step": 20200
    },
    {
      "epoch": 0.2648963221223163,
      "grad_norm": 0.723242461681366,
      "learning_rate": 0.00025267046271694015,
      "loss": 3.4499,
      "step": 20210
    },
    {
      "epoch": 0.26502739402836395,
      "grad_norm": 0.49201953411102295,
      "learning_rate": 0.0002526251267487756,
      "loss": 3.3751,
      "step": 20220
    },
    {
      "epoch": 0.2651584659344116,
      "grad_norm": 0.5864336490631104,
      "learning_rate": 0.00025257977314933325,
      "loss": 3.4293,
      "step": 20230
    },
    {
      "epoch": 0.2652895378404593,
      "grad_norm": 0.4996050000190735,
      "learning_rate": 0.0002525344019264048,
      "loss": 3.4135,
      "step": 20240
    },
    {
      "epoch": 0.26542060974650694,
      "grad_norm": 0.6004931926727295,
      "learning_rate": 0.0002524890130877852,
      "loss": 3.3747,
      "step": 20250
    },
    {
      "epoch": 0.2655516816525546,
      "grad_norm": 0.5028265714645386,
      "learning_rate": 0.0002524436066412724,
      "loss": 3.3416,
      "step": 20260
    },
    {
      "epoch": 0.26568275355860227,
      "grad_norm": 0.5191029906272888,
      "learning_rate": 0.00025239818259466734,
      "loss": 3.3809,
      "step": 20270
    },
    {
      "epoch": 0.26581382546464993,
      "grad_norm": 0.49774888157844543,
      "learning_rate": 0.00025235274095577407,
      "loss": 3.3757,
      "step": 20280
    },
    {
      "epoch": 0.26594489737069754,
      "grad_norm": 0.4886395335197449,
      "learning_rate": 0.0002523072817323994,
      "loss": 3.4093,
      "step": 20290
    },
    {
      "epoch": 0.2660759692767452,
      "grad_norm": 0.5336122512817383,
      "learning_rate": 0.00025226180493235356,
      "loss": 3.3749,
      "step": 20300
    },
    {
      "epoch": 0.26620704118279287,
      "grad_norm": 0.48177069425582886,
      "learning_rate": 0.00025221631056344945,
      "loss": 3.3987,
      "step": 20310
    },
    {
      "epoch": 0.26633811308884053,
      "grad_norm": 0.5554938316345215,
      "learning_rate": 0.0002521707986335032,
      "loss": 3.4013,
      "step": 20320
    },
    {
      "epoch": 0.2664691849948882,
      "grad_norm": 0.5745668411254883,
      "learning_rate": 0.00025212526915033387,
      "loss": 3.4073,
      "step": 20330
    },
    {
      "epoch": 0.26660025690093586,
      "grad_norm": 0.4888208210468292,
      "learning_rate": 0.00025207972212176353,
      "loss": 3.3564,
      "step": 20340
    },
    {
      "epoch": 0.2667313288069835,
      "grad_norm": 0.5924863815307617,
      "learning_rate": 0.00025203415755561725,
      "loss": 3.3591,
      "step": 20350
    },
    {
      "epoch": 0.2668624007130312,
      "grad_norm": 0.48223140835762024,
      "learning_rate": 0.00025198857545972325,
      "loss": 3.4069,
      "step": 20360
    },
    {
      "epoch": 0.26699347261907885,
      "grad_norm": 0.5194159746170044,
      "learning_rate": 0.00025194297584191255,
      "loss": 3.3763,
      "step": 20370
    },
    {
      "epoch": 0.26712454452512646,
      "grad_norm": 0.5110407471656799,
      "learning_rate": 0.00025189735871001943,
      "loss": 3.3702,
      "step": 20380
    },
    {
      "epoch": 0.2672556164311741,
      "grad_norm": 0.5254484415054321,
      "learning_rate": 0.00025185172407188083,
      "loss": 3.3706,
      "step": 20390
    },
    {
      "epoch": 0.2673866883372218,
      "grad_norm": 0.5208966135978699,
      "learning_rate": 0.00025180607193533706,
      "loss": 3.3917,
      "step": 20400
    },
    {
      "epoch": 0.26751776024326945,
      "grad_norm": 0.5291957259178162,
      "learning_rate": 0.0002517604023082313,
      "loss": 3.3954,
      "step": 20410
    },
    {
      "epoch": 0.2676488321493171,
      "grad_norm": 0.5037896633148193,
      "learning_rate": 0.0002517147151984096,
      "loss": 3.3443,
      "step": 20420
    },
    {
      "epoch": 0.2677799040553648,
      "grad_norm": 0.5346311926841736,
      "learning_rate": 0.0002516690106137212,
      "loss": 3.3648,
      "step": 20430
    },
    {
      "epoch": 0.26791097596141245,
      "grad_norm": 0.5761136412620544,
      "learning_rate": 0.0002516232885620183,
      "loss": 3.4099,
      "step": 20440
    },
    {
      "epoch": 0.2680420478674601,
      "grad_norm": 0.5631494522094727,
      "learning_rate": 0.000251577549051156,
      "loss": 3.4035,
      "step": 20450
    },
    {
      "epoch": 0.2681731197735078,
      "grad_norm": 0.47001805901527405,
      "learning_rate": 0.00025153179208899253,
      "loss": 3.3387,
      "step": 20460
    },
    {
      "epoch": 0.2683041916795554,
      "grad_norm": 0.5405745506286621,
      "learning_rate": 0.000251486017683389,
      "loss": 3.387,
      "step": 20470
    },
    {
      "epoch": 0.26843526358560305,
      "grad_norm": 0.48674580454826355,
      "learning_rate": 0.00025144022584220975,
      "loss": 3.352,
      "step": 20480
    },
    {
      "epoch": 0.2685663354916507,
      "grad_norm": 0.4968830943107605,
      "learning_rate": 0.0002513944165733217,
      "loss": 3.4127,
      "step": 20490
    },
    {
      "epoch": 0.2686974073976984,
      "grad_norm": 0.49799424409866333,
      "learning_rate": 0.00025134858988459514,
      "loss": 3.3688,
      "step": 20500
    },
    {
      "epoch": 0.2686974073976984,
      "eval_accuracy": 0.3841925708699902,
      "eval_loss": 3.356732130050659,
      "eval_runtime": 60.433,
      "eval_samples_per_second": 33.094,
      "eval_steps_per_second": 0.53,
      "step": 20500
    },
    {
      "epoch": 0.26882847930374604,
      "grad_norm": 0.5616209506988525,
      "learning_rate": 0.00025130274578390323,
      "loss": 3.367,
      "step": 20510
    },
    {
      "epoch": 0.2689595512097937,
      "grad_norm": 0.5121413469314575,
      "learning_rate": 0.00025125688427912203,
      "loss": 3.393,
      "step": 20520
    },
    {
      "epoch": 0.26909062311584137,
      "grad_norm": 0.6632899045944214,
      "learning_rate": 0.0002512110053781308,
      "loss": 3.3802,
      "step": 20530
    },
    {
      "epoch": 0.26922169502188903,
      "grad_norm": 0.4853089153766632,
      "learning_rate": 0.0002511651090888116,
      "loss": 3.3982,
      "step": 20540
    },
    {
      "epoch": 0.2693527669279367,
      "grad_norm": 0.518538773059845,
      "learning_rate": 0.0002511191954190495,
      "loss": 3.3647,
      "step": 20550
    },
    {
      "epoch": 0.2694838388339843,
      "grad_norm": 0.4814436435699463,
      "learning_rate": 0.00025107326437673257,
      "loss": 3.3676,
      "step": 20560
    },
    {
      "epoch": 0.26961491074003197,
      "grad_norm": 0.5380482077598572,
      "learning_rate": 0.000251027315969752,
      "loss": 3.4028,
      "step": 20570
    },
    {
      "epoch": 0.26974598264607963,
      "grad_norm": 0.5780704021453857,
      "learning_rate": 0.00025098135020600173,
      "loss": 3.4215,
      "step": 20580
    },
    {
      "epoch": 0.2698770545521273,
      "grad_norm": 0.5185856223106384,
      "learning_rate": 0.0002509353670933789,
      "loss": 3.3457,
      "step": 20590
    },
    {
      "epoch": 0.27000812645817496,
      "grad_norm": 0.5243133902549744,
      "learning_rate": 0.0002508893666397835,
      "loss": 3.3886,
      "step": 20600
    },
    {
      "epoch": 0.2701391983642226,
      "grad_norm": 0.5258671641349792,
      "learning_rate": 0.0002508433488531186,
      "loss": 3.3522,
      "step": 20610
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 0.545569121837616,
      "learning_rate": 0.0002507973137412901,
      "loss": 3.3869,
      "step": 20620
    },
    {
      "epoch": 0.27040134217631795,
      "grad_norm": 0.5036454796791077,
      "learning_rate": 0.0002507512613122069,
      "loss": 3.3193,
      "step": 20630
    },
    {
      "epoch": 0.2705324140823656,
      "grad_norm": 0.5290234088897705,
      "learning_rate": 0.0002507051915737811,
      "loss": 3.3942,
      "step": 20640
    },
    {
      "epoch": 0.2706634859884132,
      "grad_norm": 0.5274523496627808,
      "learning_rate": 0.0002506591045339274,
      "loss": 3.3925,
      "step": 20650
    },
    {
      "epoch": 0.2707945578944609,
      "grad_norm": 0.5356395840644836,
      "learning_rate": 0.00025061300020056386,
      "loss": 3.3518,
      "step": 20660
    },
    {
      "epoch": 0.27092562980050855,
      "grad_norm": 0.5914761424064636,
      "learning_rate": 0.0002505668785816113,
      "loss": 3.4192,
      "step": 20670
    },
    {
      "epoch": 0.2710567017065562,
      "grad_norm": 0.6294680833816528,
      "learning_rate": 0.0002505207396849935,
      "loss": 3.4405,
      "step": 20680
    },
    {
      "epoch": 0.2711877736126039,
      "grad_norm": 0.46257102489471436,
      "learning_rate": 0.0002504745835186372,
      "loss": 3.3704,
      "step": 20690
    },
    {
      "epoch": 0.27131884551865154,
      "grad_norm": 0.5353371500968933,
      "learning_rate": 0.00025042841009047224,
      "loss": 3.3939,
      "step": 20700
    },
    {
      "epoch": 0.2714499174246992,
      "grad_norm": 2.0056369304656982,
      "learning_rate": 0.0002503822194084313,
      "loss": 3.3877,
      "step": 20710
    },
    {
      "epoch": 0.27158098933074687,
      "grad_norm": 0.5806504487991333,
      "learning_rate": 0.0002503360114804501,
      "loss": 3.3946,
      "step": 20720
    },
    {
      "epoch": 0.27171206123679453,
      "grad_norm": 0.5499182939529419,
      "learning_rate": 0.00025028978631446726,
      "loss": 3.3953,
      "step": 20730
    },
    {
      "epoch": 0.27184313314284214,
      "grad_norm": 0.5258299708366394,
      "learning_rate": 0.00025024354391842444,
      "loss": 3.3719,
      "step": 20740
    },
    {
      "epoch": 0.2719742050488898,
      "grad_norm": 0.47831329703330994,
      "learning_rate": 0.0002501972843002661,
      "loss": 3.3675,
      "step": 20750
    },
    {
      "epoch": 0.27210527695493747,
      "grad_norm": 0.47481948137283325,
      "learning_rate": 0.0002501510074679399,
      "loss": 3.3506,
      "step": 20760
    },
    {
      "epoch": 0.27223634886098513,
      "grad_norm": 0.4736486077308655,
      "learning_rate": 0.00025010471342939625,
      "loss": 3.3695,
      "step": 20770
    },
    {
      "epoch": 0.2723674207670328,
      "grad_norm": 0.5486845374107361,
      "learning_rate": 0.0002500584021925886,
      "loss": 3.4127,
      "step": 20780
    },
    {
      "epoch": 0.27249849267308046,
      "grad_norm": 0.5462353229522705,
      "learning_rate": 0.0002500120737654733,
      "loss": 3.3216,
      "step": 20790
    },
    {
      "epoch": 0.2726295645791281,
      "grad_norm": 0.5956313014030457,
      "learning_rate": 0.0002499657281560098,
      "loss": 3.3532,
      "step": 20800
    },
    {
      "epoch": 0.2727606364851758,
      "grad_norm": 0.5843344926834106,
      "learning_rate": 0.0002499193653721604,
      "loss": 3.4055,
      "step": 20810
    },
    {
      "epoch": 0.2728917083912234,
      "grad_norm": 0.5147221088409424,
      "learning_rate": 0.0002498729854218903,
      "loss": 3.3521,
      "step": 20820
    },
    {
      "epoch": 0.27302278029727106,
      "grad_norm": 0.515502393245697,
      "learning_rate": 0.0002498265883131677,
      "loss": 3.3726,
      "step": 20830
    },
    {
      "epoch": 0.2731538522033187,
      "grad_norm": 0.5350995063781738,
      "learning_rate": 0.0002497801740539637,
      "loss": 3.3904,
      "step": 20840
    },
    {
      "epoch": 0.2732849241093664,
      "grad_norm": 0.5171495676040649,
      "learning_rate": 0.0002497337426522525,
      "loss": 3.3498,
      "step": 20850
    },
    {
      "epoch": 0.27341599601541405,
      "grad_norm": 0.5692862868309021,
      "learning_rate": 0.00024968729411601107,
      "loss": 3.4248,
      "step": 20860
    },
    {
      "epoch": 0.2735470679214617,
      "grad_norm": 0.7315946817398071,
      "learning_rate": 0.00024964082845321947,
      "loss": 3.3989,
      "step": 20870
    },
    {
      "epoch": 0.2736781398275094,
      "grad_norm": 0.5175470113754272,
      "learning_rate": 0.0002495943456718606,
      "loss": 3.3199,
      "step": 20880
    },
    {
      "epoch": 0.27380921173355705,
      "grad_norm": 0.47521892189979553,
      "learning_rate": 0.0002495478457799203,
      "loss": 3.3697,
      "step": 20890
    },
    {
      "epoch": 0.2739402836396047,
      "grad_norm": 0.6087992191314697,
      "learning_rate": 0.0002495013287853874,
      "loss": 3.3581,
      "step": 20900
    },
    {
      "epoch": 0.2740713555456523,
      "grad_norm": 0.5049447417259216,
      "learning_rate": 0.00024945479469625366,
      "loss": 3.3903,
      "step": 20910
    },
    {
      "epoch": 0.2742024274517,
      "grad_norm": 0.6380307674407959,
      "learning_rate": 0.0002494082435205137,
      "loss": 3.3863,
      "step": 20920
    },
    {
      "epoch": 0.27433349935774765,
      "grad_norm": 0.5244141817092896,
      "learning_rate": 0.00024936167526616517,
      "loss": 3.3562,
      "step": 20930
    },
    {
      "epoch": 0.2744645712637953,
      "grad_norm": 0.508036732673645,
      "learning_rate": 0.0002493150899412087,
      "loss": 3.4073,
      "step": 20940
    },
    {
      "epoch": 0.274595643169843,
      "grad_norm": 0.5478516221046448,
      "learning_rate": 0.00024926848755364765,
      "loss": 3.3755,
      "step": 20950
    },
    {
      "epoch": 0.27472671507589064,
      "grad_norm": 0.49120625853538513,
      "learning_rate": 0.00024922186811148856,
      "loss": 3.3912,
      "step": 20960
    },
    {
      "epoch": 0.2748577869819383,
      "grad_norm": 0.5242131948471069,
      "learning_rate": 0.0002491752316227407,
      "loss": 3.3637,
      "step": 20970
    },
    {
      "epoch": 0.27498885888798597,
      "grad_norm": 0.588809609413147,
      "learning_rate": 0.00024912857809541636,
      "loss": 3.3597,
      "step": 20980
    },
    {
      "epoch": 0.27511993079403363,
      "grad_norm": 0.523265540599823,
      "learning_rate": 0.0002490819075375308,
      "loss": 3.3987,
      "step": 20990
    },
    {
      "epoch": 0.27525100270008124,
      "grad_norm": 0.47958844900131226,
      "learning_rate": 0.000249035219957102,
      "loss": 3.4264,
      "step": 21000
    },
    {
      "epoch": 0.27525100270008124,
      "eval_accuracy": 0.38475464320625613,
      "eval_loss": 3.354015350341797,
      "eval_runtime": 59.5891,
      "eval_samples_per_second": 33.563,
      "eval_steps_per_second": 0.537,
      "step": 21000
    },
    {
      "epoch": 0.2753820746061289,
      "grad_norm": 0.512631356716156,
      "learning_rate": 0.0002489885153621512,
      "loss": 3.3347,
      "step": 21010
    },
    {
      "epoch": 0.27551314651217657,
      "grad_norm": 0.5187012553215027,
      "learning_rate": 0.00024894179376070227,
      "loss": 3.3321,
      "step": 21020
    },
    {
      "epoch": 0.27564421841822423,
      "grad_norm": 0.49427488446235657,
      "learning_rate": 0.00024889505516078215,
      "loss": 3.391,
      "step": 21030
    },
    {
      "epoch": 0.2757752903242719,
      "grad_norm": 0.5022109746932983,
      "learning_rate": 0.0002488482995704206,
      "loss": 3.3517,
      "step": 21040
    },
    {
      "epoch": 0.27590636223031956,
      "grad_norm": 0.5233657360076904,
      "learning_rate": 0.0002488015269976504,
      "loss": 3.3741,
      "step": 21050
    },
    {
      "epoch": 0.2760374341363672,
      "grad_norm": 0.5324937701225281,
      "learning_rate": 0.00024875473745050725,
      "loss": 3.3984,
      "step": 21060
    },
    {
      "epoch": 0.2761685060424149,
      "grad_norm": 0.5226995944976807,
      "learning_rate": 0.0002487079309370297,
      "loss": 3.3752,
      "step": 21070
    },
    {
      "epoch": 0.27629957794846255,
      "grad_norm": 0.5099854469299316,
      "learning_rate": 0.00024866110746525914,
      "loss": 3.375,
      "step": 21080
    },
    {
      "epoch": 0.27643064985451016,
      "grad_norm": 0.48504313826560974,
      "learning_rate": 0.00024861426704324003,
      "loss": 3.3929,
      "step": 21090
    },
    {
      "epoch": 0.2765617217605578,
      "grad_norm": 0.6852875351905823,
      "learning_rate": 0.00024856740967901976,
      "loss": 3.3798,
      "step": 21100
    },
    {
      "epoch": 0.2766927936666055,
      "grad_norm": 0.582165539264679,
      "learning_rate": 0.0002485205353806485,
      "loss": 3.3571,
      "step": 21110
    },
    {
      "epoch": 0.27682386557265315,
      "grad_norm": 0.5382540225982666,
      "learning_rate": 0.00024847364415617933,
      "loss": 3.4166,
      "step": 21120
    },
    {
      "epoch": 0.2769549374787008,
      "grad_norm": 0.6528135538101196,
      "learning_rate": 0.00024842673601366836,
      "loss": 3.4114,
      "step": 21130
    },
    {
      "epoch": 0.2770860093847485,
      "grad_norm": 0.5464576482772827,
      "learning_rate": 0.00024837981096117453,
      "loss": 3.401,
      "step": 21140
    },
    {
      "epoch": 0.27721708129079614,
      "grad_norm": 0.5541452169418335,
      "learning_rate": 0.00024833286900675965,
      "loss": 3.3502,
      "step": 21150
    },
    {
      "epoch": 0.2773481531968438,
      "grad_norm": 0.496295690536499,
      "learning_rate": 0.0002482859101584885,
      "loss": 3.3651,
      "step": 21160
    },
    {
      "epoch": 0.27747922510289147,
      "grad_norm": 0.5677711367607117,
      "learning_rate": 0.00024823893442442875,
      "loss": 3.4208,
      "step": 21170
    },
    {
      "epoch": 0.2776102970089391,
      "grad_norm": 0.5145048499107361,
      "learning_rate": 0.00024819194181265096,
      "loss": 3.3576,
      "step": 21180
    },
    {
      "epoch": 0.27774136891498674,
      "grad_norm": 0.5099654793739319,
      "learning_rate": 0.00024814493233122855,
      "loss": 3.3871,
      "step": 21190
    },
    {
      "epoch": 0.2778724408210344,
      "grad_norm": 0.5650095343589783,
      "learning_rate": 0.00024809790598823796,
      "loss": 3.3334,
      "step": 21200
    },
    {
      "epoch": 0.27800351272708207,
      "grad_norm": 0.7284607887268066,
      "learning_rate": 0.0002480508627917583,
      "loss": 3.3835,
      "step": 21210
    },
    {
      "epoch": 0.27813458463312973,
      "grad_norm": 0.5535664558410645,
      "learning_rate": 0.00024800380274987195,
      "loss": 3.3371,
      "step": 21220
    },
    {
      "epoch": 0.2782656565391774,
      "grad_norm": 0.48359134793281555,
      "learning_rate": 0.00024795672587066373,
      "loss": 3.3618,
      "step": 21230
    },
    {
      "epoch": 0.27839672844522506,
      "grad_norm": 0.5452717542648315,
      "learning_rate": 0.0002479096321622217,
      "loss": 3.3702,
      "step": 21240
    },
    {
      "epoch": 0.2785278003512727,
      "grad_norm": 0.5765952467918396,
      "learning_rate": 0.0002478625216326366,
      "loss": 3.3553,
      "step": 21250
    },
    {
      "epoch": 0.2786588722573204,
      "grad_norm": 0.5487689971923828,
      "learning_rate": 0.0002478153942900022,
      "loss": 3.4404,
      "step": 21260
    },
    {
      "epoch": 0.278789944163368,
      "grad_norm": 0.6246709227561951,
      "learning_rate": 0.0002477682501424152,
      "loss": 3.354,
      "step": 21270
    },
    {
      "epoch": 0.27892101606941566,
      "grad_norm": 0.5033740997314453,
      "learning_rate": 0.00024772108919797486,
      "loss": 3.3934,
      "step": 21280
    },
    {
      "epoch": 0.2790520879754633,
      "grad_norm": 0.4920714497566223,
      "learning_rate": 0.0002476739114647838,
      "loss": 3.4038,
      "step": 21290
    },
    {
      "epoch": 0.279183159881511,
      "grad_norm": 0.5298735499382019,
      "learning_rate": 0.00024762671695094715,
      "loss": 3.3207,
      "step": 21300
    },
    {
      "epoch": 0.27931423178755865,
      "grad_norm": 0.5849246382713318,
      "learning_rate": 0.0002475795056645731,
      "loss": 3.4242,
      "step": 21310
    },
    {
      "epoch": 0.2794453036936063,
      "grad_norm": 0.5534976720809937,
      "learning_rate": 0.00024753227761377264,
      "loss": 3.33,
      "step": 21320
    },
    {
      "epoch": 0.279576375599654,
      "grad_norm": 0.5064390301704407,
      "learning_rate": 0.0002474850328066597,
      "loss": 3.4047,
      "step": 21330
    },
    {
      "epoch": 0.27970744750570165,
      "grad_norm": 0.5181854367256165,
      "learning_rate": 0.0002474377712513511,
      "loss": 3.3678,
      "step": 21340
    },
    {
      "epoch": 0.2798385194117493,
      "grad_norm": 0.5151462554931641,
      "learning_rate": 0.0002473904929559665,
      "loss": 3.3742,
      "step": 21350
    },
    {
      "epoch": 0.2799695913177969,
      "grad_norm": 0.5190281867980957,
      "learning_rate": 0.0002473431979286284,
      "loss": 3.4225,
      "step": 21360
    },
    {
      "epoch": 0.2801006632238446,
      "grad_norm": 0.5449749231338501,
      "learning_rate": 0.0002472958861774622,
      "loss": 3.3777,
      "step": 21370
    },
    {
      "epoch": 0.28023173512989225,
      "grad_norm": 0.5760560631752014,
      "learning_rate": 0.00024724855771059633,
      "loss": 3.3319,
      "step": 21380
    },
    {
      "epoch": 0.2803628070359399,
      "grad_norm": 0.5367158055305481,
      "learning_rate": 0.00024720121253616175,
      "loss": 3.3795,
      "step": 21390
    },
    {
      "epoch": 0.2804938789419876,
      "grad_norm": 0.5466709136962891,
      "learning_rate": 0.0002471538506622926,
      "loss": 3.3739,
      "step": 21400
    },
    {
      "epoch": 0.28062495084803524,
      "grad_norm": 0.4981827437877655,
      "learning_rate": 0.0002471064720971258,
      "loss": 3.3584,
      "step": 21410
    },
    {
      "epoch": 0.2807560227540829,
      "grad_norm": 0.5874139070510864,
      "learning_rate": 0.000247059076848801,
      "loss": 3.401,
      "step": 21420
    },
    {
      "epoch": 0.28088709466013057,
      "grad_norm": 0.5961233377456665,
      "learning_rate": 0.000247011664925461,
      "loss": 3.4138,
      "step": 21430
    },
    {
      "epoch": 0.28101816656617823,
      "grad_norm": 0.5005521774291992,
      "learning_rate": 0.00024696423633525114,
      "loss": 3.3735,
      "step": 21440
    },
    {
      "epoch": 0.28114923847222584,
      "grad_norm": 0.5578511357307434,
      "learning_rate": 0.0002469167910863199,
      "loss": 3.3482,
      "step": 21450
    },
    {
      "epoch": 0.2812803103782735,
      "grad_norm": 0.5272665619850159,
      "learning_rate": 0.00024686932918681847,
      "loss": 3.3512,
      "step": 21460
    },
    {
      "epoch": 0.28141138228432117,
      "grad_norm": 0.580531895160675,
      "learning_rate": 0.0002468218506449009,
      "loss": 3.3675,
      "step": 21470
    },
    {
      "epoch": 0.28154245419036883,
      "grad_norm": 0.49747440218925476,
      "learning_rate": 0.0002467743554687241,
      "loss": 3.3932,
      "step": 21480
    },
    {
      "epoch": 0.2816735260964165,
      "grad_norm": 0.5736663937568665,
      "learning_rate": 0.000246726843666448,
      "loss": 3.3886,
      "step": 21490
    },
    {
      "epoch": 0.28180459800246416,
      "grad_norm": 0.4999784529209137,
      "learning_rate": 0.00024667931524623523,
      "loss": 3.3347,
      "step": 21500
    },
    {
      "epoch": 0.28180459800246416,
      "eval_accuracy": 0.3850791788856305,
      "eval_loss": 3.3516571521759033,
      "eval_runtime": 60.4406,
      "eval_samples_per_second": 33.09,
      "eval_steps_per_second": 0.529,
      "step": 21500
    },
    {
      "epoch": 0.2819356699085118,
      "grad_norm": 0.49472519755363464,
      "learning_rate": 0.00024663177021625116,
      "loss": 3.4098,
      "step": 21510
    },
    {
      "epoch": 0.2820667418145595,
      "grad_norm": 0.5019751191139221,
      "learning_rate": 0.0002465842085846643,
      "loss": 3.3648,
      "step": 21520
    },
    {
      "epoch": 0.28219781372060715,
      "grad_norm": 0.49514076113700867,
      "learning_rate": 0.00024653663035964583,
      "loss": 3.4214,
      "step": 21530
    },
    {
      "epoch": 0.28232888562665476,
      "grad_norm": 0.6444296836853027,
      "learning_rate": 0.0002464890355493698,
      "loss": 3.3574,
      "step": 21540
    },
    {
      "epoch": 0.2824599575327024,
      "grad_norm": 0.4834980070590973,
      "learning_rate": 0.00024644142416201315,
      "loss": 3.3572,
      "step": 21550
    },
    {
      "epoch": 0.2825910294387501,
      "grad_norm": 0.4926605522632599,
      "learning_rate": 0.00024639379620575565,
      "loss": 3.364,
      "step": 21560
    },
    {
      "epoch": 0.28272210134479775,
      "grad_norm": 0.4613964259624481,
      "learning_rate": 0.00024634615168877985,
      "loss": 3.3814,
      "step": 21570
    },
    {
      "epoch": 0.2828531732508454,
      "grad_norm": 0.5359572768211365,
      "learning_rate": 0.0002462984906192713,
      "loss": 3.3505,
      "step": 21580
    },
    {
      "epoch": 0.2829842451568931,
      "grad_norm": 0.5283089280128479,
      "learning_rate": 0.00024625081300541825,
      "loss": 3.3949,
      "step": 21590
    },
    {
      "epoch": 0.28311531706294074,
      "grad_norm": 0.5797305703163147,
      "learning_rate": 0.00024620311885541185,
      "loss": 3.3744,
      "step": 21600
    },
    {
      "epoch": 0.2832463889689884,
      "grad_norm": 0.5272244811058044,
      "learning_rate": 0.00024615540817744607,
      "loss": 3.4099,
      "step": 21610
    },
    {
      "epoch": 0.28337746087503607,
      "grad_norm": 0.5519600510597229,
      "learning_rate": 0.00024610768097971784,
      "loss": 3.341,
      "step": 21620
    },
    {
      "epoch": 0.2835085327810837,
      "grad_norm": 0.8035848736763,
      "learning_rate": 0.00024605993727042666,
      "loss": 3.392,
      "step": 21630
    },
    {
      "epoch": 0.28363960468713134,
      "grad_norm": 0.5701985359191895,
      "learning_rate": 0.00024601217705777517,
      "loss": 3.3907,
      "step": 21640
    },
    {
      "epoch": 0.283770676593179,
      "grad_norm": 0.649526834487915,
      "learning_rate": 0.0002459644003499686,
      "loss": 3.3794,
      "step": 21650
    },
    {
      "epoch": 0.28390174849922667,
      "grad_norm": 0.7397329211235046,
      "learning_rate": 0.0002459166071552152,
      "loss": 3.4101,
      "step": 21660
    },
    {
      "epoch": 0.28403282040527433,
      "grad_norm": 0.5522966980934143,
      "learning_rate": 0.00024586879748172594,
      "loss": 3.4061,
      "step": 21670
    },
    {
      "epoch": 0.284163892311322,
      "grad_norm": 0.6267199516296387,
      "learning_rate": 0.00024582097133771464,
      "loss": 3.3934,
      "step": 21680
    },
    {
      "epoch": 0.28429496421736966,
      "grad_norm": 0.5934063792228699,
      "learning_rate": 0.0002457731287313979,
      "loss": 3.3543,
      "step": 21690
    },
    {
      "epoch": 0.2844260361234173,
      "grad_norm": 0.4884076416492462,
      "learning_rate": 0.00024572526967099536,
      "loss": 3.3638,
      "step": 21700
    },
    {
      "epoch": 0.284557108029465,
      "grad_norm": 0.5215699672698975,
      "learning_rate": 0.00024567739416472926,
      "loss": 3.3406,
      "step": 21710
    },
    {
      "epoch": 0.2846881799355126,
      "grad_norm": 0.49937358498573303,
      "learning_rate": 0.00024562950222082474,
      "loss": 3.2851,
      "step": 21720
    },
    {
      "epoch": 0.28481925184156026,
      "grad_norm": 0.5448559522628784,
      "learning_rate": 0.0002455815938475098,
      "loss": 3.3474,
      "step": 21730
    },
    {
      "epoch": 0.2849503237476079,
      "grad_norm": 0.5504633188247681,
      "learning_rate": 0.0002455336690530151,
      "loss": 3.3283,
      "step": 21740
    },
    {
      "epoch": 0.2850813956536556,
      "grad_norm": 0.5158266425132751,
      "learning_rate": 0.00024548572784557443,
      "loss": 3.3604,
      "step": 21750
    },
    {
      "epoch": 0.28521246755970325,
      "grad_norm": 0.4921668469905853,
      "learning_rate": 0.0002454377702334241,
      "loss": 3.3778,
      "step": 21760
    },
    {
      "epoch": 0.2853435394657509,
      "grad_norm": 0.5138165354728699,
      "learning_rate": 0.0002453897962248035,
      "loss": 3.3576,
      "step": 21770
    },
    {
      "epoch": 0.2854746113717986,
      "grad_norm": 0.5623292326927185,
      "learning_rate": 0.0002453418058279545,
      "loss": 3.3532,
      "step": 21780
    },
    {
      "epoch": 0.28560568327784625,
      "grad_norm": 0.5229595899581909,
      "learning_rate": 0.0002452937990511221,
      "loss": 3.3467,
      "step": 21790
    },
    {
      "epoch": 0.2857367551838939,
      "grad_norm": 0.5152409076690674,
      "learning_rate": 0.000245245775902554,
      "loss": 3.3998,
      "step": 21800
    },
    {
      "epoch": 0.2858678270899415,
      "grad_norm": 0.49322009086608887,
      "learning_rate": 0.0002451977363905007,
      "loss": 3.3136,
      "step": 21810
    },
    {
      "epoch": 0.2859988989959892,
      "grad_norm": 0.505225419998169,
      "learning_rate": 0.0002451496805232155,
      "loss": 3.3487,
      "step": 21820
    },
    {
      "epoch": 0.28612997090203685,
      "grad_norm": 0.5683316588401794,
      "learning_rate": 0.00024510160830895443,
      "loss": 3.3656,
      "step": 21830
    },
    {
      "epoch": 0.2862610428080845,
      "grad_norm": 0.544196605682373,
      "learning_rate": 0.00024505351975597666,
      "loss": 3.3514,
      "step": 21840
    },
    {
      "epoch": 0.2863921147141322,
      "grad_norm": 0.647071897983551,
      "learning_rate": 0.0002450054148725438,
      "loss": 3.3201,
      "step": 21850
    },
    {
      "epoch": 0.28652318662017984,
      "grad_norm": 0.6096830368041992,
      "learning_rate": 0.0002449572936669203,
      "loss": 3.3659,
      "step": 21860
    },
    {
      "epoch": 0.2866542585262275,
      "grad_norm": 0.5538503527641296,
      "learning_rate": 0.0002449091561473737,
      "loss": 3.3428,
      "step": 21870
    },
    {
      "epoch": 0.28678533043227517,
      "grad_norm": 0.4964708685874939,
      "learning_rate": 0.0002448610023221741,
      "loss": 3.3363,
      "step": 21880
    },
    {
      "epoch": 0.28691640233832283,
      "grad_norm": 0.6618576049804688,
      "learning_rate": 0.00024481283219959435,
      "loss": 3.3533,
      "step": 21890
    },
    {
      "epoch": 0.28704747424437044,
      "grad_norm": 0.5311054587364197,
      "learning_rate": 0.0002447646457879104,
      "loss": 3.343,
      "step": 21900
    },
    {
      "epoch": 0.2871785461504181,
      "grad_norm": 0.8017573952674866,
      "learning_rate": 0.00024471644309540067,
      "loss": 3.3615,
      "step": 21910
    },
    {
      "epoch": 0.28730961805646577,
      "grad_norm": 0.5409725904464722,
      "learning_rate": 0.0002446682241303465,
      "loss": 3.3583,
      "step": 21920
    },
    {
      "epoch": 0.28744068996251343,
      "grad_norm": 0.5313457250595093,
      "learning_rate": 0.00024461998890103214,
      "loss": 3.3477,
      "step": 21930
    },
    {
      "epoch": 0.2875717618685611,
      "grad_norm": 0.5599802136421204,
      "learning_rate": 0.00024457173741574453,
      "loss": 3.3411,
      "step": 21940
    },
    {
      "epoch": 0.28770283377460876,
      "grad_norm": 0.5103743672370911,
      "learning_rate": 0.00024452346968277327,
      "loss": 3.376,
      "step": 21950
    },
    {
      "epoch": 0.2878339056806564,
      "grad_norm": 0.5034699440002441,
      "learning_rate": 0.00024447518571041104,
      "loss": 3.3169,
      "step": 21960
    },
    {
      "epoch": 0.2879649775867041,
      "grad_norm": 0.504654586315155,
      "learning_rate": 0.0002444268855069531,
      "loss": 3.3186,
      "step": 21970
    },
    {
      "epoch": 0.28809604949275175,
      "grad_norm": 0.54386967420578,
      "learning_rate": 0.0002443785690806975,
      "loss": 3.3423,
      "step": 21980
    },
    {
      "epoch": 0.28822712139879936,
      "grad_norm": 0.5279662609100342,
      "learning_rate": 0.0002443302364399453,
      "loss": 3.308,
      "step": 21990
    },
    {
      "epoch": 0.288358193304847,
      "grad_norm": 0.5992744565010071,
      "learning_rate": 0.00024428188759299997,
      "loss": 3.3716,
      "step": 22000
    },
    {
      "epoch": 0.288358193304847,
      "eval_accuracy": 0.38473362658846527,
      "eval_loss": 3.350407123565674,
      "eval_runtime": 59.9742,
      "eval_samples_per_second": 33.348,
      "eval_steps_per_second": 0.534,
      "step": 22000
    },
    {
      "epoch": 0.2884892652108947,
      "grad_norm": 0.4905395805835724,
      "learning_rate": 0.00024423352254816815,
      "loss": 3.3225,
      "step": 22010
    },
    {
      "epoch": 0.28862033711694235,
      "grad_norm": 0.5593045949935913,
      "learning_rate": 0.00024418514131375893,
      "loss": 3.3636,
      "step": 22020
    },
    {
      "epoch": 0.28875140902299,
      "grad_norm": 0.535254955291748,
      "learning_rate": 0.0002441367438980845,
      "loss": 3.3714,
      "step": 22030
    },
    {
      "epoch": 0.2888824809290377,
      "grad_norm": 0.5266671180725098,
      "learning_rate": 0.00024408833030945958,
      "loss": 3.3216,
      "step": 22040
    },
    {
      "epoch": 0.28901355283508534,
      "grad_norm": 0.5320093631744385,
      "learning_rate": 0.00024403990055620175,
      "loss": 3.341,
      "step": 22050
    },
    {
      "epoch": 0.289144624741133,
      "grad_norm": 0.5077747106552124,
      "learning_rate": 0.00024399145464663137,
      "loss": 3.3516,
      "step": 22060
    },
    {
      "epoch": 0.28927569664718067,
      "grad_norm": 0.5425472259521484,
      "learning_rate": 0.00024394299258907165,
      "loss": 3.3955,
      "step": 22070
    },
    {
      "epoch": 0.2894067685532283,
      "grad_norm": 0.49999740719795227,
      "learning_rate": 0.0002438945143918484,
      "loss": 3.349,
      "step": 22080
    },
    {
      "epoch": 0.28953784045927594,
      "grad_norm": 0.4677477478981018,
      "learning_rate": 0.0002438460200632904,
      "loss": 3.339,
      "step": 22090
    },
    {
      "epoch": 0.2896689123653236,
      "grad_norm": 0.5178941488265991,
      "learning_rate": 0.00024379750961172903,
      "loss": 3.3513,
      "step": 22100
    },
    {
      "epoch": 0.28979998427137127,
      "grad_norm": 0.49891024827957153,
      "learning_rate": 0.00024374898304549864,
      "loss": 3.3377,
      "step": 22110
    },
    {
      "epoch": 0.28993105617741893,
      "grad_norm": 0.5194788575172424,
      "learning_rate": 0.0002437004403729361,
      "loss": 3.3006,
      "step": 22120
    },
    {
      "epoch": 0.2900621280834666,
      "grad_norm": 0.5128890872001648,
      "learning_rate": 0.00024365188160238121,
      "loss": 3.357,
      "step": 22130
    },
    {
      "epoch": 0.29019319998951426,
      "grad_norm": 0.5245267748832703,
      "learning_rate": 0.00024360330674217653,
      "loss": 3.3377,
      "step": 22140
    },
    {
      "epoch": 0.2903242718955619,
      "grad_norm": 0.5327003002166748,
      "learning_rate": 0.00024355471580066732,
      "loss": 3.3114,
      "step": 22150
    },
    {
      "epoch": 0.2904553438016096,
      "grad_norm": 0.5569054484367371,
      "learning_rate": 0.0002435061087862017,
      "loss": 3.3636,
      "step": 22160
    },
    {
      "epoch": 0.2905864157076572,
      "grad_norm": 0.48410385847091675,
      "learning_rate": 0.00024345748570713037,
      "loss": 3.3399,
      "step": 22170
    },
    {
      "epoch": 0.29071748761370486,
      "grad_norm": 0.4819527566432953,
      "learning_rate": 0.00024340884657180704,
      "loss": 3.296,
      "step": 22180
    },
    {
      "epoch": 0.2908485595197525,
      "grad_norm": 0.554044783115387,
      "learning_rate": 0.00024336019138858792,
      "loss": 3.3663,
      "step": 22190
    },
    {
      "epoch": 0.2909796314258002,
      "grad_norm": 0.5095894932746887,
      "learning_rate": 0.00024331152016583224,
      "loss": 3.395,
      "step": 22200
    },
    {
      "epoch": 0.29111070333184785,
      "grad_norm": 0.5078835487365723,
      "learning_rate": 0.00024326283291190172,
      "loss": 3.3099,
      "step": 22210
    },
    {
      "epoch": 0.2912417752378955,
      "grad_norm": 0.5540070533752441,
      "learning_rate": 0.00024321412963516106,
      "loss": 3.3648,
      "step": 22220
    },
    {
      "epoch": 0.2913728471439432,
      "grad_norm": 0.5076451301574707,
      "learning_rate": 0.00024316541034397754,
      "loss": 3.3693,
      "step": 22230
    },
    {
      "epoch": 0.29150391904999085,
      "grad_norm": 0.48738306760787964,
      "learning_rate": 0.00024311667504672136,
      "loss": 3.3628,
      "step": 22240
    },
    {
      "epoch": 0.2916349909560385,
      "grad_norm": 0.5313220620155334,
      "learning_rate": 0.00024306792375176532,
      "loss": 3.312,
      "step": 22250
    },
    {
      "epoch": 0.2917660628620861,
      "grad_norm": 0.5386406779289246,
      "learning_rate": 0.000243019156467485,
      "loss": 3.3387,
      "step": 22260
    },
    {
      "epoch": 0.2918971347681338,
      "grad_norm": 0.5158824324607849,
      "learning_rate": 0.0002429703732022588,
      "loss": 3.3689,
      "step": 22270
    },
    {
      "epoch": 0.29202820667418145,
      "grad_norm": 0.9790822267532349,
      "learning_rate": 0.00024292157396446775,
      "loss": 3.414,
      "step": 22280
    },
    {
      "epoch": 0.2921592785802291,
      "grad_norm": 0.5582848191261292,
      "learning_rate": 0.0002428727587624958,
      "loss": 3.3325,
      "step": 22290
    },
    {
      "epoch": 0.2922903504862768,
      "grad_norm": 0.55227130651474,
      "learning_rate": 0.00024282392760472947,
      "loss": 3.3548,
      "step": 22300
    },
    {
      "epoch": 0.29242142239232444,
      "grad_norm": 0.5916216969490051,
      "learning_rate": 0.00024277508049955812,
      "loss": 3.3593,
      "step": 22310
    },
    {
      "epoch": 0.2925524942983721,
      "grad_norm": 0.4981248080730438,
      "learning_rate": 0.00024272621745537372,
      "loss": 3.3164,
      "step": 22320
    },
    {
      "epoch": 0.29268356620441977,
      "grad_norm": 0.5504450798034668,
      "learning_rate": 0.00024267733848057122,
      "loss": 3.3091,
      "step": 22330
    },
    {
      "epoch": 0.29281463811046743,
      "grad_norm": 0.5225972533226013,
      "learning_rate": 0.00024262844358354806,
      "loss": 3.2799,
      "step": 22340
    },
    {
      "epoch": 0.29294571001651504,
      "grad_norm": 0.6133748888969421,
      "learning_rate": 0.00024257953277270452,
      "loss": 3.3516,
      "step": 22350
    },
    {
      "epoch": 0.2930767819225627,
      "grad_norm": 0.6092876195907593,
      "learning_rate": 0.0002425306060564437,
      "loss": 3.4561,
      "step": 22360
    },
    {
      "epoch": 0.29320785382861037,
      "grad_norm": 0.5432363152503967,
      "learning_rate": 0.00024248166344317123,
      "loss": 3.3381,
      "step": 22370
    },
    {
      "epoch": 0.29333892573465803,
      "grad_norm": 0.5080586075782776,
      "learning_rate": 0.00024243270494129568,
      "loss": 3.3407,
      "step": 22380
    },
    {
      "epoch": 0.2934699976407057,
      "grad_norm": 0.5644974112510681,
      "learning_rate": 0.00024238373055922817,
      "loss": 3.3434,
      "step": 22390
    },
    {
      "epoch": 0.29360106954675336,
      "grad_norm": 0.5694626569747925,
      "learning_rate": 0.0002423347403053827,
      "loss": 3.3831,
      "step": 22400
    },
    {
      "epoch": 0.293732141452801,
      "grad_norm": 0.49748528003692627,
      "learning_rate": 0.0002422857341881759,
      "loss": 3.3449,
      "step": 22410
    },
    {
      "epoch": 0.2938632133588487,
      "grad_norm": 0.5227693915367126,
      "learning_rate": 0.00024223671221602715,
      "loss": 3.3213,
      "step": 22420
    },
    {
      "epoch": 0.29399428526489635,
      "grad_norm": 0.5170589685440063,
      "learning_rate": 0.00024218767439735856,
      "loss": 3.3665,
      "step": 22430
    },
    {
      "epoch": 0.29412535717094396,
      "grad_norm": 0.5238785743713379,
      "learning_rate": 0.000242138620740595,
      "loss": 3.3515,
      "step": 22440
    },
    {
      "epoch": 0.2942564290769916,
      "grad_norm": 0.5432807803153992,
      "learning_rate": 0.00024208955125416394,
      "loss": 3.3158,
      "step": 22450
    },
    {
      "epoch": 0.2943875009830393,
      "grad_norm": 0.5207889676094055,
      "learning_rate": 0.00024204046594649575,
      "loss": 3.3468,
      "step": 22460
    },
    {
      "epoch": 0.29451857288908695,
      "grad_norm": 0.5279417037963867,
      "learning_rate": 0.00024199136482602336,
      "loss": 3.3452,
      "step": 22470
    },
    {
      "epoch": 0.2946496447951346,
      "grad_norm": 0.5553640127182007,
      "learning_rate": 0.0002419422479011825,
      "loss": 3.3136,
      "step": 22480
    },
    {
      "epoch": 0.2947807167011823,
      "grad_norm": 0.533237636089325,
      "learning_rate": 0.00024189311518041162,
      "loss": 3.3184,
      "step": 22490
    },
    {
      "epoch": 0.29491178860722994,
      "grad_norm": 0.5006555914878845,
      "learning_rate": 0.00024184396667215177,
      "loss": 3.3404,
      "step": 22500
    },
    {
      "epoch": 0.29491178860722994,
      "eval_accuracy": 0.38572287390029325,
      "eval_loss": 3.345003128051758,
      "eval_runtime": 58.9461,
      "eval_samples_per_second": 33.929,
      "eval_steps_per_second": 0.543,
      "step": 22500
    },
    {
      "epoch": 0.2950428605132776,
      "grad_norm": 0.6142920851707458,
      "learning_rate": 0.00024179480238484686,
      "loss": 3.3558,
      "step": 22510
    },
    {
      "epoch": 0.29517393241932527,
      "grad_norm": 0.5241626501083374,
      "learning_rate": 0.00024174562232694345,
      "loss": 3.3737,
      "step": 22520
    },
    {
      "epoch": 0.2953050043253729,
      "grad_norm": 0.5529385805130005,
      "learning_rate": 0.00024169642650689077,
      "loss": 3.3687,
      "step": 22530
    },
    {
      "epoch": 0.29543607623142054,
      "grad_norm": 0.551546037197113,
      "learning_rate": 0.00024164721493314084,
      "loss": 3.3708,
      "step": 22540
    },
    {
      "epoch": 0.2955671481374682,
      "grad_norm": 0.46931543946266174,
      "learning_rate": 0.00024159798761414834,
      "loss": 3.3155,
      "step": 22550
    },
    {
      "epoch": 0.29569822004351587,
      "grad_norm": 0.510817289352417,
      "learning_rate": 0.00024154874455837064,
      "loss": 3.3485,
      "step": 22560
    },
    {
      "epoch": 0.29582929194956353,
      "grad_norm": 0.5688291788101196,
      "learning_rate": 0.00024149948577426784,
      "loss": 3.3521,
      "step": 22570
    },
    {
      "epoch": 0.2959603638556112,
      "grad_norm": 0.5152601599693298,
      "learning_rate": 0.00024145021127030267,
      "loss": 3.3588,
      "step": 22580
    },
    {
      "epoch": 0.29609143576165886,
      "grad_norm": 0.5665587782859802,
      "learning_rate": 0.00024140092105494074,
      "loss": 3.3479,
      "step": 22590
    },
    {
      "epoch": 0.2962225076677065,
      "grad_norm": 0.5446558594703674,
      "learning_rate": 0.0002413516151366502,
      "loss": 3.3129,
      "step": 22600
    },
    {
      "epoch": 0.2963535795737542,
      "grad_norm": 0.5052319169044495,
      "learning_rate": 0.00024130229352390195,
      "loss": 3.3218,
      "step": 22610
    },
    {
      "epoch": 0.2964846514798018,
      "grad_norm": 0.5465099811553955,
      "learning_rate": 0.0002412529562251695,
      "loss": 3.3396,
      "step": 22620
    },
    {
      "epoch": 0.29661572338584946,
      "grad_norm": 0.5319451093673706,
      "learning_rate": 0.00024120360324892922,
      "loss": 3.3863,
      "step": 22630
    },
    {
      "epoch": 0.2967467952918971,
      "grad_norm": 0.49311211705207825,
      "learning_rate": 0.00024115423460366008,
      "loss": 3.3074,
      "step": 22640
    },
    {
      "epoch": 0.2968778671979448,
      "grad_norm": 0.5541136860847473,
      "learning_rate": 0.0002411048502978437,
      "loss": 3.342,
      "step": 22650
    },
    {
      "epoch": 0.29700893910399245,
      "grad_norm": 0.5065752863883972,
      "learning_rate": 0.00024105545033996449,
      "loss": 3.3186,
      "step": 22660
    },
    {
      "epoch": 0.2971400110100401,
      "grad_norm": 0.5738956928253174,
      "learning_rate": 0.00024100603473850945,
      "loss": 3.3377,
      "step": 22670
    },
    {
      "epoch": 0.2972710829160878,
      "grad_norm": 0.5165664553642273,
      "learning_rate": 0.00024095660350196836,
      "loss": 3.3261,
      "step": 22680
    },
    {
      "epoch": 0.29740215482213544,
      "grad_norm": 0.5183415412902832,
      "learning_rate": 0.0002409071566388336,
      "loss": 3.3196,
      "step": 22690
    },
    {
      "epoch": 0.2975332267281831,
      "grad_norm": 0.5167087316513062,
      "learning_rate": 0.00024085769415760035,
      "loss": 3.3496,
      "step": 22700
    },
    {
      "epoch": 0.2976642986342307,
      "grad_norm": 0.4981522262096405,
      "learning_rate": 0.0002408082160667663,
      "loss": 3.3574,
      "step": 22710
    },
    {
      "epoch": 0.2977953705402784,
      "grad_norm": 0.5049541592597961,
      "learning_rate": 0.00024075872237483198,
      "loss": 3.3342,
      "step": 22720
    },
    {
      "epoch": 0.29792644244632605,
      "grad_norm": 0.5809832215309143,
      "learning_rate": 0.0002407092130903005,
      "loss": 3.3395,
      "step": 22730
    },
    {
      "epoch": 0.2980575143523737,
      "grad_norm": 0.6543958187103271,
      "learning_rate": 0.00024065968822167778,
      "loss": 3.3625,
      "step": 22740
    },
    {
      "epoch": 0.2981885862584214,
      "grad_norm": 0.555397629737854,
      "learning_rate": 0.00024061014777747224,
      "loss": 3.335,
      "step": 22750
    },
    {
      "epoch": 0.29831965816446904,
      "grad_norm": 0.4941960871219635,
      "learning_rate": 0.00024056059176619512,
      "loss": 3.3765,
      "step": 22760
    },
    {
      "epoch": 0.2984507300705167,
      "grad_norm": 0.53848797082901,
      "learning_rate": 0.00024051102019636024,
      "loss": 3.3516,
      "step": 22770
    },
    {
      "epoch": 0.29858180197656436,
      "grad_norm": 0.5042014122009277,
      "learning_rate": 0.00024046143307648415,
      "loss": 3.3173,
      "step": 22780
    },
    {
      "epoch": 0.298712873882612,
      "grad_norm": 0.6403889656066895,
      "learning_rate": 0.00024041183041508606,
      "loss": 3.3042,
      "step": 22790
    },
    {
      "epoch": 0.29884394578865964,
      "grad_norm": 0.5096739530563354,
      "learning_rate": 0.00024036221222068776,
      "loss": 3.3361,
      "step": 22800
    },
    {
      "epoch": 0.2989750176947073,
      "grad_norm": 0.5295944809913635,
      "learning_rate": 0.00024031257850181393,
      "loss": 3.4195,
      "step": 22810
    },
    {
      "epoch": 0.29910608960075497,
      "grad_norm": 0.525929868221283,
      "learning_rate": 0.00024026292926699166,
      "loss": 3.3577,
      "step": 22820
    },
    {
      "epoch": 0.29923716150680263,
      "grad_norm": 0.5381543636322021,
      "learning_rate": 0.00024021326452475095,
      "loss": 3.3232,
      "step": 22830
    },
    {
      "epoch": 0.2993682334128503,
      "grad_norm": 0.5283045768737793,
      "learning_rate": 0.00024016358428362422,
      "loss": 3.3309,
      "step": 22840
    },
    {
      "epoch": 0.29949930531889796,
      "grad_norm": 0.6148714423179626,
      "learning_rate": 0.00024011388855214673,
      "loss": 3.3127,
      "step": 22850
    },
    {
      "epoch": 0.2996303772249456,
      "grad_norm": 0.5613620281219482,
      "learning_rate": 0.00024006417733885628,
      "loss": 3.3434,
      "step": 22860
    },
    {
      "epoch": 0.2997614491309933,
      "grad_norm": 0.5173367261886597,
      "learning_rate": 0.0002400144506522935,
      "loss": 3.3345,
      "step": 22870
    },
    {
      "epoch": 0.2998925210370409,
      "grad_norm": 0.5058047771453857,
      "learning_rate": 0.0002399647085010015,
      "loss": 3.42,
      "step": 22880
    },
    {
      "epoch": 0.30002359294308856,
      "grad_norm": 0.5365564823150635,
      "learning_rate": 0.00023991495089352614,
      "loss": 3.3312,
      "step": 22890
    },
    {
      "epoch": 0.3001546648491362,
      "grad_norm": 0.5046512484550476,
      "learning_rate": 0.0002398651778384159,
      "loss": 3.3212,
      "step": 22900
    },
    {
      "epoch": 0.3002857367551839,
      "grad_norm": 0.5494726300239563,
      "learning_rate": 0.0002398153893442219,
      "loss": 3.3395,
      "step": 22910
    },
    {
      "epoch": 0.30041680866123155,
      "grad_norm": 0.5233862996101379,
      "learning_rate": 0.000239765585419498,
      "loss": 3.3416,
      "step": 22920
    },
    {
      "epoch": 0.3005478805672792,
      "grad_norm": 0.5653012990951538,
      "learning_rate": 0.00023971576607280062,
      "loss": 3.3326,
      "step": 22930
    },
    {
      "epoch": 0.3006789524733269,
      "grad_norm": 0.5929429531097412,
      "learning_rate": 0.0002396659313126889,
      "loss": 3.351,
      "step": 22940
    },
    {
      "epoch": 0.30081002437937454,
      "grad_norm": 0.5832833647727966,
      "learning_rate": 0.00023961608114772448,
      "loss": 3.3176,
      "step": 22950
    },
    {
      "epoch": 0.3009410962854222,
      "grad_norm": 0.5306656956672668,
      "learning_rate": 0.00023956621558647185,
      "loss": 3.3628,
      "step": 22960
    },
    {
      "epoch": 0.3010721681914698,
      "grad_norm": 0.569543182849884,
      "learning_rate": 0.00023951633463749803,
      "loss": 3.3272,
      "step": 22970
    },
    {
      "epoch": 0.3012032400975175,
      "grad_norm": 0.5808171033859253,
      "learning_rate": 0.00023946643830937268,
      "loss": 3.35,
      "step": 22980
    },
    {
      "epoch": 0.30133431200356514,
      "grad_norm": 0.48459044098854065,
      "learning_rate": 0.00023941652661066816,
      "loss": 3.3245,
      "step": 22990
    },
    {
      "epoch": 0.3014653839096128,
      "grad_norm": 0.6501249670982361,
      "learning_rate": 0.00023936659954995945,
      "loss": 3.4008,
      "step": 23000
    },
    {
      "epoch": 0.3014653839096128,
      "eval_accuracy": 0.3861177908113392,
      "eval_loss": 3.3397727012634277,
      "eval_runtime": 59.211,
      "eval_samples_per_second": 33.777,
      "eval_steps_per_second": 0.54,
      "step": 23000
    },
    {
      "epoch": 0.30159645581566047,
      "grad_norm": 0.5249041318893433,
      "learning_rate": 0.0002393166571358241,
      "loss": 3.357,
      "step": 23010
    },
    {
      "epoch": 0.30172752772170813,
      "grad_norm": 0.6072885990142822,
      "learning_rate": 0.00023926669937684238,
      "loss": 3.3345,
      "step": 23020
    },
    {
      "epoch": 0.3018585996277558,
      "grad_norm": 0.5100835561752319,
      "learning_rate": 0.0002392167262815972,
      "loss": 3.3566,
      "step": 23030
    },
    {
      "epoch": 0.30198967153380346,
      "grad_norm": 0.5046204328536987,
      "learning_rate": 0.00023916673785867397,
      "loss": 3.3332,
      "step": 23040
    },
    {
      "epoch": 0.3021207434398511,
      "grad_norm": 0.4916168451309204,
      "learning_rate": 0.00023911673411666098,
      "loss": 3.404,
      "step": 23050
    },
    {
      "epoch": 0.30225181534589873,
      "grad_norm": 0.4961005747318268,
      "learning_rate": 0.0002390667150641489,
      "loss": 3.3267,
      "step": 23060
    },
    {
      "epoch": 0.3023828872519464,
      "grad_norm": 0.5165991187095642,
      "learning_rate": 0.00023901668070973125,
      "loss": 3.3484,
      "step": 23070
    },
    {
      "epoch": 0.30251395915799406,
      "grad_norm": 0.5445991158485413,
      "learning_rate": 0.00023896663106200394,
      "loss": 3.3355,
      "step": 23080
    },
    {
      "epoch": 0.3026450310640417,
      "grad_norm": 0.5759246945381165,
      "learning_rate": 0.0002389165661295657,
      "loss": 3.398,
      "step": 23090
    },
    {
      "epoch": 0.3027761029700894,
      "grad_norm": 0.780782163143158,
      "learning_rate": 0.00023886648592101785,
      "loss": 3.4138,
      "step": 23100
    },
    {
      "epoch": 0.30290717487613705,
      "grad_norm": 0.5794110298156738,
      "learning_rate": 0.00023881639044496427,
      "loss": 3.3384,
      "step": 23110
    },
    {
      "epoch": 0.3030382467821847,
      "grad_norm": 0.5405265688896179,
      "learning_rate": 0.0002387662797100115,
      "loss": 3.3284,
      "step": 23120
    },
    {
      "epoch": 0.3031693186882324,
      "grad_norm": 0.524726927280426,
      "learning_rate": 0.0002387161537247687,
      "loss": 3.3686,
      "step": 23130
    },
    {
      "epoch": 0.30330039059428004,
      "grad_norm": 0.5326516032218933,
      "learning_rate": 0.00023866601249784774,
      "loss": 3.3692,
      "step": 23140
    },
    {
      "epoch": 0.30343146250032765,
      "grad_norm": 0.4874504506587982,
      "learning_rate": 0.00023861585603786285,
      "loss": 3.3209,
      "step": 23150
    },
    {
      "epoch": 0.3035625344063753,
      "grad_norm": 0.521446943283081,
      "learning_rate": 0.00023856568435343117,
      "loss": 3.3002,
      "step": 23160
    },
    {
      "epoch": 0.303693606312423,
      "grad_norm": 0.5796816945075989,
      "learning_rate": 0.0002385154974531723,
      "loss": 3.3243,
      "step": 23170
    },
    {
      "epoch": 0.30382467821847065,
      "grad_norm": 0.519051194190979,
      "learning_rate": 0.0002384652953457085,
      "loss": 3.3369,
      "step": 23180
    },
    {
      "epoch": 0.3039557501245183,
      "grad_norm": 0.5408623218536377,
      "learning_rate": 0.0002384150780396646,
      "loss": 3.3517,
      "step": 23190
    },
    {
      "epoch": 0.304086822030566,
      "grad_norm": 0.5035997629165649,
      "learning_rate": 0.0002383648455436681,
      "loss": 3.3252,
      "step": 23200
    },
    {
      "epoch": 0.30421789393661364,
      "grad_norm": 0.6559159159660339,
      "learning_rate": 0.00023831459786634903,
      "loss": 3.3038,
      "step": 23210
    },
    {
      "epoch": 0.3043489658426613,
      "grad_norm": 0.5289565324783325,
      "learning_rate": 0.00023826433501634018,
      "loss": 3.3713,
      "step": 23220
    },
    {
      "epoch": 0.30448003774870896,
      "grad_norm": 0.5206024646759033,
      "learning_rate": 0.00023821405700227678,
      "loss": 3.3536,
      "step": 23230
    },
    {
      "epoch": 0.3046111096547566,
      "grad_norm": 0.5109387040138245,
      "learning_rate": 0.00023816376383279669,
      "loss": 3.3154,
      "step": 23240
    },
    {
      "epoch": 0.30474218156080424,
      "grad_norm": 0.4832921326160431,
      "learning_rate": 0.00023811345551654048,
      "loss": 3.3122,
      "step": 23250
    },
    {
      "epoch": 0.3048732534668519,
      "grad_norm": 0.49395498633384705,
      "learning_rate": 0.00023806313206215123,
      "loss": 3.3478,
      "step": 23260
    },
    {
      "epoch": 0.30500432537289957,
      "grad_norm": 0.5116652846336365,
      "learning_rate": 0.0002380127934782747,
      "loss": 3.322,
      "step": 23270
    },
    {
      "epoch": 0.30513539727894723,
      "grad_norm": 0.5292666554450989,
      "learning_rate": 0.00023796243977355908,
      "loss": 3.4006,
      "step": 23280
    },
    {
      "epoch": 0.3052664691849949,
      "grad_norm": 0.535426914691925,
      "learning_rate": 0.00023791207095665538,
      "loss": 3.3242,
      "step": 23290
    },
    {
      "epoch": 0.30539754109104256,
      "grad_norm": 0.5790781378746033,
      "learning_rate": 0.00023786168703621703,
      "loss": 3.355,
      "step": 23300
    },
    {
      "epoch": 0.3055286129970902,
      "grad_norm": 0.49692994356155396,
      "learning_rate": 0.0002378112880209002,
      "loss": 3.3454,
      "step": 23310
    },
    {
      "epoch": 0.3056596849031379,
      "grad_norm": 0.5455918908119202,
      "learning_rate": 0.00023776087391936353,
      "loss": 3.3781,
      "step": 23320
    },
    {
      "epoch": 0.3057907568091855,
      "grad_norm": 0.6387801170349121,
      "learning_rate": 0.00023771044474026832,
      "loss": 3.3716,
      "step": 23330
    },
    {
      "epoch": 0.30592182871523316,
      "grad_norm": 0.5252436995506287,
      "learning_rate": 0.0002376600004922784,
      "loss": 3.3461,
      "step": 23340
    },
    {
      "epoch": 0.3060529006212808,
      "grad_norm": 0.5929307341575623,
      "learning_rate": 0.00023760954118406027,
      "loss": 3.3534,
      "step": 23350
    },
    {
      "epoch": 0.3061839725273285,
      "grad_norm": 0.5890244841575623,
      "learning_rate": 0.00023755906682428296,
      "loss": 3.3261,
      "step": 23360
    },
    {
      "epoch": 0.30631504443337615,
      "grad_norm": 0.5512086153030396,
      "learning_rate": 0.0002375085774216181,
      "loss": 3.3244,
      "step": 23370
    },
    {
      "epoch": 0.3064461163394238,
      "grad_norm": 0.5454772114753723,
      "learning_rate": 0.00023745807298473993,
      "loss": 3.3614,
      "step": 23380
    },
    {
      "epoch": 0.3065771882454715,
      "grad_norm": 0.613351583480835,
      "learning_rate": 0.0002374075535223253,
      "loss": 3.3557,
      "step": 23390
    },
    {
      "epoch": 0.30670826015151914,
      "grad_norm": 0.513185977935791,
      "learning_rate": 0.00023735701904305345,
      "loss": 3.3214,
      "step": 23400
    },
    {
      "epoch": 0.3068393320575668,
      "grad_norm": 0.5195509195327759,
      "learning_rate": 0.00023730646955560645,
      "loss": 3.3712,
      "step": 23410
    },
    {
      "epoch": 0.3069704039636144,
      "grad_norm": 0.49841737747192383,
      "learning_rate": 0.00023725590506866885,
      "loss": 3.318,
      "step": 23420
    },
    {
      "epoch": 0.3071014758696621,
      "grad_norm": 0.4922764301300049,
      "learning_rate": 0.0002372053255909277,
      "loss": 3.3094,
      "step": 23430
    },
    {
      "epoch": 0.30723254777570974,
      "grad_norm": 0.5359746813774109,
      "learning_rate": 0.00023715473113107268,
      "loss": 3.3381,
      "step": 23440
    },
    {
      "epoch": 0.3073636196817574,
      "grad_norm": 0.49704211950302124,
      "learning_rate": 0.00023710412169779618,
      "loss": 3.3449,
      "step": 23450
    },
    {
      "epoch": 0.30749469158780507,
      "grad_norm": 0.5008110404014587,
      "learning_rate": 0.0002370534972997929,
      "loss": 3.3208,
      "step": 23460
    },
    {
      "epoch": 0.30762576349385273,
      "grad_norm": 0.6076968908309937,
      "learning_rate": 0.0002370028579457603,
      "loss": 3.3363,
      "step": 23470
    },
    {
      "epoch": 0.3077568353999004,
      "grad_norm": 0.4801216721534729,
      "learning_rate": 0.00023695220364439843,
      "loss": 3.3393,
      "step": 23480
    },
    {
      "epoch": 0.30788790730594806,
      "grad_norm": 0.5993174314498901,
      "learning_rate": 0.00023690153440440972,
      "loss": 3.2901,
      "step": 23490
    },
    {
      "epoch": 0.3080189792119957,
      "grad_norm": 0.6075590252876282,
      "learning_rate": 0.0002368508502344994,
      "loss": 3.3227,
      "step": 23500
    },
    {
      "epoch": 0.3080189792119957,
      "eval_accuracy": 0.385836265884653,
      "eval_loss": 3.340456485748291,
      "eval_runtime": 60.2696,
      "eval_samples_per_second": 33.184,
      "eval_steps_per_second": 0.531,
      "step": 23500
    },
    {
      "epoch": 0.30815005111804333,
      "grad_norm": 0.5592019557952881,
      "learning_rate": 0.00023680015114337506,
      "loss": 3.3406,
      "step": 23510
    },
    {
      "epoch": 0.308281123024091,
      "grad_norm": 0.5206080079078674,
      "learning_rate": 0.00023674943713974697,
      "loss": 3.3387,
      "step": 23520
    },
    {
      "epoch": 0.30841219493013866,
      "grad_norm": 0.5406329035758972,
      "learning_rate": 0.00023669870823232796,
      "loss": 3.3751,
      "step": 23530
    },
    {
      "epoch": 0.3085432668361863,
      "grad_norm": 0.5391744375228882,
      "learning_rate": 0.00023664796442983334,
      "loss": 3.3684,
      "step": 23540
    },
    {
      "epoch": 0.308674338742234,
      "grad_norm": 0.5212407112121582,
      "learning_rate": 0.00023659720574098112,
      "loss": 3.3376,
      "step": 23550
    },
    {
      "epoch": 0.30880541064828165,
      "grad_norm": 0.6529204249382019,
      "learning_rate": 0.00023654643217449166,
      "loss": 3.3468,
      "step": 23560
    },
    {
      "epoch": 0.3089364825543293,
      "grad_norm": 0.560558021068573,
      "learning_rate": 0.0002364956437390881,
      "loss": 3.3493,
      "step": 23570
    },
    {
      "epoch": 0.309067554460377,
      "grad_norm": 0.5197131037712097,
      "learning_rate": 0.000236444840443496,
      "loss": 3.3209,
      "step": 23580
    },
    {
      "epoch": 0.30919862636642464,
      "grad_norm": 0.5099916458129883,
      "learning_rate": 0.00023639402229644354,
      "loss": 3.3318,
      "step": 23590
    },
    {
      "epoch": 0.30932969827247225,
      "grad_norm": 0.5440672636032104,
      "learning_rate": 0.00023634318930666135,
      "loss": 3.3013,
      "step": 23600
    },
    {
      "epoch": 0.3094607701785199,
      "grad_norm": 0.562285304069519,
      "learning_rate": 0.00023629234148288272,
      "loss": 3.4163,
      "step": 23610
    },
    {
      "epoch": 0.3095918420845676,
      "grad_norm": 0.8945772647857666,
      "learning_rate": 0.0002362414788338434,
      "loss": 3.3935,
      "step": 23620
    },
    {
      "epoch": 0.30972291399061525,
      "grad_norm": 0.5317020416259766,
      "learning_rate": 0.00023619060136828178,
      "loss": 3.3341,
      "step": 23630
    },
    {
      "epoch": 0.3098539858966629,
      "grad_norm": 0.5755566358566284,
      "learning_rate": 0.00023613970909493878,
      "loss": 3.3811,
      "step": 23640
    },
    {
      "epoch": 0.3099850578027106,
      "grad_norm": 0.5342457294464111,
      "learning_rate": 0.0002360888020225578,
      "loss": 3.3324,
      "step": 23650
    },
    {
      "epoch": 0.31011612970875824,
      "grad_norm": 0.4904495179653168,
      "learning_rate": 0.00023603788015988473,
      "loss": 3.3094,
      "step": 23660
    },
    {
      "epoch": 0.3102472016148059,
      "grad_norm": 0.5517577528953552,
      "learning_rate": 0.00023598694351566814,
      "loss": 3.3148,
      "step": 23670
    },
    {
      "epoch": 0.31037827352085356,
      "grad_norm": 0.5581772923469543,
      "learning_rate": 0.00023593599209865916,
      "loss": 3.3354,
      "step": 23680
    },
    {
      "epoch": 0.3105093454269012,
      "grad_norm": 0.5523823499679565,
      "learning_rate": 0.00023588502591761134,
      "loss": 3.3313,
      "step": 23690
    },
    {
      "epoch": 0.31064041733294884,
      "grad_norm": 0.5193866491317749,
      "learning_rate": 0.00023583404498128075,
      "loss": 3.3087,
      "step": 23700
    },
    {
      "epoch": 0.3107714892389965,
      "grad_norm": 0.5121460556983948,
      "learning_rate": 0.0002357830492984261,
      "loss": 3.2744,
      "step": 23710
    },
    {
      "epoch": 0.31090256114504417,
      "grad_norm": 0.5276694893836975,
      "learning_rate": 0.0002357320388778086,
      "loss": 3.3603,
      "step": 23720
    },
    {
      "epoch": 0.31103363305109183,
      "grad_norm": 0.5929836630821228,
      "learning_rate": 0.00023568101372819199,
      "loss": 3.354,
      "step": 23730
    },
    {
      "epoch": 0.3111647049571395,
      "grad_norm": 0.581091582775116,
      "learning_rate": 0.00023562997385834247,
      "loss": 3.2758,
      "step": 23740
    },
    {
      "epoch": 0.31129577686318716,
      "grad_norm": 0.5163605213165283,
      "learning_rate": 0.00023557891927702886,
      "loss": 3.3342,
      "step": 23750
    },
    {
      "epoch": 0.3114268487692348,
      "grad_norm": 0.536588728427887,
      "learning_rate": 0.00023552784999302252,
      "loss": 3.3238,
      "step": 23760
    },
    {
      "epoch": 0.3115579206752825,
      "grad_norm": 0.5676375031471252,
      "learning_rate": 0.00023547676601509728,
      "loss": 3.342,
      "step": 23770
    },
    {
      "epoch": 0.3116889925813301,
      "grad_norm": 0.5490791201591492,
      "learning_rate": 0.00023542566735202942,
      "loss": 3.3587,
      "step": 23780
    },
    {
      "epoch": 0.31182006448737776,
      "grad_norm": 0.8581646084785461,
      "learning_rate": 0.00023537455401259792,
      "loss": 3.346,
      "step": 23790
    },
    {
      "epoch": 0.3119511363934254,
      "grad_norm": 0.5162758231163025,
      "learning_rate": 0.00023532342600558417,
      "loss": 3.3454,
      "step": 23800
    },
    {
      "epoch": 0.3120822082994731,
      "grad_norm": 0.5362398028373718,
      "learning_rate": 0.0002352722833397721,
      "loss": 3.2928,
      "step": 23810
    },
    {
      "epoch": 0.31221328020552075,
      "grad_norm": 0.5482853651046753,
      "learning_rate": 0.00023522112602394814,
      "loss": 3.3489,
      "step": 23820
    },
    {
      "epoch": 0.3123443521115684,
      "grad_norm": 0.5353394746780396,
      "learning_rate": 0.00023516995406690133,
      "loss": 3.3352,
      "step": 23830
    },
    {
      "epoch": 0.3124754240176161,
      "grad_norm": 0.5724055171012878,
      "learning_rate": 0.00023511876747742298,
      "loss": 3.321,
      "step": 23840
    },
    {
      "epoch": 0.31260649592366374,
      "grad_norm": 0.5895740985870361,
      "learning_rate": 0.00023506756626430733,
      "loss": 3.3491,
      "step": 23850
    },
    {
      "epoch": 0.3127375678297114,
      "grad_norm": 0.5058173537254333,
      "learning_rate": 0.00023501635043635073,
      "loss": 3.3379,
      "step": 23860
    },
    {
      "epoch": 0.312868639735759,
      "grad_norm": 0.5184350609779358,
      "learning_rate": 0.0002349651200023522,
      "loss": 3.4012,
      "step": 23870
    },
    {
      "epoch": 0.3129997116418067,
      "grad_norm": 0.5318226218223572,
      "learning_rate": 0.0002349138749711133,
      "loss": 3.3657,
      "step": 23880
    },
    {
      "epoch": 0.31313078354785434,
      "grad_norm": 0.6517789363861084,
      "learning_rate": 0.00023486261535143811,
      "loss": 3.3422,
      "step": 23890
    },
    {
      "epoch": 0.313261855453902,
      "grad_norm": 0.5339535474777222,
      "learning_rate": 0.00023481134115213313,
      "loss": 3.3405,
      "step": 23900
    },
    {
      "epoch": 0.31339292735994967,
      "grad_norm": 0.4870710074901581,
      "learning_rate": 0.00023476005238200732,
      "loss": 3.3916,
      "step": 23910
    },
    {
      "epoch": 0.31352399926599733,
      "grad_norm": 0.5462631583213806,
      "learning_rate": 0.00023470874904987242,
      "loss": 3.3799,
      "step": 23920
    },
    {
      "epoch": 0.313655071172045,
      "grad_norm": 0.5240347981452942,
      "learning_rate": 0.0002346574311645423,
      "loss": 3.3286,
      "step": 23930
    },
    {
      "epoch": 0.31378614307809266,
      "grad_norm": 0.6628802418708801,
      "learning_rate": 0.00023460609873483367,
      "loss": 3.3304,
      "step": 23940
    },
    {
      "epoch": 0.3139172149841403,
      "grad_norm": 0.5082511305809021,
      "learning_rate": 0.00023455475176956548,
      "loss": 3.3665,
      "step": 23950
    },
    {
      "epoch": 0.31404828689018793,
      "grad_norm": 0.5061856508255005,
      "learning_rate": 0.0002345033902775593,
      "loss": 3.3537,
      "step": 23960
    },
    {
      "epoch": 0.3141793587962356,
      "grad_norm": 0.5664144158363342,
      "learning_rate": 0.00023445201426763916,
      "loss": 3.3615,
      "step": 23970
    },
    {
      "epoch": 0.31431043070228326,
      "grad_norm": 0.5962449908256531,
      "learning_rate": 0.0002344006237486317,
      "loss": 3.2864,
      "step": 23980
    },
    {
      "epoch": 0.3144415026083309,
      "grad_norm": 0.5322626233100891,
      "learning_rate": 0.0002343492187293658,
      "loss": 3.3641,
      "step": 23990
    },
    {
      "epoch": 0.3145725745143786,
      "grad_norm": 0.49926501512527466,
      "learning_rate": 0.00023429779921867308,
      "loss": 3.3492,
      "step": 24000
    },
    {
      "epoch": 0.3145725745143786,
      "eval_accuracy": 0.3872570869990225,
      "eval_loss": 3.331186294555664,
      "eval_runtime": 60.9983,
      "eval_samples_per_second": 32.788,
      "eval_steps_per_second": 0.525,
      "step": 24000
    },
    {
      "epoch": 0.31470364642042625,
      "grad_norm": 0.5208404660224915,
      "learning_rate": 0.0002342463652253875,
      "loss": 3.3278,
      "step": 24010
    },
    {
      "epoch": 0.3148347183264739,
      "grad_norm": 0.526390790939331,
      "learning_rate": 0.00023419491675834563,
      "loss": 3.3412,
      "step": 24020
    },
    {
      "epoch": 0.3149657902325216,
      "grad_norm": 0.5193520188331604,
      "learning_rate": 0.00023414345382638643,
      "loss": 3.3497,
      "step": 24030
    },
    {
      "epoch": 0.31509686213856924,
      "grad_norm": 0.5173376798629761,
      "learning_rate": 0.00023409197643835134,
      "loss": 3.3785,
      "step": 24040
    },
    {
      "epoch": 0.31522793404461685,
      "grad_norm": 0.4840696156024933,
      "learning_rate": 0.0002340404846030843,
      "loss": 3.3459,
      "step": 24050
    },
    {
      "epoch": 0.3153590059506645,
      "grad_norm": 0.49681025743484497,
      "learning_rate": 0.00023398897832943184,
      "loss": 3.3234,
      "step": 24060
    },
    {
      "epoch": 0.3154900778567122,
      "grad_norm": 0.6722946166992188,
      "learning_rate": 0.00023393745762624282,
      "loss": 3.3258,
      "step": 24070
    },
    {
      "epoch": 0.31562114976275985,
      "grad_norm": 0.5061906576156616,
      "learning_rate": 0.0002338859225023686,
      "loss": 3.3145,
      "step": 24080
    },
    {
      "epoch": 0.3157522216688075,
      "grad_norm": 0.8361560702323914,
      "learning_rate": 0.00023383437296666313,
      "loss": 3.3274,
      "step": 24090
    },
    {
      "epoch": 0.3158832935748552,
      "grad_norm": 0.5683650970458984,
      "learning_rate": 0.0002337828090279827,
      "loss": 3.3312,
      "step": 24100
    },
    {
      "epoch": 0.31601436548090284,
      "grad_norm": 0.5598815679550171,
      "learning_rate": 0.00023373123069518622,
      "loss": 3.3485,
      "step": 24110
    },
    {
      "epoch": 0.3161454373869505,
      "grad_norm": 0.527732789516449,
      "learning_rate": 0.00023367963797713486,
      "loss": 3.4118,
      "step": 24120
    },
    {
      "epoch": 0.31627650929299816,
      "grad_norm": 0.5513291954994202,
      "learning_rate": 0.00023362803088269256,
      "loss": 3.3139,
      "step": 24130
    },
    {
      "epoch": 0.3164075811990458,
      "grad_norm": 0.5695775151252747,
      "learning_rate": 0.0002335764094207254,
      "loss": 3.3405,
      "step": 24140
    },
    {
      "epoch": 0.31653865310509344,
      "grad_norm": 0.530193030834198,
      "learning_rate": 0.00023352477360010215,
      "loss": 3.338,
      "step": 24150
    },
    {
      "epoch": 0.3166697250111411,
      "grad_norm": 0.5665830373764038,
      "learning_rate": 0.000233473123429694,
      "loss": 3.3686,
      "step": 24160
    },
    {
      "epoch": 0.31680079691718877,
      "grad_norm": 0.5621181726455688,
      "learning_rate": 0.0002334214589183746,
      "loss": 3.3692,
      "step": 24170
    },
    {
      "epoch": 0.31693186882323643,
      "grad_norm": 0.5614228844642639,
      "learning_rate": 0.00023336978007502002,
      "loss": 3.3332,
      "step": 24180
    },
    {
      "epoch": 0.3170629407292841,
      "grad_norm": 0.6097342371940613,
      "learning_rate": 0.0002333180869085088,
      "loss": 3.3453,
      "step": 24190
    },
    {
      "epoch": 0.31719401263533176,
      "grad_norm": 0.5407678484916687,
      "learning_rate": 0.00023326637942772206,
      "loss": 3.3753,
      "step": 24200
    },
    {
      "epoch": 0.3173250845413794,
      "grad_norm": 0.5928364992141724,
      "learning_rate": 0.0002332146576415432,
      "loss": 3.3506,
      "step": 24210
    },
    {
      "epoch": 0.3174561564474271,
      "grad_norm": 0.5546523928642273,
      "learning_rate": 0.00023316292155885825,
      "loss": 3.378,
      "step": 24220
    },
    {
      "epoch": 0.3175872283534747,
      "grad_norm": 0.4982992112636566,
      "learning_rate": 0.0002331111711885555,
      "loss": 3.3599,
      "step": 24230
    },
    {
      "epoch": 0.31771830025952236,
      "grad_norm": 0.5509029626846313,
      "learning_rate": 0.0002330594065395259,
      "loss": 3.3222,
      "step": 24240
    },
    {
      "epoch": 0.31784937216557,
      "grad_norm": 0.5175657272338867,
      "learning_rate": 0.00023300762762066272,
      "loss": 3.3173,
      "step": 24250
    },
    {
      "epoch": 0.3179804440716177,
      "grad_norm": 0.4944581389427185,
      "learning_rate": 0.0002329558344408617,
      "loss": 3.3671,
      "step": 24260
    },
    {
      "epoch": 0.31811151597766535,
      "grad_norm": 0.5234600901603699,
      "learning_rate": 0.0002329040270090211,
      "loss": 3.3266,
      "step": 24270
    },
    {
      "epoch": 0.318242587883713,
      "grad_norm": 0.5395556688308716,
      "learning_rate": 0.00023285220533404157,
      "loss": 3.3575,
      "step": 24280
    },
    {
      "epoch": 0.3183736597897607,
      "grad_norm": 0.5207858085632324,
      "learning_rate": 0.00023280036942482617,
      "loss": 3.3594,
      "step": 24290
    },
    {
      "epoch": 0.31850473169580834,
      "grad_norm": 0.5298396944999695,
      "learning_rate": 0.0002327485192902805,
      "loss": 3.3601,
      "step": 24300
    },
    {
      "epoch": 0.318635803601856,
      "grad_norm": 0.5622614622116089,
      "learning_rate": 0.00023269665493931248,
      "loss": 3.3326,
      "step": 24310
    },
    {
      "epoch": 0.3187668755079036,
      "grad_norm": 0.48852986097335815,
      "learning_rate": 0.00023264477638083265,
      "loss": 3.3405,
      "step": 24320
    },
    {
      "epoch": 0.3188979474139513,
      "grad_norm": 0.5802463293075562,
      "learning_rate": 0.00023259288362375386,
      "loss": 3.3729,
      "step": 24330
    },
    {
      "epoch": 0.31902901931999894,
      "grad_norm": 0.48671597242355347,
      "learning_rate": 0.00023254097667699136,
      "loss": 3.3285,
      "step": 24340
    },
    {
      "epoch": 0.3191600912260466,
      "grad_norm": 0.5029733777046204,
      "learning_rate": 0.000232489055549463,
      "loss": 3.3004,
      "step": 24350
    },
    {
      "epoch": 0.31929116313209427,
      "grad_norm": 0.518418550491333,
      "learning_rate": 0.00023243712025008888,
      "loss": 3.3126,
      "step": 24360
    },
    {
      "epoch": 0.31942223503814193,
      "grad_norm": 0.5853186845779419,
      "learning_rate": 0.00023238517078779175,
      "loss": 3.3036,
      "step": 24370
    },
    {
      "epoch": 0.3195533069441896,
      "grad_norm": 0.5781991481781006,
      "learning_rate": 0.00023233320717149652,
      "loss": 3.2972,
      "step": 24380
    },
    {
      "epoch": 0.31968437885023726,
      "grad_norm": 0.5112811923027039,
      "learning_rate": 0.00023228122941013082,
      "loss": 3.4138,
      "step": 24390
    },
    {
      "epoch": 0.3198154507562849,
      "grad_norm": 0.5657960176467896,
      "learning_rate": 0.00023222923751262445,
      "loss": 3.351,
      "step": 24400
    },
    {
      "epoch": 0.31994652266233253,
      "grad_norm": 0.5158735513687134,
      "learning_rate": 0.0002321772314879098,
      "loss": 3.3204,
      "step": 24410
    },
    {
      "epoch": 0.3200775945683802,
      "grad_norm": 0.4983420968055725,
      "learning_rate": 0.00023212521134492172,
      "loss": 3.2999,
      "step": 24420
    },
    {
      "epoch": 0.32020866647442786,
      "grad_norm": 0.5477209091186523,
      "learning_rate": 0.00023207317709259735,
      "loss": 3.3366,
      "step": 24430
    },
    {
      "epoch": 0.3203397383804755,
      "grad_norm": 0.5082642436027527,
      "learning_rate": 0.0002320211287398763,
      "loss": 3.3496,
      "step": 24440
    },
    {
      "epoch": 0.3204708102865232,
      "grad_norm": 0.5268057584762573,
      "learning_rate": 0.00023196906629570067,
      "loss": 3.3517,
      "step": 24450
    },
    {
      "epoch": 0.32060188219257085,
      "grad_norm": 0.48722970485687256,
      "learning_rate": 0.00023191698976901488,
      "loss": 3.3448,
      "step": 24460
    },
    {
      "epoch": 0.3207329540986185,
      "grad_norm": 0.5529583692550659,
      "learning_rate": 0.00023186489916876584,
      "loss": 3.3249,
      "step": 24470
    },
    {
      "epoch": 0.3208640260046662,
      "grad_norm": 0.5457765460014343,
      "learning_rate": 0.00023181279450390288,
      "loss": 3.3119,
      "step": 24480
    },
    {
      "epoch": 0.32099509791071384,
      "grad_norm": 0.5652438402175903,
      "learning_rate": 0.00023176067578337768,
      "loss": 3.3469,
      "step": 24490
    },
    {
      "epoch": 0.32112616981676145,
      "grad_norm": 0.5004788637161255,
      "learning_rate": 0.00023170854301614449,
      "loss": 3.3151,
      "step": 24500
    },
    {
      "epoch": 0.32112616981676145,
      "eval_accuracy": 0.3871480938416422,
      "eval_loss": 3.331258773803711,
      "eval_runtime": 61.2385,
      "eval_samples_per_second": 32.659,
      "eval_steps_per_second": 0.523,
      "step": 24500
    },
    {
      "epoch": 0.3212572417228091,
      "grad_norm": 0.500078022480011,
      "learning_rate": 0.00023165639621115967,
      "loss": 3.3223,
      "step": 24510
    },
    {
      "epoch": 0.3213883136288568,
      "grad_norm": 0.5505542755126953,
      "learning_rate": 0.00023160423537738233,
      "loss": 3.2889,
      "step": 24520
    },
    {
      "epoch": 0.32151938553490444,
      "grad_norm": 0.5428082346916199,
      "learning_rate": 0.00023155206052377387,
      "loss": 3.3481,
      "step": 24530
    },
    {
      "epoch": 0.3216504574409521,
      "grad_norm": 0.6734635829925537,
      "learning_rate": 0.00023149987165929794,
      "loss": 3.3207,
      "step": 24540
    },
    {
      "epoch": 0.3217815293469998,
      "grad_norm": 0.5574280023574829,
      "learning_rate": 0.00023144766879292083,
      "loss": 3.3253,
      "step": 24550
    },
    {
      "epoch": 0.32191260125304744,
      "grad_norm": 0.5113246440887451,
      "learning_rate": 0.00023139545193361105,
      "loss": 3.3376,
      "step": 24560
    },
    {
      "epoch": 0.3220436731590951,
      "grad_norm": 0.5412176847457886,
      "learning_rate": 0.0002313432210903397,
      "loss": 3.3586,
      "step": 24570
    },
    {
      "epoch": 0.32217474506514276,
      "grad_norm": 0.5033179521560669,
      "learning_rate": 0.00023129097627208012,
      "loss": 3.3365,
      "step": 24580
    },
    {
      "epoch": 0.3223058169711904,
      "grad_norm": 0.47749605774879456,
      "learning_rate": 0.00023123871748780814,
      "loss": 3.3404,
      "step": 24590
    },
    {
      "epoch": 0.32243688887723804,
      "grad_norm": 0.6148413419723511,
      "learning_rate": 0.00023118644474650192,
      "loss": 3.2999,
      "step": 24600
    },
    {
      "epoch": 0.3225679607832857,
      "grad_norm": 0.554003894329071,
      "learning_rate": 0.00023113415805714214,
      "loss": 3.3774,
      "step": 24610
    },
    {
      "epoch": 0.32269903268933336,
      "grad_norm": 0.5046848654747009,
      "learning_rate": 0.00023108185742871172,
      "loss": 3.3191,
      "step": 24620
    },
    {
      "epoch": 0.32283010459538103,
      "grad_norm": 0.5463700890541077,
      "learning_rate": 0.00023102954287019612,
      "loss": 3.3527,
      "step": 24630
    },
    {
      "epoch": 0.3229611765014287,
      "grad_norm": 0.5026952028274536,
      "learning_rate": 0.000230977214390583,
      "loss": 3.3515,
      "step": 24640
    },
    {
      "epoch": 0.32309224840747636,
      "grad_norm": 0.6984835267066956,
      "learning_rate": 0.00023092487199886266,
      "loss": 3.3947,
      "step": 24650
    },
    {
      "epoch": 0.323223320313524,
      "grad_norm": 0.5902895331382751,
      "learning_rate": 0.00023087251570402765,
      "loss": 3.3454,
      "step": 24660
    },
    {
      "epoch": 0.3233543922195717,
      "grad_norm": 0.5040431618690491,
      "learning_rate": 0.00023082014551507285,
      "loss": 3.3218,
      "step": 24670
    },
    {
      "epoch": 0.3234854641256193,
      "grad_norm": 0.5588535070419312,
      "learning_rate": 0.0002307677614409957,
      "loss": 3.3208,
      "step": 24680
    },
    {
      "epoch": 0.32361653603166696,
      "grad_norm": 0.5326235294342041,
      "learning_rate": 0.0002307153634907959,
      "loss": 3.3343,
      "step": 24690
    },
    {
      "epoch": 0.3237476079377146,
      "grad_norm": 0.558323085308075,
      "learning_rate": 0.00023066295167347553,
      "loss": 3.2824,
      "step": 24700
    },
    {
      "epoch": 0.3238786798437623,
      "grad_norm": 0.5081027746200562,
      "learning_rate": 0.00023061052599803903,
      "loss": 3.3425,
      "step": 24710
    },
    {
      "epoch": 0.32400975174980995,
      "grad_norm": 0.4926721155643463,
      "learning_rate": 0.0002305580864734934,
      "loss": 3.342,
      "step": 24720
    },
    {
      "epoch": 0.3241408236558576,
      "grad_norm": 0.618543803691864,
      "learning_rate": 0.00023050563310884778,
      "loss": 3.4195,
      "step": 24730
    },
    {
      "epoch": 0.3242718955619053,
      "grad_norm": 0.5557615756988525,
      "learning_rate": 0.0002304531659131139,
      "loss": 3.3354,
      "step": 24740
    },
    {
      "epoch": 0.32440296746795294,
      "grad_norm": 0.5804188251495361,
      "learning_rate": 0.00023040068489530571,
      "loss": 3.3138,
      "step": 24750
    },
    {
      "epoch": 0.32453403937400055,
      "grad_norm": 0.5183318853378296,
      "learning_rate": 0.0002303481900644396,
      "loss": 3.359,
      "step": 24760
    },
    {
      "epoch": 0.3246651112800482,
      "grad_norm": 0.4941486418247223,
      "learning_rate": 0.00023029568142953437,
      "loss": 3.3006,
      "step": 24770
    },
    {
      "epoch": 0.3247961831860959,
      "grad_norm": 0.5674406886100769,
      "learning_rate": 0.00023024315899961107,
      "loss": 3.3147,
      "step": 24780
    },
    {
      "epoch": 0.32492725509214354,
      "grad_norm": 0.5291242003440857,
      "learning_rate": 0.00023019062278369326,
      "loss": 3.35,
      "step": 24790
    },
    {
      "epoch": 0.3250583269981912,
      "grad_norm": 0.6311314702033997,
      "learning_rate": 0.00023013807279080676,
      "loss": 3.3525,
      "step": 24800
    },
    {
      "epoch": 0.32518939890423887,
      "grad_norm": 0.6107289791107178,
      "learning_rate": 0.00023008550902997983,
      "loss": 3.2801,
      "step": 24810
    },
    {
      "epoch": 0.32532047081028653,
      "grad_norm": 0.5657156705856323,
      "learning_rate": 0.00023003293151024304,
      "loss": 3.3699,
      "step": 24820
    },
    {
      "epoch": 0.3254515427163342,
      "grad_norm": 0.498994916677475,
      "learning_rate": 0.00022998034024062944,
      "loss": 3.3186,
      "step": 24830
    },
    {
      "epoch": 0.32558261462238186,
      "grad_norm": 0.6595579981803894,
      "learning_rate": 0.00022992773523017424,
      "loss": 3.3154,
      "step": 24840
    },
    {
      "epoch": 0.32571368652842947,
      "grad_norm": 0.5122869610786438,
      "learning_rate": 0.00022987511648791514,
      "loss": 3.3386,
      "step": 24850
    },
    {
      "epoch": 0.32584475843447713,
      "grad_norm": 0.5577048063278198,
      "learning_rate": 0.0002298224840228923,
      "loss": 3.3356,
      "step": 24860
    },
    {
      "epoch": 0.3259758303405248,
      "grad_norm": 0.569459855556488,
      "learning_rate": 0.00022976983784414799,
      "loss": 3.3391,
      "step": 24870
    },
    {
      "epoch": 0.32610690224657246,
      "grad_norm": 0.5799359083175659,
      "learning_rate": 0.000229717177960727,
      "loss": 3.3457,
      "step": 24880
    },
    {
      "epoch": 0.3262379741526201,
      "grad_norm": 0.5512623190879822,
      "learning_rate": 0.0002296645043816765,
      "loss": 3.336,
      "step": 24890
    },
    {
      "epoch": 0.3263690460586678,
      "grad_norm": 0.532384991645813,
      "learning_rate": 0.0002296118171160459,
      "loss": 3.2899,
      "step": 24900
    },
    {
      "epoch": 0.32650011796471545,
      "grad_norm": 0.5421558022499084,
      "learning_rate": 0.00022955911617288704,
      "loss": 3.3336,
      "step": 24910
    },
    {
      "epoch": 0.3266311898707631,
      "grad_norm": 0.6003760099411011,
      "learning_rate": 0.0002295064015612541,
      "loss": 3.3364,
      "step": 24920
    },
    {
      "epoch": 0.3267622617768108,
      "grad_norm": 0.5109248757362366,
      "learning_rate": 0.0002294536732902035,
      "loss": 3.3577,
      "step": 24930
    },
    {
      "epoch": 0.3268933336828584,
      "grad_norm": 0.5502159595489502,
      "learning_rate": 0.00022940093136879425,
      "loss": 3.3969,
      "step": 24940
    },
    {
      "epoch": 0.32702440558890605,
      "grad_norm": 0.5599809288978577,
      "learning_rate": 0.00022934817580608745,
      "loss": 3.389,
      "step": 24950
    },
    {
      "epoch": 0.3271554774949537,
      "grad_norm": 0.526731014251709,
      "learning_rate": 0.00022929540661114668,
      "loss": 3.3697,
      "step": 24960
    },
    {
      "epoch": 0.3272865494010014,
      "grad_norm": 0.5166134834289551,
      "learning_rate": 0.0002292426237930378,
      "loss": 3.3852,
      "step": 24970
    },
    {
      "epoch": 0.32741762130704904,
      "grad_norm": 0.5542776584625244,
      "learning_rate": 0.00022918982736082916,
      "loss": 3.3641,
      "step": 24980
    },
    {
      "epoch": 0.3275486932130967,
      "grad_norm": 0.5829073190689087,
      "learning_rate": 0.00022913701732359115,
      "loss": 3.3197,
      "step": 24990
    },
    {
      "epoch": 0.3276797651191444,
      "grad_norm": 0.5059958100318909,
      "learning_rate": 0.00022908419369039684,
      "loss": 3.3421,
      "step": 25000
    },
    {
      "epoch": 0.3276797651191444,
      "eval_accuracy": 0.387727761485826,
      "eval_loss": 3.32719087600708,
      "eval_runtime": 60.5728,
      "eval_samples_per_second": 33.018,
      "eval_steps_per_second": 0.528,
      "step": 25000
    },
    {
      "epoch": 0.32781083702519204,
      "grad_norm": 0.5312985777854919,
      "learning_rate": 0.00022903135647032137,
      "loss": 3.3303,
      "step": 25010
    },
    {
      "epoch": 0.3279419089312397,
      "grad_norm": 0.5154162049293518,
      "learning_rate": 0.00022897850567244238,
      "loss": 3.3609,
      "step": 25020
    },
    {
      "epoch": 0.3280729808372873,
      "grad_norm": 0.49089330434799194,
      "learning_rate": 0.0002289256413058398,
      "loss": 3.3645,
      "step": 25030
    },
    {
      "epoch": 0.328204052743335,
      "grad_norm": 0.551413893699646,
      "learning_rate": 0.00022887276337959576,
      "loss": 3.3462,
      "step": 25040
    },
    {
      "epoch": 0.32833512464938264,
      "grad_norm": 0.4554513394832611,
      "learning_rate": 0.00022881987190279498,
      "loss": 3.3321,
      "step": 25050
    },
    {
      "epoch": 0.3284661965554303,
      "grad_norm": 0.5596009492874146,
      "learning_rate": 0.00022876696688452425,
      "loss": 3.368,
      "step": 25060
    },
    {
      "epoch": 0.32859726846147796,
      "grad_norm": 0.6025522351264954,
      "learning_rate": 0.00022871404833387288,
      "loss": 3.2993,
      "step": 25070
    },
    {
      "epoch": 0.32872834036752563,
      "grad_norm": 0.5435642600059509,
      "learning_rate": 0.00022866111625993233,
      "loss": 3.326,
      "step": 25080
    },
    {
      "epoch": 0.3288594122735733,
      "grad_norm": 0.5140690207481384,
      "learning_rate": 0.00022860817067179654,
      "loss": 3.3424,
      "step": 25090
    },
    {
      "epoch": 0.32899048417962096,
      "grad_norm": 0.5136452913284302,
      "learning_rate": 0.00022855521157856163,
      "loss": 3.2969,
      "step": 25100
    },
    {
      "epoch": 0.3291215560856686,
      "grad_norm": 0.565992534160614,
      "learning_rate": 0.00022850223898932627,
      "loss": 3.3378,
      "step": 25110
    },
    {
      "epoch": 0.32925262799171623,
      "grad_norm": 0.599433183670044,
      "learning_rate": 0.0002284492529131911,
      "loss": 3.3512,
      "step": 25120
    },
    {
      "epoch": 0.3293836998977639,
      "grad_norm": 0.684576690196991,
      "learning_rate": 0.00022839625335925948,
      "loss": 3.3619,
      "step": 25130
    },
    {
      "epoch": 0.32951477180381156,
      "grad_norm": 0.5458829402923584,
      "learning_rate": 0.00022834324033663676,
      "loss": 3.3115,
      "step": 25140
    },
    {
      "epoch": 0.3296458437098592,
      "grad_norm": 0.5333219170570374,
      "learning_rate": 0.00022829021385443072,
      "loss": 3.3266,
      "step": 25150
    },
    {
      "epoch": 0.3297769156159069,
      "grad_norm": 0.49438950419425964,
      "learning_rate": 0.00022823717392175146,
      "loss": 3.3352,
      "step": 25160
    },
    {
      "epoch": 0.32990798752195455,
      "grad_norm": 0.5483959913253784,
      "learning_rate": 0.0002281841205477114,
      "loss": 3.32,
      "step": 25170
    },
    {
      "epoch": 0.3300390594280022,
      "grad_norm": 0.5511313080787659,
      "learning_rate": 0.00022813105374142525,
      "loss": 3.3235,
      "step": 25180
    },
    {
      "epoch": 0.3301701313340499,
      "grad_norm": 0.5942719578742981,
      "learning_rate": 0.00022807797351201002,
      "loss": 3.3788,
      "step": 25190
    },
    {
      "epoch": 0.33030120324009754,
      "grad_norm": 0.5393945574760437,
      "learning_rate": 0.0002280248798685851,
      "loss": 3.3493,
      "step": 25200
    },
    {
      "epoch": 0.33043227514614515,
      "grad_norm": 0.5138798952102661,
      "learning_rate": 0.00022797177282027204,
      "loss": 3.316,
      "step": 25210
    },
    {
      "epoch": 0.3305633470521928,
      "grad_norm": 0.5412576198577881,
      "learning_rate": 0.00022791865237619484,
      "loss": 3.3393,
      "step": 25220
    },
    {
      "epoch": 0.3306944189582405,
      "grad_norm": 0.5028805732727051,
      "learning_rate": 0.0002278655185454797,
      "loss": 3.3046,
      "step": 25230
    },
    {
      "epoch": 0.33082549086428814,
      "grad_norm": 0.5372729897499084,
      "learning_rate": 0.00022781237133725517,
      "loss": 3.3451,
      "step": 25240
    },
    {
      "epoch": 0.3309565627703358,
      "grad_norm": 0.6325402855873108,
      "learning_rate": 0.00022775921076065205,
      "loss": 3.3823,
      "step": 25250
    },
    {
      "epoch": 0.33108763467638347,
      "grad_norm": 0.6111400127410889,
      "learning_rate": 0.00022770603682480356,
      "loss": 3.3829,
      "step": 25260
    },
    {
      "epoch": 0.33121870658243113,
      "grad_norm": 0.6283338069915771,
      "learning_rate": 0.00022765284953884507,
      "loss": 3.3643,
      "step": 25270
    },
    {
      "epoch": 0.3313497784884788,
      "grad_norm": 0.528610348701477,
      "learning_rate": 0.0002275996489119143,
      "loss": 3.3577,
      "step": 25280
    },
    {
      "epoch": 0.33148085039452646,
      "grad_norm": 0.5360432863235474,
      "learning_rate": 0.00022754643495315123,
      "loss": 3.3551,
      "step": 25290
    },
    {
      "epoch": 0.33161192230057407,
      "grad_norm": 0.6980841159820557,
      "learning_rate": 0.00022749320767169822,
      "loss": 3.396,
      "step": 25300
    },
    {
      "epoch": 0.33174299420662173,
      "grad_norm": 0.5268281102180481,
      "learning_rate": 0.0002274399670766999,
      "loss": 3.3595,
      "step": 25310
    },
    {
      "epoch": 0.3318740661126694,
      "grad_norm": 0.5570038557052612,
      "learning_rate": 0.0002273867131773031,
      "loss": 3.3218,
      "step": 25320
    },
    {
      "epoch": 0.33200513801871706,
      "grad_norm": 0.6057099103927612,
      "learning_rate": 0.00022733344598265695,
      "loss": 3.3221,
      "step": 25330
    },
    {
      "epoch": 0.3321362099247647,
      "grad_norm": 0.508512020111084,
      "learning_rate": 0.0002272801655019129,
      "loss": 3.3125,
      "step": 25340
    },
    {
      "epoch": 0.3322672818308124,
      "grad_norm": 0.571563720703125,
      "learning_rate": 0.0002272268717442248,
      "loss": 3.3536,
      "step": 25350
    },
    {
      "epoch": 0.33239835373686005,
      "grad_norm": 0.5390920639038086,
      "learning_rate": 0.0002271735647187485,
      "loss": 3.3068,
      "step": 25360
    },
    {
      "epoch": 0.3325294256429077,
      "grad_norm": 0.5680702924728394,
      "learning_rate": 0.0002271202444346424,
      "loss": 3.2836,
      "step": 25370
    },
    {
      "epoch": 0.3326604975489554,
      "grad_norm": 0.5302940011024475,
      "learning_rate": 0.00022706691090106703,
      "loss": 3.3469,
      "step": 25380
    },
    {
      "epoch": 0.332791569455003,
      "grad_norm": 0.616568386554718,
      "learning_rate": 0.0002270135641271853,
      "loss": 3.4169,
      "step": 25390
    },
    {
      "epoch": 0.33292264136105065,
      "grad_norm": 0.5214830040931702,
      "learning_rate": 0.00022696020412216224,
      "loss": 3.3364,
      "step": 25400
    },
    {
      "epoch": 0.3330537132670983,
      "grad_norm": 0.5918077230453491,
      "learning_rate": 0.00022690683089516528,
      "loss": 3.3526,
      "step": 25410
    },
    {
      "epoch": 0.333184785173146,
      "grad_norm": 0.5025444626808167,
      "learning_rate": 0.00022685344445536415,
      "loss": 3.3669,
      "step": 25420
    },
    {
      "epoch": 0.33331585707919364,
      "grad_norm": 0.5656698346138,
      "learning_rate": 0.00022680004481193068,
      "loss": 3.3463,
      "step": 25430
    },
    {
      "epoch": 0.3334469289852413,
      "grad_norm": 0.6402048468589783,
      "learning_rate": 0.00022674663197403913,
      "loss": 3.438,
      "step": 25440
    },
    {
      "epoch": 0.333578000891289,
      "grad_norm": 0.534653902053833,
      "learning_rate": 0.000226693205950866,
      "loss": 3.3242,
      "step": 25450
    },
    {
      "epoch": 0.33370907279733664,
      "grad_norm": 0.548815906047821,
      "learning_rate": 0.00022663976675158995,
      "loss": 3.3718,
      "step": 25460
    },
    {
      "epoch": 0.3338401447033843,
      "grad_norm": 0.5658780336380005,
      "learning_rate": 0.00022658631438539204,
      "loss": 3.3774,
      "step": 25470
    },
    {
      "epoch": 0.3339712166094319,
      "grad_norm": 0.5161982774734497,
      "learning_rate": 0.00022653284886145548,
      "loss": 3.3361,
      "step": 25480
    },
    {
      "epoch": 0.3341022885154796,
      "grad_norm": 0.5422012805938721,
      "learning_rate": 0.0002264793701889659,
      "loss": 3.3042,
      "step": 25490
    },
    {
      "epoch": 0.33423336042152724,
      "grad_norm": 0.5250160098075867,
      "learning_rate": 0.00022642587837711102,
      "loss": 3.343,
      "step": 25500
    },
    {
      "epoch": 0.33423336042152724,
      "eval_accuracy": 0.3876231671554252,
      "eval_loss": 3.325732946395874,
      "eval_runtime": 60.2208,
      "eval_samples_per_second": 33.211,
      "eval_steps_per_second": 0.531,
      "step": 25500
    },
    {
      "epoch": 0.3343644323275749,
      "grad_norm": 0.7093520760536194,
      "learning_rate": 0.00022637237343508084,
      "loss": 3.3406,
      "step": 25510
    },
    {
      "epoch": 0.33449550423362256,
      "grad_norm": 0.6305891871452332,
      "learning_rate": 0.00022631885537206767,
      "loss": 3.4159,
      "step": 25520
    },
    {
      "epoch": 0.33462657613967023,
      "grad_norm": 0.5617510080337524,
      "learning_rate": 0.00022626532419726614,
      "loss": 3.3112,
      "step": 25530
    },
    {
      "epoch": 0.3347576480457179,
      "grad_norm": 0.5519635677337646,
      "learning_rate": 0.00022621177991987294,
      "loss": 3.3452,
      "step": 25540
    },
    {
      "epoch": 0.33488871995176556,
      "grad_norm": 0.6788350343704224,
      "learning_rate": 0.0002261582225490872,
      "loss": 3.3136,
      "step": 25550
    },
    {
      "epoch": 0.3350197918578132,
      "grad_norm": 0.5327380895614624,
      "learning_rate": 0.00022610465209411016,
      "loss": 3.2817,
      "step": 25560
    },
    {
      "epoch": 0.33515086376386083,
      "grad_norm": 0.5404787063598633,
      "learning_rate": 0.00022605106856414547,
      "loss": 3.3653,
      "step": 25570
    },
    {
      "epoch": 0.3352819356699085,
      "grad_norm": 0.5646034479141235,
      "learning_rate": 0.00022599747196839882,
      "loss": 3.3807,
      "step": 25580
    },
    {
      "epoch": 0.33541300757595616,
      "grad_norm": 0.7309812307357788,
      "learning_rate": 0.00022594386231607834,
      "loss": 3.327,
      "step": 25590
    },
    {
      "epoch": 0.3355440794820038,
      "grad_norm": 0.6137698888778687,
      "learning_rate": 0.00022589023961639424,
      "loss": 3.317,
      "step": 25600
    },
    {
      "epoch": 0.3356751513880515,
      "grad_norm": 0.5109633207321167,
      "learning_rate": 0.0002258366038785591,
      "loss": 3.3462,
      "step": 25610
    },
    {
      "epoch": 0.33580622329409915,
      "grad_norm": 0.5439435243606567,
      "learning_rate": 0.00022578295511178762,
      "loss": 3.3167,
      "step": 25620
    },
    {
      "epoch": 0.3359372952001468,
      "grad_norm": 0.4977980852127075,
      "learning_rate": 0.00022572929332529692,
      "loss": 3.3218,
      "step": 25630
    },
    {
      "epoch": 0.3360683671061945,
      "grad_norm": 0.8358725905418396,
      "learning_rate": 0.00022567561852830613,
      "loss": 3.3495,
      "step": 25640
    },
    {
      "epoch": 0.33619943901224214,
      "grad_norm": 0.5136901140213013,
      "learning_rate": 0.00022562193073003678,
      "loss": 3.3532,
      "step": 25650
    },
    {
      "epoch": 0.33633051091828975,
      "grad_norm": 0.869697630405426,
      "learning_rate": 0.0002255682299397126,
      "loss": 3.3425,
      "step": 25660
    },
    {
      "epoch": 0.3364615828243374,
      "grad_norm": 0.5948277115821838,
      "learning_rate": 0.00022551451616655948,
      "loss": 3.3123,
      "step": 25670
    },
    {
      "epoch": 0.3365926547303851,
      "grad_norm": 0.6004504561424255,
      "learning_rate": 0.00022546078941980565,
      "loss": 3.3519,
      "step": 25680
    },
    {
      "epoch": 0.33672372663643274,
      "grad_norm": 0.5589486956596375,
      "learning_rate": 0.00022540704970868143,
      "loss": 3.4262,
      "step": 25690
    },
    {
      "epoch": 0.3368547985424804,
      "grad_norm": 0.5373388528823853,
      "learning_rate": 0.00022535329704241955,
      "loss": 3.304,
      "step": 25700
    },
    {
      "epoch": 0.33698587044852807,
      "grad_norm": 0.6315534114837646,
      "learning_rate": 0.00022529953143025477,
      "loss": 3.3082,
      "step": 25710
    },
    {
      "epoch": 0.33711694235457573,
      "grad_norm": 0.5646101236343384,
      "learning_rate": 0.00022524575288142425,
      "loss": 3.2555,
      "step": 25720
    },
    {
      "epoch": 0.3372480142606234,
      "grad_norm": 0.740477442741394,
      "learning_rate": 0.00022519196140516725,
      "loss": 3.3478,
      "step": 25730
    },
    {
      "epoch": 0.33737908616667106,
      "grad_norm": 0.5566022992134094,
      "learning_rate": 0.00022513815701072527,
      "loss": 3.3436,
      "step": 25740
    },
    {
      "epoch": 0.33751015807271867,
      "grad_norm": 0.5270554423332214,
      "learning_rate": 0.00022508433970734208,
      "loss": 3.3375,
      "step": 25750
    },
    {
      "epoch": 0.33764122997876633,
      "grad_norm": 0.6114169955253601,
      "learning_rate": 0.00022503050950426368,
      "loss": 3.3564,
      "step": 25760
    },
    {
      "epoch": 0.337772301884814,
      "grad_norm": 0.5530643463134766,
      "learning_rate": 0.00022497666641073822,
      "loss": 3.4084,
      "step": 25770
    },
    {
      "epoch": 0.33790337379086166,
      "grad_norm": 0.5578331351280212,
      "learning_rate": 0.00022492281043601605,
      "loss": 3.3375,
      "step": 25780
    },
    {
      "epoch": 0.3380344456969093,
      "grad_norm": 0.5262152552604675,
      "learning_rate": 0.00022486894158934984,
      "loss": 3.3717,
      "step": 25790
    },
    {
      "epoch": 0.338165517602957,
      "grad_norm": 0.5571235418319702,
      "learning_rate": 0.00022481505987999434,
      "loss": 3.3588,
      "step": 25800
    },
    {
      "epoch": 0.33829658950900465,
      "grad_norm": 0.5518863201141357,
      "learning_rate": 0.0002247611653172066,
      "loss": 3.3268,
      "step": 25810
    },
    {
      "epoch": 0.3384276614150523,
      "grad_norm": 0.5284160375595093,
      "learning_rate": 0.00022470725791024586,
      "loss": 3.2966,
      "step": 25820
    },
    {
      "epoch": 0.3385587333211,
      "grad_norm": 0.6205552816390991,
      "learning_rate": 0.00022465333766837357,
      "loss": 3.3027,
      "step": 25830
    },
    {
      "epoch": 0.3386898052271476,
      "grad_norm": 0.4998530149459839,
      "learning_rate": 0.00022459940460085333,
      "loss": 3.3594,
      "step": 25840
    },
    {
      "epoch": 0.33882087713319525,
      "grad_norm": 0.583716630935669,
      "learning_rate": 0.00022454545871695105,
      "loss": 3.3223,
      "step": 25850
    },
    {
      "epoch": 0.3389519490392429,
      "grad_norm": 0.5280960202217102,
      "learning_rate": 0.00022449150002593478,
      "loss": 3.3399,
      "step": 25860
    },
    {
      "epoch": 0.3390830209452906,
      "grad_norm": 0.5291954874992371,
      "learning_rate": 0.00022443752853707472,
      "loss": 3.3167,
      "step": 25870
    },
    {
      "epoch": 0.33921409285133824,
      "grad_norm": 0.5150972008705139,
      "learning_rate": 0.00022438354425964333,
      "loss": 3.3506,
      "step": 25880
    },
    {
      "epoch": 0.3393451647573859,
      "grad_norm": 0.5403798818588257,
      "learning_rate": 0.0002243295472029153,
      "loss": 3.3571,
      "step": 25890
    },
    {
      "epoch": 0.3394762366634336,
      "grad_norm": 0.5489093065261841,
      "learning_rate": 0.00022427553737616743,
      "loss": 3.3367,
      "step": 25900
    },
    {
      "epoch": 0.33960730856948124,
      "grad_norm": 0.5525540709495544,
      "learning_rate": 0.00022422151478867873,
      "loss": 3.3403,
      "step": 25910
    },
    {
      "epoch": 0.3397383804755289,
      "grad_norm": 0.5487809181213379,
      "learning_rate": 0.0002241674794497305,
      "loss": 3.3506,
      "step": 25920
    },
    {
      "epoch": 0.3398694523815765,
      "grad_norm": 0.5157622694969177,
      "learning_rate": 0.00022411343136860613,
      "loss": 3.3333,
      "step": 25930
    },
    {
      "epoch": 0.3400005242876242,
      "grad_norm": 0.5318061113357544,
      "learning_rate": 0.00022405937055459122,
      "loss": 3.3272,
      "step": 25940
    },
    {
      "epoch": 0.34013159619367184,
      "grad_norm": 0.5220856070518494,
      "learning_rate": 0.00022400529701697362,
      "loss": 3.3523,
      "step": 25950
    },
    {
      "epoch": 0.3402626680997195,
      "grad_norm": 0.6059438586235046,
      "learning_rate": 0.00022395121076504318,
      "loss": 3.3584,
      "step": 25960
    },
    {
      "epoch": 0.34039374000576716,
      "grad_norm": 0.5101557970046997,
      "learning_rate": 0.0002238971118080922,
      "loss": 3.2779,
      "step": 25970
    },
    {
      "epoch": 0.34052481191181483,
      "grad_norm": 0.5389559268951416,
      "learning_rate": 0.00022384300015541495,
      "loss": 3.318,
      "step": 25980
    },
    {
      "epoch": 0.3406558838178625,
      "grad_norm": 0.7537828683853149,
      "learning_rate": 0.00022378887581630798,
      "loss": 3.3256,
      "step": 25990
    },
    {
      "epoch": 0.34078695572391016,
      "grad_norm": 0.5684036016464233,
      "learning_rate": 0.00022373473880007007,
      "loss": 3.3811,
      "step": 26000
    },
    {
      "epoch": 0.34078695572391016,
      "eval_accuracy": 0.38715542521994134,
      "eval_loss": 3.3294451236724854,
      "eval_runtime": 59.3984,
      "eval_samples_per_second": 33.671,
      "eval_steps_per_second": 0.539,
      "step": 26000
    },
    {
      "epoch": 0.3409180276299578,
      "grad_norm": 0.5059719085693359,
      "learning_rate": 0.00022368058911600197,
      "loss": 3.3384,
      "step": 26010
    },
    {
      "epoch": 0.34104909953600543,
      "grad_norm": 0.5321153402328491,
      "learning_rate": 0.00022362642677340686,
      "loss": 3.3571,
      "step": 26020
    },
    {
      "epoch": 0.3411801714420531,
      "grad_norm": 0.5887495875358582,
      "learning_rate": 0.00022357225178158994,
      "loss": 3.3519,
      "step": 26030
    },
    {
      "epoch": 0.34131124334810076,
      "grad_norm": 0.5449793934822083,
      "learning_rate": 0.00022351806414985863,
      "loss": 3.3693,
      "step": 26040
    },
    {
      "epoch": 0.3414423152541484,
      "grad_norm": 0.5872164964675903,
      "learning_rate": 0.0002234638638875225,
      "loss": 3.3432,
      "step": 26050
    },
    {
      "epoch": 0.3415733871601961,
      "grad_norm": 0.5388712882995605,
      "learning_rate": 0.00022340965100389328,
      "loss": 3.3125,
      "step": 26060
    },
    {
      "epoch": 0.34170445906624375,
      "grad_norm": 0.5831034779548645,
      "learning_rate": 0.00022335542550828494,
      "loss": 3.3234,
      "step": 26070
    },
    {
      "epoch": 0.3418355309722914,
      "grad_norm": 0.5733153223991394,
      "learning_rate": 0.00022330118741001353,
      "loss": 3.3707,
      "step": 26080
    },
    {
      "epoch": 0.3419666028783391,
      "grad_norm": 0.4896855354309082,
      "learning_rate": 0.00022324693671839732,
      "loss": 3.3575,
      "step": 26090
    },
    {
      "epoch": 0.34209767478438674,
      "grad_norm": 0.5552607774734497,
      "learning_rate": 0.00022319267344275676,
      "loss": 3.2931,
      "step": 26100
    },
    {
      "epoch": 0.34222874669043435,
      "grad_norm": 0.5311570763587952,
      "learning_rate": 0.0002231383975924144,
      "loss": 3.3693,
      "step": 26110
    },
    {
      "epoch": 0.342359818596482,
      "grad_norm": 0.539807140827179,
      "learning_rate": 0.00022308410917669497,
      "loss": 3.2675,
      "step": 26120
    },
    {
      "epoch": 0.3424908905025297,
      "grad_norm": 0.4946378469467163,
      "learning_rate": 0.00022302980820492537,
      "loss": 3.3522,
      "step": 26130
    },
    {
      "epoch": 0.34262196240857734,
      "grad_norm": 0.6487476229667664,
      "learning_rate": 0.00022297549468643465,
      "loss": 3.3281,
      "step": 26140
    },
    {
      "epoch": 0.342753034314625,
      "grad_norm": 0.5502750873565674,
      "learning_rate": 0.00022292116863055408,
      "loss": 3.3096,
      "step": 26150
    },
    {
      "epoch": 0.34288410622067267,
      "grad_norm": 0.5484684109687805,
      "learning_rate": 0.00022286683004661692,
      "loss": 3.3497,
      "step": 26160
    },
    {
      "epoch": 0.34301517812672033,
      "grad_norm": 1.0578467845916748,
      "learning_rate": 0.00022281247894395876,
      "loss": 3.4019,
      "step": 26170
    },
    {
      "epoch": 0.343146250032768,
      "grad_norm": 0.666335940361023,
      "learning_rate": 0.00022275811533191729,
      "loss": 3.3652,
      "step": 26180
    },
    {
      "epoch": 0.34327732193881566,
      "grad_norm": 0.7150986790657043,
      "learning_rate": 0.00022270373921983228,
      "loss": 3.3526,
      "step": 26190
    },
    {
      "epoch": 0.34340839384486327,
      "grad_norm": 0.5449596047401428,
      "learning_rate": 0.0002226493506170457,
      "loss": 3.3366,
      "step": 26200
    },
    {
      "epoch": 0.34353946575091093,
      "grad_norm": 0.5346696376800537,
      "learning_rate": 0.0002225949495329017,
      "loss": 3.3026,
      "step": 26210
    },
    {
      "epoch": 0.3436705376569586,
      "grad_norm": 0.5896271467208862,
      "learning_rate": 0.00022254053597674643,
      "loss": 3.3605,
      "step": 26220
    },
    {
      "epoch": 0.34380160956300626,
      "grad_norm": 0.5122628808021545,
      "learning_rate": 0.0002224861099579284,
      "loss": 3.3309,
      "step": 26230
    },
    {
      "epoch": 0.3439326814690539,
      "grad_norm": 0.5827341675758362,
      "learning_rate": 0.00022243167148579813,
      "loss": 3.3595,
      "step": 26240
    },
    {
      "epoch": 0.3440637533751016,
      "grad_norm": 0.5335626602172852,
      "learning_rate": 0.00022237722056970825,
      "loss": 3.2798,
      "step": 26250
    },
    {
      "epoch": 0.34419482528114925,
      "grad_norm": 0.5083019733428955,
      "learning_rate": 0.00022232275721901365,
      "loss": 3.3334,
      "step": 26260
    },
    {
      "epoch": 0.3443258971871969,
      "grad_norm": 0.5267325639724731,
      "learning_rate": 0.00022226828144307118,
      "loss": 3.3444,
      "step": 26270
    },
    {
      "epoch": 0.3444569690932446,
      "grad_norm": 0.5573915243148804,
      "learning_rate": 0.00022221379325124003,
      "loss": 3.3248,
      "step": 26280
    },
    {
      "epoch": 0.3445880409992922,
      "grad_norm": 0.5412414073944092,
      "learning_rate": 0.00022215929265288138,
      "loss": 3.3753,
      "step": 26290
    },
    {
      "epoch": 0.34471911290533985,
      "grad_norm": 0.5282171368598938,
      "learning_rate": 0.0002221047796573586,
      "loss": 3.3208,
      "step": 26300
    },
    {
      "epoch": 0.3448501848113875,
      "grad_norm": 0.48896074295043945,
      "learning_rate": 0.00022205025427403713,
      "loss": 3.3505,
      "step": 26310
    },
    {
      "epoch": 0.3449812567174352,
      "grad_norm": 0.5482248663902283,
      "learning_rate": 0.00022199571651228454,
      "loss": 3.3519,
      "step": 26320
    },
    {
      "epoch": 0.34511232862348284,
      "grad_norm": 0.5481003522872925,
      "learning_rate": 0.00022194116638147072,
      "loss": 3.3586,
      "step": 26330
    },
    {
      "epoch": 0.3452434005295305,
      "grad_norm": 0.5690895915031433,
      "learning_rate": 0.00022188660389096737,
      "loss": 3.3394,
      "step": 26340
    },
    {
      "epoch": 0.3453744724355782,
      "grad_norm": 0.5522475242614746,
      "learning_rate": 0.00022183202905014858,
      "loss": 3.3374,
      "step": 26350
    },
    {
      "epoch": 0.34550554434162584,
      "grad_norm": 0.556363046169281,
      "learning_rate": 0.0002217774418683904,
      "loss": 3.3489,
      "step": 26360
    },
    {
      "epoch": 0.3456366162476735,
      "grad_norm": 0.6590280532836914,
      "learning_rate": 0.00022172284235507113,
      "loss": 3.362,
      "step": 26370
    },
    {
      "epoch": 0.3457676881537211,
      "grad_norm": 0.577272355556488,
      "learning_rate": 0.000221668230519571,
      "loss": 3.2976,
      "step": 26380
    },
    {
      "epoch": 0.3458987600597688,
      "grad_norm": 0.508924126625061,
      "learning_rate": 0.0002216136063712726,
      "loss": 3.3718,
      "step": 26390
    },
    {
      "epoch": 0.34602983196581644,
      "grad_norm": 0.5343247652053833,
      "learning_rate": 0.00022155896991956045,
      "loss": 3.3399,
      "step": 26400
    },
    {
      "epoch": 0.3461609038718641,
      "grad_norm": 0.5609007477760315,
      "learning_rate": 0.0002215043211738212,
      "loss": 3.2983,
      "step": 26410
    },
    {
      "epoch": 0.34629197577791176,
      "grad_norm": 0.5878892540931702,
      "learning_rate": 0.0002214496601434437,
      "loss": 3.3552,
      "step": 26420
    },
    {
      "epoch": 0.34642304768395943,
      "grad_norm": 0.5814974308013916,
      "learning_rate": 0.00022139498683781887,
      "loss": 3.3313,
      "step": 26430
    },
    {
      "epoch": 0.3465541195900071,
      "grad_norm": 0.8331164121627808,
      "learning_rate": 0.00022134030126633972,
      "loss": 3.3425,
      "step": 26440
    },
    {
      "epoch": 0.34668519149605476,
      "grad_norm": 0.5913481712341309,
      "learning_rate": 0.00022128560343840135,
      "loss": 3.2945,
      "step": 26450
    },
    {
      "epoch": 0.3468162634021024,
      "grad_norm": 0.5799285769462585,
      "learning_rate": 0.00022123089336340107,
      "loss": 3.3515,
      "step": 26460
    },
    {
      "epoch": 0.34694733530815003,
      "grad_norm": 0.565071165561676,
      "learning_rate": 0.00022117617105073815,
      "loss": 3.3688,
      "step": 26470
    },
    {
      "epoch": 0.3470784072141977,
      "grad_norm": 0.566055417060852,
      "learning_rate": 0.00022112143650981412,
      "loss": 3.3471,
      "step": 26480
    },
    {
      "epoch": 0.34720947912024536,
      "grad_norm": 0.5989728569984436,
      "learning_rate": 0.0002210666897500324,
      "loss": 3.3351,
      "step": 26490
    },
    {
      "epoch": 0.347340551026293,
      "grad_norm": 0.5150706171989441,
      "learning_rate": 0.00022101193078079875,
      "loss": 3.356,
      "step": 26500
    },
    {
      "epoch": 0.347340551026293,
      "eval_accuracy": 0.3891080156402737,
      "eval_loss": 3.3182857036590576,
      "eval_runtime": 62.2039,
      "eval_samples_per_second": 32.152,
      "eval_steps_per_second": 0.514,
      "step": 26500
    },
    {
      "epoch": 0.3474716229323407,
      "grad_norm": 0.5260719656944275,
      "learning_rate": 0.00022095715961152086,
      "loss": 3.3456,
      "step": 26510
    },
    {
      "epoch": 0.34760269483838835,
      "grad_norm": 0.5380013585090637,
      "learning_rate": 0.00022090237625160854,
      "loss": 3.3137,
      "step": 26520
    },
    {
      "epoch": 0.347733766744436,
      "grad_norm": 0.5639862418174744,
      "learning_rate": 0.0002208475807104738,
      "loss": 3.3364,
      "step": 26530
    },
    {
      "epoch": 0.3478648386504837,
      "grad_norm": 0.5913389325141907,
      "learning_rate": 0.00022079277299753058,
      "loss": 3.3296,
      "step": 26540
    },
    {
      "epoch": 0.34799591055653134,
      "grad_norm": 0.5360478162765503,
      "learning_rate": 0.00022073795312219503,
      "loss": 3.335,
      "step": 26550
    },
    {
      "epoch": 0.34812698246257895,
      "grad_norm": 0.6009983420372009,
      "learning_rate": 0.00022068312109388533,
      "loss": 3.3502,
      "step": 26560
    },
    {
      "epoch": 0.3482580543686266,
      "grad_norm": 0.5432466864585876,
      "learning_rate": 0.00022062827692202183,
      "loss": 3.3464,
      "step": 26570
    },
    {
      "epoch": 0.3483891262746743,
      "grad_norm": 0.5514429211616516,
      "learning_rate": 0.0002205734206160269,
      "loss": 3.3558,
      "step": 26580
    },
    {
      "epoch": 0.34852019818072194,
      "grad_norm": 0.5631410479545593,
      "learning_rate": 0.00022051855218532492,
      "loss": 3.355,
      "step": 26590
    },
    {
      "epoch": 0.3486512700867696,
      "grad_norm": 0.544377863407135,
      "learning_rate": 0.0002204636716393425,
      "loss": 3.3587,
      "step": 26600
    },
    {
      "epoch": 0.34878234199281727,
      "grad_norm": 0.5305700302124023,
      "learning_rate": 0.00022040877898750835,
      "loss": 3.3231,
      "step": 26610
    },
    {
      "epoch": 0.34891341389886493,
      "grad_norm": 0.5499244332313538,
      "learning_rate": 0.00022035387423925297,
      "loss": 3.3583,
      "step": 26620
    },
    {
      "epoch": 0.3490444858049126,
      "grad_norm": 0.5287545323371887,
      "learning_rate": 0.00022029895740400935,
      "loss": 3.3408,
      "step": 26630
    },
    {
      "epoch": 0.34917555771096026,
      "grad_norm": 0.5258705019950867,
      "learning_rate": 0.00022024402849121227,
      "loss": 3.3486,
      "step": 26640
    },
    {
      "epoch": 0.34930662961700787,
      "grad_norm": 0.5038305521011353,
      "learning_rate": 0.00022018908751029865,
      "loss": 3.3304,
      "step": 26650
    },
    {
      "epoch": 0.34943770152305553,
      "grad_norm": 0.5099867582321167,
      "learning_rate": 0.00022013413447070752,
      "loss": 3.3792,
      "step": 26660
    },
    {
      "epoch": 0.3495687734291032,
      "grad_norm": 0.5263105630874634,
      "learning_rate": 0.00022007916938187997,
      "loss": 3.3305,
      "step": 26670
    },
    {
      "epoch": 0.34969984533515086,
      "grad_norm": 0.5036457180976868,
      "learning_rate": 0.00022002419225325916,
      "loss": 3.3195,
      "step": 26680
    },
    {
      "epoch": 0.3498309172411985,
      "grad_norm": 0.6014471054077148,
      "learning_rate": 0.00021996920309429025,
      "loss": 3.4631,
      "step": 26690
    },
    {
      "epoch": 0.3499619891472462,
      "grad_norm": 0.5545088052749634,
      "learning_rate": 0.00021991420191442063,
      "loss": 3.3855,
      "step": 26700
    },
    {
      "epoch": 0.35009306105329385,
      "grad_norm": 0.5702374577522278,
      "learning_rate": 0.00021985918872309955,
      "loss": 3.3774,
      "step": 26710
    },
    {
      "epoch": 0.3502241329593415,
      "grad_norm": 0.5243615508079529,
      "learning_rate": 0.0002198041635297785,
      "loss": 3.3401,
      "step": 26720
    },
    {
      "epoch": 0.3503552048653891,
      "grad_norm": 0.5358230471611023,
      "learning_rate": 0.00021974912634391093,
      "loss": 3.376,
      "step": 26730
    },
    {
      "epoch": 0.3504862767714368,
      "grad_norm": 0.5400069952011108,
      "learning_rate": 0.0002196940771749524,
      "loss": 3.3455,
      "step": 26740
    },
    {
      "epoch": 0.35061734867748445,
      "grad_norm": 0.5308468341827393,
      "learning_rate": 0.00021963901603236046,
      "loss": 3.3384,
      "step": 26750
    },
    {
      "epoch": 0.3507484205835321,
      "grad_norm": 0.4927447438240051,
      "learning_rate": 0.00021958394292559484,
      "loss": 3.344,
      "step": 26760
    },
    {
      "epoch": 0.3508794924895798,
      "grad_norm": 0.4836474061012268,
      "learning_rate": 0.00021952885786411718,
      "loss": 3.3051,
      "step": 26770
    },
    {
      "epoch": 0.35101056439562744,
      "grad_norm": 0.5091266632080078,
      "learning_rate": 0.00021947376085739126,
      "loss": 3.2987,
      "step": 26780
    },
    {
      "epoch": 0.3511416363016751,
      "grad_norm": 0.5794975161552429,
      "learning_rate": 0.00021941865191488292,
      "loss": 3.3361,
      "step": 26790
    },
    {
      "epoch": 0.3512727082077228,
      "grad_norm": 0.5379009246826172,
      "learning_rate": 0.00021936353104606002,
      "loss": 3.3231,
      "step": 26800
    },
    {
      "epoch": 0.35140378011377044,
      "grad_norm": 0.5413955450057983,
      "learning_rate": 0.0002193083982603925,
      "loss": 3.3445,
      "step": 26810
    },
    {
      "epoch": 0.35153485201981804,
      "grad_norm": 0.7251079678535461,
      "learning_rate": 0.0002192532535673522,
      "loss": 3.3239,
      "step": 26820
    },
    {
      "epoch": 0.3516659239258657,
      "grad_norm": 0.5064908266067505,
      "learning_rate": 0.00021919809697641335,
      "loss": 3.3482,
      "step": 26830
    },
    {
      "epoch": 0.3517969958319134,
      "grad_norm": 0.5196040272712708,
      "learning_rate": 0.0002191429284970518,
      "loss": 3.3355,
      "step": 26840
    },
    {
      "epoch": 0.35192806773796104,
      "grad_norm": 0.5542014241218567,
      "learning_rate": 0.00021908774813874578,
      "loss": 3.2991,
      "step": 26850
    },
    {
      "epoch": 0.3520591396440087,
      "grad_norm": 0.541148841381073,
      "learning_rate": 0.0002190325559109753,
      "loss": 3.3137,
      "step": 26860
    },
    {
      "epoch": 0.35219021155005636,
      "grad_norm": 0.6737114787101746,
      "learning_rate": 0.00021897735182322268,
      "loss": 3.3408,
      "step": 26870
    },
    {
      "epoch": 0.35232128345610403,
      "grad_norm": 0.5837387442588806,
      "learning_rate": 0.000218922135884972,
      "loss": 3.3432,
      "step": 26880
    },
    {
      "epoch": 0.3524523553621517,
      "grad_norm": 0.6841250061988831,
      "learning_rate": 0.0002188669081057096,
      "loss": 3.3321,
      "step": 26890
    },
    {
      "epoch": 0.35258342726819936,
      "grad_norm": 0.5512218475341797,
      "learning_rate": 0.00021881166849492374,
      "loss": 3.3523,
      "step": 26900
    },
    {
      "epoch": 0.35271449917424696,
      "grad_norm": 0.5232782959938049,
      "learning_rate": 0.0002187564170621047,
      "loss": 3.4066,
      "step": 26910
    },
    {
      "epoch": 0.35284557108029463,
      "grad_norm": 0.6307722330093384,
      "learning_rate": 0.0002187011538167449,
      "loss": 3.3637,
      "step": 26920
    },
    {
      "epoch": 0.3529766429863423,
      "grad_norm": 0.582274854183197,
      "learning_rate": 0.00021864587876833867,
      "loss": 3.3064,
      "step": 26930
    },
    {
      "epoch": 0.35310771489238996,
      "grad_norm": 0.566235363483429,
      "learning_rate": 0.0002185905919263824,
      "loss": 3.4047,
      "step": 26940
    },
    {
      "epoch": 0.3532387867984376,
      "grad_norm": 0.5507163405418396,
      "learning_rate": 0.00021853529330037453,
      "loss": 3.3374,
      "step": 26950
    },
    {
      "epoch": 0.3533698587044853,
      "grad_norm": 0.5551252961158752,
      "learning_rate": 0.00021847998289981557,
      "loss": 3.37,
      "step": 26960
    },
    {
      "epoch": 0.35350093061053295,
      "grad_norm": 0.5404503345489502,
      "learning_rate": 0.00021842466073420787,
      "loss": 3.3903,
      "step": 26970
    },
    {
      "epoch": 0.3536320025165806,
      "grad_norm": 0.5261572003364563,
      "learning_rate": 0.00021836932681305608,
      "loss": 3.3312,
      "step": 26980
    },
    {
      "epoch": 0.3537630744226283,
      "grad_norm": 0.5402191877365112,
      "learning_rate": 0.00021831398114586663,
      "loss": 3.3263,
      "step": 26990
    },
    {
      "epoch": 0.3538941463286759,
      "grad_norm": 0.5918896198272705,
      "learning_rate": 0.0002182586237421481,
      "loss": 3.315,
      "step": 27000
    },
    {
      "epoch": 0.3538941463286759,
      "eval_accuracy": 0.38842717497556206,
      "eval_loss": 3.3184900283813477,
      "eval_runtime": 61.5487,
      "eval_samples_per_second": 32.495,
      "eval_steps_per_second": 0.52,
      "step": 27000
    },
    {
      "epoch": 0.35402521823472355,
      "grad_norm": 0.5640950202941895,
      "learning_rate": 0.00021820325461141104,
      "loss": 3.3639,
      "step": 27010
    },
    {
      "epoch": 0.3541562901407712,
      "grad_norm": 0.9062135815620422,
      "learning_rate": 0.00021814787376316796,
      "loss": 3.3485,
      "step": 27020
    },
    {
      "epoch": 0.3542873620468189,
      "grad_norm": 0.5680983662605286,
      "learning_rate": 0.0002180924812069335,
      "loss": 3.3821,
      "step": 27030
    },
    {
      "epoch": 0.35441843395286654,
      "grad_norm": 0.5140476226806641,
      "learning_rate": 0.00021803707695222424,
      "loss": 3.316,
      "step": 27040
    },
    {
      "epoch": 0.3545495058589142,
      "grad_norm": 0.5775889754295349,
      "learning_rate": 0.00021798166100855877,
      "loss": 3.3557,
      "step": 27050
    },
    {
      "epoch": 0.35468057776496187,
      "grad_norm": 0.5365877151489258,
      "learning_rate": 0.0002179262333854577,
      "loss": 3.3613,
      "step": 27060
    },
    {
      "epoch": 0.35481164967100953,
      "grad_norm": 0.5881848931312561,
      "learning_rate": 0.0002178707940924437,
      "loss": 3.3092,
      "step": 27070
    },
    {
      "epoch": 0.3549427215770572,
      "grad_norm": 0.783545732498169,
      "learning_rate": 0.00021781534313904132,
      "loss": 3.3918,
      "step": 27080
    },
    {
      "epoch": 0.3550737934831048,
      "grad_norm": 0.6256943345069885,
      "learning_rate": 0.00021775988053477719,
      "loss": 3.3181,
      "step": 27090
    },
    {
      "epoch": 0.35520486538915247,
      "grad_norm": 0.5850955247879028,
      "learning_rate": 0.00021770440628917997,
      "loss": 3.3256,
      "step": 27100
    },
    {
      "epoch": 0.35533593729520013,
      "grad_norm": 0.5705082416534424,
      "learning_rate": 0.00021764892041178035,
      "loss": 3.3184,
      "step": 27110
    },
    {
      "epoch": 0.3554670092012478,
      "grad_norm": 0.5762251615524292,
      "learning_rate": 0.00021759342291211085,
      "loss": 3.3224,
      "step": 27120
    },
    {
      "epoch": 0.35559808110729546,
      "grad_norm": 0.5657543540000916,
      "learning_rate": 0.0002175379137997061,
      "loss": 3.3161,
      "step": 27130
    },
    {
      "epoch": 0.3557291530133431,
      "grad_norm": 0.5625048279762268,
      "learning_rate": 0.00021748239308410275,
      "loss": 3.3321,
      "step": 27140
    },
    {
      "epoch": 0.3558602249193908,
      "grad_norm": 0.5475554466247559,
      "learning_rate": 0.00021742686077483946,
      "loss": 3.2875,
      "step": 27150
    },
    {
      "epoch": 0.35599129682543845,
      "grad_norm": 0.6351070404052734,
      "learning_rate": 0.00021737131688145673,
      "loss": 3.3747,
      "step": 27160
    },
    {
      "epoch": 0.3561223687314861,
      "grad_norm": 0.551333487033844,
      "learning_rate": 0.0002173157614134972,
      "loss": 3.4085,
      "step": 27170
    },
    {
      "epoch": 0.3562534406375337,
      "grad_norm": 0.5410186052322388,
      "learning_rate": 0.0002172601943805055,
      "loss": 3.3174,
      "step": 27180
    },
    {
      "epoch": 0.3563845125435814,
      "grad_norm": 0.5446428656578064,
      "learning_rate": 0.0002172046157920281,
      "loss": 3.3392,
      "step": 27190
    },
    {
      "epoch": 0.35651558444962905,
      "grad_norm": 0.5336136221885681,
      "learning_rate": 0.00021714902565761365,
      "loss": 3.4012,
      "step": 27200
    },
    {
      "epoch": 0.3566466563556767,
      "grad_norm": 0.5584641695022583,
      "learning_rate": 0.00021709342398681264,
      "loss": 3.3131,
      "step": 27210
    },
    {
      "epoch": 0.3567777282617244,
      "grad_norm": 0.528645932674408,
      "learning_rate": 0.00021703781078917758,
      "loss": 3.3254,
      "step": 27220
    },
    {
      "epoch": 0.35690880016777204,
      "grad_norm": 0.5452187061309814,
      "learning_rate": 0.00021698218607426294,
      "loss": 3.3291,
      "step": 27230
    },
    {
      "epoch": 0.3570398720738197,
      "grad_norm": 0.5461795926094055,
      "learning_rate": 0.00021692654985162532,
      "loss": 3.4002,
      "step": 27240
    },
    {
      "epoch": 0.3571709439798674,
      "grad_norm": 0.5468012690544128,
      "learning_rate": 0.00021687090213082306,
      "loss": 3.32,
      "step": 27250
    },
    {
      "epoch": 0.35730201588591504,
      "grad_norm": 0.5724340677261353,
      "learning_rate": 0.00021681524292141663,
      "loss": 3.3297,
      "step": 27260
    },
    {
      "epoch": 0.35743308779196264,
      "grad_norm": 0.5507761240005493,
      "learning_rate": 0.00021675957223296838,
      "loss": 3.3415,
      "step": 27270
    },
    {
      "epoch": 0.3575641596980103,
      "grad_norm": 0.5071584582328796,
      "learning_rate": 0.0002167038900750428,
      "loss": 3.3475,
      "step": 27280
    },
    {
      "epoch": 0.357695231604058,
      "grad_norm": 0.5104081630706787,
      "learning_rate": 0.0002166481964572062,
      "loss": 3.3556,
      "step": 27290
    },
    {
      "epoch": 0.35782630351010564,
      "grad_norm": 0.5347312688827515,
      "learning_rate": 0.00021659249138902687,
      "loss": 3.3291,
      "step": 27300
    },
    {
      "epoch": 0.3579573754161533,
      "grad_norm": 0.5749819278717041,
      "learning_rate": 0.00021653677488007506,
      "loss": 3.3035,
      "step": 27310
    },
    {
      "epoch": 0.35808844732220096,
      "grad_norm": 0.6626188158988953,
      "learning_rate": 0.00021648104693992303,
      "loss": 3.3261,
      "step": 27320
    },
    {
      "epoch": 0.35821951922824863,
      "grad_norm": 0.6071866154670715,
      "learning_rate": 0.00021642530757814507,
      "loss": 3.306,
      "step": 27330
    },
    {
      "epoch": 0.3583505911342963,
      "grad_norm": 0.5083776116371155,
      "learning_rate": 0.00021636955680431728,
      "loss": 3.3135,
      "step": 27340
    },
    {
      "epoch": 0.35848166304034396,
      "grad_norm": 0.4976089298725128,
      "learning_rate": 0.00021631379462801783,
      "loss": 3.3211,
      "step": 27350
    },
    {
      "epoch": 0.35861273494639156,
      "grad_norm": 0.510925829410553,
      "learning_rate": 0.0002162580210588268,
      "loss": 3.3217,
      "step": 27360
    },
    {
      "epoch": 0.35874380685243923,
      "grad_norm": 0.5436649918556213,
      "learning_rate": 0.00021620223610632625,
      "loss": 3.3473,
      "step": 27370
    },
    {
      "epoch": 0.3588748787584869,
      "grad_norm": 0.8753914833068848,
      "learning_rate": 0.00021614643978010015,
      "loss": 3.2774,
      "step": 27380
    },
    {
      "epoch": 0.35900595066453456,
      "grad_norm": 0.5142229199409485,
      "learning_rate": 0.00021609063208973456,
      "loss": 3.3332,
      "step": 27390
    },
    {
      "epoch": 0.3591370225705822,
      "grad_norm": 0.5408582091331482,
      "learning_rate": 0.00021603481304481727,
      "loss": 3.3458,
      "step": 27400
    },
    {
      "epoch": 0.3592680944766299,
      "grad_norm": 0.5829759836196899,
      "learning_rate": 0.00021597898265493824,
      "loss": 3.2966,
      "step": 27410
    },
    {
      "epoch": 0.35939916638267755,
      "grad_norm": 0.522330641746521,
      "learning_rate": 0.00021592314092968925,
      "loss": 3.3129,
      "step": 27420
    },
    {
      "epoch": 0.3595302382887252,
      "grad_norm": 0.5187906622886658,
      "learning_rate": 0.000215867287878664,
      "loss": 3.324,
      "step": 27430
    },
    {
      "epoch": 0.3596613101947729,
      "grad_norm": 0.6144665479660034,
      "learning_rate": 0.0002158114235114583,
      "loss": 3.3052,
      "step": 27440
    },
    {
      "epoch": 0.3597923821008205,
      "grad_norm": 0.5527117848396301,
      "learning_rate": 0.00021575554783766972,
      "loss": 3.3537,
      "step": 27450
    },
    {
      "epoch": 0.35992345400686815,
      "grad_norm": 0.5111397504806519,
      "learning_rate": 0.00021569966086689793,
      "loss": 3.3469,
      "step": 27460
    },
    {
      "epoch": 0.3600545259129158,
      "grad_norm": 0.551163911819458,
      "learning_rate": 0.00021564376260874445,
      "loss": 3.3196,
      "step": 27470
    },
    {
      "epoch": 0.3601855978189635,
      "grad_norm": 0.546377420425415,
      "learning_rate": 0.00021558785307281272,
      "loss": 3.3415,
      "step": 27480
    },
    {
      "epoch": 0.36031666972501114,
      "grad_norm": 0.5487934350967407,
      "learning_rate": 0.00021553193226870812,
      "loss": 3.3321,
      "step": 27490
    },
    {
      "epoch": 0.3604477416310588,
      "grad_norm": 0.5402739644050598,
      "learning_rate": 0.00021547600020603805,
      "loss": 3.3118,
      "step": 27500
    },
    {
      "epoch": 0.3604477416310588,
      "eval_accuracy": 0.38913391984359724,
      "eval_loss": 3.3116846084594727,
      "eval_runtime": 59.233,
      "eval_samples_per_second": 33.765,
      "eval_steps_per_second": 0.54,
      "step": 27500
    },
    {
      "epoch": 0.36057881353710647,
      "grad_norm": 0.543951690196991,
      "learning_rate": 0.00021542005689441178,
      "loss": 3.3309,
      "step": 27510
    },
    {
      "epoch": 0.36070988544315413,
      "grad_norm": 0.585476815700531,
      "learning_rate": 0.00021536410234344056,
      "loss": 3.3119,
      "step": 27520
    },
    {
      "epoch": 0.3608409573492018,
      "grad_norm": 0.6003909111022949,
      "learning_rate": 0.00021530813656273748,
      "loss": 3.3522,
      "step": 27530
    },
    {
      "epoch": 0.3609720292552494,
      "grad_norm": 0.5225700736045837,
      "learning_rate": 0.0002152521595619176,
      "loss": 3.3052,
      "step": 27540
    },
    {
      "epoch": 0.36110310116129707,
      "grad_norm": 0.5296647548675537,
      "learning_rate": 0.000215196171350598,
      "loss": 3.3611,
      "step": 27550
    },
    {
      "epoch": 0.36123417306734473,
      "grad_norm": 0.5320678353309631,
      "learning_rate": 0.00021514017193839755,
      "loss": 3.327,
      "step": 27560
    },
    {
      "epoch": 0.3613652449733924,
      "grad_norm": 0.5960475206375122,
      "learning_rate": 0.00021508416133493715,
      "loss": 3.3213,
      "step": 27570
    },
    {
      "epoch": 0.36149631687944006,
      "grad_norm": 0.5452954769134521,
      "learning_rate": 0.0002150281395498395,
      "loss": 3.342,
      "step": 27580
    },
    {
      "epoch": 0.3616273887854877,
      "grad_norm": 0.5819903016090393,
      "learning_rate": 0.00021497210659272934,
      "loss": 3.3227,
      "step": 27590
    },
    {
      "epoch": 0.3617584606915354,
      "grad_norm": 0.5283902883529663,
      "learning_rate": 0.00021491606247323327,
      "loss": 3.3334,
      "step": 27600
    },
    {
      "epoch": 0.36188953259758305,
      "grad_norm": 0.520560622215271,
      "learning_rate": 0.00021486000720097984,
      "loss": 3.3064,
      "step": 27610
    },
    {
      "epoch": 0.3620206045036307,
      "grad_norm": 0.49930673837661743,
      "learning_rate": 0.00021480394078559947,
      "loss": 3.2939,
      "step": 27620
    },
    {
      "epoch": 0.3621516764096783,
      "grad_norm": 0.5403222441673279,
      "learning_rate": 0.00021474786323672464,
      "loss": 3.3356,
      "step": 27630
    },
    {
      "epoch": 0.362282748315726,
      "grad_norm": 0.5281585454940796,
      "learning_rate": 0.00021469177456398948,
      "loss": 3.328,
      "step": 27640
    },
    {
      "epoch": 0.36241382022177365,
      "grad_norm": 0.5060186982154846,
      "learning_rate": 0.00021463567477703025,
      "loss": 3.3163,
      "step": 27650
    },
    {
      "epoch": 0.3625448921278213,
      "grad_norm": 0.7897487878799438,
      "learning_rate": 0.00021457956388548502,
      "loss": 3.3399,
      "step": 27660
    },
    {
      "epoch": 0.362675964033869,
      "grad_norm": 0.5786916613578796,
      "learning_rate": 0.0002145234418989938,
      "loss": 3.3461,
      "step": 27670
    },
    {
      "epoch": 0.36280703593991664,
      "grad_norm": 0.580524206161499,
      "learning_rate": 0.0002144673088271985,
      "loss": 3.3348,
      "step": 27680
    },
    {
      "epoch": 0.3629381078459643,
      "grad_norm": 0.5191983580589294,
      "learning_rate": 0.00021441116467974292,
      "loss": 3.347,
      "step": 27690
    },
    {
      "epoch": 0.363069179752012,
      "grad_norm": 0.5369277000427246,
      "learning_rate": 0.00021435500946627288,
      "loss": 3.3079,
      "step": 27700
    },
    {
      "epoch": 0.36320025165805964,
      "grad_norm": 0.5672141313552856,
      "learning_rate": 0.00021429884319643585,
      "loss": 3.3042,
      "step": 27710
    },
    {
      "epoch": 0.36333132356410724,
      "grad_norm": 0.5369529724121094,
      "learning_rate": 0.00021424266587988144,
      "loss": 3.2761,
      "step": 27720
    },
    {
      "epoch": 0.3634623954701549,
      "grad_norm": 0.5451904535293579,
      "learning_rate": 0.00021418647752626107,
      "loss": 3.3237,
      "step": 27730
    },
    {
      "epoch": 0.3635934673762026,
      "grad_norm": 0.5818816423416138,
      "learning_rate": 0.00021413027814522803,
      "loss": 3.303,
      "step": 27740
    },
    {
      "epoch": 0.36372453928225024,
      "grad_norm": 0.5838695168495178,
      "learning_rate": 0.00021407406774643747,
      "loss": 3.3427,
      "step": 27750
    },
    {
      "epoch": 0.3638556111882979,
      "grad_norm": 0.6151512861251831,
      "learning_rate": 0.00021401784633954662,
      "loss": 3.3348,
      "step": 27760
    },
    {
      "epoch": 0.36398668309434556,
      "grad_norm": 0.6470279097557068,
      "learning_rate": 0.00021396161393421435,
      "loss": 3.3449,
      "step": 27770
    },
    {
      "epoch": 0.36411775500039323,
      "grad_norm": 0.5237503051757812,
      "learning_rate": 0.00021390537054010158,
      "loss": 3.2869,
      "step": 27780
    },
    {
      "epoch": 0.3642488269064409,
      "grad_norm": 0.5565051436424255,
      "learning_rate": 0.00021384911616687113,
      "loss": 3.3607,
      "step": 27790
    },
    {
      "epoch": 0.36437989881248856,
      "grad_norm": 0.5073827505111694,
      "learning_rate": 0.00021379285082418755,
      "loss": 3.2993,
      "step": 27800
    },
    {
      "epoch": 0.36451097071853616,
      "grad_norm": 0.5319257974624634,
      "learning_rate": 0.00021373657452171754,
      "loss": 3.3182,
      "step": 27810
    },
    {
      "epoch": 0.36464204262458383,
      "grad_norm": 0.513541042804718,
      "learning_rate": 0.0002136802872691293,
      "loss": 3.362,
      "step": 27820
    },
    {
      "epoch": 0.3647731145306315,
      "grad_norm": 0.54200279712677,
      "learning_rate": 0.00021362398907609338,
      "loss": 3.3139,
      "step": 27830
    },
    {
      "epoch": 0.36490418643667916,
      "grad_norm": 0.5393368005752563,
      "learning_rate": 0.00021356767995228183,
      "loss": 3.3358,
      "step": 27840
    },
    {
      "epoch": 0.3650352583427268,
      "grad_norm": 0.6337693929672241,
      "learning_rate": 0.0002135113599073687,
      "loss": 3.311,
      "step": 27850
    },
    {
      "epoch": 0.3651663302487745,
      "grad_norm": 0.7328603267669678,
      "learning_rate": 0.00021345502895102992,
      "loss": 3.4282,
      "step": 27860
    },
    {
      "epoch": 0.36529740215482215,
      "grad_norm": 0.539165735244751,
      "learning_rate": 0.0002133986870929434,
      "loss": 3.2918,
      "step": 27870
    },
    {
      "epoch": 0.3654284740608698,
      "grad_norm": 0.5584608912467957,
      "learning_rate": 0.00021334233434278874,
      "loss": 3.3457,
      "step": 27880
    },
    {
      "epoch": 0.3655595459669175,
      "grad_norm": 0.5518494248390198,
      "learning_rate": 0.00021328597071024752,
      "loss": 3.3404,
      "step": 27890
    },
    {
      "epoch": 0.3656906178729651,
      "grad_norm": 0.5684391260147095,
      "learning_rate": 0.00021322959620500315,
      "loss": 3.3423,
      "step": 27900
    },
    {
      "epoch": 0.36582168977901275,
      "grad_norm": 0.5712750554084778,
      "learning_rate": 0.00021317321083674098,
      "loss": 3.3511,
      "step": 27910
    },
    {
      "epoch": 0.3659527616850604,
      "grad_norm": 0.5564943552017212,
      "learning_rate": 0.0002131168146151481,
      "loss": 3.328,
      "step": 27920
    },
    {
      "epoch": 0.3660838335911081,
      "grad_norm": 0.6053992509841919,
      "learning_rate": 0.00021306040754991357,
      "loss": 3.3478,
      "step": 27930
    },
    {
      "epoch": 0.36621490549715574,
      "grad_norm": 0.5572041273117065,
      "learning_rate": 0.0002130039896507283,
      "loss": 3.3169,
      "step": 27940
    },
    {
      "epoch": 0.3663459774032034,
      "grad_norm": 0.5496898889541626,
      "learning_rate": 0.000212947560927285,
      "loss": 3.3592,
      "step": 27950
    },
    {
      "epoch": 0.36647704930925107,
      "grad_norm": 0.6112011671066284,
      "learning_rate": 0.0002128911213892783,
      "loss": 3.3638,
      "step": 27960
    },
    {
      "epoch": 0.36660812121529873,
      "grad_norm": 0.5442432165145874,
      "learning_rate": 0.00021283467104640463,
      "loss": 3.288,
      "step": 27970
    },
    {
      "epoch": 0.3667391931213464,
      "grad_norm": 0.5328122973442078,
      "learning_rate": 0.00021277820990836237,
      "loss": 3.3331,
      "step": 27980
    },
    {
      "epoch": 0.366870265027394,
      "grad_norm": 0.5572381615638733,
      "learning_rate": 0.00021272173798485164,
      "loss": 3.3275,
      "step": 27990
    },
    {
      "epoch": 0.36700133693344167,
      "grad_norm": 0.5520017147064209,
      "learning_rate": 0.00021266525528557456,
      "loss": 3.3057,
      "step": 28000
    },
    {
      "epoch": 0.36700133693344167,
      "eval_accuracy": 0.3893445747800586,
      "eval_loss": 3.310812473297119,
      "eval_runtime": 61.8828,
      "eval_samples_per_second": 32.319,
      "eval_steps_per_second": 0.517,
      "step": 28000
    },
    {
      "epoch": 0.36713240883948933,
      "grad_norm": 0.5635603666305542,
      "learning_rate": 0.00021260876182023497,
      "loss": 3.3009,
      "step": 28010
    },
    {
      "epoch": 0.367263480745537,
      "grad_norm": 0.6012985110282898,
      "learning_rate": 0.00021255225759853852,
      "loss": 3.3552,
      "step": 28020
    },
    {
      "epoch": 0.36739455265158466,
      "grad_norm": 0.579226553440094,
      "learning_rate": 0.00021249574263019293,
      "loss": 3.2932,
      "step": 28030
    },
    {
      "epoch": 0.3675256245576323,
      "grad_norm": 0.5632168054580688,
      "learning_rate": 0.00021243921692490748,
      "loss": 3.3251,
      "step": 28040
    },
    {
      "epoch": 0.36765669646368,
      "grad_norm": 0.7379662394523621,
      "learning_rate": 0.00021238268049239358,
      "loss": 3.3482,
      "step": 28050
    },
    {
      "epoch": 0.36778776836972765,
      "grad_norm": 0.6342719197273254,
      "learning_rate": 0.00021232613334236424,
      "loss": 3.3567,
      "step": 28060
    },
    {
      "epoch": 0.3679188402757753,
      "grad_norm": 0.5645907521247864,
      "learning_rate": 0.00021226957548453447,
      "loss": 3.4035,
      "step": 28070
    },
    {
      "epoch": 0.3680499121818229,
      "grad_norm": 0.7663527131080627,
      "learning_rate": 0.00021221300692862106,
      "loss": 3.3255,
      "step": 28080
    },
    {
      "epoch": 0.3681809840878706,
      "grad_norm": 0.5777570009231567,
      "learning_rate": 0.00021215642768434263,
      "loss": 3.3812,
      "step": 28090
    },
    {
      "epoch": 0.36831205599391825,
      "grad_norm": 0.5560043454170227,
      "learning_rate": 0.00021209983776141966,
      "loss": 3.3505,
      "step": 28100
    },
    {
      "epoch": 0.3684431278999659,
      "grad_norm": 0.5703155398368835,
      "learning_rate": 0.00021204323716957446,
      "loss": 3.3735,
      "step": 28110
    },
    {
      "epoch": 0.3685741998060136,
      "grad_norm": 0.5798794627189636,
      "learning_rate": 0.0002119866259185311,
      "loss": 3.3225,
      "step": 28120
    },
    {
      "epoch": 0.36870527171206124,
      "grad_norm": 0.7754932641983032,
      "learning_rate": 0.00021193000401801564,
      "loss": 3.3389,
      "step": 28130
    },
    {
      "epoch": 0.3688363436181089,
      "grad_norm": 0.5348528623580933,
      "learning_rate": 0.0002118733714777558,
      "loss": 3.3279,
      "step": 28140
    },
    {
      "epoch": 0.3689674155241566,
      "grad_norm": 0.5910087823867798,
      "learning_rate": 0.0002118167283074813,
      "loss": 3.3704,
      "step": 28150
    },
    {
      "epoch": 0.36909848743020424,
      "grad_norm": 0.5643282532691956,
      "learning_rate": 0.0002117600745169235,
      "loss": 3.4,
      "step": 28160
    },
    {
      "epoch": 0.36922955933625184,
      "grad_norm": 0.5822794437408447,
      "learning_rate": 0.0002117034101158157,
      "loss": 3.3123,
      "step": 28170
    },
    {
      "epoch": 0.3693606312422995,
      "grad_norm": 0.5382328033447266,
      "learning_rate": 0.00021164673511389306,
      "loss": 3.2957,
      "step": 28180
    },
    {
      "epoch": 0.3694917031483472,
      "grad_norm": 0.7501068115234375,
      "learning_rate": 0.00021159004952089246,
      "loss": 3.3025,
      "step": 28190
    },
    {
      "epoch": 0.36962277505439484,
      "grad_norm": 0.5680118203163147,
      "learning_rate": 0.00021153335334655262,
      "loss": 3.3488,
      "step": 28200
    },
    {
      "epoch": 0.3697538469604425,
      "grad_norm": 0.6492407917976379,
      "learning_rate": 0.00021147664660061405,
      "loss": 3.307,
      "step": 28210
    },
    {
      "epoch": 0.36988491886649016,
      "grad_norm": 0.8204067945480347,
      "learning_rate": 0.00021141992929281923,
      "loss": 3.3902,
      "step": 28220
    },
    {
      "epoch": 0.37001599077253783,
      "grad_norm": 0.533372163772583,
      "learning_rate": 0.00021136320143291232,
      "loss": 3.3484,
      "step": 28230
    },
    {
      "epoch": 0.3701470626785855,
      "grad_norm": 0.6085098385810852,
      "learning_rate": 0.0002113064630306393,
      "loss": 3.3339,
      "step": 28240
    },
    {
      "epoch": 0.37027813458463316,
      "grad_norm": 0.6099501848220825,
      "learning_rate": 0.00021124971409574797,
      "loss": 3.3227,
      "step": 28250
    },
    {
      "epoch": 0.37040920649068076,
      "grad_norm": 0.5346800684928894,
      "learning_rate": 0.00021119295463798798,
      "loss": 3.3107,
      "step": 28260
    },
    {
      "epoch": 0.37054027839672843,
      "grad_norm": 0.9242376089096069,
      "learning_rate": 0.00021113618466711076,
      "loss": 3.4921,
      "step": 28270
    },
    {
      "epoch": 0.3706713503027761,
      "grad_norm": 0.5991373658180237,
      "learning_rate": 0.00021107940419286953,
      "loss": 3.4328,
      "step": 28280
    },
    {
      "epoch": 0.37080242220882376,
      "grad_norm": 0.5799722075462341,
      "learning_rate": 0.00021102261322501932,
      "loss": 3.3437,
      "step": 28290
    },
    {
      "epoch": 0.3709334941148714,
      "grad_norm": 0.5479770302772522,
      "learning_rate": 0.00021096581177331702,
      "loss": 3.3757,
      "step": 28300
    },
    {
      "epoch": 0.3710645660209191,
      "grad_norm": 0.5731379389762878,
      "learning_rate": 0.00021090899984752123,
      "loss": 3.3713,
      "step": 28310
    },
    {
      "epoch": 0.37119563792696675,
      "grad_norm": 0.690904438495636,
      "learning_rate": 0.0002108521774573924,
      "loss": 3.3884,
      "step": 28320
    },
    {
      "epoch": 0.3713267098330144,
      "grad_norm": 0.5505548119544983,
      "learning_rate": 0.00021079534461269282,
      "loss": 3.3787,
      "step": 28330
    },
    {
      "epoch": 0.3714577817390621,
      "grad_norm": 0.5653923153877258,
      "learning_rate": 0.00021073850132318647,
      "loss": 3.3137,
      "step": 28340
    },
    {
      "epoch": 0.3715888536451097,
      "grad_norm": 0.5638493299484253,
      "learning_rate": 0.0002106816475986392,
      "loss": 3.3545,
      "step": 28350
    },
    {
      "epoch": 0.37171992555115735,
      "grad_norm": 0.640590488910675,
      "learning_rate": 0.0002106247834488187,
      "loss": 3.3647,
      "step": 28360
    },
    {
      "epoch": 0.371850997457205,
      "grad_norm": 0.5456752777099609,
      "learning_rate": 0.00021056790888349428,
      "loss": 3.3411,
      "step": 28370
    },
    {
      "epoch": 0.3719820693632527,
      "grad_norm": 0.5360928177833557,
      "learning_rate": 0.00021051102391243722,
      "loss": 3.3595,
      "step": 28380
    },
    {
      "epoch": 0.37211314126930034,
      "grad_norm": 0.48407015204429626,
      "learning_rate": 0.00021045412854542046,
      "loss": 3.3151,
      "step": 28390
    },
    {
      "epoch": 0.372244213175348,
      "grad_norm": 0.5534451603889465,
      "learning_rate": 0.00021039722279221878,
      "loss": 3.2729,
      "step": 28400
    },
    {
      "epoch": 0.37237528508139567,
      "grad_norm": 0.5777167081832886,
      "learning_rate": 0.00021034030666260885,
      "loss": 3.361,
      "step": 28410
    },
    {
      "epoch": 0.37250635698744333,
      "grad_norm": 0.5372534394264221,
      "learning_rate": 0.00021028338016636892,
      "loss": 3.3221,
      "step": 28420
    },
    {
      "epoch": 0.372637428893491,
      "grad_norm": 0.5139166712760925,
      "learning_rate": 0.00021022644331327906,
      "loss": 3.2821,
      "step": 28430
    },
    {
      "epoch": 0.3727685007995386,
      "grad_norm": 0.5702176094055176,
      "learning_rate": 0.00021016949611312133,
      "loss": 3.3148,
      "step": 28440
    },
    {
      "epoch": 0.37289957270558627,
      "grad_norm": 0.581762433052063,
      "learning_rate": 0.00021011253857567934,
      "loss": 3.3235,
      "step": 28450
    },
    {
      "epoch": 0.37303064461163393,
      "grad_norm": 0.5555365681648254,
      "learning_rate": 0.00021005557071073852,
      "loss": 3.3039,
      "step": 28460
    },
    {
      "epoch": 0.3731617165176816,
      "grad_norm": 0.562186598777771,
      "learning_rate": 0.00020999859252808614,
      "loss": 3.331,
      "step": 28470
    },
    {
      "epoch": 0.37329278842372926,
      "grad_norm": 0.5326070785522461,
      "learning_rate": 0.00020994160403751125,
      "loss": 3.3261,
      "step": 28480
    },
    {
      "epoch": 0.3734238603297769,
      "grad_norm": 0.6086612939834595,
      "learning_rate": 0.00020988460524880452,
      "loss": 3.3639,
      "step": 28490
    },
    {
      "epoch": 0.3735549322358246,
      "grad_norm": 0.5010743141174316,
      "learning_rate": 0.0002098275961717586,
      "loss": 3.3436,
      "step": 28500
    },
    {
      "epoch": 0.3735549322358246,
      "eval_accuracy": 0.39019794721407625,
      "eval_loss": 3.3067052364349365,
      "eval_runtime": 59.7472,
      "eval_samples_per_second": 33.474,
      "eval_steps_per_second": 0.536,
      "step": 28500
    },
    {
      "epoch": 0.37368600414187225,
      "grad_norm": 0.5408167839050293,
      "learning_rate": 0.0002097705768161677,
      "loss": 3.3289,
      "step": 28510
    },
    {
      "epoch": 0.3738170760479199,
      "grad_norm": 0.5330541133880615,
      "learning_rate": 0.00020971354719182808,
      "loss": 3.3797,
      "step": 28520
    },
    {
      "epoch": 0.3739481479539675,
      "grad_norm": 0.5085309147834778,
      "learning_rate": 0.00020965650730853737,
      "loss": 3.3755,
      "step": 28530
    },
    {
      "epoch": 0.3740792198600152,
      "grad_norm": 0.5940837860107422,
      "learning_rate": 0.00020959945717609533,
      "loss": 3.3142,
      "step": 28540
    },
    {
      "epoch": 0.37421029176606285,
      "grad_norm": 0.5588082671165466,
      "learning_rate": 0.00020954239680430332,
      "loss": 3.3601,
      "step": 28550
    },
    {
      "epoch": 0.3743413636721105,
      "grad_norm": 0.5893656611442566,
      "learning_rate": 0.00020948532620296436,
      "loss": 3.327,
      "step": 28560
    },
    {
      "epoch": 0.3744724355781582,
      "grad_norm": 0.5307437181472778,
      "learning_rate": 0.00020942824538188346,
      "loss": 3.306,
      "step": 28570
    },
    {
      "epoch": 0.37460350748420584,
      "grad_norm": 0.6522595882415771,
      "learning_rate": 0.00020937115435086713,
      "loss": 3.3811,
      "step": 28580
    },
    {
      "epoch": 0.3747345793902535,
      "grad_norm": 0.5592530369758606,
      "learning_rate": 0.00020931405311972392,
      "loss": 3.3539,
      "step": 28590
    },
    {
      "epoch": 0.37486565129630117,
      "grad_norm": 0.5048577785491943,
      "learning_rate": 0.00020925694169826386,
      "loss": 3.3624,
      "step": 28600
    },
    {
      "epoch": 0.37499672320234884,
      "grad_norm": 0.5581935048103333,
      "learning_rate": 0.0002091998200962989,
      "loss": 3.379,
      "step": 28610
    },
    {
      "epoch": 0.37512779510839644,
      "grad_norm": 0.5909243822097778,
      "learning_rate": 0.00020914268832364263,
      "loss": 3.3181,
      "step": 28620
    },
    {
      "epoch": 0.3752588670144441,
      "grad_norm": 0.6239158511161804,
      "learning_rate": 0.00020908554639011056,
      "loss": 3.3234,
      "step": 28630
    },
    {
      "epoch": 0.3753899389204918,
      "grad_norm": 0.5379216074943542,
      "learning_rate": 0.00020902839430551973,
      "loss": 3.3342,
      "step": 28640
    },
    {
      "epoch": 0.37552101082653944,
      "grad_norm": 0.5369953513145447,
      "learning_rate": 0.0002089712320796891,
      "loss": 3.364,
      "step": 28650
    },
    {
      "epoch": 0.3756520827325871,
      "grad_norm": 0.5615555047988892,
      "learning_rate": 0.00020891405972243922,
      "loss": 3.2949,
      "step": 28660
    },
    {
      "epoch": 0.37578315463863476,
      "grad_norm": 0.5334003567695618,
      "learning_rate": 0.00020885687724359243,
      "loss": 3.3146,
      "step": 28670
    },
    {
      "epoch": 0.37591422654468243,
      "grad_norm": 0.5063300728797913,
      "learning_rate": 0.00020879968465297298,
      "loss": 3.3117,
      "step": 28680
    },
    {
      "epoch": 0.3760452984507301,
      "grad_norm": 0.5355787873268127,
      "learning_rate": 0.00020874248196040658,
      "loss": 3.3281,
      "step": 28690
    },
    {
      "epoch": 0.3761763703567777,
      "grad_norm": 0.4868262708187103,
      "learning_rate": 0.00020868526917572092,
      "loss": 3.3452,
      "step": 28700
    },
    {
      "epoch": 0.37630744226282536,
      "grad_norm": 0.5646403431892395,
      "learning_rate": 0.00020862804630874518,
      "loss": 3.3781,
      "step": 28710
    },
    {
      "epoch": 0.37643851416887303,
      "grad_norm": 0.7031459212303162,
      "learning_rate": 0.00020857081336931053,
      "loss": 3.2988,
      "step": 28720
    },
    {
      "epoch": 0.3765695860749207,
      "grad_norm": 0.5601722002029419,
      "learning_rate": 0.00020851357036724967,
      "loss": 3.2838,
      "step": 28730
    },
    {
      "epoch": 0.37670065798096836,
      "grad_norm": 0.5219257473945618,
      "learning_rate": 0.00020845631731239718,
      "loss": 3.2858,
      "step": 28740
    },
    {
      "epoch": 0.376831729887016,
      "grad_norm": 0.5558618903160095,
      "learning_rate": 0.00020839905421458915,
      "loss": 3.3368,
      "step": 28750
    },
    {
      "epoch": 0.3769628017930637,
      "grad_norm": 0.5372939705848694,
      "learning_rate": 0.0002083417810836637,
      "loss": 3.3277,
      "step": 28760
    },
    {
      "epoch": 0.37709387369911135,
      "grad_norm": 0.5972861051559448,
      "learning_rate": 0.00020828449792946043,
      "loss": 3.3448,
      "step": 28770
    },
    {
      "epoch": 0.377224945605159,
      "grad_norm": 0.6338969469070435,
      "learning_rate": 0.00020822720476182074,
      "loss": 3.3773,
      "step": 28780
    },
    {
      "epoch": 0.3773560175112066,
      "grad_norm": 0.5306142568588257,
      "learning_rate": 0.00020816990159058774,
      "loss": 3.3236,
      "step": 28790
    },
    {
      "epoch": 0.3774870894172543,
      "grad_norm": 0.5611557364463806,
      "learning_rate": 0.00020811258842560632,
      "loss": 3.3369,
      "step": 28800
    },
    {
      "epoch": 0.37761816132330195,
      "grad_norm": 0.5345104336738586,
      "learning_rate": 0.00020805526527672305,
      "loss": 3.4281,
      "step": 28810
    },
    {
      "epoch": 0.3777492332293496,
      "grad_norm": 0.525739848613739,
      "learning_rate": 0.00020799793215378612,
      "loss": 3.3426,
      "step": 28820
    },
    {
      "epoch": 0.3778803051353973,
      "grad_norm": 0.5238136649131775,
      "learning_rate": 0.00020794058906664558,
      "loss": 3.3273,
      "step": 28830
    },
    {
      "epoch": 0.37801137704144494,
      "grad_norm": 0.5301663279533386,
      "learning_rate": 0.0002078832360251531,
      "loss": 3.3953,
      "step": 28840
    },
    {
      "epoch": 0.3781424489474926,
      "grad_norm": 0.5289154648780823,
      "learning_rate": 0.00020782587303916217,
      "loss": 3.3189,
      "step": 28850
    },
    {
      "epoch": 0.37827352085354027,
      "grad_norm": 0.605006754398346,
      "learning_rate": 0.00020776850011852775,
      "loss": 3.2884,
      "step": 28860
    },
    {
      "epoch": 0.37840459275958793,
      "grad_norm": 0.5355020761489868,
      "learning_rate": 0.00020771111727310685,
      "loss": 3.3336,
      "step": 28870
    },
    {
      "epoch": 0.37853566466563554,
      "grad_norm": 0.5150967836380005,
      "learning_rate": 0.00020765372451275784,
      "loss": 3.3226,
      "step": 28880
    },
    {
      "epoch": 0.3786667365716832,
      "grad_norm": 0.549266517162323,
      "learning_rate": 0.00020759632184734106,
      "loss": 3.3387,
      "step": 28890
    },
    {
      "epoch": 0.37879780847773087,
      "grad_norm": 0.6144933104515076,
      "learning_rate": 0.0002075389092867184,
      "loss": 3.3501,
      "step": 28900
    },
    {
      "epoch": 0.37892888038377853,
      "grad_norm": 1.5179280042648315,
      "learning_rate": 0.0002074814868407535,
      "loss": 3.2851,
      "step": 28910
    },
    {
      "epoch": 0.3790599522898262,
      "grad_norm": 0.543981671333313,
      "learning_rate": 0.0002074240545193117,
      "loss": 3.3397,
      "step": 28920
    },
    {
      "epoch": 0.37919102419587386,
      "grad_norm": 0.5820293426513672,
      "learning_rate": 0.00020736661233226002,
      "loss": 3.2844,
      "step": 28930
    },
    {
      "epoch": 0.3793220961019215,
      "grad_norm": 0.5149435997009277,
      "learning_rate": 0.00020730916028946722,
      "loss": 3.328,
      "step": 28940
    },
    {
      "epoch": 0.3794531680079692,
      "grad_norm": 0.6793764233589172,
      "learning_rate": 0.00020725169840080365,
      "loss": 3.3452,
      "step": 28950
    },
    {
      "epoch": 0.37958423991401685,
      "grad_norm": 0.5280483961105347,
      "learning_rate": 0.00020719422667614146,
      "loss": 3.3456,
      "step": 28960
    },
    {
      "epoch": 0.37971531182006446,
      "grad_norm": 0.5652992725372314,
      "learning_rate": 0.00020713674512535445,
      "loss": 3.3345,
      "step": 28970
    },
    {
      "epoch": 0.3798463837261121,
      "grad_norm": 0.6117252111434937,
      "learning_rate": 0.00020707925375831814,
      "loss": 3.2896,
      "step": 28980
    },
    {
      "epoch": 0.3799774556321598,
      "grad_norm": 0.5273739099502563,
      "learning_rate": 0.00020702175258490967,
      "loss": 3.3279,
      "step": 28990
    },
    {
      "epoch": 0.38010852753820745,
      "grad_norm": 0.5240916609764099,
      "learning_rate": 0.00020696424161500786,
      "loss": 3.3338,
      "step": 29000
    },
    {
      "epoch": 0.38010852753820745,
      "eval_accuracy": 0.38975024437927663,
      "eval_loss": 3.3064444065093994,
      "eval_runtime": 59.8707,
      "eval_samples_per_second": 33.405,
      "eval_steps_per_second": 0.534,
      "step": 29000
    },
    {
      "epoch": 0.3802395994442551,
      "grad_norm": 0.5502005815505981,
      "learning_rate": 0.0002069067208584933,
      "loss": 3.3138,
      "step": 29010
    },
    {
      "epoch": 0.3803706713503028,
      "grad_norm": 0.6032023429870605,
      "learning_rate": 0.00020684919032524826,
      "loss": 3.2935,
      "step": 29020
    },
    {
      "epoch": 0.38050174325635044,
      "grad_norm": 0.6081415414810181,
      "learning_rate": 0.00020679165002515654,
      "loss": 3.294,
      "step": 29030
    },
    {
      "epoch": 0.3806328151623981,
      "grad_norm": 0.5578451156616211,
      "learning_rate": 0.00020673409996810379,
      "loss": 3.3348,
      "step": 29040
    },
    {
      "epoch": 0.38076388706844577,
      "grad_norm": 0.548667848110199,
      "learning_rate": 0.00020667654016397725,
      "loss": 3.3386,
      "step": 29050
    },
    {
      "epoch": 0.3808949589744934,
      "grad_norm": 0.5574472546577454,
      "learning_rate": 0.00020661897062266584,
      "loss": 3.2733,
      "step": 29060
    },
    {
      "epoch": 0.38102603088054104,
      "grad_norm": 0.5689435005187988,
      "learning_rate": 0.00020656139135406018,
      "loss": 3.2986,
      "step": 29070
    },
    {
      "epoch": 0.3811571027865887,
      "grad_norm": 0.5445596575737,
      "learning_rate": 0.00020650380236805253,
      "loss": 3.3123,
      "step": 29080
    },
    {
      "epoch": 0.3812881746926364,
      "grad_norm": 0.5615941286087036,
      "learning_rate": 0.00020644620367453687,
      "loss": 3.2684,
      "step": 29090
    },
    {
      "epoch": 0.38141924659868404,
      "grad_norm": 0.5612258911132812,
      "learning_rate": 0.00020638859528340878,
      "loss": 3.3033,
      "step": 29100
    },
    {
      "epoch": 0.3815503185047317,
      "grad_norm": 0.5719793438911438,
      "learning_rate": 0.00020633097720456554,
      "loss": 3.2831,
      "step": 29110
    },
    {
      "epoch": 0.38168139041077936,
      "grad_norm": 0.5364654660224915,
      "learning_rate": 0.00020627334944790609,
      "loss": 3.291,
      "step": 29120
    },
    {
      "epoch": 0.38181246231682703,
      "grad_norm": 0.5205824375152588,
      "learning_rate": 0.00020621571202333107,
      "loss": 3.2853,
      "step": 29130
    },
    {
      "epoch": 0.3819435342228747,
      "grad_norm": 0.5725723505020142,
      "learning_rate": 0.00020615806494074274,
      "loss": 3.2713,
      "step": 29140
    },
    {
      "epoch": 0.3820746061289223,
      "grad_norm": 0.6366363167762756,
      "learning_rate": 0.000206100408210045,
      "loss": 3.3441,
      "step": 29150
    },
    {
      "epoch": 0.38220567803496996,
      "grad_norm": 0.5814101696014404,
      "learning_rate": 0.00020604274184114348,
      "loss": 3.3136,
      "step": 29160
    },
    {
      "epoch": 0.38233674994101763,
      "grad_norm": 0.5764576196670532,
      "learning_rate": 0.00020598506584394534,
      "loss": 3.2725,
      "step": 29170
    },
    {
      "epoch": 0.3824678218470653,
      "grad_norm": 0.5501949191093445,
      "learning_rate": 0.00020592738022835955,
      "loss": 3.2715,
      "step": 29180
    },
    {
      "epoch": 0.38259889375311296,
      "grad_norm": 0.523466944694519,
      "learning_rate": 0.00020586968500429663,
      "loss": 3.2878,
      "step": 29190
    },
    {
      "epoch": 0.3827299656591606,
      "grad_norm": 0.5383058190345764,
      "learning_rate": 0.00020581198018166877,
      "loss": 3.2995,
      "step": 29200
    },
    {
      "epoch": 0.3828610375652083,
      "grad_norm": 0.5988202691078186,
      "learning_rate": 0.00020575426577038982,
      "loss": 3.2339,
      "step": 29210
    },
    {
      "epoch": 0.38299210947125595,
      "grad_norm": 0.5802068710327148,
      "learning_rate": 0.0002056965417803753,
      "loss": 3.302,
      "step": 29220
    },
    {
      "epoch": 0.3831231813773036,
      "grad_norm": 0.6097289323806763,
      "learning_rate": 0.00020563880822154225,
      "loss": 3.2714,
      "step": 29230
    },
    {
      "epoch": 0.3832542532833512,
      "grad_norm": 0.5340660214424133,
      "learning_rate": 0.00020558106510380961,
      "loss": 3.258,
      "step": 29240
    },
    {
      "epoch": 0.3833853251893989,
      "grad_norm": 0.5355262160301208,
      "learning_rate": 0.00020552331243709767,
      "loss": 3.3345,
      "step": 29250
    },
    {
      "epoch": 0.38351639709544655,
      "grad_norm": 0.6044197082519531,
      "learning_rate": 0.00020546555023132862,
      "loss": 3.2734,
      "step": 29260
    },
    {
      "epoch": 0.3836474690014942,
      "grad_norm": 0.5542464852333069,
      "learning_rate": 0.00020540777849642608,
      "loss": 3.2926,
      "step": 29270
    },
    {
      "epoch": 0.3837785409075419,
      "grad_norm": 0.5336845517158508,
      "learning_rate": 0.0002053499972423154,
      "loss": 3.2872,
      "step": 29280
    },
    {
      "epoch": 0.38390961281358954,
      "grad_norm": 0.536490797996521,
      "learning_rate": 0.0002052922064789236,
      "loss": 3.3193,
      "step": 29290
    },
    {
      "epoch": 0.3840406847196372,
      "grad_norm": 0.5428259968757629,
      "learning_rate": 0.0002052344062161792,
      "loss": 3.315,
      "step": 29300
    },
    {
      "epoch": 0.38417175662568487,
      "grad_norm": 0.5542668700218201,
      "learning_rate": 0.00020517659646401251,
      "loss": 3.2995,
      "step": 29310
    },
    {
      "epoch": 0.38430282853173253,
      "grad_norm": 0.5118356347084045,
      "learning_rate": 0.00020511877723235544,
      "loss": 3.2844,
      "step": 29320
    },
    {
      "epoch": 0.38443390043778014,
      "grad_norm": 0.5846819281578064,
      "learning_rate": 0.0002050609485311414,
      "loss": 3.2992,
      "step": 29330
    },
    {
      "epoch": 0.3845649723438278,
      "grad_norm": 0.6432815194129944,
      "learning_rate": 0.0002050031103703056,
      "loss": 3.3082,
      "step": 29340
    },
    {
      "epoch": 0.38469604424987547,
      "grad_norm": 0.5626099109649658,
      "learning_rate": 0.00020494526275978477,
      "loss": 3.2831,
      "step": 29350
    },
    {
      "epoch": 0.38482711615592313,
      "grad_norm": 0.5557353496551514,
      "learning_rate": 0.00020488740570951728,
      "loss": 3.3071,
      "step": 29360
    },
    {
      "epoch": 0.3849581880619708,
      "grad_norm": 0.54739910364151,
      "learning_rate": 0.00020482953922944312,
      "loss": 3.286,
      "step": 29370
    },
    {
      "epoch": 0.38508925996801846,
      "grad_norm": 0.7759668827056885,
      "learning_rate": 0.00020477166332950392,
      "loss": 3.395,
      "step": 29380
    },
    {
      "epoch": 0.3852203318740661,
      "grad_norm": 0.6002771258354187,
      "learning_rate": 0.0002047137780196429,
      "loss": 3.2695,
      "step": 29390
    },
    {
      "epoch": 0.3853514037801138,
      "grad_norm": 0.5999743938446045,
      "learning_rate": 0.00020465588330980495,
      "loss": 3.335,
      "step": 29400
    },
    {
      "epoch": 0.38548247568616145,
      "grad_norm": 0.5353532433509827,
      "learning_rate": 0.00020459797920993654,
      "loss": 3.3337,
      "step": 29410
    },
    {
      "epoch": 0.38561354759220906,
      "grad_norm": 0.6078203320503235,
      "learning_rate": 0.00020454006572998576,
      "loss": 3.2841,
      "step": 29420
    },
    {
      "epoch": 0.3857446194982567,
      "grad_norm": 0.5559561848640442,
      "learning_rate": 0.0002044821428799022,
      "loss": 3.3321,
      "step": 29430
    },
    {
      "epoch": 0.3858756914043044,
      "grad_norm": 0.5823197960853577,
      "learning_rate": 0.0002044242106696373,
      "loss": 3.2824,
      "step": 29440
    },
    {
      "epoch": 0.38600676331035205,
      "grad_norm": 0.5667377710342407,
      "learning_rate": 0.00020436626910914392,
      "loss": 3.3379,
      "step": 29450
    },
    {
      "epoch": 0.3861378352163997,
      "grad_norm": 0.553409218788147,
      "learning_rate": 0.00020430831820837656,
      "loss": 3.2901,
      "step": 29460
    },
    {
      "epoch": 0.3862689071224474,
      "grad_norm": 0.6119754314422607,
      "learning_rate": 0.00020425035797729137,
      "loss": 3.3158,
      "step": 29470
    },
    {
      "epoch": 0.38639997902849504,
      "grad_norm": 0.574988842010498,
      "learning_rate": 0.00020419238842584608,
      "loss": 3.2567,
      "step": 29480
    },
    {
      "epoch": 0.3865310509345427,
      "grad_norm": 0.559898853302002,
      "learning_rate": 0.00020413440956399995,
      "loss": 3.3088,
      "step": 29490
    },
    {
      "epoch": 0.38666212284059037,
      "grad_norm": 0.6054991483688354,
      "learning_rate": 0.000204076421401714,
      "loss": 3.2984,
      "step": 29500
    },
    {
      "epoch": 0.38666212284059037,
      "eval_accuracy": 0.3901011730205279,
      "eval_loss": 3.3054749965667725,
      "eval_runtime": 58.0519,
      "eval_samples_per_second": 34.452,
      "eval_steps_per_second": 0.551,
      "step": 29500
    },
    {
      "epoch": 0.386793194746638,
      "grad_norm": 0.5597407221794128,
      "learning_rate": 0.0002040184239489507,
      "loss": 3.3002,
      "step": 29510
    },
    {
      "epoch": 0.38692426665268564,
      "grad_norm": 0.7521269917488098,
      "learning_rate": 0.00020396041721567422,
      "loss": 3.3329,
      "step": 29520
    },
    {
      "epoch": 0.3870553385587333,
      "grad_norm": 0.5580510497093201,
      "learning_rate": 0.00020390240121185023,
      "loss": 3.3347,
      "step": 29530
    },
    {
      "epoch": 0.387186410464781,
      "grad_norm": 0.4970526695251465,
      "learning_rate": 0.00020384437594744606,
      "loss": 3.2631,
      "step": 29540
    },
    {
      "epoch": 0.38731748237082864,
      "grad_norm": 0.9105916023254395,
      "learning_rate": 0.00020378634143243058,
      "loss": 3.3043,
      "step": 29550
    },
    {
      "epoch": 0.3874485542768763,
      "grad_norm": 0.541953444480896,
      "learning_rate": 0.00020372829767677431,
      "loss": 3.2927,
      "step": 29560
    },
    {
      "epoch": 0.38757962618292396,
      "grad_norm": 0.5618892312049866,
      "learning_rate": 0.00020367024469044933,
      "loss": 3.2658,
      "step": 29570
    },
    {
      "epoch": 0.38771069808897163,
      "grad_norm": 0.5603334903717041,
      "learning_rate": 0.00020361218248342926,
      "loss": 3.2852,
      "step": 29580
    },
    {
      "epoch": 0.3878417699950193,
      "grad_norm": 0.6191147565841675,
      "learning_rate": 0.00020355411106568938,
      "loss": 3.2671,
      "step": 29590
    },
    {
      "epoch": 0.3879728419010669,
      "grad_norm": 0.6762301921844482,
      "learning_rate": 0.00020349603044720653,
      "loss": 3.3607,
      "step": 29600
    },
    {
      "epoch": 0.38810391380711456,
      "grad_norm": 0.5611785650253296,
      "learning_rate": 0.00020343794063795907,
      "loss": 3.2936,
      "step": 29610
    },
    {
      "epoch": 0.38823498571316223,
      "grad_norm": 0.5100128650665283,
      "learning_rate": 0.00020337984164792706,
      "loss": 3.2777,
      "step": 29620
    },
    {
      "epoch": 0.3883660576192099,
      "grad_norm": 0.5739402770996094,
      "learning_rate": 0.00020332173348709204,
      "loss": 3.2792,
      "step": 29630
    },
    {
      "epoch": 0.38849712952525756,
      "grad_norm": 0.5812419056892395,
      "learning_rate": 0.00020326361616543708,
      "loss": 3.2918,
      "step": 29640
    },
    {
      "epoch": 0.3886282014313052,
      "grad_norm": 0.6988663673400879,
      "learning_rate": 0.00020320548969294698,
      "loss": 3.3187,
      "step": 29650
    },
    {
      "epoch": 0.3887592733373529,
      "grad_norm": 0.5580822825431824,
      "learning_rate": 0.00020314735407960796,
      "loss": 3.3131,
      "step": 29660
    },
    {
      "epoch": 0.38889034524340055,
      "grad_norm": 0.5258876085281372,
      "learning_rate": 0.000203089209335408,
      "loss": 3.2538,
      "step": 29670
    },
    {
      "epoch": 0.3890214171494482,
      "grad_norm": 0.5898256897926331,
      "learning_rate": 0.0002030310554703364,
      "loss": 3.3059,
      "step": 29680
    },
    {
      "epoch": 0.3891524890554958,
      "grad_norm": 0.5511217713356018,
      "learning_rate": 0.0002029728924943842,
      "loss": 3.2992,
      "step": 29690
    },
    {
      "epoch": 0.3892835609615435,
      "grad_norm": 0.6298866271972656,
      "learning_rate": 0.00020291472041754395,
      "loss": 3.295,
      "step": 29700
    },
    {
      "epoch": 0.38941463286759115,
      "grad_norm": 0.5404861569404602,
      "learning_rate": 0.0002028565392498098,
      "loss": 3.2856,
      "step": 29710
    },
    {
      "epoch": 0.3895457047736388,
      "grad_norm": 0.5305551886558533,
      "learning_rate": 0.0002027983490011774,
      "loss": 3.318,
      "step": 29720
    },
    {
      "epoch": 0.3896767766796865,
      "grad_norm": 0.5419598817825317,
      "learning_rate": 0.00020274014968164398,
      "loss": 3.3721,
      "step": 29730
    },
    {
      "epoch": 0.38980784858573414,
      "grad_norm": 0.6566212177276611,
      "learning_rate": 0.00020268194130120839,
      "loss": 3.3116,
      "step": 29740
    },
    {
      "epoch": 0.3899389204917818,
      "grad_norm": 0.6159994602203369,
      "learning_rate": 0.00020262372386987097,
      "loss": 3.3536,
      "step": 29750
    },
    {
      "epoch": 0.39006999239782947,
      "grad_norm": 0.5742715001106262,
      "learning_rate": 0.00020256549739763364,
      "loss": 3.303,
      "step": 29760
    },
    {
      "epoch": 0.39020106430387713,
      "grad_norm": 0.555503785610199,
      "learning_rate": 0.00020250726189449984,
      "loss": 3.2956,
      "step": 29770
    },
    {
      "epoch": 0.39033213620992474,
      "grad_norm": 0.557773232460022,
      "learning_rate": 0.00020244901737047463,
      "loss": 3.3226,
      "step": 29780
    },
    {
      "epoch": 0.3904632081159724,
      "grad_norm": 0.7776976227760315,
      "learning_rate": 0.00020239076383556453,
      "loss": 3.2519,
      "step": 29790
    },
    {
      "epoch": 0.39059428002202007,
      "grad_norm": 0.6204597353935242,
      "learning_rate": 0.0002023325012997777,
      "loss": 3.2784,
      "step": 29800
    },
    {
      "epoch": 0.39072535192806773,
      "grad_norm": 0.5578134059906006,
      "learning_rate": 0.0002022742297731238,
      "loss": 3.301,
      "step": 29810
    },
    {
      "epoch": 0.3908564238341154,
      "grad_norm": 0.6295265555381775,
      "learning_rate": 0.000202215949265614,
      "loss": 3.3045,
      "step": 29820
    },
    {
      "epoch": 0.39098749574016306,
      "grad_norm": 0.5720866322517395,
      "learning_rate": 0.0002021576597872611,
      "loss": 3.299,
      "step": 29830
    },
    {
      "epoch": 0.3911185676462107,
      "grad_norm": 0.683104932308197,
      "learning_rate": 0.00020209936134807935,
      "loss": 3.3757,
      "step": 29840
    },
    {
      "epoch": 0.3912496395522584,
      "grad_norm": 0.5453050136566162,
      "learning_rate": 0.0002020410539580846,
      "loss": 3.3004,
      "step": 29850
    },
    {
      "epoch": 0.39138071145830605,
      "grad_norm": 0.5689142942428589,
      "learning_rate": 0.0002019827376272942,
      "loss": 3.3176,
      "step": 29860
    },
    {
      "epoch": 0.39151178336435366,
      "grad_norm": 0.5187217593193054,
      "learning_rate": 0.00020192441236572712,
      "loss": 3.3068,
      "step": 29870
    },
    {
      "epoch": 0.3916428552704013,
      "grad_norm": 0.5560576915740967,
      "learning_rate": 0.00020186607818340372,
      "loss": 3.3041,
      "step": 29880
    },
    {
      "epoch": 0.391773927176449,
      "grad_norm": 0.5727310180664062,
      "learning_rate": 0.00020180773509034607,
      "loss": 3.2895,
      "step": 29890
    },
    {
      "epoch": 0.39190499908249665,
      "grad_norm": 0.5541201233863831,
      "learning_rate": 0.00020174938309657753,
      "loss": 3.298,
      "step": 29900
    },
    {
      "epoch": 0.3920360709885443,
      "grad_norm": 0.5460637807846069,
      "learning_rate": 0.00020169102221212327,
      "loss": 3.2447,
      "step": 29910
    },
    {
      "epoch": 0.392167142894592,
      "grad_norm": 0.5416401624679565,
      "learning_rate": 0.0002016326524470098,
      "loss": 3.2862,
      "step": 29920
    },
    {
      "epoch": 0.39229821480063964,
      "grad_norm": 0.600845456123352,
      "learning_rate": 0.00020157427381126513,
      "loss": 3.2661,
      "step": 29930
    },
    {
      "epoch": 0.3924292867066873,
      "grad_norm": 0.6351855397224426,
      "learning_rate": 0.00020151588631491896,
      "loss": 3.3638,
      "step": 29940
    },
    {
      "epoch": 0.39256035861273497,
      "grad_norm": 0.5571659803390503,
      "learning_rate": 0.0002014574899680024,
      "loss": 3.3232,
      "step": 29950
    },
    {
      "epoch": 0.3926914305187826,
      "grad_norm": 0.5685449838638306,
      "learning_rate": 0.0002013990847805481,
      "loss": 3.3175,
      "step": 29960
    },
    {
      "epoch": 0.39282250242483024,
      "grad_norm": 0.5535733103752136,
      "learning_rate": 0.00020134067076259024,
      "loss": 3.2973,
      "step": 29970
    },
    {
      "epoch": 0.3929535743308779,
      "grad_norm": 0.6008982062339783,
      "learning_rate": 0.00020128224792416452,
      "loss": 3.3378,
      "step": 29980
    },
    {
      "epoch": 0.3930846462369256,
      "grad_norm": 0.5603640079498291,
      "learning_rate": 0.0002012238162753081,
      "loss": 3.2815,
      "step": 29990
    },
    {
      "epoch": 0.39321571814297324,
      "grad_norm": 0.5141309499740601,
      "learning_rate": 0.00020116537582605976,
      "loss": 3.3048,
      "step": 30000
    },
    {
      "epoch": 0.39321571814297324,
      "eval_accuracy": 0.3908616813294233,
      "eval_loss": 3.3009097576141357,
      "eval_runtime": 60.8518,
      "eval_samples_per_second": 32.867,
      "eval_steps_per_second": 0.526,
      "step": 30000
    },
    {
      "epoch": 0.3933467900490209,
      "grad_norm": 0.7090569138526917,
      "learning_rate": 0.00020110692658645963,
      "loss": 3.3858,
      "step": 30010
    },
    {
      "epoch": 0.39347786195506856,
      "grad_norm": 0.6517828702926636,
      "learning_rate": 0.00020104846856654952,
      "loss": 3.2869,
      "step": 30020
    },
    {
      "epoch": 0.39360893386111623,
      "grad_norm": 0.5829726457595825,
      "learning_rate": 0.00020099000177637268,
      "loss": 3.3573,
      "step": 30030
    },
    {
      "epoch": 0.3937400057671639,
      "grad_norm": 0.5205245614051819,
      "learning_rate": 0.0002009315262259739,
      "loss": 3.2957,
      "step": 30040
    },
    {
      "epoch": 0.3938710776732115,
      "grad_norm": 0.6187857985496521,
      "learning_rate": 0.00020087304192539934,
      "loss": 3.241,
      "step": 30050
    },
    {
      "epoch": 0.39400214957925916,
      "grad_norm": 0.5547912120819092,
      "learning_rate": 0.00020081454888469684,
      "loss": 3.3259,
      "step": 30060
    },
    {
      "epoch": 0.39413322148530683,
      "grad_norm": 0.771360456943512,
      "learning_rate": 0.00020075604711391562,
      "loss": 3.34,
      "step": 30070
    },
    {
      "epoch": 0.3942642933913545,
      "grad_norm": 0.6003206372261047,
      "learning_rate": 0.0002006975366231065,
      "loss": 3.3238,
      "step": 30080
    },
    {
      "epoch": 0.39439536529740216,
      "grad_norm": 0.5809360146522522,
      "learning_rate": 0.00020063901742232168,
      "loss": 3.2995,
      "step": 30090
    },
    {
      "epoch": 0.3945264372034498,
      "grad_norm": 1.1612625122070312,
      "learning_rate": 0.00020058048952161496,
      "loss": 3.3664,
      "step": 30100
    },
    {
      "epoch": 0.3946575091094975,
      "grad_norm": 0.6193740963935852,
      "learning_rate": 0.00020052195293104156,
      "loss": 3.3263,
      "step": 30110
    },
    {
      "epoch": 0.39478858101554515,
      "grad_norm": 0.5884212255477905,
      "learning_rate": 0.00020046340766065825,
      "loss": 3.304,
      "step": 30120
    },
    {
      "epoch": 0.3949196529215928,
      "grad_norm": 0.6232123970985413,
      "learning_rate": 0.00020040485372052326,
      "loss": 3.2587,
      "step": 30130
    },
    {
      "epoch": 0.3950507248276404,
      "grad_norm": 0.5954776406288147,
      "learning_rate": 0.00020034629112069633,
      "loss": 3.2741,
      "step": 30140
    },
    {
      "epoch": 0.3951817967336881,
      "grad_norm": 0.5500792264938354,
      "learning_rate": 0.00020028771987123868,
      "loss": 3.2973,
      "step": 30150
    },
    {
      "epoch": 0.39531286863973575,
      "grad_norm": 0.5386250019073486,
      "learning_rate": 0.000200229139982213,
      "loss": 3.3056,
      "step": 30160
    },
    {
      "epoch": 0.3954439405457834,
      "grad_norm": 0.5552341938018799,
      "learning_rate": 0.00020017055146368348,
      "loss": 3.2983,
      "step": 30170
    },
    {
      "epoch": 0.3955750124518311,
      "grad_norm": 0.5851075649261475,
      "learning_rate": 0.00020011195432571578,
      "loss": 3.3136,
      "step": 30180
    },
    {
      "epoch": 0.39570608435787874,
      "grad_norm": 0.5185765027999878,
      "learning_rate": 0.00020005334857837702,
      "loss": 3.2835,
      "step": 30190
    },
    {
      "epoch": 0.3958371562639264,
      "grad_norm": 0.5891397595405579,
      "learning_rate": 0.00019999473423173587,
      "loss": 3.2454,
      "step": 30200
    },
    {
      "epoch": 0.39596822816997407,
      "grad_norm": 0.5838890075683594,
      "learning_rate": 0.00019993611129586242,
      "loss": 3.2667,
      "step": 30210
    },
    {
      "epoch": 0.39609930007602173,
      "grad_norm": 0.6285673379898071,
      "learning_rate": 0.00019987747978082826,
      "loss": 3.292,
      "step": 30220
    },
    {
      "epoch": 0.39623037198206934,
      "grad_norm": 0.5617315769195557,
      "learning_rate": 0.00019981883969670644,
      "loss": 3.3106,
      "step": 30230
    },
    {
      "epoch": 0.396361443888117,
      "grad_norm": 1.0706000328063965,
      "learning_rate": 0.0001997601910535715,
      "loss": 3.3411,
      "step": 30240
    },
    {
      "epoch": 0.39649251579416467,
      "grad_norm": 0.6643182635307312,
      "learning_rate": 0.00019970153386149943,
      "loss": 3.3081,
      "step": 30250
    },
    {
      "epoch": 0.39662358770021233,
      "grad_norm": 0.5706015229225159,
      "learning_rate": 0.0001996428681305677,
      "loss": 3.2738,
      "step": 30260
    },
    {
      "epoch": 0.39675465960626,
      "grad_norm": 0.553134560585022,
      "learning_rate": 0.00019958419387085518,
      "loss": 3.3001,
      "step": 30270
    },
    {
      "epoch": 0.39688573151230766,
      "grad_norm": 0.5756497979164124,
      "learning_rate": 0.00019952551109244233,
      "loss": 3.3061,
      "step": 30280
    },
    {
      "epoch": 0.3970168034183553,
      "grad_norm": 0.5949447154998779,
      "learning_rate": 0.00019946681980541106,
      "loss": 3.2979,
      "step": 30290
    },
    {
      "epoch": 0.397147875324403,
      "grad_norm": 0.5374748110771179,
      "learning_rate": 0.00019940812001984458,
      "loss": 3.3178,
      "step": 30300
    },
    {
      "epoch": 0.39727894723045065,
      "grad_norm": 0.5644000768661499,
      "learning_rate": 0.00019934941174582776,
      "loss": 3.2958,
      "step": 30310
    },
    {
      "epoch": 0.39741001913649826,
      "grad_norm": 0.49704596400260925,
      "learning_rate": 0.0001992906949934468,
      "loss": 3.3091,
      "step": 30320
    },
    {
      "epoch": 0.3975410910425459,
      "grad_norm": 0.547673761844635,
      "learning_rate": 0.00019923196977278943,
      "loss": 3.2849,
      "step": 30330
    },
    {
      "epoch": 0.3976721629485936,
      "grad_norm": 0.5347675085067749,
      "learning_rate": 0.00019917323609394477,
      "loss": 3.2854,
      "step": 30340
    },
    {
      "epoch": 0.39780323485464125,
      "grad_norm": 0.6133899688720703,
      "learning_rate": 0.0001991144939670035,
      "loss": 3.2894,
      "step": 30350
    },
    {
      "epoch": 0.3979343067606889,
      "grad_norm": 0.5833730697631836,
      "learning_rate": 0.00019905574340205753,
      "loss": 3.2813,
      "step": 30360
    },
    {
      "epoch": 0.3980653786667366,
      "grad_norm": 0.5710664987564087,
      "learning_rate": 0.00019899698440920053,
      "loss": 3.2682,
      "step": 30370
    },
    {
      "epoch": 0.39819645057278424,
      "grad_norm": 0.55311119556427,
      "learning_rate": 0.00019893821699852737,
      "loss": 3.3039,
      "step": 30380
    },
    {
      "epoch": 0.3983275224788319,
      "grad_norm": 0.5700420141220093,
      "learning_rate": 0.00019887944118013445,
      "loss": 3.2808,
      "step": 30390
    },
    {
      "epoch": 0.39845859438487957,
      "grad_norm": 0.5393562316894531,
      "learning_rate": 0.00019882065696411963,
      "loss": 3.305,
      "step": 30400
    },
    {
      "epoch": 0.3985896662909272,
      "grad_norm": 0.5589162111282349,
      "learning_rate": 0.00019876186436058223,
      "loss": 3.3175,
      "step": 30410
    },
    {
      "epoch": 0.39872073819697484,
      "grad_norm": 0.5793256163597107,
      "learning_rate": 0.00019870306337962297,
      "loss": 3.2862,
      "step": 30420
    },
    {
      "epoch": 0.3988518101030225,
      "grad_norm": 0.601243257522583,
      "learning_rate": 0.00019864425403134402,
      "loss": 3.2768,
      "step": 30430
    },
    {
      "epoch": 0.39898288200907017,
      "grad_norm": 0.5618674755096436,
      "learning_rate": 0.00019858543632584892,
      "loss": 3.2983,
      "step": 30440
    },
    {
      "epoch": 0.39911395391511784,
      "grad_norm": 0.540365993976593,
      "learning_rate": 0.00019852661027324275,
      "loss": 3.2807,
      "step": 30450
    },
    {
      "epoch": 0.3992450258211655,
      "grad_norm": 0.5693140029907227,
      "learning_rate": 0.00019846777588363202,
      "loss": 3.2604,
      "step": 30460
    },
    {
      "epoch": 0.39937609772721316,
      "grad_norm": 0.6226419806480408,
      "learning_rate": 0.00019840893316712464,
      "loss": 3.2943,
      "step": 30470
    },
    {
      "epoch": 0.39950716963326083,
      "grad_norm": 0.540003776550293,
      "learning_rate": 0.00019835008213382993,
      "loss": 3.2939,
      "step": 30480
    },
    {
      "epoch": 0.3996382415393085,
      "grad_norm": 0.6361942291259766,
      "learning_rate": 0.00019829122279385863,
      "loss": 3.35,
      "step": 30490
    },
    {
      "epoch": 0.3997693134453561,
      "grad_norm": 0.5673786997795105,
      "learning_rate": 0.000198232355157323,
      "loss": 3.2865,
      "step": 30500
    },
    {
      "epoch": 0.3997693134453561,
      "eval_accuracy": 0.39088123167155425,
      "eval_loss": 3.3008861541748047,
      "eval_runtime": 60.9746,
      "eval_samples_per_second": 32.801,
      "eval_steps_per_second": 0.525,
      "step": 30500
    },
    {
      "epoch": 0.39990038535140376,
      "grad_norm": 0.6092134118080139,
      "learning_rate": 0.00019817347923433665,
      "loss": 3.3445,
      "step": 30510
    },
    {
      "epoch": 0.40003145725745143,
      "grad_norm": 0.5463336110115051,
      "learning_rate": 0.00019811459503501455,
      "loss": 3.2862,
      "step": 30520
    },
    {
      "epoch": 0.4001625291634991,
      "grad_norm": 0.6977014541625977,
      "learning_rate": 0.0001980557025694732,
      "loss": 3.3011,
      "step": 30530
    },
    {
      "epoch": 0.40029360106954676,
      "grad_norm": 0.5298960208892822,
      "learning_rate": 0.00019799680184783057,
      "loss": 3.2829,
      "step": 30540
    },
    {
      "epoch": 0.4004246729755944,
      "grad_norm": 0.5831784009933472,
      "learning_rate": 0.0001979378928802059,
      "loss": 3.2836,
      "step": 30550
    },
    {
      "epoch": 0.4005557448816421,
      "grad_norm": 0.5925511121749878,
      "learning_rate": 0.00019787897567671986,
      "loss": 3.3618,
      "step": 30560
    },
    {
      "epoch": 0.40068681678768975,
      "grad_norm": 0.541649580001831,
      "learning_rate": 0.00019782005024749467,
      "loss": 3.2707,
      "step": 30570
    },
    {
      "epoch": 0.4008178886937374,
      "grad_norm": 0.5580195188522339,
      "learning_rate": 0.0001977611166026538,
      "loss": 3.2594,
      "step": 30580
    },
    {
      "epoch": 0.400948960599785,
      "grad_norm": 0.5577034950256348,
      "learning_rate": 0.00019770217475232233,
      "loss": 3.3191,
      "step": 30590
    },
    {
      "epoch": 0.4010800325058327,
      "grad_norm": 0.545274555683136,
      "learning_rate": 0.00019764322470662652,
      "loss": 3.299,
      "step": 30600
    },
    {
      "epoch": 0.40121110441188035,
      "grad_norm": 0.5747260451316833,
      "learning_rate": 0.00019758426647569424,
      "loss": 3.3074,
      "step": 30610
    },
    {
      "epoch": 0.401342176317928,
      "grad_norm": 0.6026456952095032,
      "learning_rate": 0.00019752530006965458,
      "loss": 3.3145,
      "step": 30620
    },
    {
      "epoch": 0.4014732482239757,
      "grad_norm": 0.5632624626159668,
      "learning_rate": 0.0001974663254986382,
      "loss": 3.2815,
      "step": 30630
    },
    {
      "epoch": 0.40160432013002334,
      "grad_norm": 0.6011442542076111,
      "learning_rate": 0.00019740734277277704,
      "loss": 3.2953,
      "step": 30640
    },
    {
      "epoch": 0.401735392036071,
      "grad_norm": 0.5857046842575073,
      "learning_rate": 0.00019734835190220456,
      "loss": 3.406,
      "step": 30650
    },
    {
      "epoch": 0.40186646394211867,
      "grad_norm": 0.5670195817947388,
      "learning_rate": 0.00019728935289705548,
      "loss": 3.3178,
      "step": 30660
    },
    {
      "epoch": 0.4019975358481663,
      "grad_norm": 0.6288614273071289,
      "learning_rate": 0.00019723034576746607,
      "loss": 3.2959,
      "step": 30670
    },
    {
      "epoch": 0.40212860775421394,
      "grad_norm": 0.5568493008613586,
      "learning_rate": 0.00019717133052357387,
      "loss": 3.2768,
      "step": 30680
    },
    {
      "epoch": 0.4022596796602616,
      "grad_norm": 0.5679060816764832,
      "learning_rate": 0.00019711230717551782,
      "loss": 3.3069,
      "step": 30690
    },
    {
      "epoch": 0.40239075156630927,
      "grad_norm": 0.5857713222503662,
      "learning_rate": 0.00019705327573343837,
      "loss": 3.3305,
      "step": 30700
    },
    {
      "epoch": 0.40252182347235693,
      "grad_norm": 0.5541185736656189,
      "learning_rate": 0.00019699423620747728,
      "loss": 3.3439,
      "step": 30710
    },
    {
      "epoch": 0.4026528953784046,
      "grad_norm": 0.5638755559921265,
      "learning_rate": 0.00019693518860777764,
      "loss": 3.2714,
      "step": 30720
    },
    {
      "epoch": 0.40278396728445226,
      "grad_norm": 0.7535611391067505,
      "learning_rate": 0.000196876132944484,
      "loss": 3.3528,
      "step": 30730
    },
    {
      "epoch": 0.4029150391904999,
      "grad_norm": 0.5713750720024109,
      "learning_rate": 0.00019681706922774236,
      "loss": 3.2863,
      "step": 30740
    },
    {
      "epoch": 0.4030461110965476,
      "grad_norm": 0.5604838728904724,
      "learning_rate": 0.00019675799746769995,
      "loss": 3.3011,
      "step": 30750
    },
    {
      "epoch": 0.4031771830025952,
      "grad_norm": 0.5653800368309021,
      "learning_rate": 0.00019669891767450548,
      "loss": 3.293,
      "step": 30760
    },
    {
      "epoch": 0.40330825490864286,
      "grad_norm": 0.5883225202560425,
      "learning_rate": 0.00019663982985830905,
      "loss": 3.3228,
      "step": 30770
    },
    {
      "epoch": 0.4034393268146905,
      "grad_norm": 0.5270152688026428,
      "learning_rate": 0.00019658073402926206,
      "loss": 3.2856,
      "step": 30780
    },
    {
      "epoch": 0.4035703987207382,
      "grad_norm": 0.5802022814750671,
      "learning_rate": 0.00019652163019751736,
      "loss": 3.3004,
      "step": 30790
    },
    {
      "epoch": 0.40370147062678585,
      "grad_norm": 0.5541126728057861,
      "learning_rate": 0.00019646251837322922,
      "loss": 3.2986,
      "step": 30800
    },
    {
      "epoch": 0.4038325425328335,
      "grad_norm": 0.9400596618652344,
      "learning_rate": 0.00019640339856655308,
      "loss": 3.3109,
      "step": 30810
    },
    {
      "epoch": 0.4039636144388812,
      "grad_norm": 0.585629403591156,
      "learning_rate": 0.00019634427078764593,
      "loss": 3.3,
      "step": 30820
    },
    {
      "epoch": 0.40409468634492884,
      "grad_norm": 0.5698771476745605,
      "learning_rate": 0.00019628513504666616,
      "loss": 3.3083,
      "step": 30830
    },
    {
      "epoch": 0.4042257582509765,
      "grad_norm": 0.5789401531219482,
      "learning_rate": 0.00019622599135377334,
      "loss": 3.313,
      "step": 30840
    },
    {
      "epoch": 0.4043568301570241,
      "grad_norm": 0.6086841821670532,
      "learning_rate": 0.0001961668397191286,
      "loss": 3.3098,
      "step": 30850
    },
    {
      "epoch": 0.4044879020630718,
      "grad_norm": 0.5331345200538635,
      "learning_rate": 0.00019610768015289432,
      "loss": 3.2974,
      "step": 30860
    },
    {
      "epoch": 0.40461897396911944,
      "grad_norm": 0.619612455368042,
      "learning_rate": 0.00019604851266523428,
      "loss": 3.3601,
      "step": 30870
    },
    {
      "epoch": 0.4047500458751671,
      "grad_norm": 0.6499446630477905,
      "learning_rate": 0.00019598933726631366,
      "loss": 3.2921,
      "step": 30880
    },
    {
      "epoch": 0.40488111778121477,
      "grad_norm": 0.6291869282722473,
      "learning_rate": 0.0001959301539662989,
      "loss": 3.3028,
      "step": 30890
    },
    {
      "epoch": 0.40501218968726244,
      "grad_norm": 0.5733163952827454,
      "learning_rate": 0.00019587096277535784,
      "loss": 3.2901,
      "step": 30900
    },
    {
      "epoch": 0.4051432615933101,
      "grad_norm": 0.5586735606193542,
      "learning_rate": 0.00019581176370365972,
      "loss": 3.2994,
      "step": 30910
    },
    {
      "epoch": 0.40527433349935776,
      "grad_norm": 0.5413002371788025,
      "learning_rate": 0.00019575255676137511,
      "loss": 3.3016,
      "step": 30920
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 0.521663248538971,
      "learning_rate": 0.0001956933419586759,
      "loss": 3.2974,
      "step": 30930
    },
    {
      "epoch": 0.40553647731145304,
      "grad_norm": 0.5564540028572083,
      "learning_rate": 0.0001956341193057354,
      "loss": 3.2776,
      "step": 30940
    },
    {
      "epoch": 0.4056675492175007,
      "grad_norm": 0.5711293816566467,
      "learning_rate": 0.00019557488881272812,
      "loss": 3.2871,
      "step": 30950
    },
    {
      "epoch": 0.40579862112354836,
      "grad_norm": 0.6232595443725586,
      "learning_rate": 0.0001955156504898302,
      "loss": 3.3257,
      "step": 30960
    },
    {
      "epoch": 0.40592969302959603,
      "grad_norm": 0.5748517513275146,
      "learning_rate": 0.0001954564043472188,
      "loss": 3.2931,
      "step": 30970
    },
    {
      "epoch": 0.4060607649356437,
      "grad_norm": 0.5678707361221313,
      "learning_rate": 0.00019539715039507263,
      "loss": 3.3116,
      "step": 30980
    },
    {
      "epoch": 0.40619183684169136,
      "grad_norm": 0.5806042551994324,
      "learning_rate": 0.0001953378886435716,
      "loss": 3.2882,
      "step": 30990
    },
    {
      "epoch": 0.406322908747739,
      "grad_norm": 0.5512754321098328,
      "learning_rate": 0.00019527861910289713,
      "loss": 3.2928,
      "step": 31000
    },
    {
      "epoch": 0.406322908747739,
      "eval_accuracy": 0.3912590420332356,
      "eval_loss": 3.2954955101013184,
      "eval_runtime": 56.82,
      "eval_samples_per_second": 35.199,
      "eval_steps_per_second": 0.563,
      "step": 31000
    },
    {
      "epoch": 0.4064539806537867,
      "grad_norm": 0.5390644073486328,
      "learning_rate": 0.00019521934178323185,
      "loss": 3.3042,
      "step": 31010
    },
    {
      "epoch": 0.40658505255983435,
      "grad_norm": 0.5608130693435669,
      "learning_rate": 0.00019516005669475983,
      "loss": 3.2714,
      "step": 31020
    },
    {
      "epoch": 0.40671612446588196,
      "grad_norm": 0.5771498084068298,
      "learning_rate": 0.0001951007638476663,
      "loss": 3.3073,
      "step": 31030
    },
    {
      "epoch": 0.4068471963719296,
      "grad_norm": 0.5869618654251099,
      "learning_rate": 0.000195041463252138,
      "loss": 3.2513,
      "step": 31040
    },
    {
      "epoch": 0.4069782682779773,
      "grad_norm": 0.5522969365119934,
      "learning_rate": 0.00019498215491836298,
      "loss": 3.2646,
      "step": 31050
    },
    {
      "epoch": 0.40710934018402495,
      "grad_norm": 0.5844562649726868,
      "learning_rate": 0.00019492283885653048,
      "loss": 3.3068,
      "step": 31060
    },
    {
      "epoch": 0.4072404120900726,
      "grad_norm": 0.5554916262626648,
      "learning_rate": 0.00019486351507683124,
      "loss": 3.296,
      "step": 31070
    },
    {
      "epoch": 0.4073714839961203,
      "grad_norm": 0.5547246932983398,
      "learning_rate": 0.00019480418358945714,
      "loss": 3.3123,
      "step": 31080
    },
    {
      "epoch": 0.40750255590216794,
      "grad_norm": 0.5590881705284119,
      "learning_rate": 0.0001947448444046016,
      "loss": 3.3028,
      "step": 31090
    },
    {
      "epoch": 0.4076336278082156,
      "grad_norm": 0.5283000469207764,
      "learning_rate": 0.0001946854975324592,
      "loss": 3.2839,
      "step": 31100
    },
    {
      "epoch": 0.40776469971426327,
      "grad_norm": 0.6259887218475342,
      "learning_rate": 0.00019462614298322597,
      "loss": 3.2595,
      "step": 31110
    },
    {
      "epoch": 0.4078957716203109,
      "grad_norm": 0.6001785397529602,
      "learning_rate": 0.00019456678076709905,
      "loss": 3.2573,
      "step": 31120
    },
    {
      "epoch": 0.40802684352635854,
      "grad_norm": 0.6221110820770264,
      "learning_rate": 0.00019450741089427717,
      "loss": 3.3107,
      "step": 31130
    },
    {
      "epoch": 0.4081579154324062,
      "grad_norm": 0.6193087697029114,
      "learning_rate": 0.00019444803337496016,
      "loss": 3.3022,
      "step": 31140
    },
    {
      "epoch": 0.40828898733845387,
      "grad_norm": 0.6268954873085022,
      "learning_rate": 0.00019438864821934927,
      "loss": 3.3146,
      "step": 31150
    },
    {
      "epoch": 0.40842005924450153,
      "grad_norm": 0.559092104434967,
      "learning_rate": 0.00019432925543764695,
      "loss": 3.2918,
      "step": 31160
    },
    {
      "epoch": 0.4085511311505492,
      "grad_norm": 0.6447961330413818,
      "learning_rate": 0.00019426985504005718,
      "loss": 3.3095,
      "step": 31170
    },
    {
      "epoch": 0.40868220305659686,
      "grad_norm": 0.6030747890472412,
      "learning_rate": 0.00019421044703678503,
      "loss": 3.3166,
      "step": 31180
    },
    {
      "epoch": 0.4088132749626445,
      "grad_norm": 0.5676069259643555,
      "learning_rate": 0.00019415103143803693,
      "loss": 3.2956,
      "step": 31190
    },
    {
      "epoch": 0.4089443468686922,
      "grad_norm": 0.8105384707450867,
      "learning_rate": 0.00019409160825402074,
      "loss": 3.3204,
      "step": 31200
    },
    {
      "epoch": 0.4090754187747398,
      "grad_norm": 0.9519800543785095,
      "learning_rate": 0.0001940321774949454,
      "loss": 3.2887,
      "step": 31210
    },
    {
      "epoch": 0.40920649068078746,
      "grad_norm": 0.5562341213226318,
      "learning_rate": 0.00019397273917102142,
      "loss": 3.2796,
      "step": 31220
    },
    {
      "epoch": 0.4093375625868351,
      "grad_norm": 0.5369704961776733,
      "learning_rate": 0.00019391329329246034,
      "loss": 3.3624,
      "step": 31230
    },
    {
      "epoch": 0.4094686344928828,
      "grad_norm": 0.6438228487968445,
      "learning_rate": 0.0001938538398694752,
      "loss": 3.3222,
      "step": 31240
    },
    {
      "epoch": 0.40959970639893045,
      "grad_norm": 0.616724967956543,
      "learning_rate": 0.0001937943789122802,
      "loss": 3.3324,
      "step": 31250
    },
    {
      "epoch": 0.4097307783049781,
      "grad_norm": 0.5845969915390015,
      "learning_rate": 0.00019373491043109096,
      "loss": 3.3209,
      "step": 31260
    },
    {
      "epoch": 0.4098618502110258,
      "grad_norm": 0.5978431701660156,
      "learning_rate": 0.00019367543443612431,
      "loss": 3.2787,
      "step": 31270
    },
    {
      "epoch": 0.40999292211707344,
      "grad_norm": 0.5864400267601013,
      "learning_rate": 0.00019361595093759837,
      "loss": 3.2911,
      "step": 31280
    },
    {
      "epoch": 0.4101239940231211,
      "grad_norm": 0.5958350896835327,
      "learning_rate": 0.00019355645994573254,
      "loss": 3.3291,
      "step": 31290
    },
    {
      "epoch": 0.4102550659291687,
      "grad_norm": 0.5558182597160339,
      "learning_rate": 0.00019349696147074764,
      "loss": 3.3064,
      "step": 31300
    },
    {
      "epoch": 0.4103861378352164,
      "grad_norm": 0.5487284064292908,
      "learning_rate": 0.0001934374555228656,
      "loss": 3.2913,
      "step": 31310
    },
    {
      "epoch": 0.41051720974126404,
      "grad_norm": 0.5635449290275574,
      "learning_rate": 0.0001933779421123097,
      "loss": 3.2836,
      "step": 31320
    },
    {
      "epoch": 0.4106482816473117,
      "grad_norm": 0.5371512770652771,
      "learning_rate": 0.0001933184212493045,
      "loss": 3.2218,
      "step": 31330
    },
    {
      "epoch": 0.41077935355335937,
      "grad_norm": 0.5623146295547485,
      "learning_rate": 0.00019325889294407586,
      "loss": 3.3171,
      "step": 31340
    },
    {
      "epoch": 0.41091042545940704,
      "grad_norm": 0.5844888091087341,
      "learning_rate": 0.00019319935720685092,
      "loss": 3.2396,
      "step": 31350
    },
    {
      "epoch": 0.4110414973654547,
      "grad_norm": 0.54118412733078,
      "learning_rate": 0.00019313981404785803,
      "loss": 3.3652,
      "step": 31360
    },
    {
      "epoch": 0.41117256927150236,
      "grad_norm": 0.6698023080825806,
      "learning_rate": 0.00019308026347732697,
      "loss": 3.2693,
      "step": 31370
    },
    {
      "epoch": 0.41130364117755003,
      "grad_norm": 0.5831987261772156,
      "learning_rate": 0.00019302070550548858,
      "loss": 3.3581,
      "step": 31380
    },
    {
      "epoch": 0.41143471308359764,
      "grad_norm": 0.5527037382125854,
      "learning_rate": 0.00019296114014257515,
      "loss": 3.3313,
      "step": 31390
    },
    {
      "epoch": 0.4115657849896453,
      "grad_norm": 0.5933530926704407,
      "learning_rate": 0.00019290156739882016,
      "loss": 3.3241,
      "step": 31400
    },
    {
      "epoch": 0.41169685689569296,
      "grad_norm": 0.5625525116920471,
      "learning_rate": 0.0001928419872844584,
      "loss": 3.327,
      "step": 31410
    },
    {
      "epoch": 0.41182792880174063,
      "grad_norm": 0.5535630583763123,
      "learning_rate": 0.00019278239980972584,
      "loss": 3.269,
      "step": 31420
    },
    {
      "epoch": 0.4119590007077883,
      "grad_norm": 0.5382465124130249,
      "learning_rate": 0.0001927228049848598,
      "loss": 3.277,
      "step": 31430
    },
    {
      "epoch": 0.41209007261383596,
      "grad_norm": 0.7002916932106018,
      "learning_rate": 0.0001926632028200988,
      "loss": 3.2769,
      "step": 31440
    },
    {
      "epoch": 0.4122211445198836,
      "grad_norm": 0.6327513456344604,
      "learning_rate": 0.0001926035933256827,
      "loss": 3.2931,
      "step": 31450
    },
    {
      "epoch": 0.4123522164259313,
      "grad_norm": 0.5446491837501526,
      "learning_rate": 0.0001925439765118526,
      "loss": 3.3602,
      "step": 31460
    },
    {
      "epoch": 0.41248328833197895,
      "grad_norm": 0.5876770615577698,
      "learning_rate": 0.00019248435238885072,
      "loss": 3.2719,
      "step": 31470
    },
    {
      "epoch": 0.41261436023802656,
      "grad_norm": 0.5498236417770386,
      "learning_rate": 0.00019242472096692075,
      "loss": 3.2789,
      "step": 31480
    },
    {
      "epoch": 0.4127454321440742,
      "grad_norm": 0.5478453040122986,
      "learning_rate": 0.00019236508225630748,
      "loss": 3.3485,
      "step": 31490
    },
    {
      "epoch": 0.4128765040501219,
      "grad_norm": 0.7463825345039368,
      "learning_rate": 0.00019230543626725708,
      "loss": 3.3693,
      "step": 31500
    },
    {
      "epoch": 0.4128765040501219,
      "eval_accuracy": 0.3909115347018573,
      "eval_loss": 3.29866361618042,
      "eval_runtime": 61.4761,
      "eval_samples_per_second": 32.533,
      "eval_steps_per_second": 0.521,
      "step": 31500
    },
    {
      "epoch": 0.41300757595616955,
      "grad_norm": 0.7022485136985779,
      "learning_rate": 0.00019224578301001678,
      "loss": 3.2947,
      "step": 31510
    },
    {
      "epoch": 0.4131386478622172,
      "grad_norm": 0.5607333779335022,
      "learning_rate": 0.00019218612249483527,
      "loss": 3.2763,
      "step": 31520
    },
    {
      "epoch": 0.4132697197682649,
      "grad_norm": 0.5545626878738403,
      "learning_rate": 0.00019212645473196233,
      "loss": 3.3051,
      "step": 31530
    },
    {
      "epoch": 0.41340079167431254,
      "grad_norm": 0.7049106955528259,
      "learning_rate": 0.0001920667797316491,
      "loss": 3.2783,
      "step": 31540
    },
    {
      "epoch": 0.4135318635803602,
      "grad_norm": 0.5535250902175903,
      "learning_rate": 0.00019200709750414783,
      "loss": 3.3254,
      "step": 31550
    },
    {
      "epoch": 0.41366293548640787,
      "grad_norm": 0.8475584983825684,
      "learning_rate": 0.00019194740805971212,
      "loss": 3.3989,
      "step": 31560
    },
    {
      "epoch": 0.4137940073924555,
      "grad_norm": 0.6572781205177307,
      "learning_rate": 0.00019188771140859684,
      "loss": 3.3034,
      "step": 31570
    },
    {
      "epoch": 0.41392507929850314,
      "grad_norm": 0.5603308081626892,
      "learning_rate": 0.00019182800756105796,
      "loss": 3.2664,
      "step": 31580
    },
    {
      "epoch": 0.4140561512045508,
      "grad_norm": 0.5592421293258667,
      "learning_rate": 0.00019176829652735284,
      "loss": 3.3024,
      "step": 31590
    },
    {
      "epoch": 0.41418722311059847,
      "grad_norm": 0.57105952501297,
      "learning_rate": 0.00019170857831773993,
      "loss": 3.3499,
      "step": 31600
    },
    {
      "epoch": 0.41431829501664613,
      "grad_norm": 0.5879971981048584,
      "learning_rate": 0.00019164885294247903,
      "loss": 3.3205,
      "step": 31610
    },
    {
      "epoch": 0.4144493669226938,
      "grad_norm": 0.5788957476615906,
      "learning_rate": 0.00019158912041183107,
      "loss": 3.2731,
      "step": 31620
    },
    {
      "epoch": 0.41458043882874146,
      "grad_norm": 0.6341288685798645,
      "learning_rate": 0.0001915293807360583,
      "loss": 3.3192,
      "step": 31630
    },
    {
      "epoch": 0.4147115107347891,
      "grad_norm": 0.5827239155769348,
      "learning_rate": 0.00019146963392542416,
      "loss": 3.2546,
      "step": 31640
    },
    {
      "epoch": 0.4148425826408368,
      "grad_norm": 0.6102914810180664,
      "learning_rate": 0.00019140987999019331,
      "loss": 3.2983,
      "step": 31650
    },
    {
      "epoch": 0.4149736545468844,
      "grad_norm": 0.6544795036315918,
      "learning_rate": 0.00019135011894063164,
      "loss": 3.3504,
      "step": 31660
    },
    {
      "epoch": 0.41510472645293206,
      "grad_norm": 0.5908272862434387,
      "learning_rate": 0.0001912903507870063,
      "loss": 3.3556,
      "step": 31670
    },
    {
      "epoch": 0.4152357983589797,
      "grad_norm": 0.5318587422370911,
      "learning_rate": 0.00019123057553958552,
      "loss": 3.3113,
      "step": 31680
    },
    {
      "epoch": 0.4153668702650274,
      "grad_norm": 0.5745972394943237,
      "learning_rate": 0.00019117079320863896,
      "loss": 3.2648,
      "step": 31690
    },
    {
      "epoch": 0.41549794217107505,
      "grad_norm": 0.5546407103538513,
      "learning_rate": 0.00019111100380443736,
      "loss": 3.2458,
      "step": 31700
    },
    {
      "epoch": 0.4156290140771227,
      "grad_norm": 0.6765428185462952,
      "learning_rate": 0.00019105120733725264,
      "loss": 3.3294,
      "step": 31710
    },
    {
      "epoch": 0.4157600859831704,
      "grad_norm": 0.564892053604126,
      "learning_rate": 0.00019099140381735805,
      "loss": 3.3114,
      "step": 31720
    },
    {
      "epoch": 0.41589115788921804,
      "grad_norm": 0.7348278164863586,
      "learning_rate": 0.00019093159325502798,
      "loss": 3.3619,
      "step": 31730
    },
    {
      "epoch": 0.4160222297952657,
      "grad_norm": 0.5790043473243713,
      "learning_rate": 0.0001908717756605381,
      "loss": 3.2659,
      "step": 31740
    },
    {
      "epoch": 0.4161533017013133,
      "grad_norm": 0.6734481453895569,
      "learning_rate": 0.00019081195104416515,
      "loss": 3.3156,
      "step": 31750
    },
    {
      "epoch": 0.416284373607361,
      "grad_norm": 0.5865585207939148,
      "learning_rate": 0.00019075211941618723,
      "loss": 3.2668,
      "step": 31760
    },
    {
      "epoch": 0.41641544551340864,
      "grad_norm": 0.5697574019432068,
      "learning_rate": 0.00019069228078688358,
      "loss": 3.3098,
      "step": 31770
    },
    {
      "epoch": 0.4165465174194563,
      "grad_norm": 0.5860064029693604,
      "learning_rate": 0.00019063243516653466,
      "loss": 3.3947,
      "step": 31780
    },
    {
      "epoch": 0.41667758932550397,
      "grad_norm": 0.5524153709411621,
      "learning_rate": 0.000190572582565422,
      "loss": 3.2974,
      "step": 31790
    },
    {
      "epoch": 0.41680866123155164,
      "grad_norm": 0.5067074298858643,
      "learning_rate": 0.00019051272299382854,
      "loss": 3.302,
      "step": 31800
    },
    {
      "epoch": 0.4169397331375993,
      "grad_norm": 0.5822396874427795,
      "learning_rate": 0.00019045285646203825,
      "loss": 3.3121,
      "step": 31810
    },
    {
      "epoch": 0.41707080504364696,
      "grad_norm": 0.5487459897994995,
      "learning_rate": 0.00019039298298033643,
      "loss": 3.2973,
      "step": 31820
    },
    {
      "epoch": 0.41720187694969463,
      "grad_norm": 0.6046732068061829,
      "learning_rate": 0.0001903331025590095,
      "loss": 3.2893,
      "step": 31830
    },
    {
      "epoch": 0.41733294885574224,
      "grad_norm": 0.6521011590957642,
      "learning_rate": 0.00019027321520834504,
      "loss": 3.2907,
      "step": 31840
    },
    {
      "epoch": 0.4174640207617899,
      "grad_norm": 0.6115524172782898,
      "learning_rate": 0.00019021332093863187,
      "loss": 3.3031,
      "step": 31850
    },
    {
      "epoch": 0.41759509266783756,
      "grad_norm": 0.6073604226112366,
      "learning_rate": 0.00019015341976016004,
      "loss": 3.3274,
      "step": 31860
    },
    {
      "epoch": 0.41772616457388523,
      "grad_norm": 0.5442253351211548,
      "learning_rate": 0.0001900935116832207,
      "loss": 3.3065,
      "step": 31870
    },
    {
      "epoch": 0.4178572364799329,
      "grad_norm": 0.5556290149688721,
      "learning_rate": 0.00019003359671810618,
      "loss": 3.3049,
      "step": 31880
    },
    {
      "epoch": 0.41798830838598056,
      "grad_norm": 0.5472756028175354,
      "learning_rate": 0.00018997367487511013,
      "loss": 3.3057,
      "step": 31890
    },
    {
      "epoch": 0.4181193802920282,
      "grad_norm": 0.5596219301223755,
      "learning_rate": 0.00018991374616452716,
      "loss": 3.3065,
      "step": 31900
    },
    {
      "epoch": 0.4182504521980759,
      "grad_norm": 0.5825459957122803,
      "learning_rate": 0.0001898538105966533,
      "loss": 3.2797,
      "step": 31910
    },
    {
      "epoch": 0.41838152410412355,
      "grad_norm": 0.6987773776054382,
      "learning_rate": 0.0001897938681817856,
      "loss": 3.2726,
      "step": 31920
    },
    {
      "epoch": 0.41851259601017116,
      "grad_norm": 0.5946131348609924,
      "learning_rate": 0.00018973391893022234,
      "loss": 3.2978,
      "step": 31930
    },
    {
      "epoch": 0.4186436679162188,
      "grad_norm": 0.5593259334564209,
      "learning_rate": 0.000189673962852263,
      "loss": 3.2976,
      "step": 31940
    },
    {
      "epoch": 0.4187747398222665,
      "grad_norm": 0.7369834184646606,
      "learning_rate": 0.0001896139999582081,
      "loss": 3.3336,
      "step": 31950
    },
    {
      "epoch": 0.41890581172831415,
      "grad_norm": 0.5621054768562317,
      "learning_rate": 0.00018955403025835955,
      "loss": 3.3055,
      "step": 31960
    },
    {
      "epoch": 0.4190368836343618,
      "grad_norm": 0.5808535218238831,
      "learning_rate": 0.00018949405376302017,
      "loss": 3.2784,
      "step": 31970
    },
    {
      "epoch": 0.4191679555404095,
      "grad_norm": 0.563827395439148,
      "learning_rate": 0.0001894340704824942,
      "loss": 3.3039,
      "step": 31980
    },
    {
      "epoch": 0.41929902744645714,
      "grad_norm": 1.5291097164154053,
      "learning_rate": 0.00018937408042708686,
      "loss": 3.3136,
      "step": 31990
    },
    {
      "epoch": 0.4194300993525048,
      "grad_norm": 0.5424140691757202,
      "learning_rate": 0.00018931408360710472,
      "loss": 3.3108,
      "step": 32000
    },
    {
      "epoch": 0.4194300993525048,
      "eval_accuracy": 0.39192375366568916,
      "eval_loss": 3.2914397716522217,
      "eval_runtime": 61.5436,
      "eval_samples_per_second": 32.497,
      "eval_steps_per_second": 0.52,
      "step": 32000
    },
    {
      "epoch": 0.41956117125855247,
      "grad_norm": 0.7443562746047974,
      "learning_rate": 0.00018925408003285524,
      "loss": 3.3069,
      "step": 32010
    },
    {
      "epoch": 0.4196922431646001,
      "grad_norm": 0.5590479969978333,
      "learning_rate": 0.00018919406971464734,
      "loss": 3.2918,
      "step": 32020
    },
    {
      "epoch": 0.41982331507064774,
      "grad_norm": 0.6384625434875488,
      "learning_rate": 0.00018913405266279083,
      "loss": 3.3527,
      "step": 32030
    },
    {
      "epoch": 0.4199543869766954,
      "grad_norm": 0.580762505531311,
      "learning_rate": 0.00018907402888759692,
      "loss": 3.3058,
      "step": 32040
    },
    {
      "epoch": 0.42008545888274307,
      "grad_norm": 0.8398176431655884,
      "learning_rate": 0.00018901399839937785,
      "loss": 3.3234,
      "step": 32050
    },
    {
      "epoch": 0.42021653078879073,
      "grad_norm": 0.6487817764282227,
      "learning_rate": 0.0001889539612084469,
      "loss": 3.3133,
      "step": 32060
    },
    {
      "epoch": 0.4203476026948384,
      "grad_norm": 0.5580127239227295,
      "learning_rate": 0.00018889391732511876,
      "loss": 3.3041,
      "step": 32070
    },
    {
      "epoch": 0.42047867460088606,
      "grad_norm": 0.5487021803855896,
      "learning_rate": 0.00018883386675970902,
      "loss": 3.3155,
      "step": 32080
    },
    {
      "epoch": 0.4206097465069337,
      "grad_norm": 0.5354493856430054,
      "learning_rate": 0.00018877380952253464,
      "loss": 3.2927,
      "step": 32090
    },
    {
      "epoch": 0.4207408184129814,
      "grad_norm": 0.5311872959136963,
      "learning_rate": 0.00018871374562391352,
      "loss": 3.2853,
      "step": 32100
    },
    {
      "epoch": 0.420871890319029,
      "grad_norm": 0.5631843209266663,
      "learning_rate": 0.0001886536750741649,
      "loss": 3.3098,
      "step": 32110
    },
    {
      "epoch": 0.42100296222507666,
      "grad_norm": 0.5347350239753723,
      "learning_rate": 0.00018859359788360897,
      "loss": 3.2911,
      "step": 32120
    },
    {
      "epoch": 0.4211340341311243,
      "grad_norm": 0.5898985862731934,
      "learning_rate": 0.0001885335140625672,
      "loss": 3.2767,
      "step": 32130
    },
    {
      "epoch": 0.421265106037172,
      "grad_norm": 0.568783700466156,
      "learning_rate": 0.00018847342362136223,
      "loss": 3.32,
      "step": 32140
    },
    {
      "epoch": 0.42139617794321965,
      "grad_norm": 0.584874153137207,
      "learning_rate": 0.00018841332657031762,
      "loss": 3.2957,
      "step": 32150
    },
    {
      "epoch": 0.4215272498492673,
      "grad_norm": 0.6235558390617371,
      "learning_rate": 0.00018835322291975827,
      "loss": 3.2891,
      "step": 32160
    },
    {
      "epoch": 0.421658321755315,
      "grad_norm": 0.5638697147369385,
      "learning_rate": 0.00018829311268001017,
      "loss": 3.2901,
      "step": 32170
    },
    {
      "epoch": 0.42178939366136264,
      "grad_norm": 0.595125138759613,
      "learning_rate": 0.00018823299586140046,
      "loss": 3.3138,
      "step": 32180
    },
    {
      "epoch": 0.4219204655674103,
      "grad_norm": 0.5712145566940308,
      "learning_rate": 0.00018817287247425727,
      "loss": 3.2633,
      "step": 32190
    },
    {
      "epoch": 0.4220515374734579,
      "grad_norm": 0.6170665621757507,
      "learning_rate": 0.00018811274252891,
      "loss": 3.3042,
      "step": 32200
    },
    {
      "epoch": 0.4221826093795056,
      "grad_norm": 0.5815817713737488,
      "learning_rate": 0.0001880526060356892,
      "loss": 3.2385,
      "step": 32210
    },
    {
      "epoch": 0.42231368128555324,
      "grad_norm": 0.5885922908782959,
      "learning_rate": 0.00018799246300492644,
      "loss": 3.3202,
      "step": 32220
    },
    {
      "epoch": 0.4224447531916009,
      "grad_norm": 0.5556522607803345,
      "learning_rate": 0.00018793231344695448,
      "loss": 3.2988,
      "step": 32230
    },
    {
      "epoch": 0.42257582509764857,
      "grad_norm": 0.5432925224304199,
      "learning_rate": 0.00018787215737210713,
      "loss": 3.2877,
      "step": 32240
    },
    {
      "epoch": 0.42270689700369624,
      "grad_norm": 0.6116208434104919,
      "learning_rate": 0.00018781199479071939,
      "loss": 3.321,
      "step": 32250
    },
    {
      "epoch": 0.4228379689097439,
      "grad_norm": 0.5328592658042908,
      "learning_rate": 0.00018775182571312737,
      "loss": 3.2829,
      "step": 32260
    },
    {
      "epoch": 0.42296904081579156,
      "grad_norm": 0.634903609752655,
      "learning_rate": 0.00018769165014966826,
      "loss": 3.259,
      "step": 32270
    },
    {
      "epoch": 0.4231001127218392,
      "grad_norm": 0.5682234764099121,
      "learning_rate": 0.00018763146811068043,
      "loss": 3.2979,
      "step": 32280
    },
    {
      "epoch": 0.42323118462788684,
      "grad_norm": 0.576270341873169,
      "learning_rate": 0.00018757127960650327,
      "loss": 3.2859,
      "step": 32290
    },
    {
      "epoch": 0.4233622565339345,
      "grad_norm": 0.644291877746582,
      "learning_rate": 0.00018751108464747736,
      "loss": 3.2915,
      "step": 32300
    },
    {
      "epoch": 0.42349332843998216,
      "grad_norm": 0.6387091875076294,
      "learning_rate": 0.00018745088324394438,
      "loss": 3.3242,
      "step": 32310
    },
    {
      "epoch": 0.42362440034602983,
      "grad_norm": 0.6109904646873474,
      "learning_rate": 0.00018739067540624702,
      "loss": 3.2958,
      "step": 32320
    },
    {
      "epoch": 0.4237554722520775,
      "grad_norm": 0.5790364742279053,
      "learning_rate": 0.00018733046114472924,
      "loss": 3.2857,
      "step": 32330
    },
    {
      "epoch": 0.42388654415812516,
      "grad_norm": 0.6511567831039429,
      "learning_rate": 0.00018727024046973594,
      "loss": 3.302,
      "step": 32340
    },
    {
      "epoch": 0.4240176160641728,
      "grad_norm": 0.5495197772979736,
      "learning_rate": 0.00018721001339161322,
      "loss": 3.2821,
      "step": 32350
    },
    {
      "epoch": 0.4241486879702205,
      "grad_norm": 0.6289881467819214,
      "learning_rate": 0.0001871497799207083,
      "loss": 3.3021,
      "step": 32360
    },
    {
      "epoch": 0.42427975987626815,
      "grad_norm": 0.6040477156639099,
      "learning_rate": 0.0001870895400673694,
      "loss": 3.2995,
      "step": 32370
    },
    {
      "epoch": 0.42441083178231576,
      "grad_norm": 0.7894755005836487,
      "learning_rate": 0.0001870292938419459,
      "loss": 3.2931,
      "step": 32380
    },
    {
      "epoch": 0.4245419036883634,
      "grad_norm": 0.5537451505661011,
      "learning_rate": 0.00018696904125478831,
      "loss": 3.2603,
      "step": 32390
    },
    {
      "epoch": 0.4246729755944111,
      "grad_norm": 0.5345094203948975,
      "learning_rate": 0.00018690878231624814,
      "loss": 3.2608,
      "step": 32400
    },
    {
      "epoch": 0.42480404750045875,
      "grad_norm": 0.5613906383514404,
      "learning_rate": 0.0001868485170366781,
      "loss": 3.2365,
      "step": 32410
    },
    {
      "epoch": 0.4249351194065064,
      "grad_norm": 0.5659967660903931,
      "learning_rate": 0.00018678824542643182,
      "loss": 3.2594,
      "step": 32420
    },
    {
      "epoch": 0.4250661913125541,
      "grad_norm": 0.5804905295372009,
      "learning_rate": 0.00018672796749586427,
      "loss": 3.3077,
      "step": 32430
    },
    {
      "epoch": 0.42519726321860174,
      "grad_norm": 0.5839203000068665,
      "learning_rate": 0.00018666768325533128,
      "loss": 3.2975,
      "step": 32440
    },
    {
      "epoch": 0.4253283351246494,
      "grad_norm": 0.5865245461463928,
      "learning_rate": 0.0001866073927151898,
      "loss": 3.2862,
      "step": 32450
    },
    {
      "epoch": 0.42545940703069707,
      "grad_norm": 0.551698625087738,
      "learning_rate": 0.00018654709588579804,
      "loss": 3.2859,
      "step": 32460
    },
    {
      "epoch": 0.4255904789367447,
      "grad_norm": 0.5490029454231262,
      "learning_rate": 0.00018648679277751502,
      "loss": 3.3084,
      "step": 32470
    },
    {
      "epoch": 0.42572155084279234,
      "grad_norm": 0.6674831509590149,
      "learning_rate": 0.00018642648340070112,
      "loss": 3.2499,
      "step": 32480
    },
    {
      "epoch": 0.42585262274884,
      "grad_norm": 0.5702871680259705,
      "learning_rate": 0.00018636616776571753,
      "loss": 3.2843,
      "step": 32490
    },
    {
      "epoch": 0.42598369465488767,
      "grad_norm": 0.5689767003059387,
      "learning_rate": 0.00018630584588292675,
      "loss": 3.2915,
      "step": 32500
    },
    {
      "epoch": 0.42598369465488767,
      "eval_accuracy": 0.39197409579667647,
      "eval_loss": 3.289651870727539,
      "eval_runtime": 59.2824,
      "eval_samples_per_second": 33.737,
      "eval_steps_per_second": 0.54,
      "step": 32500
    },
    {
      "epoch": 0.42611476656093533,
      "grad_norm": 0.5832996368408203,
      "learning_rate": 0.00018624551776269215,
      "loss": 3.2725,
      "step": 32510
    },
    {
      "epoch": 0.426245838466983,
      "grad_norm": 0.567526638507843,
      "learning_rate": 0.00018618518341537832,
      "loss": 3.2607,
      "step": 32520
    },
    {
      "epoch": 0.42637691037303066,
      "grad_norm": 0.5492769479751587,
      "learning_rate": 0.00018612484285135083,
      "loss": 3.3085,
      "step": 32530
    },
    {
      "epoch": 0.4265079822790783,
      "grad_norm": 0.5762177109718323,
      "learning_rate": 0.0001860644960809764,
      "loss": 3.2953,
      "step": 32540
    },
    {
      "epoch": 0.426639054185126,
      "grad_norm": 0.5865930914878845,
      "learning_rate": 0.00018600414311462275,
      "loss": 3.3015,
      "step": 32550
    },
    {
      "epoch": 0.4267701260911736,
      "grad_norm": 0.5510077476501465,
      "learning_rate": 0.0001859437839626587,
      "loss": 3.3385,
      "step": 32560
    },
    {
      "epoch": 0.42690119799722126,
      "grad_norm": 0.6427140235900879,
      "learning_rate": 0.0001858834186354541,
      "loss": 3.2819,
      "step": 32570
    },
    {
      "epoch": 0.4270322699032689,
      "grad_norm": 0.5835079550743103,
      "learning_rate": 0.00018582304714337986,
      "loss": 3.3063,
      "step": 32580
    },
    {
      "epoch": 0.4271633418093166,
      "grad_norm": 0.561794638633728,
      "learning_rate": 0.00018576266949680797,
      "loss": 3.3246,
      "step": 32590
    },
    {
      "epoch": 0.42729441371536425,
      "grad_norm": 0.5878064632415771,
      "learning_rate": 0.00018570228570611154,
      "loss": 3.2898,
      "step": 32600
    },
    {
      "epoch": 0.4274254856214119,
      "grad_norm": 0.6068506240844727,
      "learning_rate": 0.00018564189578166456,
      "loss": 3.3095,
      "step": 32610
    },
    {
      "epoch": 0.4275565575274596,
      "grad_norm": 0.5722439885139465,
      "learning_rate": 0.00018558149973384226,
      "loss": 3.2866,
      "step": 32620
    },
    {
      "epoch": 0.42768762943350724,
      "grad_norm": 0.6016666293144226,
      "learning_rate": 0.0001855210975730209,
      "loss": 3.2718,
      "step": 32630
    },
    {
      "epoch": 0.42781870133955485,
      "grad_norm": 0.6653375029563904,
      "learning_rate": 0.0001854606893095776,
      "loss": 3.2917,
      "step": 32640
    },
    {
      "epoch": 0.4279497732456025,
      "grad_norm": 0.602748692035675,
      "learning_rate": 0.00018540027495389077,
      "loss": 3.3168,
      "step": 32650
    },
    {
      "epoch": 0.4280808451516502,
      "grad_norm": 0.5587145090103149,
      "learning_rate": 0.00018533985451633974,
      "loss": 3.2115,
      "step": 32660
    },
    {
      "epoch": 0.42821191705769784,
      "grad_norm": 0.5677564144134521,
      "learning_rate": 0.0001852794280073049,
      "loss": 3.2929,
      "step": 32670
    },
    {
      "epoch": 0.4283429889637455,
      "grad_norm": 0.7072550058364868,
      "learning_rate": 0.00018521899543716766,
      "loss": 3.2876,
      "step": 32680
    },
    {
      "epoch": 0.42847406086979317,
      "grad_norm": 0.5594938397407532,
      "learning_rate": 0.00018515855681631053,
      "loss": 3.2651,
      "step": 32690
    },
    {
      "epoch": 0.42860513277584084,
      "grad_norm": 0.5625730156898499,
      "learning_rate": 0.00018509811215511703,
      "loss": 3.283,
      "step": 32700
    },
    {
      "epoch": 0.4287362046818885,
      "grad_norm": 0.5690421462059021,
      "learning_rate": 0.00018503766146397166,
      "loss": 3.2802,
      "step": 32710
    },
    {
      "epoch": 0.42886727658793616,
      "grad_norm": 0.5841851830482483,
      "learning_rate": 0.00018497720475326012,
      "loss": 3.298,
      "step": 32720
    },
    {
      "epoch": 0.42899834849398377,
      "grad_norm": 0.5609754323959351,
      "learning_rate": 0.000184916742033369,
      "loss": 3.2805,
      "step": 32730
    },
    {
      "epoch": 0.42912942040003144,
      "grad_norm": 0.5460821986198425,
      "learning_rate": 0.00018485627331468594,
      "loss": 3.318,
      "step": 32740
    },
    {
      "epoch": 0.4292604923060791,
      "grad_norm": 0.5542404055595398,
      "learning_rate": 0.00018479579860759956,
      "loss": 3.2771,
      "step": 32750
    },
    {
      "epoch": 0.42939156421212676,
      "grad_norm": 1.4770420789718628,
      "learning_rate": 0.0001847353179224998,
      "loss": 3.2026,
      "step": 32760
    },
    {
      "epoch": 0.42952263611817443,
      "grad_norm": 0.5689968466758728,
      "learning_rate": 0.00018467483126977718,
      "loss": 3.2846,
      "step": 32770
    },
    {
      "epoch": 0.4296537080242221,
      "grad_norm": 0.5612708330154419,
      "learning_rate": 0.0001846143386598236,
      "loss": 3.2982,
      "step": 32780
    },
    {
      "epoch": 0.42978477993026976,
      "grad_norm": 0.5880599021911621,
      "learning_rate": 0.00018455384010303174,
      "loss": 3.2887,
      "step": 32790
    },
    {
      "epoch": 0.4299158518363174,
      "grad_norm": 0.5793363451957703,
      "learning_rate": 0.00018449333560979555,
      "loss": 3.2949,
      "step": 32800
    },
    {
      "epoch": 0.4300469237423651,
      "grad_norm": 0.5252099633216858,
      "learning_rate": 0.00018443282519050983,
      "loss": 3.2563,
      "step": 32810
    },
    {
      "epoch": 0.4301779956484127,
      "grad_norm": 0.5951248407363892,
      "learning_rate": 0.0001843723088555704,
      "loss": 3.3121,
      "step": 32820
    },
    {
      "epoch": 0.43030906755446036,
      "grad_norm": 0.5770452618598938,
      "learning_rate": 0.00018431178661537415,
      "loss": 3.305,
      "step": 32830
    },
    {
      "epoch": 0.430440139460508,
      "grad_norm": 0.599729597568512,
      "learning_rate": 0.00018425125848031893,
      "loss": 3.2899,
      "step": 32840
    },
    {
      "epoch": 0.4305712113665557,
      "grad_norm": 0.548062264919281,
      "learning_rate": 0.0001841907244608037,
      "loss": 3.3311,
      "step": 32850
    },
    {
      "epoch": 0.43070228327260335,
      "grad_norm": 0.578693687915802,
      "learning_rate": 0.00018413018456722838,
      "loss": 3.2866,
      "step": 32860
    },
    {
      "epoch": 0.430833355178651,
      "grad_norm": 0.7847079038619995,
      "learning_rate": 0.00018406963880999384,
      "loss": 3.3071,
      "step": 32870
    },
    {
      "epoch": 0.4309644270846987,
      "grad_norm": 0.601173996925354,
      "learning_rate": 0.000184009087199502,
      "loss": 3.2846,
      "step": 32880
    },
    {
      "epoch": 0.43109549899074634,
      "grad_norm": 0.6675735712051392,
      "learning_rate": 0.0001839485297461558,
      "loss": 3.3291,
      "step": 32890
    },
    {
      "epoch": 0.431226570896794,
      "grad_norm": 0.5435544848442078,
      "learning_rate": 0.0001838879664603592,
      "loss": 3.3784,
      "step": 32900
    },
    {
      "epoch": 0.4313576428028416,
      "grad_norm": 0.5809490084648132,
      "learning_rate": 0.00018382739735251717,
      "loss": 3.2854,
      "step": 32910
    },
    {
      "epoch": 0.4314887147088893,
      "grad_norm": 0.6012749075889587,
      "learning_rate": 0.00018376682243303555,
      "loss": 3.249,
      "step": 32920
    },
    {
      "epoch": 0.43161978661493694,
      "grad_norm": 0.6329523921012878,
      "learning_rate": 0.00018370624171232136,
      "loss": 3.2862,
      "step": 32930
    },
    {
      "epoch": 0.4317508585209846,
      "grad_norm": 0.5749862194061279,
      "learning_rate": 0.00018364565520078254,
      "loss": 3.2694,
      "step": 32940
    },
    {
      "epoch": 0.43188193042703227,
      "grad_norm": 0.5737892389297485,
      "learning_rate": 0.000183585062908828,
      "loss": 3.2851,
      "step": 32950
    },
    {
      "epoch": 0.43201300233307993,
      "grad_norm": 0.8088557124137878,
      "learning_rate": 0.0001835244648468676,
      "loss": 3.3393,
      "step": 32960
    },
    {
      "epoch": 0.4321440742391276,
      "grad_norm": 0.6186559796333313,
      "learning_rate": 0.0001834638610253123,
      "loss": 3.2969,
      "step": 32970
    },
    {
      "epoch": 0.43227514614517526,
      "grad_norm": 0.6443630456924438,
      "learning_rate": 0.00018340325145457405,
      "loss": 3.3048,
      "step": 32980
    },
    {
      "epoch": 0.4324062180512229,
      "grad_norm": 0.6378687620162964,
      "learning_rate": 0.00018334263614506565,
      "loss": 3.2567,
      "step": 32990
    },
    {
      "epoch": 0.43253728995727053,
      "grad_norm": 0.6331562399864197,
      "learning_rate": 0.00018328201510720105,
      "loss": 3.32,
      "step": 33000
    },
    {
      "epoch": 0.43253728995727053,
      "eval_accuracy": 0.3923201368523949,
      "eval_loss": 3.287951946258545,
      "eval_runtime": 62.1851,
      "eval_samples_per_second": 32.162,
      "eval_steps_per_second": 0.515,
      "step": 33000
    },
    {
      "epoch": 0.4326683618633182,
      "grad_norm": 0.5633397698402405,
      "learning_rate": 0.00018322138835139503,
      "loss": 3.3153,
      "step": 33010
    },
    {
      "epoch": 0.43279943376936586,
      "grad_norm": 0.5318943858146667,
      "learning_rate": 0.00018316075588806354,
      "loss": 3.3233,
      "step": 33020
    },
    {
      "epoch": 0.4329305056754135,
      "grad_norm": 0.5546523332595825,
      "learning_rate": 0.00018310011772762336,
      "loss": 3.3055,
      "step": 33030
    },
    {
      "epoch": 0.4330615775814612,
      "grad_norm": 0.5880576968193054,
      "learning_rate": 0.00018303947388049224,
      "loss": 3.3041,
      "step": 33040
    },
    {
      "epoch": 0.43319264948750885,
      "grad_norm": 0.5785369873046875,
      "learning_rate": 0.00018297882435708903,
      "loss": 3.3272,
      "step": 33050
    },
    {
      "epoch": 0.4333237213935565,
      "grad_norm": 0.5685509443283081,
      "learning_rate": 0.00018291816916783343,
      "loss": 3.3094,
      "step": 33060
    },
    {
      "epoch": 0.4334547932996042,
      "grad_norm": 0.5763042569160461,
      "learning_rate": 0.00018285750832314622,
      "loss": 3.2752,
      "step": 33070
    },
    {
      "epoch": 0.43358586520565184,
      "grad_norm": 0.5436571836471558,
      "learning_rate": 0.00018279684183344902,
      "loss": 3.2942,
      "step": 33080
    },
    {
      "epoch": 0.43371693711169945,
      "grad_norm": 0.7392051219940186,
      "learning_rate": 0.0001827361697091646,
      "loss": 3.2689,
      "step": 33090
    },
    {
      "epoch": 0.4338480090177471,
      "grad_norm": 0.5639039278030396,
      "learning_rate": 0.0001826754919607165,
      "loss": 3.275,
      "step": 33100
    },
    {
      "epoch": 0.4339790809237948,
      "grad_norm": 0.5614393949508667,
      "learning_rate": 0.0001826148085985294,
      "loss": 3.3131,
      "step": 33110
    },
    {
      "epoch": 0.43411015282984244,
      "grad_norm": 0.6128478646278381,
      "learning_rate": 0.00018255411963302883,
      "loss": 3.2886,
      "step": 33120
    },
    {
      "epoch": 0.4342412247358901,
      "grad_norm": 0.5725920796394348,
      "learning_rate": 0.00018249342507464138,
      "loss": 3.3024,
      "step": 33130
    },
    {
      "epoch": 0.43437229664193777,
      "grad_norm": 0.5158290863037109,
      "learning_rate": 0.00018243272493379442,
      "loss": 3.2885,
      "step": 33140
    },
    {
      "epoch": 0.43450336854798544,
      "grad_norm": 0.5443251729011536,
      "learning_rate": 0.00018237201922091648,
      "loss": 3.2708,
      "step": 33150
    },
    {
      "epoch": 0.4346344404540331,
      "grad_norm": 0.6280168890953064,
      "learning_rate": 0.00018231130794643696,
      "loss": 3.2842,
      "step": 33160
    },
    {
      "epoch": 0.43476551236008076,
      "grad_norm": 0.5514419674873352,
      "learning_rate": 0.00018225059112078626,
      "loss": 3.2721,
      "step": 33170
    },
    {
      "epoch": 0.43489658426612837,
      "grad_norm": 0.5884588360786438,
      "learning_rate": 0.00018218986875439562,
      "loss": 3.3158,
      "step": 33180
    },
    {
      "epoch": 0.43502765617217604,
      "grad_norm": 0.7262667417526245,
      "learning_rate": 0.00018212914085769734,
      "loss": 3.2938,
      "step": 33190
    },
    {
      "epoch": 0.4351587280782237,
      "grad_norm": 0.693303108215332,
      "learning_rate": 0.00018206840744112467,
      "loss": 3.2741,
      "step": 33200
    },
    {
      "epoch": 0.43528979998427136,
      "grad_norm": 0.6259666085243225,
      "learning_rate": 0.00018200766851511177,
      "loss": 3.3026,
      "step": 33210
    },
    {
      "epoch": 0.43542087189031903,
      "grad_norm": 0.6076278686523438,
      "learning_rate": 0.00018194692409009372,
      "loss": 3.2863,
      "step": 33220
    },
    {
      "epoch": 0.4355519437963667,
      "grad_norm": 0.5245094895362854,
      "learning_rate": 0.00018188617417650655,
      "loss": 3.3101,
      "step": 33230
    },
    {
      "epoch": 0.43568301570241436,
      "grad_norm": 0.6178839206695557,
      "learning_rate": 0.00018182541878478738,
      "loss": 3.3098,
      "step": 33240
    },
    {
      "epoch": 0.435814087608462,
      "grad_norm": 0.5885457992553711,
      "learning_rate": 0.00018176465792537402,
      "loss": 3.2824,
      "step": 33250
    },
    {
      "epoch": 0.4359451595145097,
      "grad_norm": 0.5895574688911438,
      "learning_rate": 0.00018170389160870545,
      "loss": 3.2746,
      "step": 33260
    },
    {
      "epoch": 0.4360762314205573,
      "grad_norm": 0.5056623816490173,
      "learning_rate": 0.00018164311984522144,
      "loss": 3.2717,
      "step": 33270
    },
    {
      "epoch": 0.43620730332660496,
      "grad_norm": 0.5212327241897583,
      "learning_rate": 0.00018158234264536278,
      "loss": 3.2725,
      "step": 33280
    },
    {
      "epoch": 0.4363383752326526,
      "grad_norm": 0.5957754254341125,
      "learning_rate": 0.00018152156001957113,
      "loss": 3.283,
      "step": 33290
    },
    {
      "epoch": 0.4364694471387003,
      "grad_norm": 0.541102409362793,
      "learning_rate": 0.00018146077197828911,
      "loss": 3.3108,
      "step": 33300
    },
    {
      "epoch": 0.43660051904474795,
      "grad_norm": 0.5514174699783325,
      "learning_rate": 0.00018139997853196027,
      "loss": 3.2842,
      "step": 33310
    },
    {
      "epoch": 0.4367315909507956,
      "grad_norm": 0.5406527519226074,
      "learning_rate": 0.00018133917969102913,
      "loss": 3.2627,
      "step": 33320
    },
    {
      "epoch": 0.4368626628568433,
      "grad_norm": 0.6051859855651855,
      "learning_rate": 0.00018127837546594102,
      "loss": 3.2927,
      "step": 33330
    },
    {
      "epoch": 0.43699373476289094,
      "grad_norm": 0.5828396081924438,
      "learning_rate": 0.00018121756586714232,
      "loss": 3.2966,
      "step": 33340
    },
    {
      "epoch": 0.4371248066689386,
      "grad_norm": 0.6372126936912537,
      "learning_rate": 0.00018115675090508033,
      "loss": 3.3219,
      "step": 33350
    },
    {
      "epoch": 0.4372558785749862,
      "grad_norm": 0.6474316716194153,
      "learning_rate": 0.0001810959305902032,
      "loss": 3.3003,
      "step": 33360
    },
    {
      "epoch": 0.4373869504810339,
      "grad_norm": 0.5485197305679321,
      "learning_rate": 0.00018103510493295996,
      "loss": 3.3143,
      "step": 33370
    },
    {
      "epoch": 0.43751802238708154,
      "grad_norm": 0.578112781047821,
      "learning_rate": 0.0001809742739438007,
      "loss": 3.3479,
      "step": 33380
    },
    {
      "epoch": 0.4376490942931292,
      "grad_norm": 0.6742129325866699,
      "learning_rate": 0.00018091343763317636,
      "loss": 3.2878,
      "step": 33390
    },
    {
      "epoch": 0.43778016619917687,
      "grad_norm": 0.5933821797370911,
      "learning_rate": 0.0001808525960115387,
      "loss": 3.2677,
      "step": 33400
    },
    {
      "epoch": 0.43791123810522453,
      "grad_norm": 0.5579009056091309,
      "learning_rate": 0.00018079174908934052,
      "loss": 3.2826,
      "step": 33410
    },
    {
      "epoch": 0.4380423100112722,
      "grad_norm": 0.7179008722305298,
      "learning_rate": 0.00018073089687703552,
      "loss": 3.2626,
      "step": 33420
    },
    {
      "epoch": 0.43817338191731986,
      "grad_norm": 0.6019567251205444,
      "learning_rate": 0.00018067003938507822,
      "loss": 3.3217,
      "step": 33430
    },
    {
      "epoch": 0.4383044538233675,
      "grad_norm": 0.5728382468223572,
      "learning_rate": 0.00018060917662392418,
      "loss": 3.2617,
      "step": 33440
    },
    {
      "epoch": 0.43843552572941513,
      "grad_norm": 0.5482255816459656,
      "learning_rate": 0.00018054830860402975,
      "loss": 3.3161,
      "step": 33450
    },
    {
      "epoch": 0.4385665976354628,
      "grad_norm": 0.5186480283737183,
      "learning_rate": 0.0001804874353358522,
      "loss": 3.274,
      "step": 33460
    },
    {
      "epoch": 0.43869766954151046,
      "grad_norm": 0.609590470790863,
      "learning_rate": 0.0001804265568298497,
      "loss": 3.2694,
      "step": 33470
    },
    {
      "epoch": 0.4388287414475581,
      "grad_norm": 0.5492480397224426,
      "learning_rate": 0.00018036567309648145,
      "loss": 3.2868,
      "step": 33480
    },
    {
      "epoch": 0.4389598133536058,
      "grad_norm": 0.6670248508453369,
      "learning_rate": 0.00018030478414620736,
      "loss": 3.3194,
      "step": 33490
    },
    {
      "epoch": 0.43909088525965345,
      "grad_norm": 0.5387709736824036,
      "learning_rate": 0.00018024388998948834,
      "loss": 3.2624,
      "step": 33500
    },
    {
      "epoch": 0.43909088525965345,
      "eval_accuracy": 0.39263831867057675,
      "eval_loss": 3.2847862243652344,
      "eval_runtime": 60.0888,
      "eval_samples_per_second": 33.284,
      "eval_steps_per_second": 0.533,
      "step": 33500
    },
    {
      "epoch": 0.4392219571657011,
      "grad_norm": 0.5845139026641846,
      "learning_rate": 0.00018018299063678615,
      "loss": 3.3147,
      "step": 33510
    },
    {
      "epoch": 0.4393530290717488,
      "grad_norm": 0.5052981376647949,
      "learning_rate": 0.00018012208609856353,
      "loss": 3.2776,
      "step": 33520
    },
    {
      "epoch": 0.43948410097779644,
      "grad_norm": 0.6061844229698181,
      "learning_rate": 0.00018006117638528392,
      "loss": 3.2836,
      "step": 33530
    },
    {
      "epoch": 0.43961517288384405,
      "grad_norm": 0.6236282587051392,
      "learning_rate": 0.00018000026150741193,
      "loss": 3.2882,
      "step": 33540
    },
    {
      "epoch": 0.4397462447898917,
      "grad_norm": 0.5939663648605347,
      "learning_rate": 0.00017993934147541278,
      "loss": 3.2507,
      "step": 33550
    },
    {
      "epoch": 0.4398773166959394,
      "grad_norm": 0.5833957195281982,
      "learning_rate": 0.00017987841629975277,
      "loss": 3.3114,
      "step": 33560
    },
    {
      "epoch": 0.44000838860198704,
      "grad_norm": 0.5389593243598938,
      "learning_rate": 0.00017981748599089903,
      "loss": 3.2806,
      "step": 33570
    },
    {
      "epoch": 0.4401394605080347,
      "grad_norm": 0.5562366843223572,
      "learning_rate": 0.00017975655055931948,
      "loss": 3.3802,
      "step": 33580
    },
    {
      "epoch": 0.44027053241408237,
      "grad_norm": 0.5952459573745728,
      "learning_rate": 0.00017969561001548307,
      "loss": 3.3056,
      "step": 33590
    },
    {
      "epoch": 0.44040160432013004,
      "grad_norm": 0.6365894675254822,
      "learning_rate": 0.00017963466436985944,
      "loss": 3.2876,
      "step": 33600
    },
    {
      "epoch": 0.4405326762261777,
      "grad_norm": 0.5639546513557434,
      "learning_rate": 0.00017957371363291934,
      "loss": 3.3259,
      "step": 33610
    },
    {
      "epoch": 0.44066374813222536,
      "grad_norm": 0.5858677625656128,
      "learning_rate": 0.00017951275781513418,
      "loss": 3.2485,
      "step": 33620
    },
    {
      "epoch": 0.44079482003827297,
      "grad_norm": 0.6949943900108337,
      "learning_rate": 0.0001794517969269764,
      "loss": 3.3153,
      "step": 33630
    },
    {
      "epoch": 0.44092589194432064,
      "grad_norm": 0.5786319375038147,
      "learning_rate": 0.00017939083097891922,
      "loss": 3.3117,
      "step": 33640
    },
    {
      "epoch": 0.4410569638503683,
      "grad_norm": 0.5830353498458862,
      "learning_rate": 0.0001793298599814368,
      "loss": 3.2876,
      "step": 33650
    },
    {
      "epoch": 0.44118803575641596,
      "grad_norm": 0.6753154397010803,
      "learning_rate": 0.0001792688839450041,
      "loss": 3.3372,
      "step": 33660
    },
    {
      "epoch": 0.44131910766246363,
      "grad_norm": 0.5761138200759888,
      "learning_rate": 0.00017920790288009693,
      "loss": 3.3485,
      "step": 33670
    },
    {
      "epoch": 0.4414501795685113,
      "grad_norm": 0.5418816804885864,
      "learning_rate": 0.00017914691679719205,
      "loss": 3.2617,
      "step": 33680
    },
    {
      "epoch": 0.44158125147455896,
      "grad_norm": 0.5578289031982422,
      "learning_rate": 0.000179085925706767,
      "loss": 3.2717,
      "step": 33690
    },
    {
      "epoch": 0.4417123233806066,
      "grad_norm": 0.5879059433937073,
      "learning_rate": 0.00017902492961930022,
      "loss": 3.3188,
      "step": 33700
    },
    {
      "epoch": 0.4418433952866543,
      "grad_norm": 0.6109389066696167,
      "learning_rate": 0.00017896392854527103,
      "loss": 3.3107,
      "step": 33710
    },
    {
      "epoch": 0.4419744671927019,
      "grad_norm": 0.5764577388763428,
      "learning_rate": 0.00017890292249515957,
      "loss": 3.3185,
      "step": 33720
    },
    {
      "epoch": 0.44210553909874956,
      "grad_norm": 0.6383343935012817,
      "learning_rate": 0.00017884191147944685,
      "loss": 3.2884,
      "step": 33730
    },
    {
      "epoch": 0.4422366110047972,
      "grad_norm": 0.6286059617996216,
      "learning_rate": 0.00017878089550861474,
      "loss": 3.3026,
      "step": 33740
    },
    {
      "epoch": 0.4423676829108449,
      "grad_norm": 0.6028517484664917,
      "learning_rate": 0.00017871987459314596,
      "loss": 3.32,
      "step": 33750
    },
    {
      "epoch": 0.44249875481689255,
      "grad_norm": 0.5556233525276184,
      "learning_rate": 0.00017865884874352405,
      "loss": 3.2799,
      "step": 33760
    },
    {
      "epoch": 0.4426298267229402,
      "grad_norm": 0.6368274688720703,
      "learning_rate": 0.00017859781797023336,
      "loss": 3.3545,
      "step": 33770
    },
    {
      "epoch": 0.4427608986289879,
      "grad_norm": 0.5965280532836914,
      "learning_rate": 0.00017853678228375925,
      "loss": 3.2857,
      "step": 33780
    },
    {
      "epoch": 0.44289197053503554,
      "grad_norm": 0.5892267227172852,
      "learning_rate": 0.00017847574169458773,
      "loss": 3.3482,
      "step": 33790
    },
    {
      "epoch": 0.4430230424410832,
      "grad_norm": 0.5952114462852478,
      "learning_rate": 0.00017841469621320582,
      "loss": 3.2686,
      "step": 33800
    },
    {
      "epoch": 0.4431541143471308,
      "grad_norm": 0.5792097449302673,
      "learning_rate": 0.00017835364585010128,
      "loss": 3.31,
      "step": 33810
    },
    {
      "epoch": 0.4432851862531785,
      "grad_norm": 0.5810233950614929,
      "learning_rate": 0.00017829259061576263,
      "loss": 3.2663,
      "step": 33820
    },
    {
      "epoch": 0.44341625815922614,
      "grad_norm": 0.6375892758369446,
      "learning_rate": 0.0001782315305206795,
      "loss": 3.3031,
      "step": 33830
    },
    {
      "epoch": 0.4435473300652738,
      "grad_norm": 0.5473793745040894,
      "learning_rate": 0.00017817046557534207,
      "loss": 3.2893,
      "step": 33840
    },
    {
      "epoch": 0.44367840197132147,
      "grad_norm": 0.6020014882087708,
      "learning_rate": 0.00017810939579024145,
      "loss": 3.3196,
      "step": 33850
    },
    {
      "epoch": 0.44380947387736913,
      "grad_norm": 0.5692124366760254,
      "learning_rate": 0.0001780483211758696,
      "loss": 3.3079,
      "step": 33860
    },
    {
      "epoch": 0.4439405457834168,
      "grad_norm": 0.5640891790390015,
      "learning_rate": 0.00017798724174271943,
      "loss": 3.2843,
      "step": 33870
    },
    {
      "epoch": 0.44407161768946446,
      "grad_norm": 0.6435897946357727,
      "learning_rate": 0.00017792615750128436,
      "loss": 3.3417,
      "step": 33880
    },
    {
      "epoch": 0.4442026895955121,
      "grad_norm": 0.6106324791908264,
      "learning_rate": 0.000177865068462059,
      "loss": 3.2521,
      "step": 33890
    },
    {
      "epoch": 0.44433376150155973,
      "grad_norm": 0.5678613185882568,
      "learning_rate": 0.0001778039746355385,
      "loss": 3.2988,
      "step": 33900
    },
    {
      "epoch": 0.4444648334076074,
      "grad_norm": 0.5533297061920166,
      "learning_rate": 0.00017774287603221904,
      "loss": 3.3596,
      "step": 33910
    },
    {
      "epoch": 0.44459590531365506,
      "grad_norm": 0.5666854977607727,
      "learning_rate": 0.00017768177266259749,
      "loss": 3.3268,
      "step": 33920
    },
    {
      "epoch": 0.4447269772197027,
      "grad_norm": 0.6100501418113708,
      "learning_rate": 0.0001776206645371715,
      "loss": 3.2569,
      "step": 33930
    },
    {
      "epoch": 0.4448580491257504,
      "grad_norm": 0.5982544422149658,
      "learning_rate": 0.00017755955166643973,
      "loss": 3.2986,
      "step": 33940
    },
    {
      "epoch": 0.44498912103179805,
      "grad_norm": 0.5499681234359741,
      "learning_rate": 0.00017749843406090145,
      "loss": 3.2762,
      "step": 33950
    },
    {
      "epoch": 0.4451201929378457,
      "grad_norm": 0.5525364875793457,
      "learning_rate": 0.00017743731173105692,
      "loss": 3.2734,
      "step": 33960
    },
    {
      "epoch": 0.4452512648438934,
      "grad_norm": 0.8356642127037048,
      "learning_rate": 0.000177376184687407,
      "loss": 3.4052,
      "step": 33970
    },
    {
      "epoch": 0.44538233674994104,
      "grad_norm": 0.6260643601417542,
      "learning_rate": 0.0001773150529404536,
      "loss": 3.3039,
      "step": 33980
    },
    {
      "epoch": 0.44551340865598865,
      "grad_norm": 0.6158179044723511,
      "learning_rate": 0.00017725391650069927,
      "loss": 3.3066,
      "step": 33990
    },
    {
      "epoch": 0.4456444805620363,
      "grad_norm": 0.577350914478302,
      "learning_rate": 0.00017719277537864742,
      "loss": 3.2626,
      "step": 34000
    },
    {
      "epoch": 0.4456444805620363,
      "eval_accuracy": 0.39269061583577713,
      "eval_loss": 3.283491849899292,
      "eval_runtime": 58.8128,
      "eval_samples_per_second": 34.006,
      "eval_steps_per_second": 0.544,
      "step": 34000
    },
    {
      "epoch": 0.445775552468084,
      "grad_norm": 0.6248023509979248,
      "learning_rate": 0.00017713162958480228,
      "loss": 3.3096,
      "step": 34010
    },
    {
      "epoch": 0.44590662437413164,
      "grad_norm": 0.5926978588104248,
      "learning_rate": 0.00017707047912966887,
      "loss": 3.34,
      "step": 34020
    },
    {
      "epoch": 0.4460376962801793,
      "grad_norm": 0.5889167189598083,
      "learning_rate": 0.00017700932402375295,
      "loss": 3.2687,
      "step": 34030
    },
    {
      "epoch": 0.44616876818622697,
      "grad_norm": 0.5497048497200012,
      "learning_rate": 0.00017694816427756121,
      "loss": 3.2481,
      "step": 34040
    },
    {
      "epoch": 0.44629984009227464,
      "grad_norm": 0.621243417263031,
      "learning_rate": 0.00017688699990160098,
      "loss": 3.3535,
      "step": 34050
    },
    {
      "epoch": 0.4464309119983223,
      "grad_norm": 0.6320370435714722,
      "learning_rate": 0.00017682583090638057,
      "loss": 3.2789,
      "step": 34060
    },
    {
      "epoch": 0.44656198390436996,
      "grad_norm": 0.5948027968406677,
      "learning_rate": 0.0001767646573024089,
      "loss": 3.2347,
      "step": 34070
    },
    {
      "epoch": 0.44669305581041757,
      "grad_norm": 0.6302648782730103,
      "learning_rate": 0.00017670347910019577,
      "loss": 3.2731,
      "step": 34080
    },
    {
      "epoch": 0.44682412771646524,
      "grad_norm": 0.5895984172821045,
      "learning_rate": 0.0001766422963102518,
      "loss": 3.3147,
      "step": 34090
    },
    {
      "epoch": 0.4469551996225129,
      "grad_norm": 0.5796096920967102,
      "learning_rate": 0.00017658110894308834,
      "loss": 3.2768,
      "step": 34100
    },
    {
      "epoch": 0.44708627152856056,
      "grad_norm": 0.5572205781936646,
      "learning_rate": 0.00017651991700921754,
      "loss": 3.2712,
      "step": 34110
    },
    {
      "epoch": 0.4472173434346082,
      "grad_norm": 0.5759187936782837,
      "learning_rate": 0.00017645872051915234,
      "loss": 3.2792,
      "step": 34120
    },
    {
      "epoch": 0.4473484153406559,
      "grad_norm": 0.5704599022865295,
      "learning_rate": 0.00017639751948340647,
      "loss": 3.3004,
      "step": 34130
    },
    {
      "epoch": 0.44747948724670356,
      "grad_norm": 0.61073237657547,
      "learning_rate": 0.00017633631391249444,
      "loss": 3.3149,
      "step": 34140
    },
    {
      "epoch": 0.4476105591527512,
      "grad_norm": 0.5733668208122253,
      "learning_rate": 0.00017627510381693158,
      "loss": 3.2333,
      "step": 34150
    },
    {
      "epoch": 0.4477416310587989,
      "grad_norm": 0.5965573787689209,
      "learning_rate": 0.00017621388920723385,
      "loss": 3.2956,
      "step": 34160
    },
    {
      "epoch": 0.4478727029648465,
      "grad_norm": 0.5670505166053772,
      "learning_rate": 0.00017615267009391816,
      "loss": 3.2683,
      "step": 34170
    },
    {
      "epoch": 0.44800377487089416,
      "grad_norm": 0.5360655188560486,
      "learning_rate": 0.00017609144648750213,
      "loss": 3.2941,
      "step": 34180
    },
    {
      "epoch": 0.4481348467769418,
      "grad_norm": 0.5907930135726929,
      "learning_rate": 0.0001760302183985041,
      "loss": 3.2526,
      "step": 34190
    },
    {
      "epoch": 0.4482659186829895,
      "grad_norm": 0.6475618481636047,
      "learning_rate": 0.00017596898583744326,
      "loss": 3.2726,
      "step": 34200
    },
    {
      "epoch": 0.44839699058903715,
      "grad_norm": 0.5865305662155151,
      "learning_rate": 0.0001759077488148395,
      "loss": 3.2744,
      "step": 34210
    },
    {
      "epoch": 0.4485280624950848,
      "grad_norm": 0.6078808903694153,
      "learning_rate": 0.0001758465073412135,
      "loss": 3.3871,
      "step": 34220
    },
    {
      "epoch": 0.4486591344011325,
      "grad_norm": 0.6118040084838867,
      "learning_rate": 0.00017578526142708676,
      "loss": 3.2817,
      "step": 34230
    },
    {
      "epoch": 0.44879020630718014,
      "grad_norm": 0.546944260597229,
      "learning_rate": 0.00017572401108298147,
      "loss": 3.316,
      "step": 34240
    },
    {
      "epoch": 0.4489212782132278,
      "grad_norm": 0.5529544949531555,
      "learning_rate": 0.0001756627563194206,
      "loss": 3.2993,
      "step": 34250
    },
    {
      "epoch": 0.4490523501192754,
      "grad_norm": 0.5408505797386169,
      "learning_rate": 0.0001756014971469279,
      "loss": 3.3075,
      "step": 34260
    },
    {
      "epoch": 0.4491834220253231,
      "grad_norm": 0.6279626488685608,
      "learning_rate": 0.0001755402335760278,
      "loss": 3.3228,
      "step": 34270
    },
    {
      "epoch": 0.44931449393137074,
      "grad_norm": 0.6133425831794739,
      "learning_rate": 0.00017547896561724568,
      "loss": 3.3633,
      "step": 34280
    },
    {
      "epoch": 0.4494455658374184,
      "grad_norm": 0.5779728293418884,
      "learning_rate": 0.0001754176932811075,
      "loss": 3.2897,
      "step": 34290
    },
    {
      "epoch": 0.44957663774346607,
      "grad_norm": 0.5970420241355896,
      "learning_rate": 0.00017535641657813995,
      "loss": 3.2983,
      "step": 34300
    },
    {
      "epoch": 0.44970770964951373,
      "grad_norm": 0.5516222715377808,
      "learning_rate": 0.00017529513551887055,
      "loss": 3.2603,
      "step": 34310
    },
    {
      "epoch": 0.4498387815555614,
      "grad_norm": 0.545790433883667,
      "learning_rate": 0.00017523385011382757,
      "loss": 3.247,
      "step": 34320
    },
    {
      "epoch": 0.44996985346160906,
      "grad_norm": 0.5734447240829468,
      "learning_rate": 0.00017517256037354002,
      "loss": 3.2964,
      "step": 34330
    },
    {
      "epoch": 0.4501009253676567,
      "grad_norm": 0.9406523108482361,
      "learning_rate": 0.00017511126630853765,
      "loss": 3.2782,
      "step": 34340
    },
    {
      "epoch": 0.45023199727370433,
      "grad_norm": 0.5874607563018799,
      "learning_rate": 0.00017504996792935097,
      "loss": 3.2805,
      "step": 34350
    },
    {
      "epoch": 0.450363069179752,
      "grad_norm": 0.6867660284042358,
      "learning_rate": 0.0001749886652465111,
      "loss": 3.4098,
      "step": 34360
    },
    {
      "epoch": 0.45049414108579966,
      "grad_norm": 0.5787460207939148,
      "learning_rate": 0.00017492735827055015,
      "loss": 3.24,
      "step": 34370
    },
    {
      "epoch": 0.4506252129918473,
      "grad_norm": 0.6159490346908569,
      "learning_rate": 0.00017486604701200074,
      "loss": 3.3161,
      "step": 34380
    },
    {
      "epoch": 0.450756284897895,
      "grad_norm": 0.5662608742713928,
      "learning_rate": 0.00017480473148139634,
      "loss": 3.2647,
      "step": 34390
    },
    {
      "epoch": 0.45088735680394265,
      "grad_norm": 0.7820384502410889,
      "learning_rate": 0.00017474341168927112,
      "loss": 3.2954,
      "step": 34400
    },
    {
      "epoch": 0.4510184287099903,
      "grad_norm": 0.5987968444824219,
      "learning_rate": 0.00017468208764616,
      "loss": 3.2913,
      "step": 34410
    },
    {
      "epoch": 0.451149500616038,
      "grad_norm": 0.555509626865387,
      "learning_rate": 0.00017462075936259858,
      "loss": 3.2393,
      "step": 34420
    },
    {
      "epoch": 0.45128057252208564,
      "grad_norm": 0.6175597906112671,
      "learning_rate": 0.0001745594268491233,
      "loss": 3.2743,
      "step": 34430
    },
    {
      "epoch": 0.45141164442813325,
      "grad_norm": 0.6339783072471619,
      "learning_rate": 0.00017449809011627117,
      "loss": 3.3066,
      "step": 34440
    },
    {
      "epoch": 0.4515427163341809,
      "grad_norm": 0.623265266418457,
      "learning_rate": 0.00017443674917458005,
      "loss": 3.3144,
      "step": 34450
    },
    {
      "epoch": 0.4516737882402286,
      "grad_norm": 0.5839799642562866,
      "learning_rate": 0.0001743754040345885,
      "loss": 3.2866,
      "step": 34460
    },
    {
      "epoch": 0.45180486014627624,
      "grad_norm": 0.7217001914978027,
      "learning_rate": 0.00017431405470683575,
      "loss": 3.3495,
      "step": 34470
    },
    {
      "epoch": 0.4519359320523239,
      "grad_norm": 0.5603193640708923,
      "learning_rate": 0.00017425270120186183,
      "loss": 3.3136,
      "step": 34480
    },
    {
      "epoch": 0.45206700395837157,
      "grad_norm": 0.5381850004196167,
      "learning_rate": 0.00017419134353020735,
      "loss": 3.317,
      "step": 34490
    },
    {
      "epoch": 0.45219807586441924,
      "grad_norm": 0.6148549318313599,
      "learning_rate": 0.0001741299817024138,
      "loss": 3.2651,
      "step": 34500
    },
    {
      "epoch": 0.45219807586441924,
      "eval_accuracy": 0.3933709677419355,
      "eval_loss": 3.2792134284973145,
      "eval_runtime": 56.719,
      "eval_samples_per_second": 35.262,
      "eval_steps_per_second": 0.564,
      "step": 34500
    },
    {
      "epoch": 0.4523291477704669,
      "grad_norm": 0.5518143177032471,
      "learning_rate": 0.00017406861572902328,
      "loss": 3.2527,
      "step": 34510
    },
    {
      "epoch": 0.45246021967651456,
      "grad_norm": 0.5118018984794617,
      "learning_rate": 0.0001740072456205787,
      "loss": 3.267,
      "step": 34520
    },
    {
      "epoch": 0.45259129158256217,
      "grad_norm": 0.5797098875045776,
      "learning_rate": 0.0001739458713876235,
      "loss": 3.2691,
      "step": 34530
    },
    {
      "epoch": 0.45272236348860984,
      "grad_norm": 0.5559801459312439,
      "learning_rate": 0.00017388449304070203,
      "loss": 3.3207,
      "step": 34540
    },
    {
      "epoch": 0.4528534353946575,
      "grad_norm": 0.5868993401527405,
      "learning_rate": 0.00017382311059035925,
      "loss": 3.3075,
      "step": 34550
    },
    {
      "epoch": 0.45298450730070516,
      "grad_norm": 0.5937730669975281,
      "learning_rate": 0.00017376172404714077,
      "loss": 3.2838,
      "step": 34560
    },
    {
      "epoch": 0.4531155792067528,
      "grad_norm": 0.5810648202896118,
      "learning_rate": 0.00017370033342159304,
      "loss": 3.3194,
      "step": 34570
    },
    {
      "epoch": 0.4532466511128005,
      "grad_norm": 0.5633224844932556,
      "learning_rate": 0.00017363893872426306,
      "loss": 3.327,
      "step": 34580
    },
    {
      "epoch": 0.45337772301884816,
      "grad_norm": 0.5773929953575134,
      "learning_rate": 0.00017357753996569865,
      "loss": 3.2394,
      "step": 34590
    },
    {
      "epoch": 0.4535087949248958,
      "grad_norm": 0.6598352193832397,
      "learning_rate": 0.0001735161371564483,
      "loss": 3.295,
      "step": 34600
    },
    {
      "epoch": 0.45363986683094343,
      "grad_norm": 0.594800591468811,
      "learning_rate": 0.0001734547303070612,
      "loss": 3.3148,
      "step": 34610
    },
    {
      "epoch": 0.4537709387369911,
      "grad_norm": 0.7706505060195923,
      "learning_rate": 0.0001733933194280871,
      "loss": 3.3313,
      "step": 34620
    },
    {
      "epoch": 0.45390201064303876,
      "grad_norm": 0.538834273815155,
      "learning_rate": 0.00017333190453007672,
      "loss": 3.2706,
      "step": 34630
    },
    {
      "epoch": 0.4540330825490864,
      "grad_norm": 0.570489227771759,
      "learning_rate": 0.0001732704856235812,
      "loss": 3.2654,
      "step": 34640
    },
    {
      "epoch": 0.4541641544551341,
      "grad_norm": 0.5883184671401978,
      "learning_rate": 0.0001732090627191525,
      "loss": 3.2994,
      "step": 34650
    },
    {
      "epoch": 0.45429522636118175,
      "grad_norm": 0.5607204437255859,
      "learning_rate": 0.0001731476358273432,
      "loss": 3.2892,
      "step": 34660
    },
    {
      "epoch": 0.4544262982672294,
      "grad_norm": 0.5869757533073425,
      "learning_rate": 0.00017308620495870675,
      "loss": 3.2893,
      "step": 34670
    },
    {
      "epoch": 0.4545573701732771,
      "grad_norm": 0.5915288329124451,
      "learning_rate": 0.00017302477012379698,
      "loss": 3.2454,
      "step": 34680
    },
    {
      "epoch": 0.45468844207932474,
      "grad_norm": 0.5763856768608093,
      "learning_rate": 0.00017296333133316863,
      "loss": 3.2575,
      "step": 34690
    },
    {
      "epoch": 0.45481951398537235,
      "grad_norm": 0.5940616130828857,
      "learning_rate": 0.0001729018885973771,
      "loss": 3.3468,
      "step": 34700
    },
    {
      "epoch": 0.45495058589142,
      "grad_norm": 0.6464433670043945,
      "learning_rate": 0.0001728404419269783,
      "loss": 3.3046,
      "step": 34710
    },
    {
      "epoch": 0.4550816577974677,
      "grad_norm": 0.5921247005462646,
      "learning_rate": 0.00017277899133252906,
      "loss": 3.294,
      "step": 34720
    },
    {
      "epoch": 0.45521272970351534,
      "grad_norm": 0.6245022416114807,
      "learning_rate": 0.0001727175368245867,
      "loss": 3.2847,
      "step": 34730
    },
    {
      "epoch": 0.455343801609563,
      "grad_norm": 0.5419604778289795,
      "learning_rate": 0.0001726560784137093,
      "loss": 3.2836,
      "step": 34740
    },
    {
      "epoch": 0.45547487351561067,
      "grad_norm": 0.5920389890670776,
      "learning_rate": 0.0001725946161104555,
      "loss": 3.2667,
      "step": 34750
    },
    {
      "epoch": 0.45560594542165833,
      "grad_norm": 0.5494979023933411,
      "learning_rate": 0.00017253314992538482,
      "loss": 3.311,
      "step": 34760
    },
    {
      "epoch": 0.455737017327706,
      "grad_norm": 0.6118741035461426,
      "learning_rate": 0.00017247167986905723,
      "loss": 3.3272,
      "step": 34770
    },
    {
      "epoch": 0.45586808923375366,
      "grad_norm": 0.6771003007888794,
      "learning_rate": 0.0001724102059520335,
      "loss": 3.298,
      "step": 34780
    },
    {
      "epoch": 0.45599916113980127,
      "grad_norm": 0.597867488861084,
      "learning_rate": 0.00017234872818487493,
      "loss": 3.3097,
      "step": 34790
    },
    {
      "epoch": 0.45613023304584893,
      "grad_norm": 0.5966109037399292,
      "learning_rate": 0.0001722872465781437,
      "loss": 3.2759,
      "step": 34800
    },
    {
      "epoch": 0.4562613049518966,
      "grad_norm": 0.590199887752533,
      "learning_rate": 0.00017222576114240247,
      "loss": 3.2048,
      "step": 34810
    },
    {
      "epoch": 0.45639237685794426,
      "grad_norm": 0.6887836456298828,
      "learning_rate": 0.00017216427188821457,
      "loss": 3.3015,
      "step": 34820
    },
    {
      "epoch": 0.4565234487639919,
      "grad_norm": 0.5979238152503967,
      "learning_rate": 0.00017210277882614403,
      "loss": 3.3043,
      "step": 34830
    },
    {
      "epoch": 0.4566545206700396,
      "grad_norm": 0.6802168488502502,
      "learning_rate": 0.00017204128196675553,
      "loss": 3.2913,
      "step": 34840
    },
    {
      "epoch": 0.45678559257608725,
      "grad_norm": 1.1250150203704834,
      "learning_rate": 0.00017197978132061444,
      "loss": 3.4005,
      "step": 34850
    },
    {
      "epoch": 0.4569166644821349,
      "grad_norm": 0.6473788022994995,
      "learning_rate": 0.00017191827689828665,
      "loss": 3.3209,
      "step": 34860
    },
    {
      "epoch": 0.4570477363881826,
      "grad_norm": 0.571444571018219,
      "learning_rate": 0.0001718567687103389,
      "loss": 3.324,
      "step": 34870
    },
    {
      "epoch": 0.4571788082942302,
      "grad_norm": 0.5877186059951782,
      "learning_rate": 0.00017179525676733838,
      "loss": 3.2621,
      "step": 34880
    },
    {
      "epoch": 0.45730988020027785,
      "grad_norm": 0.7272287607192993,
      "learning_rate": 0.00017173374107985306,
      "loss": 3.3222,
      "step": 34890
    },
    {
      "epoch": 0.4574409521063255,
      "grad_norm": 0.7038570642471313,
      "learning_rate": 0.00017167222165845146,
      "loss": 3.3316,
      "step": 34900
    },
    {
      "epoch": 0.4575720240123732,
      "grad_norm": 0.5635845065116882,
      "learning_rate": 0.00017161069851370284,
      "loss": 3.3061,
      "step": 34910
    },
    {
      "epoch": 0.45770309591842084,
      "grad_norm": 0.5489809513092041,
      "learning_rate": 0.00017154917165617705,
      "loss": 3.2824,
      "step": 34920
    },
    {
      "epoch": 0.4578341678244685,
      "grad_norm": 0.548851490020752,
      "learning_rate": 0.0001714876410964445,
      "loss": 3.2659,
      "step": 34930
    },
    {
      "epoch": 0.45796523973051617,
      "grad_norm": 0.6080682277679443,
      "learning_rate": 0.00017142610684507635,
      "loss": 3.4374,
      "step": 34940
    },
    {
      "epoch": 0.45809631163656384,
      "grad_norm": 0.5756624937057495,
      "learning_rate": 0.0001713645689126443,
      "loss": 3.2839,
      "step": 34950
    },
    {
      "epoch": 0.4582273835426115,
      "grad_norm": 0.5970150828361511,
      "learning_rate": 0.00017130302730972087,
      "loss": 3.2896,
      "step": 34960
    },
    {
      "epoch": 0.4583584554486591,
      "grad_norm": 0.7775086164474487,
      "learning_rate": 0.00017124148204687893,
      "loss": 3.2881,
      "step": 34970
    },
    {
      "epoch": 0.45848952735470677,
      "grad_norm": 0.6042128801345825,
      "learning_rate": 0.0001711799331346922,
      "loss": 3.2958,
      "step": 34980
    },
    {
      "epoch": 0.45862059926075444,
      "grad_norm": 0.5592438578605652,
      "learning_rate": 0.00017111838058373493,
      "loss": 3.329,
      "step": 34990
    },
    {
      "epoch": 0.4587516711668021,
      "grad_norm": 0.6234604716300964,
      "learning_rate": 0.00017105682440458203,
      "loss": 3.2793,
      "step": 35000
    },
    {
      "epoch": 0.4587516711668021,
      "eval_accuracy": 0.39356793743890517,
      "eval_loss": 3.2791996002197266,
      "eval_runtime": 59.0222,
      "eval_samples_per_second": 33.886,
      "eval_steps_per_second": 0.542,
      "step": 35000
    },
    {
      "epoch": 0.45888274307284976,
      "grad_norm": 0.5867906808853149,
      "learning_rate": 0.00017099526460780904,
      "loss": 3.2357,
      "step": 35010
    },
    {
      "epoch": 0.4590138149788974,
      "grad_norm": 0.5534915328025818,
      "learning_rate": 0.00017093370120399207,
      "loss": 3.2603,
      "step": 35020
    },
    {
      "epoch": 0.4591448868849451,
      "grad_norm": 0.6653027534484863,
      "learning_rate": 0.00017087213420370785,
      "loss": 3.2296,
      "step": 35030
    },
    {
      "epoch": 0.45927595879099276,
      "grad_norm": 0.570311427116394,
      "learning_rate": 0.00017081056361753383,
      "loss": 3.2621,
      "step": 35040
    },
    {
      "epoch": 0.4594070306970404,
      "grad_norm": 0.6255740523338318,
      "learning_rate": 0.00017074898945604794,
      "loss": 3.2635,
      "step": 35050
    },
    {
      "epoch": 0.45953810260308803,
      "grad_norm": 0.5617986917495728,
      "learning_rate": 0.00017068741172982885,
      "loss": 3.3353,
      "step": 35060
    },
    {
      "epoch": 0.4596691745091357,
      "grad_norm": 0.5913467407226562,
      "learning_rate": 0.00017062583044945577,
      "loss": 3.2812,
      "step": 35070
    },
    {
      "epoch": 0.45980024641518336,
      "grad_norm": 0.6337908506393433,
      "learning_rate": 0.0001705642456255085,
      "loss": 3.3025,
      "step": 35080
    },
    {
      "epoch": 0.459931318321231,
      "grad_norm": 0.7273381948471069,
      "learning_rate": 0.00017050265726856748,
      "loss": 3.3615,
      "step": 35090
    },
    {
      "epoch": 0.4600623902272787,
      "grad_norm": 0.5679564476013184,
      "learning_rate": 0.00017044106538921378,
      "loss": 3.3398,
      "step": 35100
    },
    {
      "epoch": 0.46019346213332635,
      "grad_norm": 0.857714831829071,
      "learning_rate": 0.0001703794699980291,
      "loss": 3.3262,
      "step": 35110
    },
    {
      "epoch": 0.460324534039374,
      "grad_norm": 0.6388487219810486,
      "learning_rate": 0.00017031787110559558,
      "loss": 3.2957,
      "step": 35120
    },
    {
      "epoch": 0.4604556059454217,
      "grad_norm": 0.6340341567993164,
      "learning_rate": 0.0001702562687224962,
      "loss": 3.2366,
      "step": 35130
    },
    {
      "epoch": 0.46058667785146934,
      "grad_norm": 0.6222112774848938,
      "learning_rate": 0.0001701946628593143,
      "loss": 3.3065,
      "step": 35140
    },
    {
      "epoch": 0.46071774975751695,
      "grad_norm": 0.5499728322029114,
      "learning_rate": 0.00017013305352663404,
      "loss": 3.2717,
      "step": 35150
    },
    {
      "epoch": 0.4608488216635646,
      "grad_norm": 0.5858969688415527,
      "learning_rate": 0.00017007144073504,
      "loss": 3.276,
      "step": 35160
    },
    {
      "epoch": 0.4609798935696123,
      "grad_norm": 0.7647249698638916,
      "learning_rate": 0.0001700098244951175,
      "loss": 3.332,
      "step": 35170
    },
    {
      "epoch": 0.46111096547565994,
      "grad_norm": 0.5804845690727234,
      "learning_rate": 0.00016994820481745234,
      "loss": 3.2905,
      "step": 35180
    },
    {
      "epoch": 0.4612420373817076,
      "grad_norm": 0.6168461441993713,
      "learning_rate": 0.00016988658171263092,
      "loss": 3.2846,
      "step": 35190
    },
    {
      "epoch": 0.46137310928775527,
      "grad_norm": 0.5920943021774292,
      "learning_rate": 0.00016982495519124032,
      "loss": 3.2842,
      "step": 35200
    },
    {
      "epoch": 0.46150418119380293,
      "grad_norm": 0.5898873805999756,
      "learning_rate": 0.00016976332526386808,
      "loss": 3.2608,
      "step": 35210
    },
    {
      "epoch": 0.4616352530998506,
      "grad_norm": 0.6638315916061401,
      "learning_rate": 0.0001697016919411024,
      "loss": 3.2797,
      "step": 35220
    },
    {
      "epoch": 0.46176632500589826,
      "grad_norm": 0.6184077262878418,
      "learning_rate": 0.0001696400552335321,
      "loss": 3.2654,
      "step": 35230
    },
    {
      "epoch": 0.46189739691194587,
      "grad_norm": 0.5687569379806519,
      "learning_rate": 0.00016957841515174648,
      "loss": 3.3155,
      "step": 35240
    },
    {
      "epoch": 0.46202846881799353,
      "grad_norm": 0.6163052916526794,
      "learning_rate": 0.00016951677170633553,
      "loss": 3.2924,
      "step": 35250
    },
    {
      "epoch": 0.4621595407240412,
      "grad_norm": 0.6100754141807556,
      "learning_rate": 0.00016945512490788975,
      "loss": 3.2973,
      "step": 35260
    },
    {
      "epoch": 0.46229061263008886,
      "grad_norm": 0.5634702444076538,
      "learning_rate": 0.0001693934747670002,
      "loss": 3.2942,
      "step": 35270
    },
    {
      "epoch": 0.4624216845361365,
      "grad_norm": 0.5529267191886902,
      "learning_rate": 0.0001693318212942586,
      "loss": 3.3038,
      "step": 35280
    },
    {
      "epoch": 0.4625527564421842,
      "grad_norm": 0.5566316843032837,
      "learning_rate": 0.00016927016450025707,
      "loss": 3.313,
      "step": 35290
    },
    {
      "epoch": 0.46268382834823185,
      "grad_norm": 0.684342086315155,
      "learning_rate": 0.0001692085043955885,
      "loss": 3.4263,
      "step": 35300
    },
    {
      "epoch": 0.4628149002542795,
      "grad_norm": 0.6093993782997131,
      "learning_rate": 0.00016914684099084628,
      "loss": 3.355,
      "step": 35310
    },
    {
      "epoch": 0.4629459721603272,
      "grad_norm": 0.6610050201416016,
      "learning_rate": 0.00016908517429662436,
      "loss": 3.2936,
      "step": 35320
    },
    {
      "epoch": 0.4630770440663748,
      "grad_norm": 0.578275740146637,
      "learning_rate": 0.0001690235043235172,
      "loss": 3.2703,
      "step": 35330
    },
    {
      "epoch": 0.46320811597242245,
      "grad_norm": 0.5239101648330688,
      "learning_rate": 0.00016896183108211981,
      "loss": 3.2949,
      "step": 35340
    },
    {
      "epoch": 0.4633391878784701,
      "grad_norm": 0.5935649275779724,
      "learning_rate": 0.00016890015458302798,
      "loss": 3.2818,
      "step": 35350
    },
    {
      "epoch": 0.4634702597845178,
      "grad_norm": 0.5820968151092529,
      "learning_rate": 0.0001688384748368378,
      "loss": 3.3185,
      "step": 35360
    },
    {
      "epoch": 0.46360133169056544,
      "grad_norm": 0.6952657103538513,
      "learning_rate": 0.00016877679185414604,
      "loss": 3.2772,
      "step": 35370
    },
    {
      "epoch": 0.4637324035966131,
      "grad_norm": 0.6058968305587769,
      "learning_rate": 0.00016871510564554995,
      "loss": 3.271,
      "step": 35380
    },
    {
      "epoch": 0.46386347550266077,
      "grad_norm": 0.6890294551849365,
      "learning_rate": 0.0001686534162216475,
      "loss": 3.3139,
      "step": 35390
    },
    {
      "epoch": 0.46399454740870844,
      "grad_norm": 0.6317315697669983,
      "learning_rate": 0.000168591723593037,
      "loss": 3.244,
      "step": 35400
    },
    {
      "epoch": 0.4641256193147561,
      "grad_norm": 0.5795525908470154,
      "learning_rate": 0.00016853002777031753,
      "loss": 3.3067,
      "step": 35410
    },
    {
      "epoch": 0.4642566912208037,
      "grad_norm": 0.5863848328590393,
      "learning_rate": 0.00016846832876408842,
      "loss": 3.2982,
      "step": 35420
    },
    {
      "epoch": 0.46438776312685137,
      "grad_norm": 0.5767326354980469,
      "learning_rate": 0.00016840662658494993,
      "loss": 3.2624,
      "step": 35430
    },
    {
      "epoch": 0.46451883503289904,
      "grad_norm": 0.5477858185768127,
      "learning_rate": 0.00016834492124350253,
      "loss": 3.268,
      "step": 35440
    },
    {
      "epoch": 0.4646499069389467,
      "grad_norm": 0.6284289360046387,
      "learning_rate": 0.00016828321275034742,
      "loss": 3.2929,
      "step": 35450
    },
    {
      "epoch": 0.46478097884499436,
      "grad_norm": 0.6314969658851624,
      "learning_rate": 0.0001682215011160863,
      "loss": 3.2644,
      "step": 35460
    },
    {
      "epoch": 0.464912050751042,
      "grad_norm": 0.5225383639335632,
      "learning_rate": 0.0001681597863513213,
      "loss": 3.2954,
      "step": 35470
    },
    {
      "epoch": 0.4650431226570897,
      "grad_norm": 0.8411478400230408,
      "learning_rate": 0.00016809806846665532,
      "loss": 3.2741,
      "step": 35480
    },
    {
      "epoch": 0.46517419456313736,
      "grad_norm": 0.6369122862815857,
      "learning_rate": 0.00016803634747269156,
      "loss": 3.291,
      "step": 35490
    },
    {
      "epoch": 0.465305266469185,
      "grad_norm": 0.5984206199645996,
      "learning_rate": 0.00016797462338003391,
      "loss": 3.2771,
      "step": 35500
    },
    {
      "epoch": 0.465305266469185,
      "eval_accuracy": 0.3934652981427175,
      "eval_loss": 3.2765891551971436,
      "eval_runtime": 59.5753,
      "eval_samples_per_second": 33.571,
      "eval_steps_per_second": 0.537,
      "step": 35500
    },
    {
      "epoch": 0.4654363383752326,
      "grad_norm": 0.5694512724876404,
      "learning_rate": 0.0001679128961992867,
      "loss": 3.3241,
      "step": 35510
    },
    {
      "epoch": 0.4655674102812803,
      "grad_norm": 0.7318518161773682,
      "learning_rate": 0.00016785116594105493,
      "loss": 3.3206,
      "step": 35520
    },
    {
      "epoch": 0.46569848218732796,
      "grad_norm": 0.7842200994491577,
      "learning_rate": 0.00016778943261594386,
      "loss": 3.2311,
      "step": 35530
    },
    {
      "epoch": 0.4658295540933756,
      "grad_norm": 0.5857751965522766,
      "learning_rate": 0.00016772769623455962,
      "loss": 3.2718,
      "step": 35540
    },
    {
      "epoch": 0.4659606259994233,
      "grad_norm": 0.5461675524711609,
      "learning_rate": 0.00016766595680750852,
      "loss": 3.2566,
      "step": 35550
    },
    {
      "epoch": 0.46609169790547095,
      "grad_norm": 0.6047492027282715,
      "learning_rate": 0.0001676042143453977,
      "loss": 3.2809,
      "step": 35560
    },
    {
      "epoch": 0.4662227698115186,
      "grad_norm": 0.5468048453330994,
      "learning_rate": 0.0001675424688588346,
      "loss": 3.255,
      "step": 35570
    },
    {
      "epoch": 0.4663538417175663,
      "grad_norm": 0.5782122015953064,
      "learning_rate": 0.0001674807203584273,
      "loss": 3.2758,
      "step": 35580
    },
    {
      "epoch": 0.46648491362361394,
      "grad_norm": 0.5778475403785706,
      "learning_rate": 0.0001674189688547844,
      "loss": 3.2976,
      "step": 35590
    },
    {
      "epoch": 0.46661598552966155,
      "grad_norm": 0.5811735391616821,
      "learning_rate": 0.00016735721435851486,
      "loss": 3.2749,
      "step": 35600
    },
    {
      "epoch": 0.4667470574357092,
      "grad_norm": 0.5468260049819946,
      "learning_rate": 0.0001672954568802284,
      "loss": 3.2784,
      "step": 35610
    },
    {
      "epoch": 0.4668781293417569,
      "grad_norm": 0.5909556746482849,
      "learning_rate": 0.00016723369643053502,
      "loss": 3.2957,
      "step": 35620
    },
    {
      "epoch": 0.46700920124780454,
      "grad_norm": 0.5428264737129211,
      "learning_rate": 0.00016717193302004542,
      "loss": 3.2667,
      "step": 35630
    },
    {
      "epoch": 0.4671402731538522,
      "grad_norm": 0.6294768452644348,
      "learning_rate": 0.0001671101666593707,
      "loss": 3.2807,
      "step": 35640
    },
    {
      "epoch": 0.46727134505989987,
      "grad_norm": 0.5718059539794922,
      "learning_rate": 0.00016704839735912246,
      "loss": 3.271,
      "step": 35650
    },
    {
      "epoch": 0.46740241696594753,
      "grad_norm": 0.5937339663505554,
      "learning_rate": 0.00016698662512991286,
      "loss": 3.2925,
      "step": 35660
    },
    {
      "epoch": 0.4675334888719952,
      "grad_norm": 1.0781869888305664,
      "learning_rate": 0.0001669248499823545,
      "loss": 3.3208,
      "step": 35670
    },
    {
      "epoch": 0.46766456077804286,
      "grad_norm": 0.6467334628105164,
      "learning_rate": 0.00016686307192706056,
      "loss": 3.2295,
      "step": 35680
    },
    {
      "epoch": 0.46779563268409047,
      "grad_norm": 0.6362531781196594,
      "learning_rate": 0.0001668012909746447,
      "loss": 3.2445,
      "step": 35690
    },
    {
      "epoch": 0.46792670459013813,
      "grad_norm": 0.6339471340179443,
      "learning_rate": 0.000166739507135721,
      "loss": 3.3074,
      "step": 35700
    },
    {
      "epoch": 0.4680577764961858,
      "grad_norm": 0.6322892904281616,
      "learning_rate": 0.0001666777204209041,
      "loss": 3.3025,
      "step": 35710
    },
    {
      "epoch": 0.46818884840223346,
      "grad_norm": 0.6393026113510132,
      "learning_rate": 0.0001666159308408092,
      "loss": 3.3078,
      "step": 35720
    },
    {
      "epoch": 0.4683199203082811,
      "grad_norm": 0.6014976501464844,
      "learning_rate": 0.0001665541384060518,
      "loss": 3.3133,
      "step": 35730
    },
    {
      "epoch": 0.4684509922143288,
      "grad_norm": 0.6020879745483398,
      "learning_rate": 0.00016649234312724812,
      "loss": 3.246,
      "step": 35740
    },
    {
      "epoch": 0.46858206412037645,
      "grad_norm": 0.6989703178405762,
      "learning_rate": 0.0001664305450150147,
      "loss": 3.2824,
      "step": 35750
    },
    {
      "epoch": 0.4687131360264241,
      "grad_norm": 0.5873619914054871,
      "learning_rate": 0.00016636874407996864,
      "loss": 3.2462,
      "step": 35760
    },
    {
      "epoch": 0.4688442079324718,
      "grad_norm": 0.5652187466621399,
      "learning_rate": 0.00016630694033272744,
      "loss": 3.2921,
      "step": 35770
    },
    {
      "epoch": 0.4689752798385194,
      "grad_norm": 0.6066574454307556,
      "learning_rate": 0.0001662451337839093,
      "loss": 3.2704,
      "step": 35780
    },
    {
      "epoch": 0.46910635174456705,
      "grad_norm": 0.5861733555793762,
      "learning_rate": 0.00016618332444413264,
      "loss": 3.3018,
      "step": 35790
    },
    {
      "epoch": 0.4692374236506147,
      "grad_norm": 0.7302654981613159,
      "learning_rate": 0.00016612151232401652,
      "loss": 3.3251,
      "step": 35800
    },
    {
      "epoch": 0.4693684955566624,
      "grad_norm": 0.5565829277038574,
      "learning_rate": 0.00016605969743418045,
      "loss": 3.29,
      "step": 35810
    },
    {
      "epoch": 0.46949956746271004,
      "grad_norm": 0.5997703671455383,
      "learning_rate": 0.00016599787978524433,
      "loss": 3.2263,
      "step": 35820
    },
    {
      "epoch": 0.4696306393687577,
      "grad_norm": 0.6300492286682129,
      "learning_rate": 0.00016593605938782866,
      "loss": 3.322,
      "step": 35830
    },
    {
      "epoch": 0.46976171127480537,
      "grad_norm": 0.6263577938079834,
      "learning_rate": 0.0001658742362525543,
      "loss": 3.3093,
      "step": 35840
    },
    {
      "epoch": 0.46989278318085304,
      "grad_norm": 0.5987015962600708,
      "learning_rate": 0.0001658124103900427,
      "loss": 3.3315,
      "step": 35850
    },
    {
      "epoch": 0.4700238550869007,
      "grad_norm": 0.5620373487472534,
      "learning_rate": 0.00016575058181091562,
      "loss": 3.2773,
      "step": 35860
    },
    {
      "epoch": 0.4701549269929483,
      "grad_norm": 0.6587393879890442,
      "learning_rate": 0.0001656887505257955,
      "loss": 3.2972,
      "step": 35870
    },
    {
      "epoch": 0.47028599889899597,
      "grad_norm": 0.6099607944488525,
      "learning_rate": 0.00016562691654530503,
      "loss": 3.2596,
      "step": 35880
    },
    {
      "epoch": 0.47041707080504364,
      "grad_norm": 0.6076106429100037,
      "learning_rate": 0.0001655650798800675,
      "loss": 3.2762,
      "step": 35890
    },
    {
      "epoch": 0.4705481427110913,
      "grad_norm": 0.5391535758972168,
      "learning_rate": 0.0001655032405407066,
      "loss": 3.274,
      "step": 35900
    },
    {
      "epoch": 0.47067921461713896,
      "grad_norm": 0.5848811864852905,
      "learning_rate": 0.00016544139853784655,
      "loss": 3.3094,
      "step": 35910
    },
    {
      "epoch": 0.4708102865231866,
      "grad_norm": 0.610551655292511,
      "learning_rate": 0.00016537955388211186,
      "loss": 3.2944,
      "step": 35920
    },
    {
      "epoch": 0.4709413584292343,
      "grad_norm": 0.6113791465759277,
      "learning_rate": 0.00016531770658412774,
      "loss": 3.2614,
      "step": 35930
    },
    {
      "epoch": 0.47107243033528196,
      "grad_norm": 0.5833249688148499,
      "learning_rate": 0.00016525585665451965,
      "loss": 3.2676,
      "step": 35940
    },
    {
      "epoch": 0.4712035022413296,
      "grad_norm": 0.5750905871391296,
      "learning_rate": 0.00016519400410391357,
      "loss": 3.27,
      "step": 35950
    },
    {
      "epoch": 0.4713345741473772,
      "grad_norm": 0.6123021245002747,
      "learning_rate": 0.000165132148942936,
      "loss": 3.2283,
      "step": 35960
    },
    {
      "epoch": 0.4714656460534249,
      "grad_norm": 0.5925098061561584,
      "learning_rate": 0.00016507029118221372,
      "loss": 3.2779,
      "step": 35970
    },
    {
      "epoch": 0.47159671795947256,
      "grad_norm": 0.5640482902526855,
      "learning_rate": 0.00016500843083237421,
      "loss": 3.2846,
      "step": 35980
    },
    {
      "epoch": 0.4717277898655202,
      "grad_norm": 0.6046707630157471,
      "learning_rate": 0.00016494656790404518,
      "loss": 3.2643,
      "step": 35990
    },
    {
      "epoch": 0.4718588617715679,
      "grad_norm": 0.6561214923858643,
      "learning_rate": 0.0001648847024078548,
      "loss": 3.4286,
      "step": 36000
    },
    {
      "epoch": 0.4718588617715679,
      "eval_accuracy": 0.39325806451612905,
      "eval_loss": 3.278968334197998,
      "eval_runtime": 60.1427,
      "eval_samples_per_second": 33.254,
      "eval_steps_per_second": 0.532,
      "step": 36000
    },
    {
      "epoch": 0.47198993367761555,
      "grad_norm": 0.6059043407440186,
      "learning_rate": 0.00016482283435443175,
      "loss": 3.3086,
      "step": 36010
    },
    {
      "epoch": 0.4721210055836632,
      "grad_norm": 0.6018164753913879,
      "learning_rate": 0.00016476096375440518,
      "loss": 3.2491,
      "step": 36020
    },
    {
      "epoch": 0.4722520774897109,
      "grad_norm": 0.551322340965271,
      "learning_rate": 0.00016469909061840457,
      "loss": 3.3016,
      "step": 36030
    },
    {
      "epoch": 0.47238314939575854,
      "grad_norm": 0.6495176553726196,
      "learning_rate": 0.00016463721495705997,
      "loss": 3.2984,
      "step": 36040
    },
    {
      "epoch": 0.47251422130180615,
      "grad_norm": 0.6065201759338379,
      "learning_rate": 0.0001645753367810017,
      "loss": 3.2823,
      "step": 36050
    },
    {
      "epoch": 0.4726452932078538,
      "grad_norm": 0.6097673773765564,
      "learning_rate": 0.0001645134561008607,
      "loss": 3.2732,
      "step": 36060
    },
    {
      "epoch": 0.4727763651139015,
      "grad_norm": 0.5974128246307373,
      "learning_rate": 0.00016445157292726813,
      "loss": 3.2435,
      "step": 36070
    },
    {
      "epoch": 0.47290743701994914,
      "grad_norm": 0.5930661559104919,
      "learning_rate": 0.0001643896872708558,
      "loss": 3.2872,
      "step": 36080
    },
    {
      "epoch": 0.4730385089259968,
      "grad_norm": 0.6633502244949341,
      "learning_rate": 0.00016432779914225573,
      "loss": 3.251,
      "step": 36090
    },
    {
      "epoch": 0.47316958083204447,
      "grad_norm": 0.5811805725097656,
      "learning_rate": 0.0001642659085521005,
      "loss": 3.2914,
      "step": 36100
    },
    {
      "epoch": 0.47330065273809213,
      "grad_norm": 0.59276282787323,
      "learning_rate": 0.00016420401551102315,
      "loss": 3.2913,
      "step": 36110
    },
    {
      "epoch": 0.4734317246441398,
      "grad_norm": 0.5277324318885803,
      "learning_rate": 0.000164142120029657,
      "loss": 3.2729,
      "step": 36120
    },
    {
      "epoch": 0.47356279655018746,
      "grad_norm": 0.5775078535079956,
      "learning_rate": 0.0001640802221186359,
      "loss": 3.328,
      "step": 36130
    },
    {
      "epoch": 0.47369386845623507,
      "grad_norm": 0.5655336976051331,
      "learning_rate": 0.00016401832178859402,
      "loss": 3.2462,
      "step": 36140
    },
    {
      "epoch": 0.47382494036228273,
      "grad_norm": 0.552876353263855,
      "learning_rate": 0.0001639564190501661,
      "loss": 3.2741,
      "step": 36150
    },
    {
      "epoch": 0.4739560122683304,
      "grad_norm": 0.7065843939781189,
      "learning_rate": 0.00016389451391398717,
      "loss": 3.337,
      "step": 36160
    },
    {
      "epoch": 0.47408708417437806,
      "grad_norm": 0.5963832139968872,
      "learning_rate": 0.0001638326063906927,
      "loss": 3.2718,
      "step": 36170
    },
    {
      "epoch": 0.4742181560804257,
      "grad_norm": 0.5941736698150635,
      "learning_rate": 0.0001637706964909185,
      "loss": 3.2928,
      "step": 36180
    },
    {
      "epoch": 0.4743492279864734,
      "grad_norm": 0.5963121652603149,
      "learning_rate": 0.00016370878422530095,
      "loss": 3.2784,
      "step": 36190
    },
    {
      "epoch": 0.47448029989252105,
      "grad_norm": 0.5735953450202942,
      "learning_rate": 0.00016364686960447674,
      "loss": 3.3041,
      "step": 36200
    },
    {
      "epoch": 0.4746113717985687,
      "grad_norm": 0.6216247081756592,
      "learning_rate": 0.0001635849526390829,
      "loss": 3.3718,
      "step": 36210
    },
    {
      "epoch": 0.4747424437046164,
      "grad_norm": 0.5775728225708008,
      "learning_rate": 0.000163523033339757,
      "loss": 3.2692,
      "step": 36220
    },
    {
      "epoch": 0.474873515610664,
      "grad_norm": 0.6069855093955994,
      "learning_rate": 0.00016346111171713692,
      "loss": 3.2668,
      "step": 36230
    },
    {
      "epoch": 0.47500458751671165,
      "grad_norm": 0.5821763873100281,
      "learning_rate": 0.000163399187781861,
      "loss": 3.3381,
      "step": 36240
    },
    {
      "epoch": 0.4751356594227593,
      "grad_norm": 0.8778851628303528,
      "learning_rate": 0.00016333726154456788,
      "loss": 3.3003,
      "step": 36250
    },
    {
      "epoch": 0.475266731328807,
      "grad_norm": 0.5920222997665405,
      "learning_rate": 0.00016327533301589676,
      "loss": 3.2662,
      "step": 36260
    },
    {
      "epoch": 0.47539780323485464,
      "grad_norm": 0.5626883506774902,
      "learning_rate": 0.00016321340220648697,
      "loss": 3.2771,
      "step": 36270
    },
    {
      "epoch": 0.4755288751409023,
      "grad_norm": 0.6719560623168945,
      "learning_rate": 0.00016315146912697855,
      "loss": 3.3017,
      "step": 36280
    },
    {
      "epoch": 0.47565994704694997,
      "grad_norm": 0.6170803308486938,
      "learning_rate": 0.00016308953378801162,
      "loss": 3.279,
      "step": 36290
    },
    {
      "epoch": 0.47579101895299764,
      "grad_norm": 0.7501266002655029,
      "learning_rate": 0.00016302759620022701,
      "loss": 3.2757,
      "step": 36300
    },
    {
      "epoch": 0.4759220908590453,
      "grad_norm": 0.6225187182426453,
      "learning_rate": 0.00016296565637426567,
      "loss": 3.3402,
      "step": 36310
    },
    {
      "epoch": 0.4760531627650929,
      "grad_norm": 0.8012129664421082,
      "learning_rate": 0.000162903714320769,
      "loss": 3.4064,
      "step": 36320
    },
    {
      "epoch": 0.47618423467114057,
      "grad_norm": 0.6091654896736145,
      "learning_rate": 0.00016284177005037893,
      "loss": 3.2632,
      "step": 36330
    },
    {
      "epoch": 0.47631530657718824,
      "grad_norm": 0.6420459747314453,
      "learning_rate": 0.00016277982357373755,
      "loss": 3.2786,
      "step": 36340
    },
    {
      "epoch": 0.4764463784832359,
      "grad_norm": 0.5695201754570007,
      "learning_rate": 0.0001627178749014875,
      "loss": 3.3491,
      "step": 36350
    },
    {
      "epoch": 0.47657745038928356,
      "grad_norm": 0.5885868668556213,
      "learning_rate": 0.0001626559240442717,
      "loss": 3.2767,
      "step": 36360
    },
    {
      "epoch": 0.4767085222953312,
      "grad_norm": 0.6183741092681885,
      "learning_rate": 0.00016259397101273343,
      "loss": 3.2541,
      "step": 36370
    },
    {
      "epoch": 0.4768395942013789,
      "grad_norm": 0.581429123878479,
      "learning_rate": 0.0001625320158175165,
      "loss": 3.2624,
      "step": 36380
    },
    {
      "epoch": 0.47697066610742656,
      "grad_norm": 0.5709314346313477,
      "learning_rate": 0.00016247005846926487,
      "loss": 3.2861,
      "step": 36390
    },
    {
      "epoch": 0.4771017380134742,
      "grad_norm": 0.5907177329063416,
      "learning_rate": 0.00016240809897862305,
      "loss": 3.2426,
      "step": 36400
    },
    {
      "epoch": 0.4772328099195218,
      "grad_norm": 0.580534040927887,
      "learning_rate": 0.00016234613735623585,
      "loss": 3.2856,
      "step": 36410
    },
    {
      "epoch": 0.4773638818255695,
      "grad_norm": 0.5731232166290283,
      "learning_rate": 0.0001622841736127484,
      "loss": 3.2651,
      "step": 36420
    },
    {
      "epoch": 0.47749495373161716,
      "grad_norm": 0.5386447310447693,
      "learning_rate": 0.0001622222077588063,
      "loss": 3.317,
      "step": 36430
    },
    {
      "epoch": 0.4776260256376648,
      "grad_norm": 0.5965628027915955,
      "learning_rate": 0.00016216023980505543,
      "loss": 3.3067,
      "step": 36440
    },
    {
      "epoch": 0.4777570975437125,
      "grad_norm": 0.5629191398620605,
      "learning_rate": 0.00016209826976214207,
      "loss": 3.2482,
      "step": 36450
    },
    {
      "epoch": 0.47788816944976015,
      "grad_norm": 0.5725401639938354,
      "learning_rate": 0.0001620362976407128,
      "loss": 3.292,
      "step": 36460
    },
    {
      "epoch": 0.4780192413558078,
      "grad_norm": 0.5648769736289978,
      "learning_rate": 0.0001619743234514146,
      "loss": 3.301,
      "step": 36470
    },
    {
      "epoch": 0.4781503132618555,
      "grad_norm": 0.5633532404899597,
      "learning_rate": 0.00016191234720489486,
      "loss": 3.3314,
      "step": 36480
    },
    {
      "epoch": 0.47828138516790314,
      "grad_norm": 0.6398263573646545,
      "learning_rate": 0.00016185036891180118,
      "loss": 3.2317,
      "step": 36490
    },
    {
      "epoch": 0.47841245707395075,
      "grad_norm": 0.5827578902244568,
      "learning_rate": 0.00016178838858278173,
      "loss": 3.2536,
      "step": 36500
    },
    {
      "epoch": 0.47841245707395075,
      "eval_accuracy": 0.3940542521994135,
      "eval_loss": 3.271604061126709,
      "eval_runtime": 58.7833,
      "eval_samples_per_second": 34.023,
      "eval_steps_per_second": 0.544,
      "step": 36500
    },
    {
      "epoch": 0.4785435289799984,
      "grad_norm": 0.5755452513694763,
      "learning_rate": 0.00016172640622848477,
      "loss": 3.3155,
      "step": 36510
    },
    {
      "epoch": 0.4786746008860461,
      "grad_norm": 0.5996827483177185,
      "learning_rate": 0.00016166442185955914,
      "loss": 3.3091,
      "step": 36520
    },
    {
      "epoch": 0.47880567279209374,
      "grad_norm": 0.5649058818817139,
      "learning_rate": 0.0001616024354866539,
      "loss": 3.2992,
      "step": 36530
    },
    {
      "epoch": 0.4789367446981414,
      "grad_norm": 0.5416283011436462,
      "learning_rate": 0.00016154044712041844,
      "loss": 3.2992,
      "step": 36540
    },
    {
      "epoch": 0.47906781660418907,
      "grad_norm": 0.548918604850769,
      "learning_rate": 0.00016147845677150249,
      "loss": 3.3001,
      "step": 36550
    },
    {
      "epoch": 0.47919888851023673,
      "grad_norm": 0.6181765794754028,
      "learning_rate": 0.00016141646445055628,
      "loss": 3.2881,
      "step": 36560
    },
    {
      "epoch": 0.4793299604162844,
      "grad_norm": 0.6061245799064636,
      "learning_rate": 0.00016135447016823022,
      "loss": 3.2961,
      "step": 36570
    },
    {
      "epoch": 0.479461032322332,
      "grad_norm": 0.6074725389480591,
      "learning_rate": 0.000161292473935175,
      "loss": 3.5242,
      "step": 36580
    },
    {
      "epoch": 0.47959210422837967,
      "grad_norm": 0.62363201379776,
      "learning_rate": 0.0001612304757620419,
      "loss": 3.2783,
      "step": 36590
    },
    {
      "epoch": 0.47972317613442733,
      "grad_norm": 0.6310691833496094,
      "learning_rate": 0.0001611684756594823,
      "loss": 3.2939,
      "step": 36600
    },
    {
      "epoch": 0.479854248040475,
      "grad_norm": 0.5479382872581482,
      "learning_rate": 0.00016110647363814796,
      "loss": 3.2823,
      "step": 36610
    },
    {
      "epoch": 0.47998531994652266,
      "grad_norm": 0.618272066116333,
      "learning_rate": 0.00016104446970869105,
      "loss": 3.2576,
      "step": 36620
    },
    {
      "epoch": 0.4801163918525703,
      "grad_norm": 0.6054379343986511,
      "learning_rate": 0.00016098246388176395,
      "loss": 3.2914,
      "step": 36630
    },
    {
      "epoch": 0.480247463758618,
      "grad_norm": 0.6256258487701416,
      "learning_rate": 0.00016092045616801948,
      "loss": 3.2806,
      "step": 36640
    },
    {
      "epoch": 0.48037853566466565,
      "grad_norm": 0.5725874304771423,
      "learning_rate": 0.00016085844657811078,
      "loss": 3.2854,
      "step": 36650
    },
    {
      "epoch": 0.4805096075707133,
      "grad_norm": 0.5915955305099487,
      "learning_rate": 0.0001607964351226911,
      "loss": 3.3023,
      "step": 36660
    },
    {
      "epoch": 0.4806406794767609,
      "grad_norm": 0.6082817316055298,
      "learning_rate": 0.0001607344218124144,
      "loss": 3.2748,
      "step": 36670
    },
    {
      "epoch": 0.4807717513828086,
      "grad_norm": 0.5675665736198425,
      "learning_rate": 0.00016067240665793456,
      "loss": 3.2247,
      "step": 36680
    },
    {
      "epoch": 0.48090282328885625,
      "grad_norm": 0.5712056159973145,
      "learning_rate": 0.00016061038966990604,
      "loss": 3.2768,
      "step": 36690
    },
    {
      "epoch": 0.4810338951949039,
      "grad_norm": 0.5751532912254333,
      "learning_rate": 0.00016054837085898355,
      "loss": 3.2179,
      "step": 36700
    },
    {
      "epoch": 0.4811649671009516,
      "grad_norm": 0.6264739036560059,
      "learning_rate": 0.00016048635023582197,
      "loss": 3.3076,
      "step": 36710
    },
    {
      "epoch": 0.48129603900699924,
      "grad_norm": 0.6126514077186584,
      "learning_rate": 0.00016042432781107678,
      "loss": 3.2503,
      "step": 36720
    },
    {
      "epoch": 0.4814271109130469,
      "grad_norm": 0.5923537015914917,
      "learning_rate": 0.00016036230359540345,
      "loss": 3.2528,
      "step": 36730
    },
    {
      "epoch": 0.48155818281909457,
      "grad_norm": 0.6285374760627747,
      "learning_rate": 0.000160300277599458,
      "loss": 3.2572,
      "step": 36740
    },
    {
      "epoch": 0.48168925472514224,
      "grad_norm": 0.6334313750267029,
      "learning_rate": 0.00016023824983389662,
      "loss": 3.2913,
      "step": 36750
    },
    {
      "epoch": 0.48182032663118984,
      "grad_norm": 0.636250376701355,
      "learning_rate": 0.0001601762203093759,
      "loss": 3.2717,
      "step": 36760
    },
    {
      "epoch": 0.4819513985372375,
      "grad_norm": 0.6061951518058777,
      "learning_rate": 0.0001601141890365526,
      "loss": 3.283,
      "step": 36770
    },
    {
      "epoch": 0.48208247044328517,
      "grad_norm": 0.6361709237098694,
      "learning_rate": 0.00016005215602608397,
      "loss": 3.3006,
      "step": 36780
    },
    {
      "epoch": 0.48221354234933284,
      "grad_norm": 0.5920364856719971,
      "learning_rate": 0.00015999012128862743,
      "loss": 3.3093,
      "step": 36790
    },
    {
      "epoch": 0.4823446142553805,
      "grad_norm": 0.6130295395851135,
      "learning_rate": 0.00015992808483484067,
      "loss": 3.3063,
      "step": 36800
    },
    {
      "epoch": 0.48247568616142816,
      "grad_norm": 0.6701944470405579,
      "learning_rate": 0.00015986604667538168,
      "loss": 3.2648,
      "step": 36810
    },
    {
      "epoch": 0.4826067580674758,
      "grad_norm": 0.5952348709106445,
      "learning_rate": 0.00015980400682090887,
      "loss": 3.2607,
      "step": 36820
    },
    {
      "epoch": 0.4827378299735235,
      "grad_norm": 0.6339557766914368,
      "learning_rate": 0.00015974196528208086,
      "loss": 3.275,
      "step": 36830
    },
    {
      "epoch": 0.48286890187957116,
      "grad_norm": 0.584863007068634,
      "learning_rate": 0.00015967992206955648,
      "loss": 3.2669,
      "step": 36840
    },
    {
      "epoch": 0.48299997378561876,
      "grad_norm": 0.581449031829834,
      "learning_rate": 0.000159617877193995,
      "loss": 3.2478,
      "step": 36850
    },
    {
      "epoch": 0.4831310456916664,
      "grad_norm": 0.638435959815979,
      "learning_rate": 0.00015955583066605588,
      "loss": 3.2694,
      "step": 36860
    },
    {
      "epoch": 0.4832621175977141,
      "grad_norm": 0.6741981506347656,
      "learning_rate": 0.00015949378249639887,
      "loss": 3.2719,
      "step": 36870
    },
    {
      "epoch": 0.48339318950376176,
      "grad_norm": 0.6704249978065491,
      "learning_rate": 0.00015943173269568402,
      "loss": 3.2611,
      "step": 36880
    },
    {
      "epoch": 0.4835242614098094,
      "grad_norm": 0.5973716974258423,
      "learning_rate": 0.00015936968127457164,
      "loss": 3.2442,
      "step": 36890
    },
    {
      "epoch": 0.4836553333158571,
      "grad_norm": 0.712081253528595,
      "learning_rate": 0.00015930762824372234,
      "loss": 3.2698,
      "step": 36900
    },
    {
      "epoch": 0.48378640522190475,
      "grad_norm": 0.5978334546089172,
      "learning_rate": 0.00015924557361379706,
      "loss": 3.252,
      "step": 36910
    },
    {
      "epoch": 0.4839174771279524,
      "grad_norm": 0.6146913766860962,
      "learning_rate": 0.00015918351739545685,
      "loss": 3.2694,
      "step": 36920
    },
    {
      "epoch": 0.4840485490340001,
      "grad_norm": 0.6016756296157837,
      "learning_rate": 0.00015912145959936323,
      "loss": 3.2738,
      "step": 36930
    },
    {
      "epoch": 0.4841796209400477,
      "grad_norm": 0.6393879652023315,
      "learning_rate": 0.0001590594002361778,
      "loss": 3.288,
      "step": 36940
    },
    {
      "epoch": 0.48431069284609535,
      "grad_norm": 0.6666507124900818,
      "learning_rate": 0.0001589973393165627,
      "loss": 3.2705,
      "step": 36950
    },
    {
      "epoch": 0.484441764752143,
      "grad_norm": 0.647022008895874,
      "learning_rate": 0.00015893527685117997,
      "loss": 3.2679,
      "step": 36960
    },
    {
      "epoch": 0.4845728366581907,
      "grad_norm": 0.6123931407928467,
      "learning_rate": 0.00015887321285069225,
      "loss": 3.2384,
      "step": 36970
    },
    {
      "epoch": 0.48470390856423834,
      "grad_norm": 0.921006441116333,
      "learning_rate": 0.0001588111473257623,
      "loss": 3.2895,
      "step": 36980
    },
    {
      "epoch": 0.484834980470286,
      "grad_norm": 0.6198000311851501,
      "learning_rate": 0.000158749080287053,
      "loss": 3.2649,
      "step": 36990
    },
    {
      "epoch": 0.48496605237633367,
      "grad_norm": 0.5856442451477051,
      "learning_rate": 0.00015868701174522778,
      "loss": 3.2887,
      "step": 37000
    },
    {
      "epoch": 0.48496605237633367,
      "eval_accuracy": 0.3941505376344086,
      "eval_loss": 3.270933151245117,
      "eval_runtime": 61.6318,
      "eval_samples_per_second": 32.451,
      "eval_steps_per_second": 0.519,
      "step": 37000
    },
    {
      "epoch": 0.48509712428238133,
      "grad_norm": 0.708427906036377,
      "learning_rate": 0.0001586249417109501,
      "loss": 3.2544,
      "step": 37010
    },
    {
      "epoch": 0.485228196188429,
      "grad_norm": 0.6856189966201782,
      "learning_rate": 0.00015856287019488387,
      "loss": 3.2395,
      "step": 37020
    },
    {
      "epoch": 0.4853592680944766,
      "grad_norm": 0.6416229605674744,
      "learning_rate": 0.00015850079720769304,
      "loss": 3.2497,
      "step": 37030
    },
    {
      "epoch": 0.48549034000052427,
      "grad_norm": 0.630643367767334,
      "learning_rate": 0.00015843872276004203,
      "loss": 3.2379,
      "step": 37040
    },
    {
      "epoch": 0.48562141190657193,
      "grad_norm": 0.655580997467041,
      "learning_rate": 0.00015837664686259526,
      "loss": 3.2117,
      "step": 37050
    },
    {
      "epoch": 0.4857524838126196,
      "grad_norm": 0.6320750713348389,
      "learning_rate": 0.00015831456952601763,
      "loss": 3.2486,
      "step": 37060
    },
    {
      "epoch": 0.48588355571866726,
      "grad_norm": 0.6632112860679626,
      "learning_rate": 0.0001582524907609742,
      "loss": 3.2292,
      "step": 37070
    },
    {
      "epoch": 0.4860146276247149,
      "grad_norm": 0.672859251499176,
      "learning_rate": 0.00015819041057813028,
      "loss": 3.2941,
      "step": 37080
    },
    {
      "epoch": 0.4861456995307626,
      "grad_norm": 0.666069746017456,
      "learning_rate": 0.00015812832898815134,
      "loss": 3.2219,
      "step": 37090
    },
    {
      "epoch": 0.48627677143681025,
      "grad_norm": 0.7113033533096313,
      "learning_rate": 0.00015806624600170322,
      "loss": 3.2218,
      "step": 37100
    },
    {
      "epoch": 0.4864078433428579,
      "grad_norm": 0.5925906896591187,
      "learning_rate": 0.00015800416162945196,
      "loss": 3.2491,
      "step": 37110
    },
    {
      "epoch": 0.4865389152489055,
      "grad_norm": 0.5970014929771423,
      "learning_rate": 0.00015794207588206376,
      "loss": 3.2614,
      "step": 37120
    },
    {
      "epoch": 0.4866699871549532,
      "grad_norm": 0.6119125485420227,
      "learning_rate": 0.00015787998877020518,
      "loss": 3.2322,
      "step": 37130
    },
    {
      "epoch": 0.48680105906100085,
      "grad_norm": 0.7256065607070923,
      "learning_rate": 0.00015781790030454295,
      "loss": 3.2707,
      "step": 37140
    },
    {
      "epoch": 0.4869321309670485,
      "grad_norm": 0.63951176404953,
      "learning_rate": 0.00015775581049574404,
      "loss": 3.2803,
      "step": 37150
    },
    {
      "epoch": 0.4870632028730962,
      "grad_norm": 0.6368032097816467,
      "learning_rate": 0.00015769371935447565,
      "loss": 3.2729,
      "step": 37160
    },
    {
      "epoch": 0.48719427477914384,
      "grad_norm": 0.6407690644264221,
      "learning_rate": 0.0001576316268914052,
      "loss": 3.2234,
      "step": 37170
    },
    {
      "epoch": 0.4873253466851915,
      "grad_norm": 0.7055342197418213,
      "learning_rate": 0.0001575695331172003,
      "loss": 3.26,
      "step": 37180
    },
    {
      "epoch": 0.48745641859123917,
      "grad_norm": 0.64448082447052,
      "learning_rate": 0.00015750743804252893,
      "loss": 3.2632,
      "step": 37190
    },
    {
      "epoch": 0.48758749049728684,
      "grad_norm": 0.6167199015617371,
      "learning_rate": 0.00015744534167805912,
      "loss": 3.3083,
      "step": 37200
    },
    {
      "epoch": 0.48771856240333444,
      "grad_norm": 0.6284044981002808,
      "learning_rate": 0.0001573832440344592,
      "loss": 3.2774,
      "step": 37210
    },
    {
      "epoch": 0.4878496343093821,
      "grad_norm": 0.627443253993988,
      "learning_rate": 0.00015732114512239775,
      "loss": 3.2356,
      "step": 37220
    },
    {
      "epoch": 0.48798070621542977,
      "grad_norm": 0.5931061506271362,
      "learning_rate": 0.00015725904495254346,
      "loss": 3.2591,
      "step": 37230
    },
    {
      "epoch": 0.48811177812147744,
      "grad_norm": 0.579247772693634,
      "learning_rate": 0.00015719694353556542,
      "loss": 3.274,
      "step": 37240
    },
    {
      "epoch": 0.4882428500275251,
      "grad_norm": 0.5992260575294495,
      "learning_rate": 0.00015713484088213274,
      "loss": 3.2685,
      "step": 37250
    },
    {
      "epoch": 0.48837392193357276,
      "grad_norm": 0.6046287417411804,
      "learning_rate": 0.00015707273700291488,
      "loss": 3.2322,
      "step": 37260
    },
    {
      "epoch": 0.4885049938396204,
      "grad_norm": 0.5924400091171265,
      "learning_rate": 0.00015701063190858139,
      "loss": 3.278,
      "step": 37270
    },
    {
      "epoch": 0.4886360657456681,
      "grad_norm": 0.6171079874038696,
      "learning_rate": 0.00015694852560980214,
      "loss": 3.2512,
      "step": 37280
    },
    {
      "epoch": 0.48876713765171576,
      "grad_norm": 1.216789960861206,
      "learning_rate": 0.00015688641811724717,
      "loss": 3.224,
      "step": 37290
    },
    {
      "epoch": 0.48889820955776336,
      "grad_norm": 0.6501765251159668,
      "learning_rate": 0.0001568243094415867,
      "loss": 3.2501,
      "step": 37300
    },
    {
      "epoch": 0.489029281463811,
      "grad_norm": 0.6024908423423767,
      "learning_rate": 0.00015676219959349116,
      "loss": 3.2645,
      "step": 37310
    },
    {
      "epoch": 0.4891603533698587,
      "grad_norm": 0.610381543636322,
      "learning_rate": 0.00015670008858363122,
      "loss": 3.2196,
      "step": 37320
    },
    {
      "epoch": 0.48929142527590636,
      "grad_norm": 0.5619906783103943,
      "learning_rate": 0.0001566379764226777,
      "loss": 3.3084,
      "step": 37330
    },
    {
      "epoch": 0.489422497181954,
      "grad_norm": 0.606250524520874,
      "learning_rate": 0.0001565758631213017,
      "loss": 3.2672,
      "step": 37340
    },
    {
      "epoch": 0.4895535690880017,
      "grad_norm": 0.6172598004341125,
      "learning_rate": 0.00015651374869017433,
      "loss": 3.2343,
      "step": 37350
    },
    {
      "epoch": 0.48968464099404935,
      "grad_norm": 0.7312554121017456,
      "learning_rate": 0.0001564516331399671,
      "loss": 3.277,
      "step": 37360
    },
    {
      "epoch": 0.489815712900097,
      "grad_norm": 0.6577658653259277,
      "learning_rate": 0.00015638951648135166,
      "loss": 3.2475,
      "step": 37370
    },
    {
      "epoch": 0.4899467848061447,
      "grad_norm": 0.6053968667984009,
      "learning_rate": 0.0001563273987249997,
      "loss": 3.2914,
      "step": 37380
    },
    {
      "epoch": 0.4900778567121923,
      "grad_norm": 0.8270416855812073,
      "learning_rate": 0.0001562652798815834,
      "loss": 3.3286,
      "step": 37390
    },
    {
      "epoch": 0.49020892861823995,
      "grad_norm": 0.620601236820221,
      "learning_rate": 0.0001562031599617748,
      "loss": 3.2303,
      "step": 37400
    },
    {
      "epoch": 0.4903400005242876,
      "grad_norm": 0.6468815207481384,
      "learning_rate": 0.00015614103897624638,
      "loss": 3.2638,
      "step": 37410
    },
    {
      "epoch": 0.4904710724303353,
      "grad_norm": 0.6048716306686401,
      "learning_rate": 0.0001560789169356706,
      "loss": 3.2626,
      "step": 37420
    },
    {
      "epoch": 0.49060214433638294,
      "grad_norm": 0.6043686866760254,
      "learning_rate": 0.00015601679385072025,
      "loss": 3.2839,
      "step": 37430
    },
    {
      "epoch": 0.4907332162424306,
      "grad_norm": 0.5899128317832947,
      "learning_rate": 0.00015595466973206823,
      "loss": 3.2542,
      "step": 37440
    },
    {
      "epoch": 0.49086428814847827,
      "grad_norm": 0.5496528744697571,
      "learning_rate": 0.00015589254459038763,
      "loss": 3.2622,
      "step": 37450
    },
    {
      "epoch": 0.49099536005452593,
      "grad_norm": 0.6141208410263062,
      "learning_rate": 0.00015583041843635173,
      "loss": 3.2823,
      "step": 37460
    },
    {
      "epoch": 0.4911264319605736,
      "grad_norm": 0.5974559783935547,
      "learning_rate": 0.00015576829128063397,
      "loss": 3.2851,
      "step": 37470
    },
    {
      "epoch": 0.4912575038666212,
      "grad_norm": 0.6064287424087524,
      "learning_rate": 0.00015570616313390797,
      "loss": 3.2597,
      "step": 37480
    },
    {
      "epoch": 0.49138857577266887,
      "grad_norm": 0.5973837971687317,
      "learning_rate": 0.00015564403400684753,
      "loss": 3.3269,
      "step": 37490
    },
    {
      "epoch": 0.49151964767871653,
      "grad_norm": 0.586722195148468,
      "learning_rate": 0.00015558190391012656,
      "loss": 3.2977,
      "step": 37500
    },
    {
      "epoch": 0.49151964767871653,
      "eval_accuracy": 0.3948343108504399,
      "eval_loss": 3.2665812969207764,
      "eval_runtime": 60.1149,
      "eval_samples_per_second": 33.27,
      "eval_steps_per_second": 0.532,
      "step": 37500
    },
    {
      "epoch": 0.4916507195847642,
      "grad_norm": 0.7739748358726501,
      "learning_rate": 0.00015551977285441923,
      "loss": 3.3223,
      "step": 37510
    },
    {
      "epoch": 0.49178179149081186,
      "grad_norm": 0.6026442646980286,
      "learning_rate": 0.0001554576408503998,
      "loss": 3.2571,
      "step": 37520
    },
    {
      "epoch": 0.4919128633968595,
      "grad_norm": 0.6100154519081116,
      "learning_rate": 0.0001553955079087427,
      "loss": 3.2414,
      "step": 37530
    },
    {
      "epoch": 0.4920439353029072,
      "grad_norm": 0.6126632690429688,
      "learning_rate": 0.0001553333740401226,
      "loss": 3.2573,
      "step": 37540
    },
    {
      "epoch": 0.49217500720895485,
      "grad_norm": 0.5916719436645508,
      "learning_rate": 0.00015527123925521415,
      "loss": 3.2381,
      "step": 37550
    },
    {
      "epoch": 0.4923060791150025,
      "grad_norm": 0.5856756567955017,
      "learning_rate": 0.00015520910356469242,
      "loss": 3.2332,
      "step": 37560
    },
    {
      "epoch": 0.4924371510210501,
      "grad_norm": 0.5991889834403992,
      "learning_rate": 0.00015514696697923244,
      "loss": 3.2712,
      "step": 37570
    },
    {
      "epoch": 0.4925682229270978,
      "grad_norm": 0.6295677423477173,
      "learning_rate": 0.0001550848295095094,
      "loss": 3.2928,
      "step": 37580
    },
    {
      "epoch": 0.49269929483314545,
      "grad_norm": 0.6851557493209839,
      "learning_rate": 0.00015502269116619873,
      "loss": 3.2656,
      "step": 37590
    },
    {
      "epoch": 0.4928303667391931,
      "grad_norm": 0.6387191414833069,
      "learning_rate": 0.00015496055195997596,
      "loss": 3.349,
      "step": 37600
    },
    {
      "epoch": 0.4929614386452408,
      "grad_norm": 0.621385395526886,
      "learning_rate": 0.0001548984119015168,
      "loss": 3.245,
      "step": 37610
    },
    {
      "epoch": 0.49309251055128844,
      "grad_norm": 0.5905426740646362,
      "learning_rate": 0.000154836271001497,
      "loss": 3.2533,
      "step": 37620
    },
    {
      "epoch": 0.4932235824573361,
      "grad_norm": 0.6103399991989136,
      "learning_rate": 0.00015477412927059265,
      "loss": 3.2785,
      "step": 37630
    },
    {
      "epoch": 0.49335465436338377,
      "grad_norm": 0.5693601369857788,
      "learning_rate": 0.00015471198671947975,
      "loss": 3.2931,
      "step": 37640
    },
    {
      "epoch": 0.49348572626943144,
      "grad_norm": 0.5823444128036499,
      "learning_rate": 0.00015464984335883468,
      "loss": 3.2673,
      "step": 37650
    },
    {
      "epoch": 0.49361679817547904,
      "grad_norm": 0.6745126247406006,
      "learning_rate": 0.00015458769919933371,
      "loss": 3.2888,
      "step": 37660
    },
    {
      "epoch": 0.4937478700815267,
      "grad_norm": 0.6344786882400513,
      "learning_rate": 0.00015452555425165353,
      "loss": 3.233,
      "step": 37670
    },
    {
      "epoch": 0.49387894198757437,
      "grad_norm": 0.5846032500267029,
      "learning_rate": 0.0001544634085264707,
      "loss": 3.2755,
      "step": 37680
    },
    {
      "epoch": 0.49401001389362204,
      "grad_norm": 0.6851807236671448,
      "learning_rate": 0.0001544012620344621,
      "loss": 3.2426,
      "step": 37690
    },
    {
      "epoch": 0.4941410857996697,
      "grad_norm": 0.5842995643615723,
      "learning_rate": 0.0001543391147863046,
      "loss": 3.2158,
      "step": 37700
    },
    {
      "epoch": 0.49427215770571736,
      "grad_norm": 0.627249002456665,
      "learning_rate": 0.00015427696679267528,
      "loss": 3.2605,
      "step": 37710
    },
    {
      "epoch": 0.494403229611765,
      "grad_norm": 0.5863708257675171,
      "learning_rate": 0.00015421481806425136,
      "loss": 3.2257,
      "step": 37720
    },
    {
      "epoch": 0.4945343015178127,
      "grad_norm": 0.5877141952514648,
      "learning_rate": 0.00015415266861171012,
      "loss": 3.2069,
      "step": 37730
    },
    {
      "epoch": 0.49466537342386036,
      "grad_norm": 0.5912908911705017,
      "learning_rate": 0.00015409051844572904,
      "loss": 3.2571,
      "step": 37740
    },
    {
      "epoch": 0.49479644532990796,
      "grad_norm": 0.6409209966659546,
      "learning_rate": 0.0001540283675769857,
      "loss": 3.292,
      "step": 37750
    },
    {
      "epoch": 0.4949275172359556,
      "grad_norm": 0.6512109041213989,
      "learning_rate": 0.0001539662160161578,
      "loss": 3.253,
      "step": 37760
    },
    {
      "epoch": 0.4950585891420033,
      "grad_norm": 0.7133368253707886,
      "learning_rate": 0.00015390406377392305,
      "loss": 3.2762,
      "step": 37770
    },
    {
      "epoch": 0.49518966104805096,
      "grad_norm": 0.6098271608352661,
      "learning_rate": 0.00015384191086095954,
      "loss": 3.2764,
      "step": 37780
    },
    {
      "epoch": 0.4953207329540986,
      "grad_norm": 0.5924927592277527,
      "learning_rate": 0.00015377975728794518,
      "loss": 3.261,
      "step": 37790
    },
    {
      "epoch": 0.4954518048601463,
      "grad_norm": 0.6280924081802368,
      "learning_rate": 0.00015371760306555818,
      "loss": 3.2395,
      "step": 37800
    },
    {
      "epoch": 0.49558287676619395,
      "grad_norm": 0.7646282315254211,
      "learning_rate": 0.00015365544820447675,
      "loss": 3.31,
      "step": 37810
    },
    {
      "epoch": 0.4957139486722416,
      "grad_norm": 0.6185669898986816,
      "learning_rate": 0.00015359329271537935,
      "loss": 3.2581,
      "step": 37820
    },
    {
      "epoch": 0.4958450205782893,
      "grad_norm": 0.6210280060768127,
      "learning_rate": 0.0001535311366089444,
      "loss": 3.2455,
      "step": 37830
    },
    {
      "epoch": 0.4959760924843369,
      "grad_norm": 1.2351998090744019,
      "learning_rate": 0.0001534689798958505,
      "loss": 3.2537,
      "step": 37840
    },
    {
      "epoch": 0.49610716439038455,
      "grad_norm": 0.5883873105049133,
      "learning_rate": 0.00015340682258677634,
      "loss": 3.2352,
      "step": 37850
    },
    {
      "epoch": 0.4962382362964322,
      "grad_norm": 0.6110877394676208,
      "learning_rate": 0.00015334466469240073,
      "loss": 3.3281,
      "step": 37860
    },
    {
      "epoch": 0.4963693082024799,
      "grad_norm": 0.679965615272522,
      "learning_rate": 0.0001532825062234026,
      "loss": 3.2601,
      "step": 37870
    },
    {
      "epoch": 0.49650038010852754,
      "grad_norm": 0.6191540360450745,
      "learning_rate": 0.00015322034719046087,
      "loss": 3.2925,
      "step": 37880
    },
    {
      "epoch": 0.4966314520145752,
      "grad_norm": 0.6600185632705688,
      "learning_rate": 0.00015315818760425472,
      "loss": 3.3305,
      "step": 37890
    },
    {
      "epoch": 0.49676252392062287,
      "grad_norm": 0.6230777502059937,
      "learning_rate": 0.00015309602747546323,
      "loss": 3.292,
      "step": 37900
    },
    {
      "epoch": 0.49689359582667053,
      "grad_norm": 0.6368182301521301,
      "learning_rate": 0.00015303386681476578,
      "loss": 3.2735,
      "step": 37910
    },
    {
      "epoch": 0.4970246677327182,
      "grad_norm": 0.5717352628707886,
      "learning_rate": 0.00015297170563284166,
      "loss": 3.2792,
      "step": 37920
    },
    {
      "epoch": 0.4971557396387658,
      "grad_norm": 0.6473208069801331,
      "learning_rate": 0.0001529095439403704,
      "loss": 3.3271,
      "step": 37930
    },
    {
      "epoch": 0.49728681154481347,
      "grad_norm": 0.6040807366371155,
      "learning_rate": 0.0001528473817480315,
      "loss": 3.2251,
      "step": 37940
    },
    {
      "epoch": 0.49741788345086113,
      "grad_norm": 0.6010756492614746,
      "learning_rate": 0.0001527852190665046,
      "loss": 3.2689,
      "step": 37950
    },
    {
      "epoch": 0.4975489553569088,
      "grad_norm": 0.6081663370132446,
      "learning_rate": 0.00015272305590646948,
      "loss": 3.2978,
      "step": 37960
    },
    {
      "epoch": 0.49768002726295646,
      "grad_norm": 0.5802282691001892,
      "learning_rate": 0.00015266089227860592,
      "loss": 3.222,
      "step": 37970
    },
    {
      "epoch": 0.4978110991690041,
      "grad_norm": 0.6232720613479614,
      "learning_rate": 0.00015259872819359374,
      "loss": 3.2841,
      "step": 37980
    },
    {
      "epoch": 0.4979421710750518,
      "grad_norm": 0.6499055027961731,
      "learning_rate": 0.00015253656366211293,
      "loss": 3.2665,
      "step": 37990
    },
    {
      "epoch": 0.49807324298109945,
      "grad_norm": 0.6103366017341614,
      "learning_rate": 0.00015247439869484354,
      "loss": 3.2829,
      "step": 38000
    },
    {
      "epoch": 0.49807324298109945,
      "eval_accuracy": 0.39496823069403714,
      "eval_loss": 3.265247106552124,
      "eval_runtime": 60.2035,
      "eval_samples_per_second": 33.221,
      "eval_steps_per_second": 0.532,
      "step": 38000
    },
    {
      "epoch": 0.4982043148871471,
      "grad_norm": 0.6250506043434143,
      "learning_rate": 0.00015241223330246563,
      "loss": 3.2525,
      "step": 38010
    },
    {
      "epoch": 0.4983353867931947,
      "grad_norm": 0.6130933165550232,
      "learning_rate": 0.00015235006749565952,
      "loss": 3.301,
      "step": 38020
    },
    {
      "epoch": 0.4984664586992424,
      "grad_norm": 0.6740614175796509,
      "learning_rate": 0.0001522879012851053,
      "loss": 3.259,
      "step": 38030
    },
    {
      "epoch": 0.49859753060529005,
      "grad_norm": 0.6438813805580139,
      "learning_rate": 0.00015222573468148343,
      "loss": 3.2915,
      "step": 38040
    },
    {
      "epoch": 0.4987286025113377,
      "grad_norm": 0.636891782283783,
      "learning_rate": 0.00015216356769547423,
      "loss": 3.2659,
      "step": 38050
    },
    {
      "epoch": 0.4988596744173854,
      "grad_norm": 0.5767593383789062,
      "learning_rate": 0.00015210140033775817,
      "loss": 3.2593,
      "step": 38060
    },
    {
      "epoch": 0.49899074632343304,
      "grad_norm": 0.6000739932060242,
      "learning_rate": 0.00015203923261901574,
      "loss": 3.2371,
      "step": 38070
    },
    {
      "epoch": 0.4991218182294807,
      "grad_norm": 0.6195977330207825,
      "learning_rate": 0.00015197706454992758,
      "loss": 3.3058,
      "step": 38080
    },
    {
      "epoch": 0.49925289013552837,
      "grad_norm": 0.6868708729743958,
      "learning_rate": 0.00015191489614117432,
      "loss": 3.3037,
      "step": 38090
    },
    {
      "epoch": 0.49938396204157603,
      "grad_norm": 0.6040688157081604,
      "learning_rate": 0.00015185272740343667,
      "loss": 3.2535,
      "step": 38100
    },
    {
      "epoch": 0.49951503394762364,
      "grad_norm": 0.5792253017425537,
      "learning_rate": 0.00015179055834739537,
      "loss": 3.2701,
      "step": 38110
    },
    {
      "epoch": 0.4996461058536713,
      "grad_norm": 0.5699061155319214,
      "learning_rate": 0.00015172838898373122,
      "loss": 3.2437,
      "step": 38120
    },
    {
      "epoch": 0.49977717775971897,
      "grad_norm": 0.696327805519104,
      "learning_rate": 0.00015166621932312513,
      "loss": 3.2379,
      "step": 38130
    },
    {
      "epoch": 0.49990824966576664,
      "grad_norm": 0.6414421200752258,
      "learning_rate": 0.00015160404937625803,
      "loss": 3.2356,
      "step": 38140
    },
    {
      "epoch": 0.5000393215718143,
      "grad_norm": 0.6192728877067566,
      "learning_rate": 0.0001515418791538108,
      "loss": 3.3023,
      "step": 38150
    },
    {
      "epoch": 0.500170393477862,
      "grad_norm": 0.5632935166358948,
      "learning_rate": 0.0001514797086664645,
      "loss": 3.3288,
      "step": 38160
    },
    {
      "epoch": 0.5003014653839096,
      "grad_norm": 1.0463660955429077,
      "learning_rate": 0.00015141753792490022,
      "loss": 3.2566,
      "step": 38170
    },
    {
      "epoch": 0.5004325372899573,
      "grad_norm": 0.6227558851242065,
      "learning_rate": 0.000151355366939799,
      "loss": 3.2283,
      "step": 38180
    },
    {
      "epoch": 0.500563609196005,
      "grad_norm": 0.695426344871521,
      "learning_rate": 0.0001512931957218421,
      "loss": 3.2878,
      "step": 38190
    },
    {
      "epoch": 0.5006946811020526,
      "grad_norm": 0.6402430534362793,
      "learning_rate": 0.0001512310242817106,
      "loss": 3.2436,
      "step": 38200
    },
    {
      "epoch": 0.5008257530081003,
      "grad_norm": 0.5775047540664673,
      "learning_rate": 0.00015116885263008576,
      "loss": 3.2801,
      "step": 38210
    },
    {
      "epoch": 0.500956824914148,
      "grad_norm": 0.642344057559967,
      "learning_rate": 0.00015110668077764883,
      "loss": 3.2511,
      "step": 38220
    },
    {
      "epoch": 0.5010878968201956,
      "grad_norm": 0.5989440083503723,
      "learning_rate": 0.00015104450873508113,
      "loss": 3.2383,
      "step": 38230
    },
    {
      "epoch": 0.5012189687262432,
      "grad_norm": 0.6457329988479614,
      "learning_rate": 0.00015098233651306394,
      "loss": 3.2381,
      "step": 38240
    },
    {
      "epoch": 0.5013500406322908,
      "grad_norm": 0.6108090877532959,
      "learning_rate": 0.00015092016412227865,
      "loss": 3.261,
      "step": 38250
    },
    {
      "epoch": 0.5014811125383385,
      "grad_norm": 0.6937341094017029,
      "learning_rate": 0.00015085799157340664,
      "loss": 3.3061,
      "step": 38260
    },
    {
      "epoch": 0.5016121844443862,
      "grad_norm": 0.7011916637420654,
      "learning_rate": 0.00015079581887712934,
      "loss": 3.2856,
      "step": 38270
    },
    {
      "epoch": 0.5017432563504338,
      "grad_norm": 0.6247754096984863,
      "learning_rate": 0.0001507336460441282,
      "loss": 3.2718,
      "step": 38280
    },
    {
      "epoch": 0.5018743282564815,
      "grad_norm": 0.6372054219245911,
      "learning_rate": 0.0001506714730850846,
      "loss": 3.2453,
      "step": 38290
    },
    {
      "epoch": 0.5020054001625291,
      "grad_norm": 0.6016072630882263,
      "learning_rate": 0.00015060930001068012,
      "loss": 3.2957,
      "step": 38300
    },
    {
      "epoch": 0.5021364720685768,
      "grad_norm": 0.575780987739563,
      "learning_rate": 0.00015054712683159623,
      "loss": 3.2666,
      "step": 38310
    },
    {
      "epoch": 0.5022675439746245,
      "grad_norm": 0.5816180109977722,
      "learning_rate": 0.00015048495355851448,
      "loss": 3.2184,
      "step": 38320
    },
    {
      "epoch": 0.5023986158806721,
      "grad_norm": 1.2581263780593872,
      "learning_rate": 0.00015042278020211634,
      "loss": 3.2132,
      "step": 38330
    },
    {
      "epoch": 0.5025296877867198,
      "grad_norm": 0.6293199062347412,
      "learning_rate": 0.0001503606067730834,
      "loss": 3.2309,
      "step": 38340
    },
    {
      "epoch": 0.5026607596927675,
      "grad_norm": 0.6730409860610962,
      "learning_rate": 0.00015029843328209723,
      "loss": 3.26,
      "step": 38350
    },
    {
      "epoch": 0.5027918315988151,
      "grad_norm": 0.6247910857200623,
      "learning_rate": 0.0001502362597398394,
      "loss": 3.2303,
      "step": 38360
    },
    {
      "epoch": 0.5029229035048628,
      "grad_norm": 0.6220833659172058,
      "learning_rate": 0.00015017408615699152,
      "loss": 3.298,
      "step": 38370
    },
    {
      "epoch": 0.5030539754109105,
      "grad_norm": 0.5771218538284302,
      "learning_rate": 0.00015011191254423516,
      "loss": 3.2867,
      "step": 38380
    },
    {
      "epoch": 0.5031850473169581,
      "grad_norm": 0.5973936319351196,
      "learning_rate": 0.00015004973891225191,
      "loss": 3.2424,
      "step": 38390
    },
    {
      "epoch": 0.5033161192230058,
      "grad_norm": 0.7079201340675354,
      "learning_rate": 0.0001499875652717234,
      "loss": 3.2444,
      "step": 38400
    },
    {
      "epoch": 0.5034471911290535,
      "grad_norm": 0.6678962707519531,
      "learning_rate": 0.0001499253916333312,
      "loss": 3.278,
      "step": 38410
    },
    {
      "epoch": 0.503578263035101,
      "grad_norm": 0.6563135385513306,
      "learning_rate": 0.0001498632180077569,
      "loss": 3.2548,
      "step": 38420
    },
    {
      "epoch": 0.5037093349411487,
      "grad_norm": 0.6201413869857788,
      "learning_rate": 0.0001498010444056821,
      "loss": 3.2785,
      "step": 38430
    },
    {
      "epoch": 0.5038404068471963,
      "grad_norm": 0.5744729042053223,
      "learning_rate": 0.0001497388708377885,
      "loss": 3.2789,
      "step": 38440
    },
    {
      "epoch": 0.503971478753244,
      "grad_norm": 0.6150601506233215,
      "learning_rate": 0.00014967669731475754,
      "loss": 3.2221,
      "step": 38450
    },
    {
      "epoch": 0.5041025506592917,
      "grad_norm": 0.6147163510322571,
      "learning_rate": 0.00014961452384727085,
      "loss": 3.2521,
      "step": 38460
    },
    {
      "epoch": 0.5042336225653393,
      "grad_norm": 0.6126275062561035,
      "learning_rate": 0.0001495523504460101,
      "loss": 3.2438,
      "step": 38470
    },
    {
      "epoch": 0.504364694471387,
      "grad_norm": 0.6457476615905762,
      "learning_rate": 0.0001494901771216567,
      "loss": 3.2233,
      "step": 38480
    },
    {
      "epoch": 0.5044957663774347,
      "grad_norm": 0.6577040553092957,
      "learning_rate": 0.00014942800388489228,
      "loss": 3.2448,
      "step": 38490
    },
    {
      "epoch": 0.5046268382834823,
      "grad_norm": 0.6282755732536316,
      "learning_rate": 0.00014936583074639837,
      "loss": 3.2183,
      "step": 38500
    },
    {
      "epoch": 0.5046268382834823,
      "eval_accuracy": 0.3951681329423265,
      "eval_loss": 3.2636630535125732,
      "eval_runtime": 60.8271,
      "eval_samples_per_second": 32.88,
      "eval_steps_per_second": 0.526,
      "step": 38500
    },
    {
      "epoch": 0.50475791018953,
      "grad_norm": 0.6190617680549622,
      "learning_rate": 0.0001493036577168565,
      "loss": 3.2365,
      "step": 38510
    },
    {
      "epoch": 0.5048889820955776,
      "grad_norm": 0.6034877896308899,
      "learning_rate": 0.0001492414848069481,
      "loss": 3.2219,
      "step": 38520
    },
    {
      "epoch": 0.5050200540016253,
      "grad_norm": 0.5828790664672852,
      "learning_rate": 0.0001491793120273547,
      "loss": 3.2751,
      "step": 38530
    },
    {
      "epoch": 0.505151125907673,
      "grad_norm": 0.5916118025779724,
      "learning_rate": 0.0001491171393887577,
      "loss": 3.2591,
      "step": 38540
    },
    {
      "epoch": 0.5052821978137206,
      "grad_norm": 0.6635059118270874,
      "learning_rate": 0.00014905496690183868,
      "loss": 3.2609,
      "step": 38550
    },
    {
      "epoch": 0.5054132697197683,
      "grad_norm": 0.6622881889343262,
      "learning_rate": 0.00014899279457727886,
      "loss": 3.2737,
      "step": 38560
    },
    {
      "epoch": 0.505544341625816,
      "grad_norm": 0.6123888492584229,
      "learning_rate": 0.00014893062242575967,
      "loss": 3.3248,
      "step": 38570
    },
    {
      "epoch": 0.5056754135318636,
      "grad_norm": 0.5767281651496887,
      "learning_rate": 0.00014886845045796257,
      "loss": 3.2755,
      "step": 38580
    },
    {
      "epoch": 0.5058064854379113,
      "grad_norm": 0.6128004789352417,
      "learning_rate": 0.0001488062786845687,
      "loss": 3.2982,
      "step": 38590
    },
    {
      "epoch": 0.5059375573439588,
      "grad_norm": 0.6065090894699097,
      "learning_rate": 0.00014874410711625944,
      "loss": 3.2592,
      "step": 38600
    },
    {
      "epoch": 0.5060686292500065,
      "grad_norm": 0.6065571904182434,
      "learning_rate": 0.000148681935763716,
      "loss": 3.2763,
      "step": 38610
    },
    {
      "epoch": 0.5061997011560542,
      "grad_norm": 0.6307502388954163,
      "learning_rate": 0.00014861976463761967,
      "loss": 3.3503,
      "step": 38620
    },
    {
      "epoch": 0.5063307730621018,
      "grad_norm": 0.585487961769104,
      "learning_rate": 0.00014855759374865152,
      "loss": 3.2394,
      "step": 38630
    },
    {
      "epoch": 0.5064618449681495,
      "grad_norm": 0.8138853311538696,
      "learning_rate": 0.00014849542310749265,
      "loss": 3.2573,
      "step": 38640
    },
    {
      "epoch": 0.5065929168741972,
      "grad_norm": 0.6118040084838867,
      "learning_rate": 0.00014843325272482426,
      "loss": 3.3333,
      "step": 38650
    },
    {
      "epoch": 0.5067239887802448,
      "grad_norm": 0.595526933670044,
      "learning_rate": 0.0001483710826113274,
      "loss": 3.2862,
      "step": 38660
    },
    {
      "epoch": 0.5068550606862925,
      "grad_norm": 0.6313866972923279,
      "learning_rate": 0.00014830891277768294,
      "loss": 3.2315,
      "step": 38670
    },
    {
      "epoch": 0.5069861325923402,
      "grad_norm": 0.6278272867202759,
      "learning_rate": 0.00014824674323457194,
      "loss": 3.2781,
      "step": 38680
    },
    {
      "epoch": 0.5071172044983878,
      "grad_norm": 0.6671293377876282,
      "learning_rate": 0.0001481845739926752,
      "loss": 3.3008,
      "step": 38690
    },
    {
      "epoch": 0.5072482764044355,
      "grad_norm": 0.7043506503105164,
      "learning_rate": 0.0001481224050626737,
      "loss": 3.2556,
      "step": 38700
    },
    {
      "epoch": 0.5073793483104831,
      "grad_norm": 0.6064945459365845,
      "learning_rate": 0.0001480602364552481,
      "loss": 3.2571,
      "step": 38710
    },
    {
      "epoch": 0.5075104202165308,
      "grad_norm": 0.6753714084625244,
      "learning_rate": 0.0001479980681810792,
      "loss": 3.2772,
      "step": 38720
    },
    {
      "epoch": 0.5076414921225785,
      "grad_norm": 0.5752739310264587,
      "learning_rate": 0.0001479359002508477,
      "loss": 3.2729,
      "step": 38730
    },
    {
      "epoch": 0.5077725640286261,
      "grad_norm": 0.666944682598114,
      "learning_rate": 0.00014787373267523415,
      "loss": 3.2873,
      "step": 38740
    },
    {
      "epoch": 0.5079036359346738,
      "grad_norm": 0.5669562220573425,
      "learning_rate": 0.00014781156546491914,
      "loss": 3.2572,
      "step": 38750
    },
    {
      "epoch": 0.5080347078407215,
      "grad_norm": 0.6077606081962585,
      "learning_rate": 0.0001477493986305832,
      "loss": 3.2545,
      "step": 38760
    },
    {
      "epoch": 0.5081657797467691,
      "grad_norm": 1.1567224264144897,
      "learning_rate": 0.00014768723218290677,
      "loss": 3.4157,
      "step": 38770
    },
    {
      "epoch": 0.5082968516528167,
      "grad_norm": 0.6085394024848938,
      "learning_rate": 0.00014762506613257017,
      "loss": 3.277,
      "step": 38780
    },
    {
      "epoch": 0.5084279235588643,
      "grad_norm": 0.5669998526573181,
      "learning_rate": 0.0001475629004902537,
      "loss": 3.2659,
      "step": 38790
    },
    {
      "epoch": 0.508558995464912,
      "grad_norm": 0.6231284737586975,
      "learning_rate": 0.00014750073526663757,
      "loss": 3.2505,
      "step": 38800
    },
    {
      "epoch": 0.5086900673709597,
      "grad_norm": 0.6040982604026794,
      "learning_rate": 0.00014743857047240204,
      "loss": 3.1939,
      "step": 38810
    },
    {
      "epoch": 0.5088211392770073,
      "grad_norm": 0.5718399286270142,
      "learning_rate": 0.00014737640611822708,
      "loss": 3.2719,
      "step": 38820
    },
    {
      "epoch": 0.508952211183055,
      "grad_norm": 0.5968985557556152,
      "learning_rate": 0.00014731424221479272,
      "loss": 3.2832,
      "step": 38830
    },
    {
      "epoch": 0.5090832830891027,
      "grad_norm": 0.630395770072937,
      "learning_rate": 0.00014725207877277897,
      "loss": 3.2596,
      "step": 38840
    },
    {
      "epoch": 0.5092143549951503,
      "grad_norm": 0.5740984678268433,
      "learning_rate": 0.00014718991580286552,
      "loss": 3.2388,
      "step": 38850
    },
    {
      "epoch": 0.509345426901198,
      "grad_norm": 0.6189412474632263,
      "learning_rate": 0.00014712775331573232,
      "loss": 3.2032,
      "step": 38860
    },
    {
      "epoch": 0.5094764988072457,
      "grad_norm": 0.6192337870597839,
      "learning_rate": 0.0001470655913220589,
      "loss": 3.2434,
      "step": 38870
    },
    {
      "epoch": 0.5096075707132933,
      "grad_norm": 0.6012256145477295,
      "learning_rate": 0.000147003429832525,
      "loss": 3.2989,
      "step": 38880
    },
    {
      "epoch": 0.509738642619341,
      "grad_norm": 0.6251506805419922,
      "learning_rate": 0.00014694126885781,
      "loss": 3.2844,
      "step": 38890
    },
    {
      "epoch": 0.5098697145253887,
      "grad_norm": 0.5768140554428101,
      "learning_rate": 0.0001468791084085934,
      "loss": 3.2901,
      "step": 38900
    },
    {
      "epoch": 0.5100007864314363,
      "grad_norm": 0.6365700960159302,
      "learning_rate": 0.0001468169484955545,
      "loss": 3.2961,
      "step": 38910
    },
    {
      "epoch": 0.510131858337484,
      "grad_norm": 0.5880820155143738,
      "learning_rate": 0.00014675478912937266,
      "loss": 3.2525,
      "step": 38920
    },
    {
      "epoch": 0.5102629302435316,
      "grad_norm": 0.6588606238365173,
      "learning_rate": 0.00014669263032072687,
      "loss": 3.2971,
      "step": 38930
    },
    {
      "epoch": 0.5103940021495793,
      "grad_norm": 0.5971477031707764,
      "learning_rate": 0.0001466304720802963,
      "loss": 3.2751,
      "step": 38940
    },
    {
      "epoch": 0.510525074055627,
      "grad_norm": 0.6300888061523438,
      "learning_rate": 0.00014656831441875977,
      "loss": 3.2792,
      "step": 38950
    },
    {
      "epoch": 0.5106561459616745,
      "grad_norm": 0.6269235014915466,
      "learning_rate": 0.00014650615734679631,
      "loss": 3.2063,
      "step": 38960
    },
    {
      "epoch": 0.5107872178677222,
      "grad_norm": 0.8022274374961853,
      "learning_rate": 0.00014644400087508453,
      "loss": 3.2095,
      "step": 38970
    },
    {
      "epoch": 0.5109182897737699,
      "grad_norm": 0.5562174916267395,
      "learning_rate": 0.00014638184501430317,
      "loss": 3.246,
      "step": 38980
    },
    {
      "epoch": 0.5110493616798175,
      "grad_norm": 0.6875340342521667,
      "learning_rate": 0.0001463196897751308,
      "loss": 3.2305,
      "step": 38990
    },
    {
      "epoch": 0.5111804335858652,
      "grad_norm": 0.597612202167511,
      "learning_rate": 0.0001462575351682457,
      "loss": 3.2719,
      "step": 39000
    },
    {
      "epoch": 0.5111804335858652,
      "eval_accuracy": 0.3955200391006843,
      "eval_loss": 3.2611618041992188,
      "eval_runtime": 60.7959,
      "eval_samples_per_second": 32.897,
      "eval_steps_per_second": 0.526,
      "step": 39000
    },
    {
      "epoch": 0.5113115054919128,
      "grad_norm": 0.8473418354988098,
      "learning_rate": 0.00014619538120432636,
      "loss": 3.2849,
      "step": 39010
    },
    {
      "epoch": 0.5114425773979605,
      "grad_norm": 0.6229843497276306,
      "learning_rate": 0.00014613322789405094,
      "loss": 3.2593,
      "step": 39020
    },
    {
      "epoch": 0.5115736493040082,
      "grad_norm": 0.6048614382743835,
      "learning_rate": 0.0001460710752480976,
      "loss": 3.2379,
      "step": 39030
    },
    {
      "epoch": 0.5117047212100558,
      "grad_norm": 0.6123390197753906,
      "learning_rate": 0.00014600892327714426,
      "loss": 3.2464,
      "step": 39040
    },
    {
      "epoch": 0.5118357931161035,
      "grad_norm": 0.6584635972976685,
      "learning_rate": 0.00014594677199186883,
      "loss": 3.2743,
      "step": 39050
    },
    {
      "epoch": 0.5119668650221512,
      "grad_norm": 0.6096651554107666,
      "learning_rate": 0.00014588462140294908,
      "loss": 3.255,
      "step": 39060
    },
    {
      "epoch": 0.5120979369281988,
      "grad_norm": 0.6031484007835388,
      "learning_rate": 0.0001458224715210627,
      "loss": 3.2465,
      "step": 39070
    },
    {
      "epoch": 0.5122290088342465,
      "grad_norm": 0.6681720018386841,
      "learning_rate": 0.0001457603223568871,
      "loss": 3.2535,
      "step": 39080
    },
    {
      "epoch": 0.5123600807402942,
      "grad_norm": 0.6506447196006775,
      "learning_rate": 0.0001456981739210997,
      "loss": 3.3065,
      "step": 39090
    },
    {
      "epoch": 0.5124911526463418,
      "grad_norm": 0.6275690793991089,
      "learning_rate": 0.0001456360262243779,
      "loss": 3.2792,
      "step": 39100
    },
    {
      "epoch": 0.5126222245523895,
      "grad_norm": 0.5496920943260193,
      "learning_rate": 0.00014557387927739866,
      "loss": 3.3102,
      "step": 39110
    },
    {
      "epoch": 0.5127532964584371,
      "grad_norm": 0.6101159453392029,
      "learning_rate": 0.00014551173309083911,
      "loss": 3.2849,
      "step": 39120
    },
    {
      "epoch": 0.5128843683644848,
      "grad_norm": 0.5766636729240417,
      "learning_rate": 0.00014544958767537613,
      "loss": 3.2531,
      "step": 39130
    },
    {
      "epoch": 0.5130154402705324,
      "grad_norm": 0.6894168257713318,
      "learning_rate": 0.0001453874430416865,
      "loss": 3.2637,
      "step": 39140
    },
    {
      "epoch": 0.51314651217658,
      "grad_norm": 0.6624734997749329,
      "learning_rate": 0.00014532529920044672,
      "loss": 3.2815,
      "step": 39150
    },
    {
      "epoch": 0.5132775840826277,
      "grad_norm": 0.5939466953277588,
      "learning_rate": 0.00014526315616233333,
      "loss": 3.2754,
      "step": 39160
    },
    {
      "epoch": 0.5134086559886754,
      "grad_norm": 0.7537423968315125,
      "learning_rate": 0.00014520101393802271,
      "loss": 3.2387,
      "step": 39170
    },
    {
      "epoch": 0.513539727894723,
      "grad_norm": 0.589716911315918,
      "learning_rate": 0.0001451388725381911,
      "loss": 3.2666,
      "step": 39180
    },
    {
      "epoch": 0.5136707998007707,
      "grad_norm": 0.6316945552825928,
      "learning_rate": 0.0001450767319735144,
      "loss": 3.2709,
      "step": 39190
    },
    {
      "epoch": 0.5138018717068183,
      "grad_norm": 0.6056029200553894,
      "learning_rate": 0.00014501459225466868,
      "loss": 3.2256,
      "step": 39200
    },
    {
      "epoch": 0.513932943612866,
      "grad_norm": 0.6052190065383911,
      "learning_rate": 0.00014495245339232973,
      "loss": 3.2917,
      "step": 39210
    },
    {
      "epoch": 0.5140640155189137,
      "grad_norm": 0.5750093460083008,
      "learning_rate": 0.00014489031539717304,
      "loss": 3.2373,
      "step": 39220
    },
    {
      "epoch": 0.5141950874249613,
      "grad_norm": 0.6958158612251282,
      "learning_rate": 0.00014482817827987418,
      "loss": 3.2718,
      "step": 39230
    },
    {
      "epoch": 0.514326159331009,
      "grad_norm": 0.5868870615959167,
      "learning_rate": 0.00014476604205110845,
      "loss": 3.2642,
      "step": 39240
    },
    {
      "epoch": 0.5144572312370567,
      "grad_norm": 0.6224728226661682,
      "learning_rate": 0.00014470390672155107,
      "loss": 3.226,
      "step": 39250
    },
    {
      "epoch": 0.5145883031431043,
      "grad_norm": 0.6068905591964722,
      "learning_rate": 0.000144641772301877,
      "loss": 3.2375,
      "step": 39260
    },
    {
      "epoch": 0.514719375049152,
      "grad_norm": 0.6114330887794495,
      "learning_rate": 0.0001445796388027611,
      "loss": 3.2645,
      "step": 39270
    },
    {
      "epoch": 0.5148504469551997,
      "grad_norm": 0.5785655379295349,
      "learning_rate": 0.0001445175062348781,
      "loss": 3.2775,
      "step": 39280
    },
    {
      "epoch": 0.5149815188612473,
      "grad_norm": 0.5678136348724365,
      "learning_rate": 0.00014445537460890264,
      "loss": 3.2528,
      "step": 39290
    },
    {
      "epoch": 0.515112590767295,
      "grad_norm": 0.6057736873626709,
      "learning_rate": 0.0001443932439355089,
      "loss": 3.236,
      "step": 39300
    },
    {
      "epoch": 0.5152436626733427,
      "grad_norm": 0.5800579786300659,
      "learning_rate": 0.00014433111422537128,
      "loss": 3.2594,
      "step": 39310
    },
    {
      "epoch": 0.5153747345793902,
      "grad_norm": 0.6195803880691528,
      "learning_rate": 0.0001442689854891637,
      "loss": 3.2963,
      "step": 39320
    },
    {
      "epoch": 0.5155058064854379,
      "grad_norm": 0.6012951135635376,
      "learning_rate": 0.0001442068577375602,
      "loss": 3.2372,
      "step": 39330
    },
    {
      "epoch": 0.5156368783914855,
      "grad_norm": 0.6839767098426819,
      "learning_rate": 0.00014414473098123436,
      "loss": 3.2658,
      "step": 39340
    },
    {
      "epoch": 0.5157679502975332,
      "grad_norm": 0.6235365867614746,
      "learning_rate": 0.00014408260523085975,
      "loss": 3.2849,
      "step": 39350
    },
    {
      "epoch": 0.5158990222035809,
      "grad_norm": 0.6644881367683411,
      "learning_rate": 0.00014402048049710987,
      "loss": 3.264,
      "step": 39360
    },
    {
      "epoch": 0.5160300941096285,
      "grad_norm": 0.6473128199577332,
      "learning_rate": 0.00014395835679065772,
      "loss": 3.3558,
      "step": 39370
    },
    {
      "epoch": 0.5161611660156762,
      "grad_norm": 0.6243932843208313,
      "learning_rate": 0.00014389623412217642,
      "loss": 3.254,
      "step": 39380
    },
    {
      "epoch": 0.5162922379217239,
      "grad_norm": 0.6157949566841125,
      "learning_rate": 0.0001438341125023389,
      "loss": 3.2637,
      "step": 39390
    },
    {
      "epoch": 0.5164233098277715,
      "grad_norm": 0.7104302644729614,
      "learning_rate": 0.0001437719919418177,
      "loss": 3.3113,
      "step": 39400
    },
    {
      "epoch": 0.5165543817338192,
      "grad_norm": 0.600967288017273,
      "learning_rate": 0.00014370987245128535,
      "loss": 3.246,
      "step": 39410
    },
    {
      "epoch": 0.5166854536398668,
      "grad_norm": 0.6634412407875061,
      "learning_rate": 0.00014364775404141414,
      "loss": 3.2526,
      "step": 39420
    },
    {
      "epoch": 0.5168165255459145,
      "grad_norm": 0.6410297155380249,
      "learning_rate": 0.00014358563672287614,
      "loss": 3.2577,
      "step": 39430
    },
    {
      "epoch": 0.5169475974519622,
      "grad_norm": 0.70206218957901,
      "learning_rate": 0.0001435235205063434,
      "loss": 3.2659,
      "step": 39440
    },
    {
      "epoch": 0.5170786693580098,
      "grad_norm": 0.6507130861282349,
      "learning_rate": 0.00014346140540248754,
      "loss": 3.2841,
      "step": 39450
    },
    {
      "epoch": 0.5172097412640575,
      "grad_norm": 0.7133651375770569,
      "learning_rate": 0.00014339929142198014,
      "loss": 3.2882,
      "step": 39460
    },
    {
      "epoch": 0.5173408131701052,
      "grad_norm": 0.6466358304023743,
      "learning_rate": 0.00014333717857549258,
      "loss": 3.2523,
      "step": 39470
    },
    {
      "epoch": 0.5174718850761528,
      "grad_norm": 0.5666990876197815,
      "learning_rate": 0.00014327506687369598,
      "loss": 3.2593,
      "step": 39480
    },
    {
      "epoch": 0.5176029569822005,
      "grad_norm": 0.5855979919433594,
      "learning_rate": 0.00014321295632726134,
      "loss": 3.237,
      "step": 39490
    },
    {
      "epoch": 0.517734028888248,
      "grad_norm": 0.5952074527740479,
      "learning_rate": 0.0001431508469468593,
      "loss": 3.2023,
      "step": 39500
    },
    {
      "epoch": 0.517734028888248,
      "eval_accuracy": 0.3955449657869013,
      "eval_loss": 3.2585537433624268,
      "eval_runtime": 56.2396,
      "eval_samples_per_second": 35.562,
      "eval_steps_per_second": 0.569,
      "step": 39500
    },
    {
      "epoch": 0.5178651007942957,
      "grad_norm": 0.621511697769165,
      "learning_rate": 0.00014308873874316068,
      "loss": 3.2313,
      "step": 39510
    },
    {
      "epoch": 0.5179961727003434,
      "grad_norm": 0.6036537289619446,
      "learning_rate": 0.00014302663172683554,
      "loss": 3.2743,
      "step": 39520
    },
    {
      "epoch": 0.518127244606391,
      "grad_norm": 0.6458509564399719,
      "learning_rate": 0.00014296452590855418,
      "loss": 3.306,
      "step": 39530
    },
    {
      "epoch": 0.5182583165124387,
      "grad_norm": 0.8501524925231934,
      "learning_rate": 0.00014290242129898656,
      "loss": 3.2692,
      "step": 39540
    },
    {
      "epoch": 0.5183893884184864,
      "grad_norm": 0.5972505807876587,
      "learning_rate": 0.00014284031790880248,
      "loss": 3.2212,
      "step": 39550
    },
    {
      "epoch": 0.518520460324534,
      "grad_norm": 0.6167404055595398,
      "learning_rate": 0.0001427782157486713,
      "loss": 3.2893,
      "step": 39560
    },
    {
      "epoch": 0.5186515322305817,
      "grad_norm": 0.614136815071106,
      "learning_rate": 0.0001427161148292625,
      "loss": 3.2726,
      "step": 39570
    },
    {
      "epoch": 0.5187826041366294,
      "grad_norm": 0.5966436862945557,
      "learning_rate": 0.0001426540151612451,
      "loss": 3.2443,
      "step": 39580
    },
    {
      "epoch": 0.518913676042677,
      "grad_norm": 0.6172049641609192,
      "learning_rate": 0.00014259191675528802,
      "loss": 3.2957,
      "step": 39590
    },
    {
      "epoch": 0.5190447479487247,
      "grad_norm": 0.6189588308334351,
      "learning_rate": 0.00014252981962205994,
      "loss": 3.2353,
      "step": 39600
    },
    {
      "epoch": 0.5191758198547723,
      "grad_norm": 0.6259549260139465,
      "learning_rate": 0.00014246772377222926,
      "loss": 3.2588,
      "step": 39610
    },
    {
      "epoch": 0.51930689176082,
      "grad_norm": 0.6664966940879822,
      "learning_rate": 0.00014240562921646435,
      "loss": 3.255,
      "step": 39620
    },
    {
      "epoch": 0.5194379636668677,
      "grad_norm": 0.6063348650932312,
      "learning_rate": 0.0001423435359654331,
      "loss": 3.2918,
      "step": 39630
    },
    {
      "epoch": 0.5195690355729153,
      "grad_norm": 0.6203917264938354,
      "learning_rate": 0.00014228144402980334,
      "loss": 3.2516,
      "step": 39640
    },
    {
      "epoch": 0.519700107478963,
      "grad_norm": 0.6124438643455505,
      "learning_rate": 0.00014221935342024266,
      "loss": 3.2744,
      "step": 39650
    },
    {
      "epoch": 0.5198311793850107,
      "grad_norm": 0.7025836706161499,
      "learning_rate": 0.0001421572641474184,
      "loss": 3.2185,
      "step": 39660
    },
    {
      "epoch": 0.5199622512910582,
      "grad_norm": 0.5851163864135742,
      "learning_rate": 0.00014209517622199762,
      "loss": 3.2494,
      "step": 39670
    },
    {
      "epoch": 0.5200933231971059,
      "grad_norm": 0.6389443874359131,
      "learning_rate": 0.0001420330896546472,
      "loss": 3.2351,
      "step": 39680
    },
    {
      "epoch": 0.5202243951031535,
      "grad_norm": 0.6480221152305603,
      "learning_rate": 0.00014197100445603378,
      "loss": 3.2582,
      "step": 39690
    },
    {
      "epoch": 0.5203554670092012,
      "grad_norm": 0.6606339812278748,
      "learning_rate": 0.00014190892063682388,
      "loss": 3.2703,
      "step": 39700
    },
    {
      "epoch": 0.5204865389152489,
      "grad_norm": 0.6400294303894043,
      "learning_rate": 0.0001418468382076835,
      "loss": 3.2256,
      "step": 39710
    },
    {
      "epoch": 0.5206176108212965,
      "grad_norm": 0.5747137665748596,
      "learning_rate": 0.00014178475717927864,
      "loss": 3.3067,
      "step": 39720
    },
    {
      "epoch": 0.5207486827273442,
      "grad_norm": 0.5531978011131287,
      "learning_rate": 0.0001417226775622751,
      "loss": 3.2612,
      "step": 39730
    },
    {
      "epoch": 0.5208797546333919,
      "grad_norm": 0.7633252739906311,
      "learning_rate": 0.00014166059936733815,
      "loss": 3.2699,
      "step": 39740
    },
    {
      "epoch": 0.5210108265394395,
      "grad_norm": 0.6878849864006042,
      "learning_rate": 0.00014159852260513307,
      "loss": 3.3033,
      "step": 39750
    },
    {
      "epoch": 0.5211418984454872,
      "grad_norm": 0.6514809727668762,
      "learning_rate": 0.00014153644728632484,
      "loss": 3.2325,
      "step": 39760
    },
    {
      "epoch": 0.5212729703515349,
      "grad_norm": 0.5714231729507446,
      "learning_rate": 0.00014147437342157819,
      "loss": 3.25,
      "step": 39770
    },
    {
      "epoch": 0.5214040422575825,
      "grad_norm": 0.586419403553009,
      "learning_rate": 0.0001414123010215575,
      "loss": 3.2799,
      "step": 39780
    },
    {
      "epoch": 0.5215351141636302,
      "grad_norm": 0.5830458402633667,
      "learning_rate": 0.000141350230096927,
      "loss": 3.2434,
      "step": 39790
    },
    {
      "epoch": 0.5216661860696779,
      "grad_norm": 0.6032927632331848,
      "learning_rate": 0.00014128816065835066,
      "loss": 3.2742,
      "step": 39800
    },
    {
      "epoch": 0.5217972579757255,
      "grad_norm": 0.6430414915084839,
      "learning_rate": 0.00014122609271649227,
      "loss": 3.3224,
      "step": 39810
    },
    {
      "epoch": 0.5219283298817732,
      "grad_norm": 0.5808059573173523,
      "learning_rate": 0.0001411640262820151,
      "loss": 3.2451,
      "step": 39820
    },
    {
      "epoch": 0.5220594017878208,
      "grad_norm": 0.6092066168785095,
      "learning_rate": 0.00014110196136558241,
      "loss": 3.2353,
      "step": 39830
    },
    {
      "epoch": 0.5221904736938685,
      "grad_norm": 0.5547782778739929,
      "learning_rate": 0.00014103989797785722,
      "loss": 3.2464,
      "step": 39840
    },
    {
      "epoch": 0.5223215455999161,
      "grad_norm": 0.5687118172645569,
      "learning_rate": 0.0001409778361295021,
      "loss": 3.2595,
      "step": 39850
    },
    {
      "epoch": 0.5224526175059637,
      "grad_norm": 0.6073830127716064,
      "learning_rate": 0.00014091577583117936,
      "loss": 3.258,
      "step": 39860
    },
    {
      "epoch": 0.5225836894120114,
      "grad_norm": 0.5705782175064087,
      "learning_rate": 0.00014085371709355123,
      "loss": 3.288,
      "step": 39870
    },
    {
      "epoch": 0.522714761318059,
      "grad_norm": 0.5984293222427368,
      "learning_rate": 0.00014079165992727963,
      "loss": 3.2996,
      "step": 39880
    },
    {
      "epoch": 0.5228458332241067,
      "grad_norm": 0.7587903738021851,
      "learning_rate": 0.00014072960434302604,
      "loss": 3.2144,
      "step": 39890
    },
    {
      "epoch": 0.5229769051301544,
      "grad_norm": 0.7040389776229858,
      "learning_rate": 0.00014066755035145178,
      "loss": 3.2738,
      "step": 39900
    },
    {
      "epoch": 0.523107977036202,
      "grad_norm": 0.6026336550712585,
      "learning_rate": 0.00014060549796321794,
      "loss": 3.2861,
      "step": 39910
    },
    {
      "epoch": 0.5232390489422497,
      "grad_norm": 0.6209523677825928,
      "learning_rate": 0.00014054344718898539,
      "loss": 3.3001,
      "step": 39920
    },
    {
      "epoch": 0.5233701208482974,
      "grad_norm": 0.5972778797149658,
      "learning_rate": 0.00014048139803941442,
      "loss": 3.2758,
      "step": 39930
    },
    {
      "epoch": 0.523501192754345,
      "grad_norm": 0.8541239500045776,
      "learning_rate": 0.00014041935052516537,
      "loss": 3.2436,
      "step": 39940
    },
    {
      "epoch": 0.5236322646603927,
      "grad_norm": 0.6476089954376221,
      "learning_rate": 0.0001403573046568981,
      "loss": 3.297,
      "step": 39950
    },
    {
      "epoch": 0.5237633365664404,
      "grad_norm": 0.5969424247741699,
      "learning_rate": 0.00014029526044527242,
      "loss": 3.2458,
      "step": 39960
    },
    {
      "epoch": 0.523894408472488,
      "grad_norm": 0.6824296712875366,
      "learning_rate": 0.00014023321790094747,
      "loss": 3.2765,
      "step": 39970
    },
    {
      "epoch": 0.5240254803785357,
      "grad_norm": 0.6093080043792725,
      "learning_rate": 0.00014017117703458247,
      "loss": 3.2435,
      "step": 39980
    },
    {
      "epoch": 0.5241565522845834,
      "grad_norm": 0.6269532442092896,
      "learning_rate": 0.0001401091378568362,
      "loss": 3.2801,
      "step": 39990
    },
    {
      "epoch": 0.524287624190631,
      "grad_norm": 0.6195786595344543,
      "learning_rate": 0.00014004710037836713,
      "loss": 3.2521,
      "step": 40000
    },
    {
      "epoch": 0.524287624190631,
      "eval_accuracy": 0.3957590420332356,
      "eval_loss": 3.2567973136901855,
      "eval_runtime": 59.4889,
      "eval_samples_per_second": 33.62,
      "eval_steps_per_second": 0.538,
      "step": 40000
    },
    {
      "epoch": 0.5244186960966787,
      "grad_norm": 0.6164523363113403,
      "learning_rate": 0.00013998506460983346,
      "loss": 3.2764,
      "step": 40010
    },
    {
      "epoch": 0.5245497680027263,
      "grad_norm": 0.5962275266647339,
      "learning_rate": 0.00013992303056189315,
      "loss": 3.2653,
      "step": 40020
    },
    {
      "epoch": 0.5246808399087739,
      "grad_norm": 0.5822033286094666,
      "learning_rate": 0.00013986099824520382,
      "loss": 3.2522,
      "step": 40030
    },
    {
      "epoch": 0.5248119118148216,
      "grad_norm": 0.793688952922821,
      "learning_rate": 0.00013979896767042272,
      "loss": 3.2884,
      "step": 40040
    },
    {
      "epoch": 0.5249429837208692,
      "grad_norm": 0.5755127668380737,
      "learning_rate": 0.00013973693884820692,
      "loss": 3.262,
      "step": 40050
    },
    {
      "epoch": 0.5250740556269169,
      "grad_norm": 0.6299104690551758,
      "learning_rate": 0.00013967491178921315,
      "loss": 3.2532,
      "step": 40060
    },
    {
      "epoch": 0.5252051275329646,
      "grad_norm": 0.6203197240829468,
      "learning_rate": 0.00013961288650409784,
      "loss": 3.2348,
      "step": 40070
    },
    {
      "epoch": 0.5253361994390122,
      "grad_norm": 0.7371894121170044,
      "learning_rate": 0.00013955086300351707,
      "loss": 3.3056,
      "step": 40080
    },
    {
      "epoch": 0.5254672713450599,
      "grad_norm": 0.5978001356124878,
      "learning_rate": 0.00013948884129812663,
      "loss": 3.2753,
      "step": 40090
    },
    {
      "epoch": 0.5255983432511075,
      "grad_norm": 0.601982831954956,
      "learning_rate": 0.0001394268213985821,
      "loss": 3.2318,
      "step": 40100
    },
    {
      "epoch": 0.5257294151571552,
      "grad_norm": 0.6611925959587097,
      "learning_rate": 0.00013936480331553858,
      "loss": 3.2985,
      "step": 40110
    },
    {
      "epoch": 0.5258604870632029,
      "grad_norm": 0.6053532361984253,
      "learning_rate": 0.000139302787059651,
      "loss": 3.2817,
      "step": 40120
    },
    {
      "epoch": 0.5259915589692505,
      "grad_norm": 0.5689846277236938,
      "learning_rate": 0.0001392407726415739,
      "loss": 3.2456,
      "step": 40130
    },
    {
      "epoch": 0.5261226308752982,
      "grad_norm": 0.6141642928123474,
      "learning_rate": 0.00013917876007196156,
      "loss": 3.2658,
      "step": 40140
    },
    {
      "epoch": 0.5262537027813459,
      "grad_norm": 0.6064952611923218,
      "learning_rate": 0.00013911674936146784,
      "loss": 3.2268,
      "step": 40150
    },
    {
      "epoch": 0.5263847746873935,
      "grad_norm": 0.6465510129928589,
      "learning_rate": 0.00013905474052074642,
      "loss": 3.2459,
      "step": 40160
    },
    {
      "epoch": 0.5265158465934412,
      "grad_norm": 0.6389144062995911,
      "learning_rate": 0.00013899273356045053,
      "loss": 3.2549,
      "step": 40170
    },
    {
      "epoch": 0.5266469184994889,
      "grad_norm": 0.5890321731567383,
      "learning_rate": 0.00013893072849123324,
      "loss": 3.235,
      "step": 40180
    },
    {
      "epoch": 0.5267779904055365,
      "grad_norm": 0.617202877998352,
      "learning_rate": 0.00013886872532374705,
      "loss": 3.2611,
      "step": 40190
    },
    {
      "epoch": 0.5269090623115842,
      "grad_norm": 0.6476760506629944,
      "learning_rate": 0.00013880672406864437,
      "loss": 3.2643,
      "step": 40200
    },
    {
      "epoch": 0.5270401342176317,
      "grad_norm": 0.6779279708862305,
      "learning_rate": 0.00013874472473657717,
      "loss": 3.2749,
      "step": 40210
    },
    {
      "epoch": 0.5271712061236794,
      "grad_norm": 0.6242055892944336,
      "learning_rate": 0.00013868272733819707,
      "loss": 3.2753,
      "step": 40220
    },
    {
      "epoch": 0.5273022780297271,
      "grad_norm": 0.5542540550231934,
      "learning_rate": 0.0001386207318841554,
      "loss": 3.2508,
      "step": 40230
    },
    {
      "epoch": 0.5274333499357747,
      "grad_norm": 0.5979219079017639,
      "learning_rate": 0.00013855873838510317,
      "loss": 3.2338,
      "step": 40240
    },
    {
      "epoch": 0.5275644218418224,
      "grad_norm": 0.6378568410873413,
      "learning_rate": 0.00013849674685169108,
      "loss": 3.2782,
      "step": 40250
    },
    {
      "epoch": 0.5276954937478701,
      "grad_norm": 0.6537618637084961,
      "learning_rate": 0.00013843475729456933,
      "loss": 3.2173,
      "step": 40260
    },
    {
      "epoch": 0.5278265656539177,
      "grad_norm": 0.6029495596885681,
      "learning_rate": 0.00013837276972438798,
      "loss": 3.2358,
      "step": 40270
    },
    {
      "epoch": 0.5279576375599654,
      "grad_norm": 0.6118044853210449,
      "learning_rate": 0.00013831078415179666,
      "loss": 3.2524,
      "step": 40280
    },
    {
      "epoch": 0.528088709466013,
      "grad_norm": 0.9317169189453125,
      "learning_rate": 0.00013824880058744464,
      "loss": 3.2589,
      "step": 40290
    },
    {
      "epoch": 0.5282197813720607,
      "grad_norm": 0.7903556823730469,
      "learning_rate": 0.00013818681904198088,
      "loss": 3.2412,
      "step": 40300
    },
    {
      "epoch": 0.5283508532781084,
      "grad_norm": 0.5778988003730774,
      "learning_rate": 0.00013812483952605392,
      "loss": 3.2692,
      "step": 40310
    },
    {
      "epoch": 0.528481925184156,
      "grad_norm": 0.778652012348175,
      "learning_rate": 0.0001380628620503121,
      "loss": 3.3096,
      "step": 40320
    },
    {
      "epoch": 0.5286129970902037,
      "grad_norm": 0.6303300857543945,
      "learning_rate": 0.0001380008866254033,
      "loss": 3.2146,
      "step": 40330
    },
    {
      "epoch": 0.5287440689962514,
      "grad_norm": 0.6045466661453247,
      "learning_rate": 0.000137938913261975,
      "loss": 3.2419,
      "step": 40340
    },
    {
      "epoch": 0.528875140902299,
      "grad_norm": 0.6248611807823181,
      "learning_rate": 0.00013787694197067441,
      "loss": 3.3101,
      "step": 40350
    },
    {
      "epoch": 0.5290062128083467,
      "grad_norm": 0.5898770689964294,
      "learning_rate": 0.00013781497276214852,
      "loss": 3.2639,
      "step": 40360
    },
    {
      "epoch": 0.5291372847143944,
      "grad_norm": 0.6508750915527344,
      "learning_rate": 0.0001377530056470436,
      "loss": 3.2498,
      "step": 40370
    },
    {
      "epoch": 0.529268356620442,
      "grad_norm": 0.5999598503112793,
      "learning_rate": 0.0001376910406360059,
      "loss": 3.2282,
      "step": 40380
    },
    {
      "epoch": 0.5293994285264896,
      "grad_norm": 0.6038385033607483,
      "learning_rate": 0.0001376290777396811,
      "loss": 3.2593,
      "step": 40390
    },
    {
      "epoch": 0.5295305004325372,
      "grad_norm": 0.6478764414787292,
      "learning_rate": 0.0001375671169687147,
      "loss": 3.2217,
      "step": 40400
    },
    {
      "epoch": 0.5296615723385849,
      "grad_norm": 0.6402465105056763,
      "learning_rate": 0.00013750515833375158,
      "loss": 3.3028,
      "step": 40410
    },
    {
      "epoch": 0.5297926442446326,
      "grad_norm": 0.6059333086013794,
      "learning_rate": 0.0001374432018454365,
      "loss": 3.2761,
      "step": 40420
    },
    {
      "epoch": 0.5299237161506802,
      "grad_norm": 0.6214873194694519,
      "learning_rate": 0.00013738124751441373,
      "loss": 3.2718,
      "step": 40430
    },
    {
      "epoch": 0.5300547880567279,
      "grad_norm": 0.6892498135566711,
      "learning_rate": 0.00013731929535132728,
      "loss": 3.2699,
      "step": 40440
    },
    {
      "epoch": 0.5301858599627756,
      "grad_norm": 0.6364609599113464,
      "learning_rate": 0.00013725734536682057,
      "loss": 3.2962,
      "step": 40450
    },
    {
      "epoch": 0.5303169318688232,
      "grad_norm": 0.6636871695518494,
      "learning_rate": 0.00013719539757153684,
      "loss": 3.2144,
      "step": 40460
    },
    {
      "epoch": 0.5304480037748709,
      "grad_norm": 0.6393410563468933,
      "learning_rate": 0.00013713345197611892,
      "loss": 3.2588,
      "step": 40470
    },
    {
      "epoch": 0.5305790756809186,
      "grad_norm": 0.5930665731430054,
      "learning_rate": 0.00013707150859120916,
      "loss": 3.2865,
      "step": 40480
    },
    {
      "epoch": 0.5307101475869662,
      "grad_norm": 0.643663763999939,
      "learning_rate": 0.00013700956742744962,
      "loss": 3.268,
      "step": 40490
    },
    {
      "epoch": 0.5308412194930139,
      "grad_norm": 0.5799185633659363,
      "learning_rate": 0.000136947628495482,
      "loss": 3.2427,
      "step": 40500
    },
    {
      "epoch": 0.5308412194930139,
      "eval_accuracy": 0.3962761485826002,
      "eval_loss": 3.2536983489990234,
      "eval_runtime": 58.6639,
      "eval_samples_per_second": 34.092,
      "eval_steps_per_second": 0.545,
      "step": 40500
    },
    {
      "epoch": 0.5309722913990615,
      "grad_norm": 0.6084781289100647,
      "learning_rate": 0.0001368856918059476,
      "loss": 3.2341,
      "step": 40510
    },
    {
      "epoch": 0.5311033633051092,
      "grad_norm": 0.6013733148574829,
      "learning_rate": 0.00013682375736948722,
      "loss": 3.2695,
      "step": 40520
    },
    {
      "epoch": 0.5312344352111569,
      "grad_norm": 0.6095032691955566,
      "learning_rate": 0.00013676182519674142,
      "loss": 3.2725,
      "step": 40530
    },
    {
      "epoch": 0.5313655071172045,
      "grad_norm": 0.6071112751960754,
      "learning_rate": 0.00013669989529835032,
      "loss": 3.2689,
      "step": 40540
    },
    {
      "epoch": 0.5314965790232522,
      "grad_norm": 0.592881977558136,
      "learning_rate": 0.0001366379676849537,
      "loss": 3.2315,
      "step": 40550
    },
    {
      "epoch": 0.5316276509292999,
      "grad_norm": 0.633018434047699,
      "learning_rate": 0.00013657604236719077,
      "loss": 3.2559,
      "step": 40560
    },
    {
      "epoch": 0.5317587228353474,
      "grad_norm": 0.6054872870445251,
      "learning_rate": 0.00013651411935570058,
      "loss": 3.2625,
      "step": 40570
    },
    {
      "epoch": 0.5318897947413951,
      "grad_norm": 0.6395840644836426,
      "learning_rate": 0.00013645219866112158,
      "loss": 3.2727,
      "step": 40580
    },
    {
      "epoch": 0.5320208666474427,
      "grad_norm": 0.6797610521316528,
      "learning_rate": 0.00013639028029409201,
      "loss": 3.2433,
      "step": 40590
    },
    {
      "epoch": 0.5321519385534904,
      "grad_norm": 0.73389732837677,
      "learning_rate": 0.00013632836426524954,
      "loss": 3.2751,
      "step": 40600
    },
    {
      "epoch": 0.5322830104595381,
      "grad_norm": 0.5791915059089661,
      "learning_rate": 0.00013626645058523155,
      "loss": 3.2627,
      "step": 40610
    },
    {
      "epoch": 0.5324140823655857,
      "grad_norm": 0.6207979917526245,
      "learning_rate": 0.000136204539264675,
      "loss": 3.2775,
      "step": 40620
    },
    {
      "epoch": 0.5325451542716334,
      "grad_norm": 0.6308601498603821,
      "learning_rate": 0.00013614263031421633,
      "loss": 3.2365,
      "step": 40630
    },
    {
      "epoch": 0.5326762261776811,
      "grad_norm": 0.6045910716056824,
      "learning_rate": 0.00013608072374449176,
      "loss": 3.27,
      "step": 40640
    },
    {
      "epoch": 0.5328072980837287,
      "grad_norm": 0.7687103748321533,
      "learning_rate": 0.000136018819566137,
      "loss": 3.3351,
      "step": 40650
    },
    {
      "epoch": 0.5329383699897764,
      "grad_norm": 0.6634029746055603,
      "learning_rate": 0.00013595691778978732,
      "loss": 3.2709,
      "step": 40660
    },
    {
      "epoch": 0.5330694418958241,
      "grad_norm": 0.637937068939209,
      "learning_rate": 0.00013589501842607767,
      "loss": 3.2687,
      "step": 40670
    },
    {
      "epoch": 0.5332005138018717,
      "grad_norm": 0.6726875901222229,
      "learning_rate": 0.00013583312148564243,
      "loss": 3.2554,
      "step": 40680
    },
    {
      "epoch": 0.5333315857079194,
      "grad_norm": 0.5952921509742737,
      "learning_rate": 0.00013577122697911576,
      "loss": 3.2241,
      "step": 40690
    },
    {
      "epoch": 0.533462657613967,
      "grad_norm": 0.6052809357643127,
      "learning_rate": 0.0001357093349171313,
      "loss": 3.2632,
      "step": 40700
    },
    {
      "epoch": 0.5335937295200147,
      "grad_norm": 0.9151293039321899,
      "learning_rate": 0.0001356474453103222,
      "loss": 3.3075,
      "step": 40710
    },
    {
      "epoch": 0.5337248014260624,
      "grad_norm": 0.6131239533424377,
      "learning_rate": 0.00013558555816932133,
      "loss": 3.2544,
      "step": 40720
    },
    {
      "epoch": 0.53385587333211,
      "grad_norm": 0.6432704925537109,
      "learning_rate": 0.0001355236735047611,
      "loss": 3.2325,
      "step": 40730
    },
    {
      "epoch": 0.5339869452381577,
      "grad_norm": 0.6007055044174194,
      "learning_rate": 0.00013546179132727334,
      "loss": 3.2513,
      "step": 40740
    },
    {
      "epoch": 0.5341180171442053,
      "grad_norm": 0.660437822341919,
      "learning_rate": 0.00013539991164748973,
      "loss": 3.2108,
      "step": 40750
    },
    {
      "epoch": 0.5342490890502529,
      "grad_norm": 0.6928515434265137,
      "learning_rate": 0.0001353380344760412,
      "loss": 3.2365,
      "step": 40760
    },
    {
      "epoch": 0.5343801609563006,
      "grad_norm": 0.6171672940254211,
      "learning_rate": 0.00013527615982355864,
      "loss": 3.211,
      "step": 40770
    },
    {
      "epoch": 0.5345112328623483,
      "grad_norm": 0.6461727619171143,
      "learning_rate": 0.00013521428770067205,
      "loss": 3.2479,
      "step": 40780
    },
    {
      "epoch": 0.5346423047683959,
      "grad_norm": 0.6844959259033203,
      "learning_rate": 0.00013515241811801136,
      "loss": 3.2756,
      "step": 40790
    },
    {
      "epoch": 0.5347733766744436,
      "grad_norm": 0.6081936955451965,
      "learning_rate": 0.0001350905510862059,
      "loss": 3.25,
      "step": 40800
    },
    {
      "epoch": 0.5349044485804912,
      "grad_norm": 0.6095540523529053,
      "learning_rate": 0.00013502868661588468,
      "loss": 3.2247,
      "step": 40810
    },
    {
      "epoch": 0.5350355204865389,
      "grad_norm": 0.572963535785675,
      "learning_rate": 0.00013496682471767605,
      "loss": 3.255,
      "step": 40820
    },
    {
      "epoch": 0.5351665923925866,
      "grad_norm": 0.5922354459762573,
      "learning_rate": 0.0001349049654022081,
      "loss": 3.2567,
      "step": 40830
    },
    {
      "epoch": 0.5352976642986342,
      "grad_norm": 0.828615128993988,
      "learning_rate": 0.0001348431086801085,
      "loss": 3.2163,
      "step": 40840
    },
    {
      "epoch": 0.5354287362046819,
      "grad_norm": 0.6285271048545837,
      "learning_rate": 0.00013478125456200434,
      "loss": 3.2259,
      "step": 40850
    },
    {
      "epoch": 0.5355598081107296,
      "grad_norm": 0.636303186416626,
      "learning_rate": 0.0001347194030585223,
      "loss": 3.2342,
      "step": 40860
    },
    {
      "epoch": 0.5356908800167772,
      "grad_norm": 0.5920599102973938,
      "learning_rate": 0.00013465755418028868,
      "loss": 3.2698,
      "step": 40870
    },
    {
      "epoch": 0.5358219519228249,
      "grad_norm": 0.586889386177063,
      "learning_rate": 0.00013459570793792933,
      "loss": 3.2719,
      "step": 40880
    },
    {
      "epoch": 0.5359530238288726,
      "grad_norm": 0.5995627641677856,
      "learning_rate": 0.0001345338643420695,
      "loss": 3.2865,
      "step": 40890
    },
    {
      "epoch": 0.5360840957349202,
      "grad_norm": 0.6544495820999146,
      "learning_rate": 0.00013447202340333413,
      "loss": 3.2694,
      "step": 40900
    },
    {
      "epoch": 0.5362151676409679,
      "grad_norm": 0.5733885169029236,
      "learning_rate": 0.00013441018513234772,
      "loss": 3.2617,
      "step": 40910
    },
    {
      "epoch": 0.5363462395470155,
      "grad_norm": 0.6325270533561707,
      "learning_rate": 0.00013434834953973422,
      "loss": 3.2293,
      "step": 40920
    },
    {
      "epoch": 0.5364773114530631,
      "grad_norm": 0.5904917120933533,
      "learning_rate": 0.00013428651663611714,
      "loss": 3.2596,
      "step": 40930
    },
    {
      "epoch": 0.5366083833591108,
      "grad_norm": 0.6005881428718567,
      "learning_rate": 0.0001342246864321195,
      "loss": 3.2506,
      "step": 40940
    },
    {
      "epoch": 0.5367394552651584,
      "grad_norm": 0.6770060658454895,
      "learning_rate": 0.00013416285893836394,
      "loss": 3.2681,
      "step": 40950
    },
    {
      "epoch": 0.5368705271712061,
      "grad_norm": 0.6743752360343933,
      "learning_rate": 0.0001341010341654727,
      "loss": 3.2568,
      "step": 40960
    },
    {
      "epoch": 0.5370015990772538,
      "grad_norm": 0.6093729734420776,
      "learning_rate": 0.00013403921212406725,
      "loss": 3.2494,
      "step": 40970
    },
    {
      "epoch": 0.5371326709833014,
      "grad_norm": 0.6392437219619751,
      "learning_rate": 0.00013397739282476888,
      "loss": 3.1921,
      "step": 40980
    },
    {
      "epoch": 0.5372637428893491,
      "grad_norm": 0.6208516359329224,
      "learning_rate": 0.00013391557627819841,
      "loss": 3.248,
      "step": 40990
    },
    {
      "epoch": 0.5373948147953967,
      "grad_norm": 0.5871934294700623,
      "learning_rate": 0.0001338537624949759,
      "loss": 3.2603,
      "step": 41000
    },
    {
      "epoch": 0.5373948147953967,
      "eval_accuracy": 0.3960293255131965,
      "eval_loss": 3.2537145614624023,
      "eval_runtime": 56.128,
      "eval_samples_per_second": 35.633,
      "eval_steps_per_second": 0.57,
      "step": 41000
    },
    {
      "epoch": 0.5375258867014444,
      "grad_norm": 0.5616732835769653,
      "learning_rate": 0.00013379195148572126,
      "loss": 3.2031,
      "step": 41010
    },
    {
      "epoch": 0.5376569586074921,
      "grad_norm": 0.5734233856201172,
      "learning_rate": 0.00013373014326105374,
      "loss": 3.2347,
      "step": 41020
    },
    {
      "epoch": 0.5377880305135397,
      "grad_norm": 0.6649495363235474,
      "learning_rate": 0.00013366833783159222,
      "loss": 3.2098,
      "step": 41030
    },
    {
      "epoch": 0.5379191024195874,
      "grad_norm": 0.6217477917671204,
      "learning_rate": 0.00013360653520795495,
      "loss": 3.2315,
      "step": 41040
    },
    {
      "epoch": 0.5380501743256351,
      "grad_norm": 0.6610840559005737,
      "learning_rate": 0.00013354473540075983,
      "loss": 3.2798,
      "step": 41050
    },
    {
      "epoch": 0.5381812462316827,
      "grad_norm": 0.6237424612045288,
      "learning_rate": 0.00013348293842062426,
      "loss": 3.2744,
      "step": 41060
    },
    {
      "epoch": 0.5383123181377304,
      "grad_norm": 0.6571800112724304,
      "learning_rate": 0.00013342114427816518,
      "loss": 3.2184,
      "step": 41070
    },
    {
      "epoch": 0.5384433900437781,
      "grad_norm": 0.7015886306762695,
      "learning_rate": 0.00013335935298399885,
      "loss": 3.2756,
      "step": 41080
    },
    {
      "epoch": 0.5385744619498257,
      "grad_norm": 0.6043084859848022,
      "learning_rate": 0.0001332975645487413,
      "loss": 3.2363,
      "step": 41090
    },
    {
      "epoch": 0.5387055338558734,
      "grad_norm": 0.5931277871131897,
      "learning_rate": 0.00013323577898300792,
      "loss": 3.2539,
      "step": 41100
    },
    {
      "epoch": 0.5388366057619209,
      "grad_norm": 0.7113502025604248,
      "learning_rate": 0.00013317399629741362,
      "loss": 3.2448,
      "step": 41110
    },
    {
      "epoch": 0.5389676776679686,
      "grad_norm": 0.6668540239334106,
      "learning_rate": 0.00013311221650257282,
      "loss": 3.2802,
      "step": 41120
    },
    {
      "epoch": 0.5390987495740163,
      "grad_norm": 0.6150121688842773,
      "learning_rate": 0.0001330504396090995,
      "loss": 3.2603,
      "step": 41130
    },
    {
      "epoch": 0.5392298214800639,
      "grad_norm": 0.5837573409080505,
      "learning_rate": 0.00013298866562760716,
      "loss": 3.2493,
      "step": 41140
    },
    {
      "epoch": 0.5393608933861116,
      "grad_norm": 1.095885992050171,
      "learning_rate": 0.00013292689456870859,
      "loss": 3.2178,
      "step": 41150
    },
    {
      "epoch": 0.5394919652921593,
      "grad_norm": 0.6420198678970337,
      "learning_rate": 0.0001328651264430163,
      "loss": 3.2715,
      "step": 41160
    },
    {
      "epoch": 0.5396230371982069,
      "grad_norm": 0.5848408341407776,
      "learning_rate": 0.00013280336126114223,
      "loss": 3.263,
      "step": 41170
    },
    {
      "epoch": 0.5397541091042546,
      "grad_norm": 0.6630277037620544,
      "learning_rate": 0.00013274159903369783,
      "loss": 3.2068,
      "step": 41180
    },
    {
      "epoch": 0.5398851810103023,
      "grad_norm": 0.6046363711357117,
      "learning_rate": 0.00013267983977129403,
      "loss": 3.2639,
      "step": 41190
    },
    {
      "epoch": 0.5400162529163499,
      "grad_norm": 0.6718875169754028,
      "learning_rate": 0.00013261808348454112,
      "loss": 3.2555,
      "step": 41200
    },
    {
      "epoch": 0.5401473248223976,
      "grad_norm": 0.6379103660583496,
      "learning_rate": 0.00013255633018404917,
      "loss": 3.2351,
      "step": 41210
    },
    {
      "epoch": 0.5402783967284452,
      "grad_norm": 0.6016038656234741,
      "learning_rate": 0.00013249457988042743,
      "loss": 3.241,
      "step": 41220
    },
    {
      "epoch": 0.5404094686344929,
      "grad_norm": 0.6195015907287598,
      "learning_rate": 0.00013243283258428478,
      "loss": 3.2573,
      "step": 41230
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 0.6217765808105469,
      "learning_rate": 0.00013237108830622962,
      "loss": 3.2962,
      "step": 41240
    },
    {
      "epoch": 0.5406716124465882,
      "grad_norm": 0.9454050064086914,
      "learning_rate": 0.00013230934705686987,
      "loss": 3.2292,
      "step": 41250
    },
    {
      "epoch": 0.5408026843526359,
      "grad_norm": 0.690099835395813,
      "learning_rate": 0.0001322476088468127,
      "loss": 3.2839,
      "step": 41260
    },
    {
      "epoch": 0.5409337562586836,
      "grad_norm": 0.7889778017997742,
      "learning_rate": 0.00013218587368666495,
      "loss": 3.2639,
      "step": 41270
    },
    {
      "epoch": 0.5410648281647312,
      "grad_norm": 0.6488425135612488,
      "learning_rate": 0.0001321241415870329,
      "loss": 3.2742,
      "step": 41280
    },
    {
      "epoch": 0.5411959000707788,
      "grad_norm": 0.6279201507568359,
      "learning_rate": 0.00013206241255852238,
      "loss": 3.2584,
      "step": 41290
    },
    {
      "epoch": 0.5413269719768264,
      "grad_norm": 0.5601656436920166,
      "learning_rate": 0.00013200068661173845,
      "loss": 3.2777,
      "step": 41300
    },
    {
      "epoch": 0.5414580438828741,
      "grad_norm": 0.7063508629798889,
      "learning_rate": 0.00013193896375728582,
      "loss": 3.2648,
      "step": 41310
    },
    {
      "epoch": 0.5415891157889218,
      "grad_norm": 0.5962551832199097,
      "learning_rate": 0.0001318772440057687,
      "loss": 3.2662,
      "step": 41320
    },
    {
      "epoch": 0.5417201876949694,
      "grad_norm": 0.7267482280731201,
      "learning_rate": 0.00013181552736779082,
      "loss": 3.2445,
      "step": 41330
    },
    {
      "epoch": 0.5418512596010171,
      "grad_norm": 0.6155030727386475,
      "learning_rate": 0.00013175381385395506,
      "loss": 3.2613,
      "step": 41340
    },
    {
      "epoch": 0.5419823315070648,
      "grad_norm": 0.5728044509887695,
      "learning_rate": 0.00013169210347486405,
      "loss": 3.2155,
      "step": 41350
    },
    {
      "epoch": 0.5421134034131124,
      "grad_norm": 0.6135300993919373,
      "learning_rate": 0.00013163039624111992,
      "loss": 3.2813,
      "step": 41360
    },
    {
      "epoch": 0.5422444753191601,
      "grad_norm": 0.5919345617294312,
      "learning_rate": 0.00013156869216332393,
      "loss": 3.2435,
      "step": 41370
    },
    {
      "epoch": 0.5423755472252078,
      "grad_norm": 0.5709983110427856,
      "learning_rate": 0.00013150699125207718,
      "loss": 3.2291,
      "step": 41380
    },
    {
      "epoch": 0.5425066191312554,
      "grad_norm": 0.5726591348648071,
      "learning_rate": 0.00013144529351797994,
      "loss": 3.2615,
      "step": 41390
    },
    {
      "epoch": 0.5426376910373031,
      "grad_norm": 0.6353119015693665,
      "learning_rate": 0.00013138359897163217,
      "loss": 3.2665,
      "step": 41400
    },
    {
      "epoch": 0.5427687629433507,
      "grad_norm": 0.5664894580841064,
      "learning_rate": 0.00013132190762363301,
      "loss": 3.286,
      "step": 41410
    },
    {
      "epoch": 0.5428998348493984,
      "grad_norm": 0.5909487009048462,
      "learning_rate": 0.00013126021948458132,
      "loss": 3.209,
      "step": 41420
    },
    {
      "epoch": 0.5430309067554461,
      "grad_norm": 0.6598386764526367,
      "learning_rate": 0.00013119853456507524,
      "loss": 3.2605,
      "step": 41430
    },
    {
      "epoch": 0.5431619786614937,
      "grad_norm": 0.6545625329017639,
      "learning_rate": 0.0001311368528757125,
      "loss": 3.3414,
      "step": 41440
    },
    {
      "epoch": 0.5432930505675414,
      "grad_norm": 0.6128087043762207,
      "learning_rate": 0.00013107517442709004,
      "loss": 3.227,
      "step": 41450
    },
    {
      "epoch": 0.5434241224735891,
      "grad_norm": 0.6808581948280334,
      "learning_rate": 0.0001310134992298045,
      "loss": 3.3066,
      "step": 41460
    },
    {
      "epoch": 0.5435551943796366,
      "grad_norm": 0.6551191806793213,
      "learning_rate": 0.0001309518272944518,
      "loss": 3.2488,
      "step": 41470
    },
    {
      "epoch": 0.5436862662856843,
      "grad_norm": 0.5872228741645813,
      "learning_rate": 0.00013089015863162735,
      "loss": 3.2618,
      "step": 41480
    },
    {
      "epoch": 0.543817338191732,
      "grad_norm": 0.6446829438209534,
      "learning_rate": 0.00013082849325192595,
      "loss": 3.238,
      "step": 41490
    },
    {
      "epoch": 0.5439484100977796,
      "grad_norm": 0.6068676710128784,
      "learning_rate": 0.00013076683116594194,
      "loss": 3.2323,
      "step": 41500
    },
    {
      "epoch": 0.5439484100977796,
      "eval_accuracy": 0.39682551319648096,
      "eval_loss": 3.2515103816986084,
      "eval_runtime": 57.462,
      "eval_samples_per_second": 34.806,
      "eval_steps_per_second": 0.557,
      "step": 41500
    },
    {
      "epoch": 0.5440794820038273,
      "grad_norm": 0.5923199653625488,
      "learning_rate": 0.00013070517238426908,
      "loss": 3.2745,
      "step": 41510
    },
    {
      "epoch": 0.5442105539098749,
      "grad_norm": 0.5671214461326599,
      "learning_rate": 0.00013064351691750042,
      "loss": 3.2255,
      "step": 41520
    },
    {
      "epoch": 0.5443416258159226,
      "grad_norm": 0.6648491024971008,
      "learning_rate": 0.00013058186477622855,
      "loss": 3.2678,
      "step": 41530
    },
    {
      "epoch": 0.5444726977219703,
      "grad_norm": 0.6379991769790649,
      "learning_rate": 0.00013052021597104556,
      "loss": 3.2649,
      "step": 41540
    },
    {
      "epoch": 0.5446037696280179,
      "grad_norm": 0.5739851593971252,
      "learning_rate": 0.0001304585705125428,
      "loss": 3.2544,
      "step": 41550
    },
    {
      "epoch": 0.5447348415340656,
      "grad_norm": 0.6842323541641235,
      "learning_rate": 0.0001303969284113112,
      "loss": 3.288,
      "step": 41560
    },
    {
      "epoch": 0.5448659134401133,
      "grad_norm": 0.6580625176429749,
      "learning_rate": 0.00013033528967794093,
      "loss": 3.2571,
      "step": 41570
    },
    {
      "epoch": 0.5449969853461609,
      "grad_norm": 0.597370982170105,
      "learning_rate": 0.00013027365432302177,
      "loss": 3.2683,
      "step": 41580
    },
    {
      "epoch": 0.5451280572522086,
      "grad_norm": 0.6272065043449402,
      "learning_rate": 0.0001302120223571429,
      "loss": 3.2457,
      "step": 41590
    },
    {
      "epoch": 0.5452591291582563,
      "grad_norm": 0.8991113901138306,
      "learning_rate": 0.00013015039379089274,
      "loss": 3.2931,
      "step": 41600
    },
    {
      "epoch": 0.5453902010643039,
      "grad_norm": 0.6045796275138855,
      "learning_rate": 0.00013008876863485928,
      "loss": 3.2683,
      "step": 41610
    },
    {
      "epoch": 0.5455212729703516,
      "grad_norm": 0.6040561199188232,
      "learning_rate": 0.00013002714689963,
      "loss": 3.2263,
      "step": 41620
    },
    {
      "epoch": 0.5456523448763992,
      "grad_norm": 0.6558725833892822,
      "learning_rate": 0.00012996552859579155,
      "loss": 3.2772,
      "step": 41630
    },
    {
      "epoch": 0.5457834167824468,
      "grad_norm": 0.7071288824081421,
      "learning_rate": 0.00012990391373393016,
      "loss": 3.2722,
      "step": 41640
    },
    {
      "epoch": 0.5459144886884945,
      "grad_norm": 0.627366304397583,
      "learning_rate": 0.00012984230232463144,
      "loss": 3.2763,
      "step": 41650
    },
    {
      "epoch": 0.5460455605945421,
      "grad_norm": 0.5999047756195068,
      "learning_rate": 0.00012978069437848044,
      "loss": 3.2716,
      "step": 41660
    },
    {
      "epoch": 0.5461766325005898,
      "grad_norm": 0.6636313796043396,
      "learning_rate": 0.00012971908990606147,
      "loss": 3.2863,
      "step": 41670
    },
    {
      "epoch": 0.5463077044066375,
      "grad_norm": 0.6659784317016602,
      "learning_rate": 0.00012965748891795846,
      "loss": 3.2136,
      "step": 41680
    },
    {
      "epoch": 0.5464387763126851,
      "grad_norm": 0.6297033429145813,
      "learning_rate": 0.00012959589142475452,
      "loss": 3.252,
      "step": 41690
    },
    {
      "epoch": 0.5465698482187328,
      "grad_norm": 0.6677568554878235,
      "learning_rate": 0.0001295342974370324,
      "loss": 3.2517,
      "step": 41700
    },
    {
      "epoch": 0.5467009201247804,
      "grad_norm": 0.6205234527587891,
      "learning_rate": 0.00012947270696537402,
      "loss": 3.214,
      "step": 41710
    },
    {
      "epoch": 0.5468319920308281,
      "grad_norm": 0.6269914507865906,
      "learning_rate": 0.00012941112002036082,
      "loss": 3.2499,
      "step": 41720
    },
    {
      "epoch": 0.5469630639368758,
      "grad_norm": 0.6594983339309692,
      "learning_rate": 0.00012934953661257361,
      "loss": 3.2511,
      "step": 41730
    },
    {
      "epoch": 0.5470941358429234,
      "grad_norm": 0.6539808511734009,
      "learning_rate": 0.00012928795675259258,
      "loss": 3.2599,
      "step": 41740
    },
    {
      "epoch": 0.5472252077489711,
      "grad_norm": 0.5673391819000244,
      "learning_rate": 0.00012922638045099728,
      "loss": 3.2117,
      "step": 41750
    },
    {
      "epoch": 0.5473562796550188,
      "grad_norm": 0.5948545932769775,
      "learning_rate": 0.00012916480771836674,
      "loss": 3.2719,
      "step": 41760
    },
    {
      "epoch": 0.5474873515610664,
      "grad_norm": 0.6278615593910217,
      "learning_rate": 0.00012910323856527933,
      "loss": 3.2563,
      "step": 41770
    },
    {
      "epoch": 0.5476184234671141,
      "grad_norm": 0.5801980495452881,
      "learning_rate": 0.00012904167300231277,
      "loss": 3.2512,
      "step": 41780
    },
    {
      "epoch": 0.5477494953731618,
      "grad_norm": 0.6033198237419128,
      "learning_rate": 0.00012898011104004419,
      "loss": 3.1749,
      "step": 41790
    },
    {
      "epoch": 0.5478805672792094,
      "grad_norm": 0.5982435345649719,
      "learning_rate": 0.00012891855268905013,
      "loss": 3.2371,
      "step": 41800
    },
    {
      "epoch": 0.5480116391852571,
      "grad_norm": 0.570986807346344,
      "learning_rate": 0.00012885699795990647,
      "loss": 3.2663,
      "step": 41810
    },
    {
      "epoch": 0.5481427110913046,
      "grad_norm": 0.6369062066078186,
      "learning_rate": 0.00012879544686318848,
      "loss": 3.2231,
      "step": 41820
    },
    {
      "epoch": 0.5482737829973523,
      "grad_norm": 0.589508593082428,
      "learning_rate": 0.00012873389940947077,
      "loss": 3.3175,
      "step": 41830
    },
    {
      "epoch": 0.5484048549034,
      "grad_norm": 0.6211560964584351,
      "learning_rate": 0.00012867235560932747,
      "loss": 3.2403,
      "step": 41840
    },
    {
      "epoch": 0.5485359268094476,
      "grad_norm": 0.6722044944763184,
      "learning_rate": 0.00012861081547333185,
      "loss": 3.2482,
      "step": 41850
    },
    {
      "epoch": 0.5486669987154953,
      "grad_norm": 0.6856252551078796,
      "learning_rate": 0.0001285492790120567,
      "loss": 3.2227,
      "step": 41860
    },
    {
      "epoch": 0.548798070621543,
      "grad_norm": 0.7625269293785095,
      "learning_rate": 0.0001284877462360742,
      "loss": 3.2196,
      "step": 41870
    },
    {
      "epoch": 0.5489291425275906,
      "grad_norm": 0.6388958692550659,
      "learning_rate": 0.00012842621715595588,
      "loss": 3.247,
      "step": 41880
    },
    {
      "epoch": 0.5490602144336383,
      "grad_norm": 0.6898766160011292,
      "learning_rate": 0.00012836469178227248,
      "loss": 3.2546,
      "step": 41890
    },
    {
      "epoch": 0.549191286339686,
      "grad_norm": 0.8217582106590271,
      "learning_rate": 0.00012830317012559435,
      "loss": 3.3226,
      "step": 41900
    },
    {
      "epoch": 0.5493223582457336,
      "grad_norm": 0.6592479944229126,
      "learning_rate": 0.00012824165219649098,
      "loss": 3.2844,
      "step": 41910
    },
    {
      "epoch": 0.5494534301517813,
      "grad_norm": 0.6289352178573608,
      "learning_rate": 0.00012818013800553144,
      "loss": 3.2938,
      "step": 41920
    },
    {
      "epoch": 0.5495845020578289,
      "grad_norm": 0.5944625735282898,
      "learning_rate": 0.0001281186275632839,
      "loss": 3.2264,
      "step": 41930
    },
    {
      "epoch": 0.5497155739638766,
      "grad_norm": 0.5700408816337585,
      "learning_rate": 0.0001280571208803161,
      "loss": 3.2558,
      "step": 41940
    },
    {
      "epoch": 0.5498466458699243,
      "grad_norm": 0.6599850654602051,
      "learning_rate": 0.000127995617967195,
      "loss": 3.2473,
      "step": 41950
    },
    {
      "epoch": 0.5499777177759719,
      "grad_norm": 0.6289823651313782,
      "learning_rate": 0.0001279341188344871,
      "loss": 3.2391,
      "step": 41960
    },
    {
      "epoch": 0.5501087896820196,
      "grad_norm": 0.6064648628234863,
      "learning_rate": 0.00012787262349275798,
      "loss": 3.2466,
      "step": 41970
    },
    {
      "epoch": 0.5502398615880673,
      "grad_norm": 0.6136929988861084,
      "learning_rate": 0.00012781113195257276,
      "loss": 3.2845,
      "step": 41980
    },
    {
      "epoch": 0.5503709334941149,
      "grad_norm": 0.6335753798484802,
      "learning_rate": 0.00012774964422449588,
      "loss": 3.3567,
      "step": 41990
    },
    {
      "epoch": 0.5505020054001625,
      "grad_norm": 0.7499961853027344,
      "learning_rate": 0.00012768816031909103,
      "loss": 3.2502,
      "step": 42000
    },
    {
      "epoch": 0.5505020054001625,
      "eval_accuracy": 0.39674046920821116,
      "eval_loss": 3.2500317096710205,
      "eval_runtime": 60.3553,
      "eval_samples_per_second": 33.137,
      "eval_steps_per_second": 0.53,
      "step": 42000
    },
    {
      "epoch": 0.5506330773062101,
      "grad_norm": 0.5668957829475403,
      "learning_rate": 0.00012762668024692141,
      "loss": 3.2465,
      "step": 42010
    },
    {
      "epoch": 0.5507641492122578,
      "grad_norm": 0.6025965213775635,
      "learning_rate": 0.00012756520401854937,
      "loss": 3.227,
      "step": 42020
    },
    {
      "epoch": 0.5508952211183055,
      "grad_norm": 0.654167652130127,
      "learning_rate": 0.00012750373164453681,
      "loss": 3.2451,
      "step": 42030
    },
    {
      "epoch": 0.5510262930243531,
      "grad_norm": 0.5426486134529114,
      "learning_rate": 0.00012744226313544468,
      "loss": 3.2952,
      "step": 42040
    },
    {
      "epoch": 0.5511573649304008,
      "grad_norm": 0.5884668231010437,
      "learning_rate": 0.00012738079850183357,
      "loss": 3.2581,
      "step": 42050
    },
    {
      "epoch": 0.5512884368364485,
      "grad_norm": 0.6027165651321411,
      "learning_rate": 0.00012731933775426323,
      "loss": 3.2127,
      "step": 42060
    },
    {
      "epoch": 0.5514195087424961,
      "grad_norm": 0.5953251719474792,
      "learning_rate": 0.0001272578809032929,
      "loss": 3.2426,
      "step": 42070
    },
    {
      "epoch": 0.5515505806485438,
      "grad_norm": 0.582270085811615,
      "learning_rate": 0.0001271964279594808,
      "loss": 3.2416,
      "step": 42080
    },
    {
      "epoch": 0.5516816525545915,
      "grad_norm": 0.6011871099472046,
      "learning_rate": 0.00012713497893338492,
      "loss": 3.2046,
      "step": 42090
    },
    {
      "epoch": 0.5518127244606391,
      "grad_norm": 0.6342535614967346,
      "learning_rate": 0.00012707353383556227,
      "loss": 3.2712,
      "step": 42100
    },
    {
      "epoch": 0.5519437963666868,
      "grad_norm": 0.5777013301849365,
      "learning_rate": 0.00012701209267656932,
      "loss": 3.2503,
      "step": 42110
    },
    {
      "epoch": 0.5520748682727344,
      "grad_norm": 0.6216441988945007,
      "learning_rate": 0.0001269506554669618,
      "loss": 3.2698,
      "step": 42120
    },
    {
      "epoch": 0.5522059401787821,
      "grad_norm": 0.6401820182800293,
      "learning_rate": 0.0001268892222172948,
      "loss": 3.2423,
      "step": 42130
    },
    {
      "epoch": 0.5523370120848298,
      "grad_norm": 0.6452779173851013,
      "learning_rate": 0.0001268277929381228,
      "loss": 3.2193,
      "step": 42140
    },
    {
      "epoch": 0.5524680839908774,
      "grad_norm": 0.8983054161071777,
      "learning_rate": 0.0001267663676399994,
      "loss": 3.2844,
      "step": 42150
    },
    {
      "epoch": 0.5525991558969251,
      "grad_norm": 0.6589357256889343,
      "learning_rate": 0.00012670494633347767,
      "loss": 3.2694,
      "step": 42160
    },
    {
      "epoch": 0.5527302278029728,
      "grad_norm": 0.7605552673339844,
      "learning_rate": 0.00012664352902911,
      "loss": 3.3128,
      "step": 42170
    },
    {
      "epoch": 0.5528612997090203,
      "grad_norm": 0.6697857975959778,
      "learning_rate": 0.00012658211573744803,
      "loss": 3.2648,
      "step": 42180
    },
    {
      "epoch": 0.552992371615068,
      "grad_norm": 0.6709420680999756,
      "learning_rate": 0.00012652070646904275,
      "loss": 3.2622,
      "step": 42190
    },
    {
      "epoch": 0.5531234435211156,
      "grad_norm": 0.5938488841056824,
      "learning_rate": 0.00012645930123444436,
      "loss": 3.2539,
      "step": 42200
    },
    {
      "epoch": 0.5532545154271633,
      "grad_norm": 0.5825215578079224,
      "learning_rate": 0.00012639790004420251,
      "loss": 3.2031,
      "step": 42210
    },
    {
      "epoch": 0.553385587333211,
      "grad_norm": 0.6275821328163147,
      "learning_rate": 0.00012633650290886617,
      "loss": 3.2644,
      "step": 42220
    },
    {
      "epoch": 0.5535166592392586,
      "grad_norm": 0.5718841552734375,
      "learning_rate": 0.00012627510983898336,
      "loss": 3.2649,
      "step": 42230
    },
    {
      "epoch": 0.5536477311453063,
      "grad_norm": 1.0546420812606812,
      "learning_rate": 0.0001262137208451017,
      "loss": 3.2535,
      "step": 42240
    },
    {
      "epoch": 0.553778803051354,
      "grad_norm": 0.618228554725647,
      "learning_rate": 0.000126152335937768,
      "loss": 3.2932,
      "step": 42250
    },
    {
      "epoch": 0.5539098749574016,
      "grad_norm": 0.6282578110694885,
      "learning_rate": 0.00012609095512752826,
      "loss": 3.2531,
      "step": 42260
    },
    {
      "epoch": 0.5540409468634493,
      "grad_norm": 0.6590080857276917,
      "learning_rate": 0.00012602957842492796,
      "loss": 3.2438,
      "step": 42270
    },
    {
      "epoch": 0.554172018769497,
      "grad_norm": 0.6815877556800842,
      "learning_rate": 0.00012596820584051172,
      "loss": 3.2459,
      "step": 42280
    },
    {
      "epoch": 0.5543030906755446,
      "grad_norm": 0.6659173965454102,
      "learning_rate": 0.0001259068373848236,
      "loss": 3.244,
      "step": 42290
    },
    {
      "epoch": 0.5544341625815923,
      "grad_norm": 0.653017520904541,
      "learning_rate": 0.00012584547306840677,
      "loss": 3.303,
      "step": 42300
    },
    {
      "epoch": 0.55456523448764,
      "grad_norm": 0.6339126229286194,
      "learning_rate": 0.00012578411290180383,
      "loss": 3.1987,
      "step": 42310
    },
    {
      "epoch": 0.5546963063936876,
      "grad_norm": 0.6108158230781555,
      "learning_rate": 0.0001257227568955566,
      "loss": 3.2189,
      "step": 42320
    },
    {
      "epoch": 0.5548273782997353,
      "grad_norm": 0.6724685430526733,
      "learning_rate": 0.00012566140506020635,
      "loss": 3.2496,
      "step": 42330
    },
    {
      "epoch": 0.5549584502057829,
      "grad_norm": 0.6441592574119568,
      "learning_rate": 0.0001256000574062933,
      "loss": 3.3037,
      "step": 42340
    },
    {
      "epoch": 0.5550895221118306,
      "grad_norm": 0.8015444278717041,
      "learning_rate": 0.00012553871394435726,
      "loss": 3.2333,
      "step": 42350
    },
    {
      "epoch": 0.5552205940178782,
      "grad_norm": 0.6025912165641785,
      "learning_rate": 0.0001254773746849372,
      "loss": 3.2388,
      "step": 42360
    },
    {
      "epoch": 0.5553516659239258,
      "grad_norm": 0.6571113467216492,
      "learning_rate": 0.0001254160396385713,
      "loss": 3.1798,
      "step": 42370
    },
    {
      "epoch": 0.5554827378299735,
      "grad_norm": 0.5984496474266052,
      "learning_rate": 0.00012535470881579712,
      "loss": 3.2481,
      "step": 42380
    },
    {
      "epoch": 0.5556138097360211,
      "grad_norm": 0.5860995650291443,
      "learning_rate": 0.0001252933822271515,
      "loss": 3.2421,
      "step": 42390
    },
    {
      "epoch": 0.5557448816420688,
      "grad_norm": 0.6499239802360535,
      "learning_rate": 0.00012523205988317055,
      "loss": 3.2459,
      "step": 42400
    },
    {
      "epoch": 0.5558759535481165,
      "grad_norm": 0.622852623462677,
      "learning_rate": 0.00012517074179438947,
      "loss": 3.2514,
      "step": 42410
    },
    {
      "epoch": 0.5560070254541641,
      "grad_norm": 0.6441624164581299,
      "learning_rate": 0.000125109427971343,
      "loss": 3.2258,
      "step": 42420
    },
    {
      "epoch": 0.5561380973602118,
      "grad_norm": 0.5953484773635864,
      "learning_rate": 0.000125048118424565,
      "loss": 3.2851,
      "step": 42430
    },
    {
      "epoch": 0.5562691692662595,
      "grad_norm": 0.5982539057731628,
      "learning_rate": 0.00012498681316458861,
      "loss": 3.246,
      "step": 42440
    },
    {
      "epoch": 0.5564002411723071,
      "grad_norm": 0.5919554233551025,
      "learning_rate": 0.00012492551220194624,
      "loss": 3.2654,
      "step": 42450
    },
    {
      "epoch": 0.5565313130783548,
      "grad_norm": 0.6753155589103699,
      "learning_rate": 0.00012486421554716955,
      "loss": 3.2384,
      "step": 42460
    },
    {
      "epoch": 0.5566623849844025,
      "grad_norm": 0.5889798998832703,
      "learning_rate": 0.00012480292321078953,
      "loss": 3.3218,
      "step": 42470
    },
    {
      "epoch": 0.5567934568904501,
      "grad_norm": 0.6753676533699036,
      "learning_rate": 0.00012474163520333626,
      "loss": 3.3055,
      "step": 42480
    },
    {
      "epoch": 0.5569245287964978,
      "grad_norm": 0.6083866357803345,
      "learning_rate": 0.00012468035153533926,
      "loss": 3.2761,
      "step": 42490
    },
    {
      "epoch": 0.5570556007025455,
      "grad_norm": 0.6314940452575684,
      "learning_rate": 0.00012461907221732722,
      "loss": 3.2959,
      "step": 42500
    },
    {
      "epoch": 0.5570556007025455,
      "eval_accuracy": 0.3973299120234604,
      "eval_loss": 3.24623441696167,
      "eval_runtime": 60.5554,
      "eval_samples_per_second": 33.028,
      "eval_steps_per_second": 0.528,
      "step": 42500
    },
    {
      "epoch": 0.5571866726085931,
      "grad_norm": 0.9002912640571594,
      "learning_rate": 0.0001245577972598282,
      "loss": 3.2788,
      "step": 42510
    },
    {
      "epoch": 0.5573177445146408,
      "grad_norm": 0.6064646244049072,
      "learning_rate": 0.00012449652667336923,
      "loss": 3.2274,
      "step": 42520
    },
    {
      "epoch": 0.5574488164206884,
      "grad_norm": 0.6044290661811829,
      "learning_rate": 0.00012443526046847684,
      "loss": 3.2314,
      "step": 42530
    },
    {
      "epoch": 0.557579888326736,
      "grad_norm": 0.7704267501831055,
      "learning_rate": 0.00012437399865567673,
      "loss": 3.242,
      "step": 42540
    },
    {
      "epoch": 0.5577109602327837,
      "grad_norm": 0.6443463563919067,
      "learning_rate": 0.0001243127412454939,
      "loss": 3.2548,
      "step": 42550
    },
    {
      "epoch": 0.5578420321388313,
      "grad_norm": 0.6372406482696533,
      "learning_rate": 0.00012425148824845244,
      "loss": 3.2528,
      "step": 42560
    },
    {
      "epoch": 0.557973104044879,
      "grad_norm": 0.5984323620796204,
      "learning_rate": 0.00012419023967507584,
      "loss": 3.2446,
      "step": 42570
    },
    {
      "epoch": 0.5581041759509267,
      "grad_norm": 0.6088480949401855,
      "learning_rate": 0.00012412899553588676,
      "loss": 3.2469,
      "step": 42580
    },
    {
      "epoch": 0.5582352478569743,
      "grad_norm": 0.5886126160621643,
      "learning_rate": 0.00012406775584140715,
      "loss": 3.2421,
      "step": 42590
    },
    {
      "epoch": 0.558366319763022,
      "grad_norm": 0.60127192735672,
      "learning_rate": 0.0001240065206021581,
      "loss": 3.2205,
      "step": 42600
    },
    {
      "epoch": 0.5584973916690696,
      "grad_norm": 0.6054568290710449,
      "learning_rate": 0.00012394528982866002,
      "loss": 3.2271,
      "step": 42610
    },
    {
      "epoch": 0.5586284635751173,
      "grad_norm": 0.6944029927253723,
      "learning_rate": 0.00012388406353143255,
      "loss": 3.2501,
      "step": 42620
    },
    {
      "epoch": 0.558759535481165,
      "grad_norm": 0.6145868897438049,
      "learning_rate": 0.00012382284172099452,
      "loss": 3.2593,
      "step": 42630
    },
    {
      "epoch": 0.5588906073872126,
      "grad_norm": 0.5607081055641174,
      "learning_rate": 0.00012376162440786394,
      "loss": 3.2428,
      "step": 42640
    },
    {
      "epoch": 0.5590216792932603,
      "grad_norm": 0.6626426577568054,
      "learning_rate": 0.00012370041160255816,
      "loss": 3.2407,
      "step": 42650
    },
    {
      "epoch": 0.559152751199308,
      "grad_norm": 0.6469648480415344,
      "learning_rate": 0.00012363920331559376,
      "loss": 3.243,
      "step": 42660
    },
    {
      "epoch": 0.5592838231053556,
      "grad_norm": 0.585746705532074,
      "learning_rate": 0.0001235779995574864,
      "loss": 3.275,
      "step": 42670
    },
    {
      "epoch": 0.5594148950114033,
      "grad_norm": 0.6247645020484924,
      "learning_rate": 0.00012351680033875108,
      "loss": 3.2837,
      "step": 42680
    },
    {
      "epoch": 0.559545966917451,
      "grad_norm": 0.6373127698898315,
      "learning_rate": 0.00012345560566990203,
      "loss": 3.2394,
      "step": 42690
    },
    {
      "epoch": 0.5596770388234986,
      "grad_norm": 0.6944007277488708,
      "learning_rate": 0.00012339441556145269,
      "loss": 3.21,
      "step": 42700
    },
    {
      "epoch": 0.5598081107295463,
      "grad_norm": 0.6206851601600647,
      "learning_rate": 0.0001233332300239156,
      "loss": 3.2481,
      "step": 42710
    },
    {
      "epoch": 0.5599391826355938,
      "grad_norm": 0.6568619012832642,
      "learning_rate": 0.00012327204906780265,
      "loss": 3.2375,
      "step": 42720
    },
    {
      "epoch": 0.5600702545416415,
      "grad_norm": 0.637696385383606,
      "learning_rate": 0.00012321087270362492,
      "loss": 3.2461,
      "step": 42730
    },
    {
      "epoch": 0.5602013264476892,
      "grad_norm": 0.6820730566978455,
      "learning_rate": 0.0001231497009418926,
      "loss": 3.2136,
      "step": 42740
    },
    {
      "epoch": 0.5603323983537368,
      "grad_norm": 0.5873591899871826,
      "learning_rate": 0.0001230885337931152,
      "loss": 3.2467,
      "step": 42750
    },
    {
      "epoch": 0.5604634702597845,
      "grad_norm": 0.7100130915641785,
      "learning_rate": 0.00012302737126780141,
      "loss": 3.3194,
      "step": 42760
    },
    {
      "epoch": 0.5605945421658322,
      "grad_norm": 0.7396765351295471,
      "learning_rate": 0.0001229662133764592,
      "loss": 3.2795,
      "step": 42770
    },
    {
      "epoch": 0.5607256140718798,
      "grad_norm": 0.6243323087692261,
      "learning_rate": 0.0001229050601295955,
      "loss": 3.2391,
      "step": 42780
    },
    {
      "epoch": 0.5608566859779275,
      "grad_norm": 0.6129120588302612,
      "learning_rate": 0.00012284391153771673,
      "loss": 3.2013,
      "step": 42790
    },
    {
      "epoch": 0.5609877578839751,
      "grad_norm": 0.6071087718009949,
      "learning_rate": 0.00012278276761132836,
      "loss": 3.2839,
      "step": 42800
    },
    {
      "epoch": 0.5611188297900228,
      "grad_norm": 0.6373593807220459,
      "learning_rate": 0.0001227216283609351,
      "loss": 3.2178,
      "step": 42810
    },
    {
      "epoch": 0.5612499016960705,
      "grad_norm": 0.6481397151947021,
      "learning_rate": 0.00012266049379704077,
      "loss": 3.2419,
      "step": 42820
    },
    {
      "epoch": 0.5613809736021181,
      "grad_norm": 0.6197222471237183,
      "learning_rate": 0.00012259936393014848,
      "loss": 3.2851,
      "step": 42830
    },
    {
      "epoch": 0.5615120455081658,
      "grad_norm": 0.6300904154777527,
      "learning_rate": 0.00012253823877076054,
      "loss": 3.2666,
      "step": 42840
    },
    {
      "epoch": 0.5616431174142135,
      "grad_norm": 0.6319248676300049,
      "learning_rate": 0.00012247711832937845,
      "loss": 3.2107,
      "step": 42850
    },
    {
      "epoch": 0.5617741893202611,
      "grad_norm": 0.5837290287017822,
      "learning_rate": 0.0001224160026165028,
      "loss": 3.2149,
      "step": 42860
    },
    {
      "epoch": 0.5619052612263088,
      "grad_norm": 0.5593276023864746,
      "learning_rate": 0.00012235489164263345,
      "loss": 3.2511,
      "step": 42870
    },
    {
      "epoch": 0.5620363331323565,
      "grad_norm": 0.5967832803726196,
      "learning_rate": 0.0001222937854182695,
      "loss": 3.2665,
      "step": 42880
    },
    {
      "epoch": 0.5621674050384041,
      "grad_norm": 0.6447561979293823,
      "learning_rate": 0.00012223268395390906,
      "loss": 3.2785,
      "step": 42890
    },
    {
      "epoch": 0.5622984769444517,
      "grad_norm": 0.6525065898895264,
      "learning_rate": 0.0001221715872600496,
      "loss": 3.2998,
      "step": 42900
    },
    {
      "epoch": 0.5624295488504993,
      "grad_norm": 0.6282928586006165,
      "learning_rate": 0.00012211049534718767,
      "loss": 3.2233,
      "step": 42910
    },
    {
      "epoch": 0.562560620756547,
      "grad_norm": 0.6120992302894592,
      "learning_rate": 0.00012204940822581908,
      "loss": 3.2545,
      "step": 42920
    },
    {
      "epoch": 0.5626916926625947,
      "grad_norm": 0.595097005367279,
      "learning_rate": 0.00012198832590643866,
      "loss": 3.257,
      "step": 42930
    },
    {
      "epoch": 0.5628227645686423,
      "grad_norm": 0.6346606612205505,
      "learning_rate": 0.0001219272483995406,
      "loss": 3.2517,
      "step": 42940
    },
    {
      "epoch": 0.56295383647469,
      "grad_norm": 0.6321684122085571,
      "learning_rate": 0.00012186617571561816,
      "loss": 3.2447,
      "step": 42950
    },
    {
      "epoch": 0.5630849083807377,
      "grad_norm": 0.5866535305976868,
      "learning_rate": 0.00012180510786516388,
      "loss": 3.2099,
      "step": 42960
    },
    {
      "epoch": 0.5632159802867853,
      "grad_norm": 0.6268247365951538,
      "learning_rate": 0.00012174404485866922,
      "loss": 3.2787,
      "step": 42970
    },
    {
      "epoch": 0.563347052192833,
      "grad_norm": 0.6316028237342834,
      "learning_rate": 0.00012168298670662512,
      "loss": 3.2094,
      "step": 42980
    },
    {
      "epoch": 0.5634781240988807,
      "grad_norm": 0.6275749206542969,
      "learning_rate": 0.00012162193341952145,
      "loss": 3.2446,
      "step": 42990
    },
    {
      "epoch": 0.5636091960049283,
      "grad_norm": 0.6610090136528015,
      "learning_rate": 0.00012156088500784737,
      "loss": 3.2233,
      "step": 43000
    },
    {
      "epoch": 0.5636091960049283,
      "eval_accuracy": 0.39718817204301077,
      "eval_loss": 3.2444655895233154,
      "eval_runtime": 59.3398,
      "eval_samples_per_second": 33.704,
      "eval_steps_per_second": 0.539,
      "step": 43000
    },
    {
      "epoch": 0.563740267910976,
      "grad_norm": 0.6213541626930237,
      "learning_rate": 0.0001214998414820911,
      "loss": 3.272,
      "step": 43010
    },
    {
      "epoch": 0.5638713398170236,
      "grad_norm": 0.7418097853660583,
      "learning_rate": 0.00012143880285274015,
      "loss": 3.2694,
      "step": 43020
    },
    {
      "epoch": 0.5640024117230713,
      "grad_norm": 0.6527386903762817,
      "learning_rate": 0.00012137776913028117,
      "loss": 3.1938,
      "step": 43030
    },
    {
      "epoch": 0.564133483629119,
      "grad_norm": 0.7256929874420166,
      "learning_rate": 0.0001213167403251998,
      "loss": 3.2594,
      "step": 43040
    },
    {
      "epoch": 0.5642645555351666,
      "grad_norm": 0.6325649619102478,
      "learning_rate": 0.00012125571644798101,
      "loss": 3.2726,
      "step": 43050
    },
    {
      "epoch": 0.5643956274412143,
      "grad_norm": 0.9402995109558105,
      "learning_rate": 0.00012119469750910893,
      "loss": 3.2868,
      "step": 43060
    },
    {
      "epoch": 0.564526699347262,
      "grad_norm": 0.67952960729599,
      "learning_rate": 0.0001211336835190667,
      "loss": 3.235,
      "step": 43070
    },
    {
      "epoch": 0.5646577712533095,
      "grad_norm": 0.6465554237365723,
      "learning_rate": 0.00012107267448833675,
      "loss": 3.2616,
      "step": 43080
    },
    {
      "epoch": 0.5647888431593572,
      "grad_norm": 1.083726406097412,
      "learning_rate": 0.00012101167042740051,
      "loss": 3.2264,
      "step": 43090
    },
    {
      "epoch": 0.5649199150654048,
      "grad_norm": 0.5495240092277527,
      "learning_rate": 0.00012095067134673875,
      "loss": 3.2628,
      "step": 43100
    },
    {
      "epoch": 0.5650509869714525,
      "grad_norm": 0.7268524169921875,
      "learning_rate": 0.0001208896772568312,
      "loss": 3.2529,
      "step": 43110
    },
    {
      "epoch": 0.5651820588775002,
      "grad_norm": 0.6181415319442749,
      "learning_rate": 0.00012082868816815682,
      "loss": 3.3322,
      "step": 43120
    },
    {
      "epoch": 0.5653131307835478,
      "grad_norm": 0.8651639223098755,
      "learning_rate": 0.00012076770409119375,
      "loss": 3.2786,
      "step": 43130
    },
    {
      "epoch": 0.5654442026895955,
      "grad_norm": 0.6164896488189697,
      "learning_rate": 0.00012070672503641923,
      "loss": 3.277,
      "step": 43140
    },
    {
      "epoch": 0.5655752745956432,
      "grad_norm": 0.7646924257278442,
      "learning_rate": 0.00012064575101430953,
      "loss": 3.2511,
      "step": 43150
    },
    {
      "epoch": 0.5657063465016908,
      "grad_norm": 0.5828595757484436,
      "learning_rate": 0.00012058478203534028,
      "loss": 3.2324,
      "step": 43160
    },
    {
      "epoch": 0.5658374184077385,
      "grad_norm": 0.6382690668106079,
      "learning_rate": 0.000120523818109986,
      "loss": 3.2458,
      "step": 43170
    },
    {
      "epoch": 0.5659684903137862,
      "grad_norm": 0.6081956624984741,
      "learning_rate": 0.00012046285924872058,
      "loss": 3.2345,
      "step": 43180
    },
    {
      "epoch": 0.5660995622198338,
      "grad_norm": 0.716952383518219,
      "learning_rate": 0.0001204019054620168,
      "loss": 3.2913,
      "step": 43190
    },
    {
      "epoch": 0.5662306341258815,
      "grad_norm": 0.6723486185073853,
      "learning_rate": 0.00012034095676034675,
      "loss": 3.2534,
      "step": 43200
    },
    {
      "epoch": 0.5663617060319291,
      "grad_norm": 0.5785861015319824,
      "learning_rate": 0.00012028001315418155,
      "loss": 3.2725,
      "step": 43210
    },
    {
      "epoch": 0.5664927779379768,
      "grad_norm": 0.6231510639190674,
      "learning_rate": 0.00012021907465399159,
      "loss": 3.2744,
      "step": 43220
    },
    {
      "epoch": 0.5666238498440245,
      "grad_norm": 0.6003397703170776,
      "learning_rate": 0.00012015814127024612,
      "loss": 3.2267,
      "step": 43230
    },
    {
      "epoch": 0.5667549217500721,
      "grad_norm": 0.6370406746864319,
      "learning_rate": 0.00012009721301341375,
      "loss": 3.2198,
      "step": 43240
    },
    {
      "epoch": 0.5668859936561198,
      "grad_norm": 0.5604175329208374,
      "learning_rate": 0.00012003628989396211,
      "loss": 3.2311,
      "step": 43250
    },
    {
      "epoch": 0.5670170655621674,
      "grad_norm": 0.7046313285827637,
      "learning_rate": 0.00011997537192235794,
      "loss": 3.3137,
      "step": 43260
    },
    {
      "epoch": 0.567148137468215,
      "grad_norm": 0.6208400130271912,
      "learning_rate": 0.00011991445910906709,
      "loss": 3.2613,
      "step": 43270
    },
    {
      "epoch": 0.5672792093742627,
      "grad_norm": 0.6924196481704712,
      "learning_rate": 0.00011985355146455461,
      "loss": 3.2797,
      "step": 43280
    },
    {
      "epoch": 0.5674102812803103,
      "grad_norm": 0.6309049129486084,
      "learning_rate": 0.00011979264899928459,
      "loss": 3.214,
      "step": 43290
    },
    {
      "epoch": 0.567541353186358,
      "grad_norm": 0.6157112717628479,
      "learning_rate": 0.00011973175172372019,
      "loss": 3.2656,
      "step": 43300
    },
    {
      "epoch": 0.5676724250924057,
      "grad_norm": 0.6638853549957275,
      "learning_rate": 0.00011967085964832376,
      "loss": 3.1937,
      "step": 43310
    },
    {
      "epoch": 0.5678034969984533,
      "grad_norm": 0.5795360803604126,
      "learning_rate": 0.00011960997278355671,
      "loss": 3.2498,
      "step": 43320
    },
    {
      "epoch": 0.567934568904501,
      "grad_norm": 1.2784022092819214,
      "learning_rate": 0.00011954909113987965,
      "loss": 3.3026,
      "step": 43330
    },
    {
      "epoch": 0.5680656408105487,
      "grad_norm": 0.6048783659934998,
      "learning_rate": 0.00011948821472775213,
      "loss": 3.293,
      "step": 43340
    },
    {
      "epoch": 0.5681967127165963,
      "grad_norm": 0.6757849454879761,
      "learning_rate": 0.00011942734355763287,
      "loss": 3.2655,
      "step": 43350
    },
    {
      "epoch": 0.568327784622644,
      "grad_norm": 0.6158847808837891,
      "learning_rate": 0.0001193664776399798,
      "loss": 3.255,
      "step": 43360
    },
    {
      "epoch": 0.5684588565286917,
      "grad_norm": 0.7038004398345947,
      "learning_rate": 0.00011930561698524971,
      "loss": 3.2474,
      "step": 43370
    },
    {
      "epoch": 0.5685899284347393,
      "grad_norm": 0.6013663411140442,
      "learning_rate": 0.00011924476160389872,
      "loss": 3.238,
      "step": 43380
    },
    {
      "epoch": 0.568721000340787,
      "grad_norm": 0.620312511920929,
      "learning_rate": 0.00011918391150638194,
      "loss": 3.192,
      "step": 43390
    },
    {
      "epoch": 0.5688520722468347,
      "grad_norm": 0.5915484428405762,
      "learning_rate": 0.00011912306670315365,
      "loss": 3.2049,
      "step": 43400
    },
    {
      "epoch": 0.5689831441528823,
      "grad_norm": 0.6437747478485107,
      "learning_rate": 0.00011906222720466699,
      "loss": 3.2986,
      "step": 43410
    },
    {
      "epoch": 0.56911421605893,
      "grad_norm": 0.6576946377754211,
      "learning_rate": 0.00011900139302137447,
      "loss": 3.2759,
      "step": 43420
    },
    {
      "epoch": 0.5692452879649776,
      "grad_norm": 0.6345543265342712,
      "learning_rate": 0.00011894056416372759,
      "loss": 3.2864,
      "step": 43430
    },
    {
      "epoch": 0.5693763598710252,
      "grad_norm": 0.6375799775123596,
      "learning_rate": 0.00011887974064217683,
      "loss": 3.2344,
      "step": 43440
    },
    {
      "epoch": 0.5695074317770729,
      "grad_norm": 0.6860445141792297,
      "learning_rate": 0.00011881892246717192,
      "loss": 3.2574,
      "step": 43450
    },
    {
      "epoch": 0.5696385036831205,
      "grad_norm": 1.2995883226394653,
      "learning_rate": 0.0001187581096491615,
      "loss": 3.2888,
      "step": 43460
    },
    {
      "epoch": 0.5697695755891682,
      "grad_norm": 0.6449354887008667,
      "learning_rate": 0.00011869730219859348,
      "loss": 3.2381,
      "step": 43470
    },
    {
      "epoch": 0.5699006474952159,
      "grad_norm": 0.6677840352058411,
      "learning_rate": 0.00011863650012591463,
      "loss": 3.2961,
      "step": 43480
    },
    {
      "epoch": 0.5700317194012635,
      "grad_norm": 0.6202585101127625,
      "learning_rate": 0.00011857570344157099,
      "loss": 3.2326,
      "step": 43490
    },
    {
      "epoch": 0.5701627913073112,
      "grad_norm": 0.6145714521408081,
      "learning_rate": 0.00011851491215600758,
      "loss": 3.2613,
      "step": 43500
    },
    {
      "epoch": 0.5701627913073112,
      "eval_accuracy": 0.3975381231671554,
      "eval_loss": 3.243159294128418,
      "eval_runtime": 61.2596,
      "eval_samples_per_second": 32.648,
      "eval_steps_per_second": 0.522,
      "step": 43500
    },
    {
      "epoch": 0.5702938632133588,
      "grad_norm": 0.6366761922836304,
      "learning_rate": 0.00011845412627966858,
      "loss": 3.2273,
      "step": 43510
    },
    {
      "epoch": 0.5704249351194065,
      "grad_norm": 0.5727043747901917,
      "learning_rate": 0.00011839334582299703,
      "loss": 3.263,
      "step": 43520
    },
    {
      "epoch": 0.5705560070254542,
      "grad_norm": 0.6051214337348938,
      "learning_rate": 0.00011833257079643528,
      "loss": 3.2239,
      "step": 43530
    },
    {
      "epoch": 0.5706870789315018,
      "grad_norm": 0.6732583045959473,
      "learning_rate": 0.00011827180121042458,
      "loss": 3.2369,
      "step": 43540
    },
    {
      "epoch": 0.5708181508375495,
      "grad_norm": 0.5960179567337036,
      "learning_rate": 0.00011821103707540542,
      "loss": 3.2433,
      "step": 43550
    },
    {
      "epoch": 0.5709492227435972,
      "grad_norm": 0.6697182059288025,
      "learning_rate": 0.0001181502784018171,
      "loss": 3.1666,
      "step": 43560
    },
    {
      "epoch": 0.5710802946496448,
      "grad_norm": 0.658229410648346,
      "learning_rate": 0.00011808952520009821,
      "loss": 3.2425,
      "step": 43570
    },
    {
      "epoch": 0.5712113665556925,
      "grad_norm": 0.899276852607727,
      "learning_rate": 0.00011802877748068629,
      "loss": 3.2724,
      "step": 43580
    },
    {
      "epoch": 0.5713424384617402,
      "grad_norm": 0.6311741471290588,
      "learning_rate": 0.00011796803525401808,
      "loss": 3.2569,
      "step": 43590
    },
    {
      "epoch": 0.5714735103677878,
      "grad_norm": 0.6979262828826904,
      "learning_rate": 0.00011790729853052907,
      "loss": 3.2655,
      "step": 43600
    },
    {
      "epoch": 0.5716045822738354,
      "grad_norm": 0.6784321665763855,
      "learning_rate": 0.00011784656732065413,
      "loss": 3.2323,
      "step": 43610
    },
    {
      "epoch": 0.571735654179883,
      "grad_norm": 0.5697076320648193,
      "learning_rate": 0.00011778584163482699,
      "loss": 3.2807,
      "step": 43620
    },
    {
      "epoch": 0.5718667260859307,
      "grad_norm": 0.6218394637107849,
      "learning_rate": 0.00011772512148348052,
      "loss": 3.2307,
      "step": 43630
    },
    {
      "epoch": 0.5719977979919784,
      "grad_norm": 0.6381251215934753,
      "learning_rate": 0.00011766440687704656,
      "loss": 3.263,
      "step": 43640
    },
    {
      "epoch": 0.572128869898026,
      "grad_norm": 0.5995039343833923,
      "learning_rate": 0.00011760369782595607,
      "loss": 3.2647,
      "step": 43650
    },
    {
      "epoch": 0.5722599418040737,
      "grad_norm": 0.6439694166183472,
      "learning_rate": 0.0001175429943406391,
      "loss": 3.2332,
      "step": 43660
    },
    {
      "epoch": 0.5723910137101214,
      "grad_norm": 0.5705400109291077,
      "learning_rate": 0.00011748229643152455,
      "loss": 3.2642,
      "step": 43670
    },
    {
      "epoch": 0.572522085616169,
      "grad_norm": 0.618656575679779,
      "learning_rate": 0.00011742160410904057,
      "loss": 3.2545,
      "step": 43680
    },
    {
      "epoch": 0.5726531575222167,
      "grad_norm": 0.6085860729217529,
      "learning_rate": 0.00011736091738361427,
      "loss": 3.2777,
      "step": 43690
    },
    {
      "epoch": 0.5727842294282643,
      "grad_norm": 0.6229195594787598,
      "learning_rate": 0.00011730023626567177,
      "loss": 3.2389,
      "step": 43700
    },
    {
      "epoch": 0.572915301334312,
      "grad_norm": 0.6317970752716064,
      "learning_rate": 0.00011723956076563825,
      "loss": 3.2701,
      "step": 43710
    },
    {
      "epoch": 0.5730463732403597,
      "grad_norm": 0.5994996428489685,
      "learning_rate": 0.00011717889089393792,
      "loss": 3.2596,
      "step": 43720
    },
    {
      "epoch": 0.5731774451464073,
      "grad_norm": 0.6655870676040649,
      "learning_rate": 0.0001171182266609941,
      "loss": 3.2381,
      "step": 43730
    },
    {
      "epoch": 0.573308517052455,
      "grad_norm": 1.0828248262405396,
      "learning_rate": 0.00011705756807722896,
      "loss": 3.2781,
      "step": 43740
    },
    {
      "epoch": 0.5734395889585027,
      "grad_norm": 0.6480745673179626,
      "learning_rate": 0.00011699691515306386,
      "loss": 3.2219,
      "step": 43750
    },
    {
      "epoch": 0.5735706608645503,
      "grad_norm": 0.721631646156311,
      "learning_rate": 0.00011693626789891913,
      "loss": 3.267,
      "step": 43760
    },
    {
      "epoch": 0.573701732770598,
      "grad_norm": 0.8623172640800476,
      "learning_rate": 0.00011687562632521425,
      "loss": 3.2556,
      "step": 43770
    },
    {
      "epoch": 0.5738328046766457,
      "grad_norm": 0.6004065871238708,
      "learning_rate": 0.00011681499044236741,
      "loss": 3.2519,
      "step": 43780
    },
    {
      "epoch": 0.5739638765826932,
      "grad_norm": 0.5865307450294495,
      "learning_rate": 0.00011675436026079617,
      "loss": 3.2901,
      "step": 43790
    },
    {
      "epoch": 0.5740949484887409,
      "grad_norm": 0.6044079661369324,
      "learning_rate": 0.00011669373579091687,
      "loss": 3.2271,
      "step": 43800
    },
    {
      "epoch": 0.5742260203947885,
      "grad_norm": 0.5992538332939148,
      "learning_rate": 0.00011663311704314507,
      "loss": 3.2476,
      "step": 43810
    },
    {
      "epoch": 0.5743570923008362,
      "grad_norm": 0.6272064447402954,
      "learning_rate": 0.00011657250402789511,
      "loss": 3.2468,
      "step": 43820
    },
    {
      "epoch": 0.5744881642068839,
      "grad_norm": 0.6651163101196289,
      "learning_rate": 0.00011651189675558052,
      "loss": 3.2449,
      "step": 43830
    },
    {
      "epoch": 0.5746192361129315,
      "grad_norm": 0.674713134765625,
      "learning_rate": 0.00011645129523661379,
      "loss": 3.1796,
      "step": 43840
    },
    {
      "epoch": 0.5747503080189792,
      "grad_norm": 0.6734214425086975,
      "learning_rate": 0.00011639069948140653,
      "loss": 3.2686,
      "step": 43850
    },
    {
      "epoch": 0.5748813799250269,
      "grad_norm": 0.615215539932251,
      "learning_rate": 0.00011633010950036908,
      "loss": 3.2385,
      "step": 43860
    },
    {
      "epoch": 0.5750124518310745,
      "grad_norm": 0.6813642978668213,
      "learning_rate": 0.00011626952530391111,
      "loss": 3.2678,
      "step": 43870
    },
    {
      "epoch": 0.5751435237371222,
      "grad_norm": 0.6137657165527344,
      "learning_rate": 0.00011620894690244106,
      "loss": 3.2657,
      "step": 43880
    },
    {
      "epoch": 0.5752745956431699,
      "grad_norm": 0.5825940370559692,
      "learning_rate": 0.00011614837430636653,
      "loss": 3.1976,
      "step": 43890
    },
    {
      "epoch": 0.5754056675492175,
      "grad_norm": 0.8284161686897278,
      "learning_rate": 0.00011608780752609398,
      "loss": 3.2812,
      "step": 43900
    },
    {
      "epoch": 0.5755367394552652,
      "grad_norm": 0.7353925108909607,
      "learning_rate": 0.00011602724657202901,
      "loss": 3.2052,
      "step": 43910
    },
    {
      "epoch": 0.5756678113613128,
      "grad_norm": 0.6302970051765442,
      "learning_rate": 0.0001159666914545762,
      "loss": 3.2397,
      "step": 43920
    },
    {
      "epoch": 0.5757988832673605,
      "grad_norm": 0.6776291728019714,
      "learning_rate": 0.00011590614218413895,
      "loss": 3.2815,
      "step": 43930
    },
    {
      "epoch": 0.5759299551734082,
      "grad_norm": 0.6846403479576111,
      "learning_rate": 0.00011584559877111988,
      "loss": 3.1895,
      "step": 43940
    },
    {
      "epoch": 0.5760610270794558,
      "grad_norm": 0.5829125642776489,
      "learning_rate": 0.00011578506122592055,
      "loss": 3.2676,
      "step": 43950
    },
    {
      "epoch": 0.5761920989855035,
      "grad_norm": 0.6091601252555847,
      "learning_rate": 0.00011572452955894141,
      "loss": 3.1885,
      "step": 43960
    },
    {
      "epoch": 0.576323170891551,
      "grad_norm": 0.6871894001960754,
      "learning_rate": 0.000115664003780582,
      "loss": 3.2926,
      "step": 43970
    },
    {
      "epoch": 0.5764542427975987,
      "grad_norm": 0.6193121075630188,
      "learning_rate": 0.00011560348390124076,
      "loss": 3.3302,
      "step": 43980
    },
    {
      "epoch": 0.5765853147036464,
      "grad_norm": 0.6359044313430786,
      "learning_rate": 0.00011554296993131526,
      "loss": 3.2165,
      "step": 43990
    },
    {
      "epoch": 0.576716386609694,
      "grad_norm": 0.6618209481239319,
      "learning_rate": 0.0001154824618812019,
      "loss": 3.2372,
      "step": 44000
    },
    {
      "epoch": 0.576716386609694,
      "eval_accuracy": 0.39772482893450634,
      "eval_loss": 3.241494655609131,
      "eval_runtime": 60.0316,
      "eval_samples_per_second": 33.316,
      "eval_steps_per_second": 0.533,
      "step": 44000
    },
    {
      "epoch": 0.5768474585157417,
      "grad_norm": 0.5905164480209351,
      "learning_rate": 0.00011542195976129609,
      "loss": 3.2369,
      "step": 44010
    },
    {
      "epoch": 0.5769785304217894,
      "grad_norm": 0.6310632824897766,
      "learning_rate": 0.00011536146358199236,
      "loss": 3.2326,
      "step": 44020
    },
    {
      "epoch": 0.577109602327837,
      "grad_norm": 0.6369888782501221,
      "learning_rate": 0.00011530097335368412,
      "loss": 3.2605,
      "step": 44030
    },
    {
      "epoch": 0.5772406742338847,
      "grad_norm": 0.6033466458320618,
      "learning_rate": 0.00011524048908676364,
      "loss": 3.2231,
      "step": 44040
    },
    {
      "epoch": 0.5773717461399324,
      "grad_norm": 0.6440674066543579,
      "learning_rate": 0.00011518001079162236,
      "loss": 3.2564,
      "step": 44050
    },
    {
      "epoch": 0.57750281804598,
      "grad_norm": 0.6155526041984558,
      "learning_rate": 0.00011511953847865066,
      "loss": 3.3448,
      "step": 44060
    },
    {
      "epoch": 0.5776338899520277,
      "grad_norm": 0.763528048992157,
      "learning_rate": 0.00011505907215823782,
      "loss": 3.2417,
      "step": 44070
    },
    {
      "epoch": 0.5777649618580754,
      "grad_norm": 0.6444329619407654,
      "learning_rate": 0.00011499861184077203,
      "loss": 3.2923,
      "step": 44080
    },
    {
      "epoch": 0.577896033764123,
      "grad_norm": 0.656024694442749,
      "learning_rate": 0.00011493815753664061,
      "loss": 3.2498,
      "step": 44090
    },
    {
      "epoch": 0.5780271056701707,
      "grad_norm": 0.6203871965408325,
      "learning_rate": 0.00011487770925622986,
      "loss": 3.3096,
      "step": 44100
    },
    {
      "epoch": 0.5781581775762183,
      "grad_norm": 0.6460731029510498,
      "learning_rate": 0.00011481726700992479,
      "loss": 3.2609,
      "step": 44110
    },
    {
      "epoch": 0.578289249482266,
      "grad_norm": 0.6372194290161133,
      "learning_rate": 0.00011475683080810963,
      "loss": 3.2982,
      "step": 44120
    },
    {
      "epoch": 0.5784203213883137,
      "grad_norm": 0.5933446884155273,
      "learning_rate": 0.00011469640066116747,
      "loss": 3.2533,
      "step": 44130
    },
    {
      "epoch": 0.5785513932943613,
      "grad_norm": 0.6525600552558899,
      "learning_rate": 0.00011463597657948047,
      "loss": 3.2547,
      "step": 44140
    },
    {
      "epoch": 0.5786824652004089,
      "grad_norm": 0.6417176127433777,
      "learning_rate": 0.00011457555857342949,
      "loss": 3.2152,
      "step": 44150
    },
    {
      "epoch": 0.5788135371064566,
      "grad_norm": 0.5923622846603394,
      "learning_rate": 0.00011451514665339463,
      "loss": 3.2418,
      "step": 44160
    },
    {
      "epoch": 0.5789446090125042,
      "grad_norm": 0.6038665175437927,
      "learning_rate": 0.00011445474082975475,
      "loss": 3.2791,
      "step": 44170
    },
    {
      "epoch": 0.5790756809185519,
      "grad_norm": 0.5869840383529663,
      "learning_rate": 0.00011439434111288783,
      "loss": 3.2342,
      "step": 44180
    },
    {
      "epoch": 0.5792067528245995,
      "grad_norm": 0.687934398651123,
      "learning_rate": 0.00011433394751317056,
      "loss": 3.2319,
      "step": 44190
    },
    {
      "epoch": 0.5793378247306472,
      "grad_norm": 0.6892923712730408,
      "learning_rate": 0.00011427356004097884,
      "loss": 3.2556,
      "step": 44200
    },
    {
      "epoch": 0.5794688966366949,
      "grad_norm": 0.6072307825088501,
      "learning_rate": 0.00011421317870668737,
      "loss": 3.2554,
      "step": 44210
    },
    {
      "epoch": 0.5795999685427425,
      "grad_norm": 0.6325558423995972,
      "learning_rate": 0.00011415280352066988,
      "loss": 3.2581,
      "step": 44220
    },
    {
      "epoch": 0.5797310404487902,
      "grad_norm": 0.6215746402740479,
      "learning_rate": 0.0001140924344932989,
      "loss": 3.2728,
      "step": 44230
    },
    {
      "epoch": 0.5798621123548379,
      "grad_norm": 0.8204739093780518,
      "learning_rate": 0.00011403207163494606,
      "loss": 3.2894,
      "step": 44240
    },
    {
      "epoch": 0.5799931842608855,
      "grad_norm": 0.6109630465507507,
      "learning_rate": 0.00011397171495598185,
      "loss": 3.2948,
      "step": 44250
    },
    {
      "epoch": 0.5801242561669332,
      "grad_norm": 0.6577589511871338,
      "learning_rate": 0.0001139113644667757,
      "loss": 3.2497,
      "step": 44260
    },
    {
      "epoch": 0.5802553280729809,
      "grad_norm": 0.5761979222297668,
      "learning_rate": 0.00011385102017769596,
      "loss": 3.232,
      "step": 44270
    },
    {
      "epoch": 0.5803863999790285,
      "grad_norm": 0.7664276957511902,
      "learning_rate": 0.00011379068209910999,
      "loss": 3.2581,
      "step": 44280
    },
    {
      "epoch": 0.5805174718850762,
      "grad_norm": 0.638039767742157,
      "learning_rate": 0.00011373035024138408,
      "loss": 3.3077,
      "step": 44290
    },
    {
      "epoch": 0.5806485437911239,
      "grad_norm": 0.6024799346923828,
      "learning_rate": 0.00011367002461488331,
      "loss": 3.2342,
      "step": 44300
    },
    {
      "epoch": 0.5807796156971715,
      "grad_norm": 0.6250893473625183,
      "learning_rate": 0.00011360970522997181,
      "loss": 3.1964,
      "step": 44310
    },
    {
      "epoch": 0.5809106876032192,
      "grad_norm": 0.6229225397109985,
      "learning_rate": 0.00011354939209701269,
      "loss": 3.2362,
      "step": 44320
    },
    {
      "epoch": 0.5810417595092667,
      "grad_norm": 0.6373034119606018,
      "learning_rate": 0.00011348908522636786,
      "loss": 3.1928,
      "step": 44330
    },
    {
      "epoch": 0.5811728314153144,
      "grad_norm": 0.6969342827796936,
      "learning_rate": 0.0001134287846283982,
      "loss": 3.2793,
      "step": 44340
    },
    {
      "epoch": 0.5813039033213621,
      "grad_norm": 0.6626304984092712,
      "learning_rate": 0.00011336849031346347,
      "loss": 3.2103,
      "step": 44350
    },
    {
      "epoch": 0.5814349752274097,
      "grad_norm": 0.6306712627410889,
      "learning_rate": 0.00011330820229192253,
      "loss": 3.2126,
      "step": 44360
    },
    {
      "epoch": 0.5815660471334574,
      "grad_norm": 0.6077147126197815,
      "learning_rate": 0.00011324792057413291,
      "loss": 3.2027,
      "step": 44370
    },
    {
      "epoch": 0.581697119039505,
      "grad_norm": 0.637656033039093,
      "learning_rate": 0.00011318764517045119,
      "loss": 3.1874,
      "step": 44380
    },
    {
      "epoch": 0.5818281909455527,
      "grad_norm": 0.6672019958496094,
      "learning_rate": 0.00011312737609123288,
      "loss": 3.2195,
      "step": 44390
    },
    {
      "epoch": 0.5819592628516004,
      "grad_norm": 0.6666444540023804,
      "learning_rate": 0.00011306711334683243,
      "loss": 3.21,
      "step": 44400
    },
    {
      "epoch": 0.582090334757648,
      "grad_norm": 0.6972216367721558,
      "learning_rate": 0.00011300685694760302,
      "loss": 3.2308,
      "step": 44410
    },
    {
      "epoch": 0.5822214066636957,
      "grad_norm": 0.6605156064033508,
      "learning_rate": 0.00011294660690389696,
      "loss": 3.2216,
      "step": 44420
    },
    {
      "epoch": 0.5823524785697434,
      "grad_norm": 0.6402026414871216,
      "learning_rate": 0.00011288636322606528,
      "loss": 3.2394,
      "step": 44430
    },
    {
      "epoch": 0.582483550475791,
      "grad_norm": 0.7068410515785217,
      "learning_rate": 0.00011282612592445815,
      "loss": 3.2168,
      "step": 44440
    },
    {
      "epoch": 0.5826146223818387,
      "grad_norm": 0.7361351847648621,
      "learning_rate": 0.00011276589500942436,
      "loss": 3.2891,
      "step": 44450
    },
    {
      "epoch": 0.5827456942878864,
      "grad_norm": 0.7006919980049133,
      "learning_rate": 0.00011270567049131179,
      "loss": 3.1797,
      "step": 44460
    },
    {
      "epoch": 0.582876766193934,
      "grad_norm": 0.6827424764633179,
      "learning_rate": 0.00011264545238046721,
      "loss": 3.1865,
      "step": 44470
    },
    {
      "epoch": 0.5830078380999817,
      "grad_norm": 0.6968122124671936,
      "learning_rate": 0.00011258524068723629,
      "loss": 3.24,
      "step": 44480
    },
    {
      "epoch": 0.5831389100060294,
      "grad_norm": 0.7153684496879578,
      "learning_rate": 0.00011252503542196347,
      "loss": 3.2213,
      "step": 44490
    },
    {
      "epoch": 0.583269981912077,
      "grad_norm": 0.6850215196609497,
      "learning_rate": 0.00011246483659499224,
      "loss": 3.2397,
      "step": 44500
    },
    {
      "epoch": 0.583269981912077,
      "eval_accuracy": 0.398297165200391,
      "eval_loss": 3.238992214202881,
      "eval_runtime": 61.4001,
      "eval_samples_per_second": 32.573,
      "eval_steps_per_second": 0.521,
      "step": 44500
    },
    {
      "epoch": 0.5834010538181246,
      "grad_norm": 0.6450862884521484,
      "learning_rate": 0.00011240464421666491,
      "loss": 3.2092,
      "step": 44510
    },
    {
      "epoch": 0.5835321257241722,
      "grad_norm": 0.6891219615936279,
      "learning_rate": 0.00011234445829732269,
      "loss": 3.2113,
      "step": 44520
    },
    {
      "epoch": 0.5836631976302199,
      "grad_norm": 0.6447392106056213,
      "learning_rate": 0.00011228427884730565,
      "loss": 3.2471,
      "step": 44530
    },
    {
      "epoch": 0.5837942695362676,
      "grad_norm": 0.5949611663818359,
      "learning_rate": 0.00011222410587695284,
      "loss": 3.2318,
      "step": 44540
    },
    {
      "epoch": 0.5839253414423152,
      "grad_norm": 0.7082892060279846,
      "learning_rate": 0.00011216393939660218,
      "loss": 3.2211,
      "step": 44550
    },
    {
      "epoch": 0.5840564133483629,
      "grad_norm": 0.712929904460907,
      "learning_rate": 0.0001121037794165903,
      "loss": 3.2773,
      "step": 44560
    },
    {
      "epoch": 0.5841874852544106,
      "grad_norm": 0.6628018617630005,
      "learning_rate": 0.00011204362594725296,
      "loss": 3.2183,
      "step": 44570
    },
    {
      "epoch": 0.5843185571604582,
      "grad_norm": 0.6733047366142273,
      "learning_rate": 0.00011198347899892466,
      "loss": 3.2258,
      "step": 44580
    },
    {
      "epoch": 0.5844496290665059,
      "grad_norm": 0.7088468074798584,
      "learning_rate": 0.00011192333858193883,
      "loss": 3.2163,
      "step": 44590
    },
    {
      "epoch": 0.5845807009725535,
      "grad_norm": 0.6599271297454834,
      "learning_rate": 0.00011186320470662773,
      "loss": 3.148,
      "step": 44600
    },
    {
      "epoch": 0.5847117728786012,
      "grad_norm": 0.6286370754241943,
      "learning_rate": 0.0001118030773833225,
      "loss": 3.2324,
      "step": 44610
    },
    {
      "epoch": 0.5848428447846489,
      "grad_norm": 0.6679529547691345,
      "learning_rate": 0.00011174295662235329,
      "loss": 3.2783,
      "step": 44620
    },
    {
      "epoch": 0.5849739166906965,
      "grad_norm": 0.6306409239768982,
      "learning_rate": 0.00011168284243404885,
      "loss": 3.1994,
      "step": 44630
    },
    {
      "epoch": 0.5851049885967442,
      "grad_norm": 0.697668731212616,
      "learning_rate": 0.00011162273482873704,
      "loss": 3.2201,
      "step": 44640
    },
    {
      "epoch": 0.5852360605027919,
      "grad_norm": 0.622584342956543,
      "learning_rate": 0.00011156263381674453,
      "loss": 3.2528,
      "step": 44650
    },
    {
      "epoch": 0.5853671324088395,
      "grad_norm": 0.6450229287147522,
      "learning_rate": 0.0001115025394083969,
      "loss": 3.253,
      "step": 44660
    },
    {
      "epoch": 0.5854982043148872,
      "grad_norm": 0.60419762134552,
      "learning_rate": 0.00011144245161401838,
      "loss": 3.2338,
      "step": 44670
    },
    {
      "epoch": 0.5856292762209349,
      "grad_norm": 0.6282832622528076,
      "learning_rate": 0.00011138237044393234,
      "loss": 3.2676,
      "step": 44680
    },
    {
      "epoch": 0.5857603481269824,
      "grad_norm": 0.6811118721961975,
      "learning_rate": 0.0001113222959084608,
      "loss": 3.1963,
      "step": 44690
    },
    {
      "epoch": 0.5858914200330301,
      "grad_norm": 1.0865787267684937,
      "learning_rate": 0.00011126222801792485,
      "loss": 3.2251,
      "step": 44700
    },
    {
      "epoch": 0.5860224919390777,
      "grad_norm": 0.6359794735908508,
      "learning_rate": 0.00011120216678264417,
      "loss": 3.1995,
      "step": 44710
    },
    {
      "epoch": 0.5861535638451254,
      "grad_norm": 0.6867290139198303,
      "learning_rate": 0.00011114211221293755,
      "loss": 3.2397,
      "step": 44720
    },
    {
      "epoch": 0.5862846357511731,
      "grad_norm": 0.6383755207061768,
      "learning_rate": 0.00011108206431912252,
      "loss": 3.2526,
      "step": 44730
    },
    {
      "epoch": 0.5864157076572207,
      "grad_norm": 0.6334186792373657,
      "learning_rate": 0.00011102202311151544,
      "loss": 3.2405,
      "step": 44740
    },
    {
      "epoch": 0.5865467795632684,
      "grad_norm": 0.622083842754364,
      "learning_rate": 0.00011096198860043152,
      "loss": 3.2643,
      "step": 44750
    },
    {
      "epoch": 0.5866778514693161,
      "grad_norm": 0.6280596256256104,
      "learning_rate": 0.00011090196079618492,
      "loss": 3.2298,
      "step": 44760
    },
    {
      "epoch": 0.5868089233753637,
      "grad_norm": 0.6764287352561951,
      "learning_rate": 0.00011084193970908861,
      "loss": 3.2328,
      "step": 44770
    },
    {
      "epoch": 0.5869399952814114,
      "grad_norm": 0.7042009234428406,
      "learning_rate": 0.0001107819253494543,
      "loss": 3.2204,
      "step": 44780
    },
    {
      "epoch": 0.587071067187459,
      "grad_norm": 0.6416953206062317,
      "learning_rate": 0.00011072191772759261,
      "loss": 3.2227,
      "step": 44790
    },
    {
      "epoch": 0.5872021390935067,
      "grad_norm": 0.7152805924415588,
      "learning_rate": 0.00011066191685381308,
      "loss": 3.2203,
      "step": 44800
    },
    {
      "epoch": 0.5873332109995544,
      "grad_norm": 0.7257099151611328,
      "learning_rate": 0.00011060192273842405,
      "loss": 3.1585,
      "step": 44810
    },
    {
      "epoch": 0.587464282905602,
      "grad_norm": 0.6220877766609192,
      "learning_rate": 0.00011054193539173253,
      "loss": 3.2009,
      "step": 44820
    },
    {
      "epoch": 0.5875953548116497,
      "grad_norm": 0.7538883686065674,
      "learning_rate": 0.00011048195482404462,
      "loss": 3.2703,
      "step": 44830
    },
    {
      "epoch": 0.5877264267176974,
      "grad_norm": 0.582646369934082,
      "learning_rate": 0.00011042198104566513,
      "loss": 3.2065,
      "step": 44840
    },
    {
      "epoch": 0.587857498623745,
      "grad_norm": 0.6036754250526428,
      "learning_rate": 0.00011036201406689779,
      "loss": 3.2197,
      "step": 44850
    },
    {
      "epoch": 0.5879885705297927,
      "grad_norm": 0.6361536979675293,
      "learning_rate": 0.00011030205389804494,
      "loss": 3.2356,
      "step": 44860
    },
    {
      "epoch": 0.5881196424358403,
      "grad_norm": 0.6381674408912659,
      "learning_rate": 0.000110242100549408,
      "loss": 3.2201,
      "step": 44870
    },
    {
      "epoch": 0.5882507143418879,
      "grad_norm": 0.6409244537353516,
      "learning_rate": 0.00011018215403128711,
      "loss": 3.2135,
      "step": 44880
    },
    {
      "epoch": 0.5883817862479356,
      "grad_norm": 0.6465046405792236,
      "learning_rate": 0.00011012221435398123,
      "loss": 3.2258,
      "step": 44890
    },
    {
      "epoch": 0.5885128581539832,
      "grad_norm": 0.6357307434082031,
      "learning_rate": 0.00011006228152778813,
      "loss": 3.2068,
      "step": 44900
    },
    {
      "epoch": 0.5886439300600309,
      "grad_norm": 0.6043647527694702,
      "learning_rate": 0.00011000235556300449,
      "loss": 3.2516,
      "step": 44910
    },
    {
      "epoch": 0.5887750019660786,
      "grad_norm": 0.5827827453613281,
      "learning_rate": 0.00010994243646992577,
      "loss": 3.1988,
      "step": 44920
    },
    {
      "epoch": 0.5889060738721262,
      "grad_norm": 0.6246683597564697,
      "learning_rate": 0.00010988252425884616,
      "loss": 3.2076,
      "step": 44930
    },
    {
      "epoch": 0.5890371457781739,
      "grad_norm": 0.6957534551620483,
      "learning_rate": 0.00010982261894005876,
      "loss": 3.2656,
      "step": 44940
    },
    {
      "epoch": 0.5891682176842216,
      "grad_norm": 0.6880134344100952,
      "learning_rate": 0.00010976272052385552,
      "loss": 3.2634,
      "step": 44950
    },
    {
      "epoch": 0.5892992895902692,
      "grad_norm": 0.6150718331336975,
      "learning_rate": 0.00010970282902052714,
      "loss": 3.2244,
      "step": 44960
    },
    {
      "epoch": 0.5894303614963169,
      "grad_norm": 0.6648474335670471,
      "learning_rate": 0.0001096429444403631,
      "loss": 3.2503,
      "step": 44970
    },
    {
      "epoch": 0.5895614334023646,
      "grad_norm": 0.6725148558616638,
      "learning_rate": 0.00010958306679365174,
      "loss": 3.1912,
      "step": 44980
    },
    {
      "epoch": 0.5896925053084122,
      "grad_norm": 0.6671502590179443,
      "learning_rate": 0.00010952319609068026,
      "loss": 3.2171,
      "step": 44990
    },
    {
      "epoch": 0.5898235772144599,
      "grad_norm": 0.6187174916267395,
      "learning_rate": 0.00010946333234173454,
      "loss": 3.2357,
      "step": 45000
    },
    {
      "epoch": 0.5898235772144599,
      "eval_accuracy": 0.3985596285434995,
      "eval_loss": 3.236715316772461,
      "eval_runtime": 60.3684,
      "eval_samples_per_second": 33.13,
      "eval_steps_per_second": 0.53,
      "step": 45000
    },
    {
      "epoch": 0.5899546491205075,
      "grad_norm": 0.6462904810905457,
      "learning_rate": 0.00010940347555709937,
      "loss": 3.2322,
      "step": 45010
    },
    {
      "epoch": 0.5900857210265552,
      "grad_norm": 0.6593872904777527,
      "learning_rate": 0.0001093436257470583,
      "loss": 3.2432,
      "step": 45020
    },
    {
      "epoch": 0.5902167929326029,
      "grad_norm": 0.9725048542022705,
      "learning_rate": 0.00010928378292189378,
      "loss": 3.2152,
      "step": 45030
    },
    {
      "epoch": 0.5903478648386505,
      "grad_norm": 0.6592382788658142,
      "learning_rate": 0.0001092239470918868,
      "loss": 3.2233,
      "step": 45040
    },
    {
      "epoch": 0.5904789367446981,
      "grad_norm": 0.7442682385444641,
      "learning_rate": 0.00010916411826731746,
      "loss": 3.2805,
      "step": 45050
    },
    {
      "epoch": 0.5906100086507458,
      "grad_norm": 0.6761847138404846,
      "learning_rate": 0.00010910429645846443,
      "loss": 3.2302,
      "step": 45060
    },
    {
      "epoch": 0.5907410805567934,
      "grad_norm": 0.6358495950698853,
      "learning_rate": 0.00010904448167560536,
      "loss": 3.2247,
      "step": 45070
    },
    {
      "epoch": 0.5908721524628411,
      "grad_norm": 0.639390230178833,
      "learning_rate": 0.00010898467392901646,
      "loss": 3.2003,
      "step": 45080
    },
    {
      "epoch": 0.5910032243688887,
      "grad_norm": 0.7471226453781128,
      "learning_rate": 0.00010892487322897298,
      "loss": 3.2677,
      "step": 45090
    },
    {
      "epoch": 0.5911342962749364,
      "grad_norm": 0.7535632252693176,
      "learning_rate": 0.00010886507958574875,
      "loss": 3.2019,
      "step": 45100
    },
    {
      "epoch": 0.5912653681809841,
      "grad_norm": 0.637824535369873,
      "learning_rate": 0.00010880529300961662,
      "loss": 3.2225,
      "step": 45110
    },
    {
      "epoch": 0.5913964400870317,
      "grad_norm": 0.6344773173332214,
      "learning_rate": 0.00010874551351084791,
      "loss": 3.2277,
      "step": 45120
    },
    {
      "epoch": 0.5915275119930794,
      "grad_norm": 0.6142618656158447,
      "learning_rate": 0.00010868574109971306,
      "loss": 3.2281,
      "step": 45130
    },
    {
      "epoch": 0.5916585838991271,
      "grad_norm": 0.6277701258659363,
      "learning_rate": 0.00010862597578648103,
      "loss": 3.2728,
      "step": 45140
    },
    {
      "epoch": 0.5917896558051747,
      "grad_norm": 0.6181873679161072,
      "learning_rate": 0.00010856621758141972,
      "loss": 3.2338,
      "step": 45150
    },
    {
      "epoch": 0.5919207277112224,
      "grad_norm": 0.6512242555618286,
      "learning_rate": 0.0001085064664947957,
      "loss": 3.233,
      "step": 45160
    },
    {
      "epoch": 0.5920517996172701,
      "grad_norm": 0.806002676486969,
      "learning_rate": 0.0001084467225368744,
      "loss": 3.2202,
      "step": 45170
    },
    {
      "epoch": 0.5921828715233177,
      "grad_norm": 0.6520618796348572,
      "learning_rate": 0.00010838698571792008,
      "loss": 3.2675,
      "step": 45180
    },
    {
      "epoch": 0.5923139434293654,
      "grad_norm": 0.658032238483429,
      "learning_rate": 0.00010832725604819551,
      "loss": 3.2514,
      "step": 45190
    },
    {
      "epoch": 0.592445015335413,
      "grad_norm": 0.7252050638198853,
      "learning_rate": 0.00010826753353796253,
      "loss": 3.32,
      "step": 45200
    },
    {
      "epoch": 0.5925760872414607,
      "grad_norm": 0.6357904672622681,
      "learning_rate": 0.00010820781819748164,
      "loss": 3.2295,
      "step": 45210
    },
    {
      "epoch": 0.5927071591475084,
      "grad_norm": 0.7491424083709717,
      "learning_rate": 0.00010814811003701206,
      "loss": 3.1749,
      "step": 45220
    },
    {
      "epoch": 0.5928382310535559,
      "grad_norm": 1.6011720895767212,
      "learning_rate": 0.00010808840906681184,
      "loss": 3.2261,
      "step": 45230
    },
    {
      "epoch": 0.5929693029596036,
      "grad_norm": 0.6518107056617737,
      "learning_rate": 0.00010802871529713773,
      "loss": 3.2361,
      "step": 45240
    },
    {
      "epoch": 0.5931003748656513,
      "grad_norm": 0.6279640197753906,
      "learning_rate": 0.00010796902873824533,
      "loss": 3.2569,
      "step": 45250
    },
    {
      "epoch": 0.5932314467716989,
      "grad_norm": 0.6637915372848511,
      "learning_rate": 0.00010790934940038891,
      "loss": 3.2331,
      "step": 45260
    },
    {
      "epoch": 0.5933625186777466,
      "grad_norm": 0.6248162388801575,
      "learning_rate": 0.00010784967729382156,
      "loss": 3.224,
      "step": 45270
    },
    {
      "epoch": 0.5934935905837943,
      "grad_norm": 0.625615656375885,
      "learning_rate": 0.0001077900124287951,
      "loss": 3.2201,
      "step": 45280
    },
    {
      "epoch": 0.5936246624898419,
      "grad_norm": 0.598606288433075,
      "learning_rate": 0.0001077303548155602,
      "loss": 3.2099,
      "step": 45290
    },
    {
      "epoch": 0.5937557343958896,
      "grad_norm": 0.6347043514251709,
      "learning_rate": 0.00010767070446436606,
      "loss": 3.2566,
      "step": 45300
    },
    {
      "epoch": 0.5938868063019372,
      "grad_norm": 0.6267805695533752,
      "learning_rate": 0.00010761106138546092,
      "loss": 3.1916,
      "step": 45310
    },
    {
      "epoch": 0.5940178782079849,
      "grad_norm": 0.622682511806488,
      "learning_rate": 0.00010755142558909149,
      "loss": 3.246,
      "step": 45320
    },
    {
      "epoch": 0.5941489501140326,
      "grad_norm": 0.6557353138923645,
      "learning_rate": 0.00010749179708550352,
      "loss": 3.2429,
      "step": 45330
    },
    {
      "epoch": 0.5942800220200802,
      "grad_norm": 0.6622795462608337,
      "learning_rate": 0.00010743217588494117,
      "loss": 3.1969,
      "step": 45340
    },
    {
      "epoch": 0.5944110939261279,
      "grad_norm": 0.7286412119865417,
      "learning_rate": 0.00010737256199764763,
      "loss": 3.2015,
      "step": 45350
    },
    {
      "epoch": 0.5945421658321756,
      "grad_norm": 0.711043655872345,
      "learning_rate": 0.00010731295543386475,
      "loss": 3.2784,
      "step": 45360
    },
    {
      "epoch": 0.5946732377382232,
      "grad_norm": 0.6559299826622009,
      "learning_rate": 0.00010725335620383302,
      "loss": 3.2224,
      "step": 45370
    },
    {
      "epoch": 0.5948043096442709,
      "grad_norm": 0.6475300788879395,
      "learning_rate": 0.00010719376431779178,
      "loss": 3.2054,
      "step": 45380
    },
    {
      "epoch": 0.5949353815503186,
      "grad_norm": 0.8150238990783691,
      "learning_rate": 0.00010713417978597914,
      "loss": 3.2589,
      "step": 45390
    },
    {
      "epoch": 0.5950664534563662,
      "grad_norm": 0.6795294880867004,
      "learning_rate": 0.00010707460261863183,
      "loss": 3.2312,
      "step": 45400
    },
    {
      "epoch": 0.5951975253624138,
      "grad_norm": 0.7004638314247131,
      "learning_rate": 0.00010701503282598537,
      "loss": 3.2818,
      "step": 45410
    },
    {
      "epoch": 0.5953285972684614,
      "grad_norm": 0.8439451456069946,
      "learning_rate": 0.000106955470418274,
      "loss": 3.235,
      "step": 45420
    },
    {
      "epoch": 0.5954596691745091,
      "grad_norm": 0.6194049715995789,
      "learning_rate": 0.00010689591540573075,
      "loss": 3.284,
      "step": 45430
    },
    {
      "epoch": 0.5955907410805568,
      "grad_norm": 0.6306747198104858,
      "learning_rate": 0.00010683636779858734,
      "loss": 3.2745,
      "step": 45440
    },
    {
      "epoch": 0.5957218129866044,
      "grad_norm": 0.6995223164558411,
      "learning_rate": 0.00010677682760707415,
      "loss": 3.3255,
      "step": 45450
    },
    {
      "epoch": 0.5958528848926521,
      "grad_norm": 0.5937404632568359,
      "learning_rate": 0.00010671729484142036,
      "loss": 3.2184,
      "step": 45460
    },
    {
      "epoch": 0.5959839567986998,
      "grad_norm": 0.6267967820167542,
      "learning_rate": 0.0001066577695118539,
      "loss": 3.2326,
      "step": 45470
    },
    {
      "epoch": 0.5961150287047474,
      "grad_norm": 0.6738429069519043,
      "learning_rate": 0.00010659825162860142,
      "loss": 3.2218,
      "step": 45480
    },
    {
      "epoch": 0.5962461006107951,
      "grad_norm": 0.6594774127006531,
      "learning_rate": 0.00010653874120188817,
      "loss": 3.2703,
      "step": 45490
    },
    {
      "epoch": 0.5963771725168427,
      "grad_norm": 0.6727324724197388,
      "learning_rate": 0.00010647923824193823,
      "loss": 3.2919,
      "step": 45500
    },
    {
      "epoch": 0.5963771725168427,
      "eval_accuracy": 0.3984066471163245,
      "eval_loss": 3.235849142074585,
      "eval_runtime": 58.6107,
      "eval_samples_per_second": 34.123,
      "eval_steps_per_second": 0.546,
      "step": 45500
    },
    {
      "epoch": 0.5965082444228904,
      "grad_norm": 0.6523264050483704,
      "learning_rate": 0.00010641974275897444,
      "loss": 3.245,
      "step": 45510
    },
    {
      "epoch": 0.5966393163289381,
      "grad_norm": 0.6391008496284485,
      "learning_rate": 0.00010636025476321815,
      "loss": 3.2167,
      "step": 45520
    },
    {
      "epoch": 0.5967703882349857,
      "grad_norm": 0.7676718235015869,
      "learning_rate": 0.00010630077426488964,
      "loss": 3.2597,
      "step": 45530
    },
    {
      "epoch": 0.5969014601410334,
      "grad_norm": 0.646519124507904,
      "learning_rate": 0.00010624130127420784,
      "loss": 3.2352,
      "step": 45540
    },
    {
      "epoch": 0.5970325320470811,
      "grad_norm": 0.6474140286445618,
      "learning_rate": 0.0001061818358013904,
      "loss": 3.2431,
      "step": 45550
    },
    {
      "epoch": 0.5971636039531287,
      "grad_norm": 0.6045957803726196,
      "learning_rate": 0.00010612237785665354,
      "loss": 3.2224,
      "step": 45560
    },
    {
      "epoch": 0.5972946758591764,
      "grad_norm": 0.6328110694885254,
      "learning_rate": 0.00010606292745021236,
      "loss": 3.2193,
      "step": 45570
    },
    {
      "epoch": 0.597425747765224,
      "grad_norm": 0.6645777225494385,
      "learning_rate": 0.00010600348459228064,
      "loss": 3.2143,
      "step": 45580
    },
    {
      "epoch": 0.5975568196712716,
      "grad_norm": 0.6423099637031555,
      "learning_rate": 0.00010594404929307079,
      "loss": 3.2115,
      "step": 45590
    },
    {
      "epoch": 0.5976878915773193,
      "grad_norm": 0.6486244797706604,
      "learning_rate": 0.00010588462156279397,
      "loss": 3.2286,
      "step": 45600
    },
    {
      "epoch": 0.5978189634833669,
      "grad_norm": 0.6645761132240295,
      "learning_rate": 0.00010582520141165997,
      "loss": 3.1991,
      "step": 45610
    },
    {
      "epoch": 0.5979500353894146,
      "grad_norm": 0.6308107376098633,
      "learning_rate": 0.00010576578884987746,
      "loss": 3.2329,
      "step": 45620
    },
    {
      "epoch": 0.5980811072954623,
      "grad_norm": 0.6584576964378357,
      "learning_rate": 0.00010570638388765351,
      "loss": 3.2397,
      "step": 45630
    },
    {
      "epoch": 0.5982121792015099,
      "grad_norm": 0.624386727809906,
      "learning_rate": 0.00010564698653519415,
      "loss": 3.1902,
      "step": 45640
    },
    {
      "epoch": 0.5983432511075576,
      "grad_norm": 0.7340356111526489,
      "learning_rate": 0.00010558759680270402,
      "loss": 3.2489,
      "step": 45650
    },
    {
      "epoch": 0.5984743230136053,
      "grad_norm": 0.6534768342971802,
      "learning_rate": 0.00010552821470038648,
      "loss": 3.2248,
      "step": 45660
    },
    {
      "epoch": 0.5986053949196529,
      "grad_norm": 0.6578162312507629,
      "learning_rate": 0.00010546884023844341,
      "loss": 3.2442,
      "step": 45670
    },
    {
      "epoch": 0.5987364668257006,
      "grad_norm": 0.6339426636695862,
      "learning_rate": 0.00010540947342707561,
      "loss": 3.2052,
      "step": 45680
    },
    {
      "epoch": 0.5988675387317483,
      "grad_norm": 0.6208382248878479,
      "learning_rate": 0.00010535011427648239,
      "loss": 3.2425,
      "step": 45690
    },
    {
      "epoch": 0.5989986106377959,
      "grad_norm": 0.5985956192016602,
      "learning_rate": 0.00010529076279686193,
      "loss": 3.2303,
      "step": 45700
    },
    {
      "epoch": 0.5991296825438436,
      "grad_norm": 0.6076611876487732,
      "learning_rate": 0.00010523141899841083,
      "loss": 3.2712,
      "step": 45710
    },
    {
      "epoch": 0.5992607544498912,
      "grad_norm": 0.6446940898895264,
      "learning_rate": 0.00010517208289132458,
      "loss": 3.1997,
      "step": 45720
    },
    {
      "epoch": 0.5993918263559389,
      "grad_norm": 0.6885919570922852,
      "learning_rate": 0.00010511275448579736,
      "loss": 3.2407,
      "step": 45730
    },
    {
      "epoch": 0.5995228982619866,
      "grad_norm": 0.6789318323135376,
      "learning_rate": 0.0001050534337920218,
      "loss": 3.2104,
      "step": 45740
    },
    {
      "epoch": 0.5996539701680342,
      "grad_norm": 0.619826078414917,
      "learning_rate": 0.00010499412082018947,
      "loss": 3.2586,
      "step": 45750
    },
    {
      "epoch": 0.5997850420740818,
      "grad_norm": 0.7688161134719849,
      "learning_rate": 0.00010493481558049048,
      "loss": 3.2275,
      "step": 45760
    },
    {
      "epoch": 0.5999161139801295,
      "grad_norm": 0.5811552405357361,
      "learning_rate": 0.00010487551808311364,
      "loss": 3.2332,
      "step": 45770
    },
    {
      "epoch": 0.6000471858861771,
      "grad_norm": 0.6393649578094482,
      "learning_rate": 0.00010481622833824639,
      "loss": 3.1941,
      "step": 45780
    },
    {
      "epoch": 0.6001782577922248,
      "grad_norm": 0.8356134295463562,
      "learning_rate": 0.00010475694635607486,
      "loss": 3.2014,
      "step": 45790
    },
    {
      "epoch": 0.6003093296982724,
      "grad_norm": 0.6784019470214844,
      "learning_rate": 0.0001046976721467839,
      "loss": 3.2397,
      "step": 45800
    },
    {
      "epoch": 0.6004404016043201,
      "grad_norm": 0.6365929841995239,
      "learning_rate": 0.00010463840572055701,
      "loss": 3.1868,
      "step": 45810
    },
    {
      "epoch": 0.6005714735103678,
      "grad_norm": 0.657372236251831,
      "learning_rate": 0.00010457914708757624,
      "loss": 3.2061,
      "step": 45820
    },
    {
      "epoch": 0.6007025454164154,
      "grad_norm": 0.6439395546913147,
      "learning_rate": 0.0001045198962580224,
      "loss": 3.2128,
      "step": 45830
    },
    {
      "epoch": 0.6008336173224631,
      "grad_norm": 0.6316313743591309,
      "learning_rate": 0.00010446065324207502,
      "loss": 3.1899,
      "step": 45840
    },
    {
      "epoch": 0.6009646892285108,
      "grad_norm": 0.6790297627449036,
      "learning_rate": 0.00010440141804991216,
      "loss": 3.2323,
      "step": 45850
    },
    {
      "epoch": 0.6010957611345584,
      "grad_norm": 0.7063311338424683,
      "learning_rate": 0.00010434219069171059,
      "loss": 3.1964,
      "step": 45860
    },
    {
      "epoch": 0.6012268330406061,
      "grad_norm": 0.591114342212677,
      "learning_rate": 0.0001042829711776457,
      "loss": 3.2506,
      "step": 45870
    },
    {
      "epoch": 0.6013579049466538,
      "grad_norm": 0.6354190707206726,
      "learning_rate": 0.00010422375951789168,
      "loss": 3.2263,
      "step": 45880
    },
    {
      "epoch": 0.6014889768527014,
      "grad_norm": 0.6239058971405029,
      "learning_rate": 0.00010416455572262108,
      "loss": 3.2682,
      "step": 45890
    },
    {
      "epoch": 0.6016200487587491,
      "grad_norm": 0.7289928197860718,
      "learning_rate": 0.00010410535980200538,
      "loss": 3.2227,
      "step": 45900
    },
    {
      "epoch": 0.6017511206647967,
      "grad_norm": 0.737095057964325,
      "learning_rate": 0.00010404617176621459,
      "loss": 3.2698,
      "step": 45910
    },
    {
      "epoch": 0.6018821925708444,
      "grad_norm": 0.6657622456550598,
      "learning_rate": 0.00010398699162541744,
      "loss": 3.2454,
      "step": 45920
    },
    {
      "epoch": 0.6020132644768921,
      "grad_norm": 0.6507141590118408,
      "learning_rate": 0.0001039278193897811,
      "loss": 3.2146,
      "step": 45930
    },
    {
      "epoch": 0.6021443363829396,
      "grad_norm": 0.6198927164077759,
      "learning_rate": 0.00010386865506947166,
      "loss": 3.2247,
      "step": 45940
    },
    {
      "epoch": 0.6022754082889873,
      "grad_norm": 0.6042706966400146,
      "learning_rate": 0.00010380949867465361,
      "loss": 3.2705,
      "step": 45950
    },
    {
      "epoch": 0.602406480195035,
      "grad_norm": 0.7463015913963318,
      "learning_rate": 0.00010375035021549026,
      "loss": 3.2525,
      "step": 45960
    },
    {
      "epoch": 0.6025375521010826,
      "grad_norm": 0.5954564809799194,
      "learning_rate": 0.00010369120970214342,
      "loss": 3.2459,
      "step": 45970
    },
    {
      "epoch": 0.6026686240071303,
      "grad_norm": 0.6506423354148865,
      "learning_rate": 0.00010363207714477359,
      "loss": 3.1874,
      "step": 45980
    },
    {
      "epoch": 0.602799695913178,
      "grad_norm": 0.8579496741294861,
      "learning_rate": 0.00010357295255353997,
      "loss": 3.2692,
      "step": 45990
    },
    {
      "epoch": 0.6029307678192256,
      "grad_norm": 0.6938508152961731,
      "learning_rate": 0.00010351383593860024,
      "loss": 3.2239,
      "step": 46000
    },
    {
      "epoch": 0.6029307678192256,
      "eval_accuracy": 0.3989496578690127,
      "eval_loss": 3.2324368953704834,
      "eval_runtime": 59.071,
      "eval_samples_per_second": 33.858,
      "eval_steps_per_second": 0.542,
      "step": 46000
    },
    {
      "epoch": 0.6030618397252733,
      "grad_norm": 0.6219264268875122,
      "learning_rate": 0.00010345472731011084,
      "loss": 3.2245,
      "step": 46010
    },
    {
      "epoch": 0.6031929116313209,
      "grad_norm": 0.614022433757782,
      "learning_rate": 0.0001033956266782268,
      "loss": 3.2619,
      "step": 46020
    },
    {
      "epoch": 0.6033239835373686,
      "grad_norm": 0.6726503968238831,
      "learning_rate": 0.0001033365340531018,
      "loss": 3.2604,
      "step": 46030
    },
    {
      "epoch": 0.6034550554434163,
      "grad_norm": 0.7042705416679382,
      "learning_rate": 0.00010327744944488807,
      "loss": 3.2621,
      "step": 46040
    },
    {
      "epoch": 0.6035861273494639,
      "grad_norm": 0.5855175852775574,
      "learning_rate": 0.00010321837286373647,
      "loss": 3.2132,
      "step": 46050
    },
    {
      "epoch": 0.6037171992555116,
      "grad_norm": 0.6653004884719849,
      "learning_rate": 0.00010315930431979658,
      "loss": 3.2278,
      "step": 46060
    },
    {
      "epoch": 0.6038482711615593,
      "grad_norm": 0.6612849235534668,
      "learning_rate": 0.00010310024382321658,
      "loss": 3.2559,
      "step": 46070
    },
    {
      "epoch": 0.6039793430676069,
      "grad_norm": 0.6311576962471008,
      "learning_rate": 0.0001030411913841431,
      "loss": 3.2318,
      "step": 46080
    },
    {
      "epoch": 0.6041104149736546,
      "grad_norm": 0.6075425744056702,
      "learning_rate": 0.00010298214701272159,
      "loss": 3.2248,
      "step": 46090
    },
    {
      "epoch": 0.6042414868797022,
      "grad_norm": 0.6190316677093506,
      "learning_rate": 0.00010292311071909602,
      "loss": 3.2141,
      "step": 46100
    },
    {
      "epoch": 0.6043725587857499,
      "grad_norm": 0.6562777161598206,
      "learning_rate": 0.00010286408251340905,
      "loss": 3.2349,
      "step": 46110
    },
    {
      "epoch": 0.6045036306917975,
      "grad_norm": 0.608486533164978,
      "learning_rate": 0.00010280506240580179,
      "loss": 3.2307,
      "step": 46120
    },
    {
      "epoch": 0.6046347025978451,
      "grad_norm": 0.6395454406738281,
      "learning_rate": 0.00010274605040641408,
      "loss": 3.2475,
      "step": 46130
    },
    {
      "epoch": 0.6047657745038928,
      "grad_norm": 0.6653308868408203,
      "learning_rate": 0.0001026870465253844,
      "loss": 3.2326,
      "step": 46140
    },
    {
      "epoch": 0.6048968464099405,
      "grad_norm": 0.7694252133369446,
      "learning_rate": 0.00010262805077284969,
      "loss": 3.2535,
      "step": 46150
    },
    {
      "epoch": 0.6050279183159881,
      "grad_norm": 0.6712005734443665,
      "learning_rate": 0.00010256906315894562,
      "loss": 3.242,
      "step": 46160
    },
    {
      "epoch": 0.6051589902220358,
      "grad_norm": 0.6094689965248108,
      "learning_rate": 0.00010251008369380645,
      "loss": 3.2578,
      "step": 46170
    },
    {
      "epoch": 0.6052900621280835,
      "grad_norm": 0.6993538737297058,
      "learning_rate": 0.00010245111238756503,
      "loss": 3.2303,
      "step": 46180
    },
    {
      "epoch": 0.6054211340341311,
      "grad_norm": 0.6670473217964172,
      "learning_rate": 0.0001023921492503527,
      "loss": 3.218,
      "step": 46190
    },
    {
      "epoch": 0.6055522059401788,
      "grad_norm": 0.6516269445419312,
      "learning_rate": 0.00010233319429229955,
      "loss": 3.2395,
      "step": 46200
    },
    {
      "epoch": 0.6056832778462264,
      "grad_norm": 0.6479690074920654,
      "learning_rate": 0.00010227424752353422,
      "loss": 3.2595,
      "step": 46210
    },
    {
      "epoch": 0.6058143497522741,
      "grad_norm": 0.606497585773468,
      "learning_rate": 0.00010221530895418396,
      "loss": 3.2286,
      "step": 46220
    },
    {
      "epoch": 0.6059454216583218,
      "grad_norm": 0.6444209218025208,
      "learning_rate": 0.00010215637859437445,
      "loss": 3.2202,
      "step": 46230
    },
    {
      "epoch": 0.6060764935643694,
      "grad_norm": 0.6833163499832153,
      "learning_rate": 0.00010209745645423016,
      "loss": 3.287,
      "step": 46240
    },
    {
      "epoch": 0.6062075654704171,
      "grad_norm": 0.6393874287605286,
      "learning_rate": 0.00010203854254387413,
      "loss": 3.2517,
      "step": 46250
    },
    {
      "epoch": 0.6063386373764648,
      "grad_norm": 0.7110936045646667,
      "learning_rate": 0.00010197963687342781,
      "loss": 3.2121,
      "step": 46260
    },
    {
      "epoch": 0.6064697092825124,
      "grad_norm": 0.6357977986335754,
      "learning_rate": 0.00010192073945301145,
      "loss": 3.2764,
      "step": 46270
    },
    {
      "epoch": 0.6066007811885601,
      "grad_norm": 0.6365695595741272,
      "learning_rate": 0.00010186185029274372,
      "loss": 3.2677,
      "step": 46280
    },
    {
      "epoch": 0.6067318530946078,
      "grad_norm": 0.649030864238739,
      "learning_rate": 0.00010180296940274207,
      "loss": 3.2464,
      "step": 46290
    },
    {
      "epoch": 0.6068629250006553,
      "grad_norm": 0.651207447052002,
      "learning_rate": 0.00010174409679312222,
      "loss": 3.2355,
      "step": 46300
    },
    {
      "epoch": 0.606993996906703,
      "grad_norm": 0.6729286313056946,
      "learning_rate": 0.0001016852324739988,
      "loss": 3.2191,
      "step": 46310
    },
    {
      "epoch": 0.6071250688127506,
      "grad_norm": 0.737335741519928,
      "learning_rate": 0.00010162637645548474,
      "loss": 3.1865,
      "step": 46320
    },
    {
      "epoch": 0.6072561407187983,
      "grad_norm": 0.660658597946167,
      "learning_rate": 0.00010156752874769177,
      "loss": 3.2212,
      "step": 46330
    },
    {
      "epoch": 0.607387212624846,
      "grad_norm": 0.6110531091690063,
      "learning_rate": 0.00010150868936072999,
      "loss": 3.2376,
      "step": 46340
    },
    {
      "epoch": 0.6075182845308936,
      "grad_norm": 0.5958556532859802,
      "learning_rate": 0.00010144985830470821,
      "loss": 3.2181,
      "step": 46350
    },
    {
      "epoch": 0.6076493564369413,
      "grad_norm": 0.7129707336425781,
      "learning_rate": 0.00010139103558973384,
      "loss": 3.2596,
      "step": 46360
    },
    {
      "epoch": 0.607780428342989,
      "grad_norm": 0.665770411491394,
      "learning_rate": 0.00010133222122591265,
      "loss": 3.2223,
      "step": 46370
    },
    {
      "epoch": 0.6079115002490366,
      "grad_norm": 0.6651104688644409,
      "learning_rate": 0.00010127341522334918,
      "loss": 3.2264,
      "step": 46380
    },
    {
      "epoch": 0.6080425721550843,
      "grad_norm": 1.0241354703903198,
      "learning_rate": 0.00010121461759214649,
      "loss": 3.2714,
      "step": 46390
    },
    {
      "epoch": 0.608173644061132,
      "grad_norm": 0.7431330680847168,
      "learning_rate": 0.0001011558283424061,
      "loss": 3.2469,
      "step": 46400
    },
    {
      "epoch": 0.6083047159671796,
      "grad_norm": 0.6433840990066528,
      "learning_rate": 0.00010109704748422824,
      "loss": 3.2413,
      "step": 46410
    },
    {
      "epoch": 0.6084357878732273,
      "grad_norm": 0.6310520768165588,
      "learning_rate": 0.00010103827502771156,
      "loss": 3.2671,
      "step": 46420
    },
    {
      "epoch": 0.6085668597792749,
      "grad_norm": 0.6456714868545532,
      "learning_rate": 0.00010097951098295333,
      "loss": 3.2625,
      "step": 46430
    },
    {
      "epoch": 0.6086979316853226,
      "grad_norm": 0.6332817673683167,
      "learning_rate": 0.0001009207553600495,
      "loss": 3.2386,
      "step": 46440
    },
    {
      "epoch": 0.6088290035913703,
      "grad_norm": 0.7312588095664978,
      "learning_rate": 0.00010086200816909425,
      "loss": 3.2637,
      "step": 46450
    },
    {
      "epoch": 0.6089600754974179,
      "grad_norm": 0.622722864151001,
      "learning_rate": 0.00010080326942018063,
      "loss": 3.2665,
      "step": 46460
    },
    {
      "epoch": 0.6090911474034656,
      "grad_norm": 0.6822323799133301,
      "learning_rate": 0.00010074453912340011,
      "loss": 3.2037,
      "step": 46470
    },
    {
      "epoch": 0.6092222193095131,
      "grad_norm": 0.7424899339675903,
      "learning_rate": 0.0001006858172888427,
      "loss": 3.2416,
      "step": 46480
    },
    {
      "epoch": 0.6093532912155608,
      "grad_norm": 0.6434851288795471,
      "learning_rate": 0.00010062710392659695,
      "loss": 3.2346,
      "step": 46490
    },
    {
      "epoch": 0.6094843631216085,
      "grad_norm": 0.7062909007072449,
      "learning_rate": 0.00010056839904675,
      "loss": 3.2432,
      "step": 46500
    },
    {
      "epoch": 0.6094843631216085,
      "eval_accuracy": 0.39903616813294235,
      "eval_loss": 3.2309646606445312,
      "eval_runtime": 60.7899,
      "eval_samples_per_second": 32.9,
      "eval_steps_per_second": 0.526,
      "step": 46500
    },
    {
      "epoch": 0.6096154350276561,
      "grad_norm": 0.7248334288597107,
      "learning_rate": 0.00010050970265938755,
      "loss": 3.2191,
      "step": 46510
    },
    {
      "epoch": 0.6097465069337038,
      "grad_norm": 0.6939072608947754,
      "learning_rate": 0.0001004510147745937,
      "loss": 3.2081,
      "step": 46520
    },
    {
      "epoch": 0.6098775788397515,
      "grad_norm": 0.6874194145202637,
      "learning_rate": 0.00010039233540245125,
      "loss": 3.2095,
      "step": 46530
    },
    {
      "epoch": 0.6100086507457991,
      "grad_norm": 0.6661149859428406,
      "learning_rate": 0.00010033366455304145,
      "loss": 3.212,
      "step": 46540
    },
    {
      "epoch": 0.6101397226518468,
      "grad_norm": 0.6120688915252686,
      "learning_rate": 0.00010027500223644424,
      "loss": 3.2682,
      "step": 46550
    },
    {
      "epoch": 0.6102707945578945,
      "grad_norm": 0.6188989281654358,
      "learning_rate": 0.00010021634846273775,
      "loss": 3.2536,
      "step": 46560
    },
    {
      "epoch": 0.6104018664639421,
      "grad_norm": 0.677006721496582,
      "learning_rate": 0.00010015770324199903,
      "loss": 3.2352,
      "step": 46570
    },
    {
      "epoch": 0.6105329383699898,
      "grad_norm": 0.5996061563491821,
      "learning_rate": 0.00010009906658430341,
      "loss": 3.235,
      "step": 46580
    },
    {
      "epoch": 0.6106640102760374,
      "grad_norm": 0.7179942727088928,
      "learning_rate": 0.00010004043849972489,
      "loss": 3.1978,
      "step": 46590
    },
    {
      "epoch": 0.6107950821820851,
      "grad_norm": 0.6874227523803711,
      "learning_rate": 9.998181899833583e-05,
      "loss": 3.1724,
      "step": 46600
    },
    {
      "epoch": 0.6109261540881328,
      "grad_norm": 0.7165815830230713,
      "learning_rate": 9.99232080902073e-05,
      "loss": 3.1896,
      "step": 46610
    },
    {
      "epoch": 0.6110572259941804,
      "grad_norm": 0.6888381242752075,
      "learning_rate": 9.986460578540886e-05,
      "loss": 3.2641,
      "step": 46620
    },
    {
      "epoch": 0.6111882979002281,
      "grad_norm": 0.7022085785865784,
      "learning_rate": 9.980601209400843e-05,
      "loss": 3.2582,
      "step": 46630
    },
    {
      "epoch": 0.6113193698062758,
      "grad_norm": 0.6943024396896362,
      "learning_rate": 9.974742702607262e-05,
      "loss": 3.1946,
      "step": 46640
    },
    {
      "epoch": 0.6114504417123234,
      "grad_norm": 0.6320481896400452,
      "learning_rate": 9.968885059166653e-05,
      "loss": 3.2411,
      "step": 46650
    },
    {
      "epoch": 0.611581513618371,
      "grad_norm": 0.6720809936523438,
      "learning_rate": 9.963028280085373e-05,
      "loss": 3.2235,
      "step": 46660
    },
    {
      "epoch": 0.6117125855244187,
      "grad_norm": 0.6375182867050171,
      "learning_rate": 9.957172366369632e-05,
      "loss": 3.2466,
      "step": 46670
    },
    {
      "epoch": 0.6118436574304663,
      "grad_norm": 0.6096304655075073,
      "learning_rate": 9.951317319025491e-05,
      "loss": 3.228,
      "step": 46680
    },
    {
      "epoch": 0.611974729336514,
      "grad_norm": 0.6518574953079224,
      "learning_rate": 9.945463139058867e-05,
      "loss": 3.2361,
      "step": 46690
    },
    {
      "epoch": 0.6121058012425616,
      "grad_norm": 0.7310232520103455,
      "learning_rate": 9.939609827475524e-05,
      "loss": 3.2229,
      "step": 46700
    },
    {
      "epoch": 0.6122368731486093,
      "grad_norm": 0.5949954986572266,
      "learning_rate": 9.933757385281073e-05,
      "loss": 3.2449,
      "step": 46710
    },
    {
      "epoch": 0.612367945054657,
      "grad_norm": 0.6023111343383789,
      "learning_rate": 9.927905813480977e-05,
      "loss": 3.2233,
      "step": 46720
    },
    {
      "epoch": 0.6124990169607046,
      "grad_norm": 0.6577599048614502,
      "learning_rate": 9.922055113080565e-05,
      "loss": 3.2113,
      "step": 46730
    },
    {
      "epoch": 0.6126300888667523,
      "grad_norm": 0.6531931161880493,
      "learning_rate": 9.91620528508499e-05,
      "loss": 3.2246,
      "step": 46740
    },
    {
      "epoch": 0.6127611607728,
      "grad_norm": 0.7038547396659851,
      "learning_rate": 9.910356330499277e-05,
      "loss": 3.2321,
      "step": 46750
    },
    {
      "epoch": 0.6128922326788476,
      "grad_norm": 0.838652491569519,
      "learning_rate": 9.904508250328285e-05,
      "loss": 3.2985,
      "step": 46760
    },
    {
      "epoch": 0.6130233045848953,
      "grad_norm": 0.6685378551483154,
      "learning_rate": 9.89866104557674e-05,
      "loss": 3.2357,
      "step": 46770
    },
    {
      "epoch": 0.613154376490943,
      "grad_norm": 0.7856571078300476,
      "learning_rate": 9.892814717249197e-05,
      "loss": 3.2264,
      "step": 46780
    },
    {
      "epoch": 0.6132854483969906,
      "grad_norm": 0.6753517389297485,
      "learning_rate": 9.886969266350074e-05,
      "loss": 3.2112,
      "step": 46790
    },
    {
      "epoch": 0.6134165203030383,
      "grad_norm": 0.6568011045455933,
      "learning_rate": 9.881124693883639e-05,
      "loss": 3.2281,
      "step": 46800
    },
    {
      "epoch": 0.613547592209086,
      "grad_norm": 0.707349956035614,
      "learning_rate": 9.875281000854009e-05,
      "loss": 3.2872,
      "step": 46810
    },
    {
      "epoch": 0.6136786641151336,
      "grad_norm": 0.7462494373321533,
      "learning_rate": 9.869438188265137e-05,
      "loss": 3.2561,
      "step": 46820
    },
    {
      "epoch": 0.6138097360211813,
      "grad_norm": 0.6671332716941833,
      "learning_rate": 9.863596257120843e-05,
      "loss": 3.2462,
      "step": 46830
    },
    {
      "epoch": 0.6139408079272288,
      "grad_norm": 0.7925904393196106,
      "learning_rate": 9.857755208424777e-05,
      "loss": 3.2392,
      "step": 46840
    },
    {
      "epoch": 0.6140718798332765,
      "grad_norm": 0.6601688861846924,
      "learning_rate": 9.851915043180461e-05,
      "loss": 3.2442,
      "step": 46850
    },
    {
      "epoch": 0.6142029517393242,
      "grad_norm": 0.7439512610435486,
      "learning_rate": 9.846075762391234e-05,
      "loss": 3.2471,
      "step": 46860
    },
    {
      "epoch": 0.6143340236453718,
      "grad_norm": 0.5951361656188965,
      "learning_rate": 9.840237367060314e-05,
      "loss": 3.2039,
      "step": 46870
    },
    {
      "epoch": 0.6144650955514195,
      "grad_norm": 0.5695250034332275,
      "learning_rate": 9.834399858190753e-05,
      "loss": 3.2369,
      "step": 46880
    },
    {
      "epoch": 0.6145961674574671,
      "grad_norm": 0.6092811822891235,
      "learning_rate": 9.828563236785441e-05,
      "loss": 3.2454,
      "step": 46890
    },
    {
      "epoch": 0.6147272393635148,
      "grad_norm": 0.6412503719329834,
      "learning_rate": 9.822727503847131e-05,
      "loss": 3.2488,
      "step": 46900
    },
    {
      "epoch": 0.6148583112695625,
      "grad_norm": 0.6511025428771973,
      "learning_rate": 9.816892660378418e-05,
      "loss": 3.2716,
      "step": 46910
    },
    {
      "epoch": 0.6149893831756101,
      "grad_norm": 0.6001058220863342,
      "learning_rate": 9.81105870738175e-05,
      "loss": 3.2538,
      "step": 46920
    },
    {
      "epoch": 0.6151204550816578,
      "grad_norm": 0.6340818405151367,
      "learning_rate": 9.805225645859406e-05,
      "loss": 3.1722,
      "step": 46930
    },
    {
      "epoch": 0.6152515269877055,
      "grad_norm": 0.6574063897132874,
      "learning_rate": 9.79939347681353e-05,
      "loss": 3.2109,
      "step": 46940
    },
    {
      "epoch": 0.6153825988937531,
      "grad_norm": 0.6378859877586365,
      "learning_rate": 9.793562201246096e-05,
      "loss": 3.1795,
      "step": 46950
    },
    {
      "epoch": 0.6155136707998008,
      "grad_norm": 0.6317182779312134,
      "learning_rate": 9.787731820158948e-05,
      "loss": 3.2079,
      "step": 46960
    },
    {
      "epoch": 0.6156447427058485,
      "grad_norm": 0.6875243782997131,
      "learning_rate": 9.781902334553744e-05,
      "loss": 3.1947,
      "step": 46970
    },
    {
      "epoch": 0.6157758146118961,
      "grad_norm": 0.6557679772377014,
      "learning_rate": 9.776073745432013e-05,
      "loss": 3.2715,
      "step": 46980
    },
    {
      "epoch": 0.6159068865179438,
      "grad_norm": 0.7047168612480164,
      "learning_rate": 9.770246053795127e-05,
      "loss": 3.3205,
      "step": 46990
    },
    {
      "epoch": 0.6160379584239914,
      "grad_norm": 0.627339243888855,
      "learning_rate": 9.764419260644292e-05,
      "loss": 3.2086,
      "step": 47000
    },
    {
      "epoch": 0.6160379584239914,
      "eval_accuracy": 0.39909970674486805,
      "eval_loss": 3.2312915325164795,
      "eval_runtime": 60.0193,
      "eval_samples_per_second": 33.323,
      "eval_steps_per_second": 0.533,
      "step": 47000
    },
    {
      "epoch": 0.6161690303300391,
      "grad_norm": 0.6777347922325134,
      "learning_rate": 9.75859336698057e-05,
      "loss": 3.2387,
      "step": 47010
    },
    {
      "epoch": 0.6163001022360867,
      "grad_norm": 0.6595043540000916,
      "learning_rate": 9.75276837380487e-05,
      "loss": 3.2311,
      "step": 47020
    },
    {
      "epoch": 0.6164311741421343,
      "grad_norm": 0.6593804359436035,
      "learning_rate": 9.746944282117935e-05,
      "loss": 3.2421,
      "step": 47030
    },
    {
      "epoch": 0.616562246048182,
      "grad_norm": 0.6537586450576782,
      "learning_rate": 9.741121092920363e-05,
      "loss": 3.2346,
      "step": 47040
    },
    {
      "epoch": 0.6166933179542297,
      "grad_norm": 0.6575263738632202,
      "learning_rate": 9.735298807212591e-05,
      "loss": 3.2277,
      "step": 47050
    },
    {
      "epoch": 0.6168243898602773,
      "grad_norm": 0.6552338004112244,
      "learning_rate": 9.729477425994905e-05,
      "loss": 3.216,
      "step": 47060
    },
    {
      "epoch": 0.616955461766325,
      "grad_norm": 0.6585009694099426,
      "learning_rate": 9.723656950267443e-05,
      "loss": 3.2176,
      "step": 47070
    },
    {
      "epoch": 0.6170865336723727,
      "grad_norm": 0.6404207944869995,
      "learning_rate": 9.717837381030165e-05,
      "loss": 3.2365,
      "step": 47080
    },
    {
      "epoch": 0.6172176055784203,
      "grad_norm": 0.699341893196106,
      "learning_rate": 9.712018719282892e-05,
      "loss": 3.2519,
      "step": 47090
    },
    {
      "epoch": 0.617348677484468,
      "grad_norm": 0.6689814329147339,
      "learning_rate": 9.706200966025296e-05,
      "loss": 3.2566,
      "step": 47100
    },
    {
      "epoch": 0.6174797493905156,
      "grad_norm": 0.7128313183784485,
      "learning_rate": 9.700384122256873e-05,
      "loss": 3.2061,
      "step": 47110
    },
    {
      "epoch": 0.6176108212965633,
      "grad_norm": 0.6400073766708374,
      "learning_rate": 9.694568188976976e-05,
      "loss": 3.3091,
      "step": 47120
    },
    {
      "epoch": 0.617741893202611,
      "grad_norm": 0.6701743602752686,
      "learning_rate": 9.688753167184794e-05,
      "loss": 3.2532,
      "step": 47130
    },
    {
      "epoch": 0.6178729651086586,
      "grad_norm": 0.685852587223053,
      "learning_rate": 9.682939057879376e-05,
      "loss": 3.1667,
      "step": 47140
    },
    {
      "epoch": 0.6180040370147063,
      "grad_norm": 0.6643427610397339,
      "learning_rate": 9.677125862059586e-05,
      "loss": 3.2072,
      "step": 47150
    },
    {
      "epoch": 0.618135108920754,
      "grad_norm": 0.6366466879844666,
      "learning_rate": 9.671313580724153e-05,
      "loss": 3.2467,
      "step": 47160
    },
    {
      "epoch": 0.6182661808268016,
      "grad_norm": 0.6131079792976379,
      "learning_rate": 9.665502214871647e-05,
      "loss": 3.2638,
      "step": 47170
    },
    {
      "epoch": 0.6183972527328493,
      "grad_norm": 0.6192038655281067,
      "learning_rate": 9.659691765500477e-05,
      "loss": 3.2206,
      "step": 47180
    },
    {
      "epoch": 0.618528324638897,
      "grad_norm": 0.836887001991272,
      "learning_rate": 9.653882233608886e-05,
      "loss": 3.2195,
      "step": 47190
    },
    {
      "epoch": 0.6186593965449445,
      "grad_norm": 0.6292768120765686,
      "learning_rate": 9.648073620194976e-05,
      "loss": 3.2351,
      "step": 47200
    },
    {
      "epoch": 0.6187904684509922,
      "grad_norm": 0.6945095658302307,
      "learning_rate": 9.642265926256674e-05,
      "loss": 3.246,
      "step": 47210
    },
    {
      "epoch": 0.6189215403570398,
      "grad_norm": 0.655971348285675,
      "learning_rate": 9.636459152791771e-05,
      "loss": 3.2329,
      "step": 47220
    },
    {
      "epoch": 0.6190526122630875,
      "grad_norm": 0.7090550065040588,
      "learning_rate": 9.630653300797872e-05,
      "loss": 3.2289,
      "step": 47230
    },
    {
      "epoch": 0.6191836841691352,
      "grad_norm": 0.8033362627029419,
      "learning_rate": 9.624848371272445e-05,
      "loss": 3.2499,
      "step": 47240
    },
    {
      "epoch": 0.6193147560751828,
      "grad_norm": 0.6385515332221985,
      "learning_rate": 9.619044365212796e-05,
      "loss": 3.1959,
      "step": 47250
    },
    {
      "epoch": 0.6194458279812305,
      "grad_norm": 0.6910477876663208,
      "learning_rate": 9.613241283616064e-05,
      "loss": 3.2026,
      "step": 47260
    },
    {
      "epoch": 0.6195768998872782,
      "grad_norm": 0.7156989574432373,
      "learning_rate": 9.607439127479233e-05,
      "loss": 3.2119,
      "step": 47270
    },
    {
      "epoch": 0.6197079717933258,
      "grad_norm": 0.6280362606048584,
      "learning_rate": 9.601637897799133e-05,
      "loss": 3.1895,
      "step": 47280
    },
    {
      "epoch": 0.6198390436993735,
      "grad_norm": 0.6325256824493408,
      "learning_rate": 9.595837595572435e-05,
      "loss": 3.2549,
      "step": 47290
    },
    {
      "epoch": 0.6199701156054211,
      "grad_norm": 0.6439722180366516,
      "learning_rate": 9.590038221795641e-05,
      "loss": 3.2438,
      "step": 47300
    },
    {
      "epoch": 0.6201011875114688,
      "grad_norm": 0.6495147347450256,
      "learning_rate": 9.584239777465095e-05,
      "loss": 3.2181,
      "step": 47310
    },
    {
      "epoch": 0.6202322594175165,
      "grad_norm": 0.6981726884841919,
      "learning_rate": 9.578442263576993e-05,
      "loss": 3.1962,
      "step": 47320
    },
    {
      "epoch": 0.6203633313235641,
      "grad_norm": 0.7679027915000916,
      "learning_rate": 9.572645681127369e-05,
      "loss": 3.2245,
      "step": 47330
    },
    {
      "epoch": 0.6204944032296118,
      "grad_norm": 0.632557213306427,
      "learning_rate": 9.566850031112077e-05,
      "loss": 3.2287,
      "step": 47340
    },
    {
      "epoch": 0.6206254751356595,
      "grad_norm": 0.7152566313743591,
      "learning_rate": 9.561055314526832e-05,
      "loss": 3.2371,
      "step": 47350
    },
    {
      "epoch": 0.6207565470417071,
      "grad_norm": 0.6541335582733154,
      "learning_rate": 9.55526153236719e-05,
      "loss": 3.2353,
      "step": 47360
    },
    {
      "epoch": 0.6208876189477548,
      "grad_norm": 0.661189079284668,
      "learning_rate": 9.549468685628527e-05,
      "loss": 3.2355,
      "step": 47370
    },
    {
      "epoch": 0.6210186908538023,
      "grad_norm": 0.6386539936065674,
      "learning_rate": 9.54367677530608e-05,
      "loss": 3.2366,
      "step": 47380
    },
    {
      "epoch": 0.62114976275985,
      "grad_norm": 0.6600014567375183,
      "learning_rate": 9.537885802394904e-05,
      "loss": 3.2082,
      "step": 47390
    },
    {
      "epoch": 0.6212808346658977,
      "grad_norm": 0.5996801257133484,
      "learning_rate": 9.532095767889917e-05,
      "loss": 3.2152,
      "step": 47400
    },
    {
      "epoch": 0.6214119065719453,
      "grad_norm": 0.6419047713279724,
      "learning_rate": 9.52630667278585e-05,
      "loss": 3.2638,
      "step": 47410
    },
    {
      "epoch": 0.621542978477993,
      "grad_norm": 0.6197593808174133,
      "learning_rate": 9.520518518077292e-05,
      "loss": 3.2536,
      "step": 47420
    },
    {
      "epoch": 0.6216740503840407,
      "grad_norm": 0.7054881453514099,
      "learning_rate": 9.514731304758661e-05,
      "loss": 3.2818,
      "step": 47430
    },
    {
      "epoch": 0.6218051222900883,
      "grad_norm": 0.6788120865821838,
      "learning_rate": 9.508945033824225e-05,
      "loss": 3.2121,
      "step": 47440
    },
    {
      "epoch": 0.621936194196136,
      "grad_norm": 0.6038095355033875,
      "learning_rate": 9.503159706268069e-05,
      "loss": 3.2567,
      "step": 47450
    },
    {
      "epoch": 0.6220672661021837,
      "grad_norm": 0.6851884722709656,
      "learning_rate": 9.497375323084136e-05,
      "loss": 3.2728,
      "step": 47460
    },
    {
      "epoch": 0.6221983380082313,
      "grad_norm": 0.6265565752983093,
      "learning_rate": 9.491591885266192e-05,
      "loss": 3.2341,
      "step": 47470
    },
    {
      "epoch": 0.622329409914279,
      "grad_norm": 0.6092013716697693,
      "learning_rate": 9.485809393807861e-05,
      "loss": 3.2166,
      "step": 47480
    },
    {
      "epoch": 0.6224604818203266,
      "grad_norm": 0.6962762475013733,
      "learning_rate": 9.480027849702572e-05,
      "loss": 3.1965,
      "step": 47490
    },
    {
      "epoch": 0.6225915537263743,
      "grad_norm": 0.6299521923065186,
      "learning_rate": 9.474247253943623e-05,
      "loss": 3.2099,
      "step": 47500
    },
    {
      "epoch": 0.6225915537263743,
      "eval_accuracy": 0.3998724340175953,
      "eval_loss": 3.2263271808624268,
      "eval_runtime": 59.7564,
      "eval_samples_per_second": 33.469,
      "eval_steps_per_second": 0.536,
      "step": 47500
    },
    {
      "epoch": 0.622722625632422,
      "grad_norm": 0.6280195713043213,
      "learning_rate": 9.468467607524137e-05,
      "loss": 3.25,
      "step": 47510
    },
    {
      "epoch": 0.6228536975384696,
      "grad_norm": 0.6984219551086426,
      "learning_rate": 9.462688911437063e-05,
      "loss": 3.2415,
      "step": 47520
    },
    {
      "epoch": 0.6229847694445173,
      "grad_norm": 0.7601839303970337,
      "learning_rate": 9.456911166675204e-05,
      "loss": 3.3118,
      "step": 47530
    },
    {
      "epoch": 0.623115841350565,
      "grad_norm": 0.6929161548614502,
      "learning_rate": 9.451134374231191e-05,
      "loss": 3.2198,
      "step": 47540
    },
    {
      "epoch": 0.6232469132566125,
      "grad_norm": 0.6816897392272949,
      "learning_rate": 9.445358535097499e-05,
      "loss": 3.1733,
      "step": 47550
    },
    {
      "epoch": 0.6233779851626602,
      "grad_norm": 0.6594375371932983,
      "learning_rate": 9.439583650266424e-05,
      "loss": 3.2467,
      "step": 47560
    },
    {
      "epoch": 0.6235090570687079,
      "grad_norm": 0.6885610818862915,
      "learning_rate": 9.433809720730105e-05,
      "loss": 3.2176,
      "step": 47570
    },
    {
      "epoch": 0.6236401289747555,
      "grad_norm": 0.927130937576294,
      "learning_rate": 9.428036747480529e-05,
      "loss": 3.2081,
      "step": 47580
    },
    {
      "epoch": 0.6237712008808032,
      "grad_norm": 0.6482477188110352,
      "learning_rate": 9.422264731509505e-05,
      "loss": 3.2077,
      "step": 47590
    },
    {
      "epoch": 0.6239022727868508,
      "grad_norm": 0.6166397929191589,
      "learning_rate": 9.416493673808673e-05,
      "loss": 3.2161,
      "step": 47600
    },
    {
      "epoch": 0.6240333446928985,
      "grad_norm": 0.7229884266853333,
      "learning_rate": 9.410723575369526e-05,
      "loss": 3.1857,
      "step": 47610
    },
    {
      "epoch": 0.6241644165989462,
      "grad_norm": 0.6214064955711365,
      "learning_rate": 9.404954437183386e-05,
      "loss": 3.2443,
      "step": 47620
    },
    {
      "epoch": 0.6242954885049938,
      "grad_norm": 0.6719332933425903,
      "learning_rate": 9.399186260241394e-05,
      "loss": 3.4056,
      "step": 47630
    },
    {
      "epoch": 0.6244265604110415,
      "grad_norm": 0.6824233531951904,
      "learning_rate": 9.393419045534546e-05,
      "loss": 3.2983,
      "step": 47640
    },
    {
      "epoch": 0.6245576323170892,
      "grad_norm": 0.6507546901702881,
      "learning_rate": 9.387652794053667e-05,
      "loss": 3.2472,
      "step": 47650
    },
    {
      "epoch": 0.6246887042231368,
      "grad_norm": 0.6942393183708191,
      "learning_rate": 9.381887506789415e-05,
      "loss": 3.2285,
      "step": 47660
    },
    {
      "epoch": 0.6248197761291845,
      "grad_norm": 0.648705005645752,
      "learning_rate": 9.376123184732274e-05,
      "loss": 3.2121,
      "step": 47670
    },
    {
      "epoch": 0.6249508480352322,
      "grad_norm": 0.5787950754165649,
      "learning_rate": 9.370359828872578e-05,
      "loss": 3.2125,
      "step": 47680
    },
    {
      "epoch": 0.6250819199412798,
      "grad_norm": 0.753471314907074,
      "learning_rate": 9.364597440200482e-05,
      "loss": 3.2391,
      "step": 47690
    },
    {
      "epoch": 0.6252129918473275,
      "grad_norm": 0.6960231065750122,
      "learning_rate": 9.358836019705991e-05,
      "loss": 3.2272,
      "step": 47700
    },
    {
      "epoch": 0.6253440637533751,
      "grad_norm": 0.6731109023094177,
      "learning_rate": 9.353075568378918e-05,
      "loss": 3.2203,
      "step": 47710
    },
    {
      "epoch": 0.6254751356594228,
      "grad_norm": 0.7097241282463074,
      "learning_rate": 9.347316087208932e-05,
      "loss": 3.2546,
      "step": 47720
    },
    {
      "epoch": 0.6256062075654704,
      "grad_norm": 0.6814481019973755,
      "learning_rate": 9.341557577185528e-05,
      "loss": 3.1959,
      "step": 47730
    },
    {
      "epoch": 0.625737279471518,
      "grad_norm": 0.6176202893257141,
      "learning_rate": 9.335800039298038e-05,
      "loss": 3.2576,
      "step": 47740
    },
    {
      "epoch": 0.6258683513775657,
      "grad_norm": 0.6454567313194275,
      "learning_rate": 9.330043474535613e-05,
      "loss": 3.1762,
      "step": 47750
    },
    {
      "epoch": 0.6259994232836134,
      "grad_norm": 0.6443287134170532,
      "learning_rate": 9.32428788388725e-05,
      "loss": 3.2684,
      "step": 47760
    },
    {
      "epoch": 0.626130495189661,
      "grad_norm": 0.6694857478141785,
      "learning_rate": 9.318533268341784e-05,
      "loss": 3.2302,
      "step": 47770
    },
    {
      "epoch": 0.6262615670957087,
      "grad_norm": 0.6364569067955017,
      "learning_rate": 9.312779628887859e-05,
      "loss": 3.2618,
      "step": 47780
    },
    {
      "epoch": 0.6263926390017563,
      "grad_norm": 0.646737813949585,
      "learning_rate": 9.307026966513973e-05,
      "loss": 3.224,
      "step": 47790
    },
    {
      "epoch": 0.626523710907804,
      "grad_norm": 0.6261652112007141,
      "learning_rate": 9.30127528220845e-05,
      "loss": 3.2228,
      "step": 47800
    },
    {
      "epoch": 0.6266547828138517,
      "grad_norm": 0.7065747976303101,
      "learning_rate": 9.29552457695945e-05,
      "loss": 3.2767,
      "step": 47810
    },
    {
      "epoch": 0.6267858547198993,
      "grad_norm": 0.617780864238739,
      "learning_rate": 9.289774851754949e-05,
      "loss": 3.2327,
      "step": 47820
    },
    {
      "epoch": 0.626916926625947,
      "grad_norm": 0.615584671497345,
      "learning_rate": 9.284026107582774e-05,
      "loss": 3.3459,
      "step": 47830
    },
    {
      "epoch": 0.6270479985319947,
      "grad_norm": 0.6487901210784912,
      "learning_rate": 9.27827834543057e-05,
      "loss": 3.2493,
      "step": 47840
    },
    {
      "epoch": 0.6271790704380423,
      "grad_norm": 0.711365818977356,
      "learning_rate": 9.272531566285824e-05,
      "loss": 3.245,
      "step": 47850
    },
    {
      "epoch": 0.62731014234409,
      "grad_norm": 0.7267693877220154,
      "learning_rate": 9.26678577113584e-05,
      "loss": 3.1991,
      "step": 47860
    },
    {
      "epoch": 0.6274412142501377,
      "grad_norm": 0.610774040222168,
      "learning_rate": 9.261040960967766e-05,
      "loss": 3.2446,
      "step": 47870
    },
    {
      "epoch": 0.6275722861561853,
      "grad_norm": 0.600073516368866,
      "learning_rate": 9.255297136768582e-05,
      "loss": 3.2238,
      "step": 47880
    },
    {
      "epoch": 0.627703358062233,
      "grad_norm": 0.6462284326553345,
      "learning_rate": 9.249554299525081e-05,
      "loss": 3.2325,
      "step": 47890
    },
    {
      "epoch": 0.6278344299682806,
      "grad_norm": 0.6960959434509277,
      "learning_rate": 9.243812450223904e-05,
      "loss": 3.2073,
      "step": 47900
    },
    {
      "epoch": 0.6279655018743282,
      "grad_norm": 0.6119141578674316,
      "learning_rate": 9.23807158985152e-05,
      "loss": 3.2906,
      "step": 47910
    },
    {
      "epoch": 0.6280965737803759,
      "grad_norm": 0.6240313649177551,
      "learning_rate": 9.23233171939422e-05,
      "loss": 3.23,
      "step": 47920
    },
    {
      "epoch": 0.6282276456864235,
      "grad_norm": 0.6524478197097778,
      "learning_rate": 9.226592839838131e-05,
      "loss": 3.2276,
      "step": 47930
    },
    {
      "epoch": 0.6283587175924712,
      "grad_norm": 0.64365553855896,
      "learning_rate": 9.220854952169205e-05,
      "loss": 3.243,
      "step": 47940
    },
    {
      "epoch": 0.6284897894985189,
      "grad_norm": 0.6872357130050659,
      "learning_rate": 9.21511805737323e-05,
      "loss": 3.202,
      "step": 47950
    },
    {
      "epoch": 0.6286208614045665,
      "grad_norm": 0.6357784867286682,
      "learning_rate": 9.209382156435824e-05,
      "loss": 3.237,
      "step": 47960
    },
    {
      "epoch": 0.6287519333106142,
      "grad_norm": 0.7168740034103394,
      "learning_rate": 9.203647250342423e-05,
      "loss": 3.2618,
      "step": 47970
    },
    {
      "epoch": 0.6288830052166619,
      "grad_norm": 0.6022654175758362,
      "learning_rate": 9.1979133400783e-05,
      "loss": 3.2212,
      "step": 47980
    },
    {
      "epoch": 0.6290140771227095,
      "grad_norm": 0.6093465089797974,
      "learning_rate": 9.192180426628567e-05,
      "loss": 3.2531,
      "step": 47990
    },
    {
      "epoch": 0.6291451490287572,
      "grad_norm": 0.6737045645713806,
      "learning_rate": 9.18644851097814e-05,
      "loss": 3.2229,
      "step": 48000
    },
    {
      "epoch": 0.6291451490287572,
      "eval_accuracy": 0.39996041055718473,
      "eval_loss": 3.2245070934295654,
      "eval_runtime": 62.1503,
      "eval_samples_per_second": 32.18,
      "eval_steps_per_second": 0.515,
      "step": 48000
    },
    {
      "epoch": 0.6292762209348048,
      "grad_norm": 0.61720871925354,
      "learning_rate": 9.18071759411179e-05,
      "loss": 3.257,
      "step": 48010
    },
    {
      "epoch": 0.6294072928408525,
      "grad_norm": 0.6512932181358337,
      "learning_rate": 9.174987677014092e-05,
      "loss": 3.2082,
      "step": 48020
    },
    {
      "epoch": 0.6295383647469002,
      "grad_norm": 0.5734454393386841,
      "learning_rate": 9.169258760669475e-05,
      "loss": 3.2046,
      "step": 48030
    },
    {
      "epoch": 0.6296694366529478,
      "grad_norm": 0.6138554215431213,
      "learning_rate": 9.163530846062171e-05,
      "loss": 3.2253,
      "step": 48040
    },
    {
      "epoch": 0.6298005085589955,
      "grad_norm": 0.6955382823944092,
      "learning_rate": 9.157803934176255e-05,
      "loss": 3.2078,
      "step": 48050
    },
    {
      "epoch": 0.6299315804650432,
      "grad_norm": 0.6082448959350586,
      "learning_rate": 9.152078025995627e-05,
      "loss": 3.2257,
      "step": 48060
    },
    {
      "epoch": 0.6300626523710908,
      "grad_norm": 0.6566722393035889,
      "learning_rate": 9.146353122504021e-05,
      "loss": 3.2281,
      "step": 48070
    },
    {
      "epoch": 0.6301937242771385,
      "grad_norm": 0.7833299040794373,
      "learning_rate": 9.140629224684977e-05,
      "loss": 3.2515,
      "step": 48080
    },
    {
      "epoch": 0.630324796183186,
      "grad_norm": 0.5989523530006409,
      "learning_rate": 9.134906333521885e-05,
      "loss": 3.2036,
      "step": 48090
    },
    {
      "epoch": 0.6304558680892337,
      "grad_norm": 0.660858154296875,
      "learning_rate": 9.129184449997951e-05,
      "loss": 3.2844,
      "step": 48100
    },
    {
      "epoch": 0.6305869399952814,
      "grad_norm": 0.5794146656990051,
      "learning_rate": 9.123463575096217e-05,
      "loss": 3.2037,
      "step": 48110
    },
    {
      "epoch": 0.630718011901329,
      "grad_norm": 0.6321814060211182,
      "learning_rate": 9.117743709799532e-05,
      "loss": 3.2015,
      "step": 48120
    },
    {
      "epoch": 0.6308490838073767,
      "grad_norm": 0.6882767081260681,
      "learning_rate": 9.11202485509059e-05,
      "loss": 3.226,
      "step": 48130
    },
    {
      "epoch": 0.6309801557134244,
      "grad_norm": 0.6049200892448425,
      "learning_rate": 9.106307011951916e-05,
      "loss": 3.2323,
      "step": 48140
    },
    {
      "epoch": 0.631111227619472,
      "grad_norm": 0.8645472526550293,
      "learning_rate": 9.100590181365835e-05,
      "loss": 3.1927,
      "step": 48150
    },
    {
      "epoch": 0.6312422995255197,
      "grad_norm": 0.8259637951850891,
      "learning_rate": 9.094874364314522e-05,
      "loss": 3.2303,
      "step": 48160
    },
    {
      "epoch": 0.6313733714315674,
      "grad_norm": 0.6406517624855042,
      "learning_rate": 9.089159561779973e-05,
      "loss": 3.2797,
      "step": 48170
    },
    {
      "epoch": 0.631504443337615,
      "grad_norm": 0.6976808309555054,
      "learning_rate": 9.083445774744003e-05,
      "loss": 3.238,
      "step": 48180
    },
    {
      "epoch": 0.6316355152436627,
      "grad_norm": 0.5939189791679382,
      "learning_rate": 9.077733004188258e-05,
      "loss": 3.2294,
      "step": 48190
    },
    {
      "epoch": 0.6317665871497103,
      "grad_norm": 0.7672045826911926,
      "learning_rate": 9.0720212510942e-05,
      "loss": 3.2622,
      "step": 48200
    },
    {
      "epoch": 0.631897659055758,
      "grad_norm": 0.6824083924293518,
      "learning_rate": 9.066310516443134e-05,
      "loss": 3.2238,
      "step": 48210
    },
    {
      "epoch": 0.6320287309618057,
      "grad_norm": 0.6701219081878662,
      "learning_rate": 9.060600801216181e-05,
      "loss": 3.2594,
      "step": 48220
    },
    {
      "epoch": 0.6321598028678533,
      "grad_norm": 0.6923904418945312,
      "learning_rate": 9.054892106394273e-05,
      "loss": 3.2413,
      "step": 48230
    },
    {
      "epoch": 0.632290874773901,
      "grad_norm": 0.673308253288269,
      "learning_rate": 9.049184432958188e-05,
      "loss": 3.2146,
      "step": 48240
    },
    {
      "epoch": 0.6324219466799487,
      "grad_norm": 0.6733179688453674,
      "learning_rate": 9.043477781888525e-05,
      "loss": 3.2437,
      "step": 48250
    },
    {
      "epoch": 0.6325530185859963,
      "grad_norm": 0.6185832619667053,
      "learning_rate": 9.03777215416569e-05,
      "loss": 3.2093,
      "step": 48260
    },
    {
      "epoch": 0.6326840904920439,
      "grad_norm": 0.6424225568771362,
      "learning_rate": 9.032067550769935e-05,
      "loss": 3.2125,
      "step": 48270
    },
    {
      "epoch": 0.6328151623980915,
      "grad_norm": 0.6579079031944275,
      "learning_rate": 9.026363972681322e-05,
      "loss": 3.2508,
      "step": 48280
    },
    {
      "epoch": 0.6329462343041392,
      "grad_norm": 0.6600950360298157,
      "learning_rate": 9.020661420879747e-05,
      "loss": 3.2248,
      "step": 48290
    },
    {
      "epoch": 0.6330773062101869,
      "grad_norm": 0.6777704358100891,
      "learning_rate": 9.014959896344912e-05,
      "loss": 3.2315,
      "step": 48300
    },
    {
      "epoch": 0.6332083781162345,
      "grad_norm": 0.7116653323173523,
      "learning_rate": 9.009259400056364e-05,
      "loss": 3.2174,
      "step": 48310
    },
    {
      "epoch": 0.6333394500222822,
      "grad_norm": 0.661887526512146,
      "learning_rate": 9.003559932993463e-05,
      "loss": 3.1994,
      "step": 48320
    },
    {
      "epoch": 0.6334705219283299,
      "grad_norm": 0.71802818775177,
      "learning_rate": 8.997861496135397e-05,
      "loss": 3.2778,
      "step": 48330
    },
    {
      "epoch": 0.6336015938343775,
      "grad_norm": 0.6205228567123413,
      "learning_rate": 8.992164090461163e-05,
      "loss": 3.2194,
      "step": 48340
    },
    {
      "epoch": 0.6337326657404252,
      "grad_norm": 0.7064113020896912,
      "learning_rate": 8.986467716949595e-05,
      "loss": 3.2451,
      "step": 48350
    },
    {
      "epoch": 0.6338637376464729,
      "grad_norm": 0.6913325190544128,
      "learning_rate": 8.98077237657935e-05,
      "loss": 3.2257,
      "step": 48360
    },
    {
      "epoch": 0.6339948095525205,
      "grad_norm": 0.663932204246521,
      "learning_rate": 8.9750780703289e-05,
      "loss": 3.2532,
      "step": 48370
    },
    {
      "epoch": 0.6341258814585682,
      "grad_norm": 1.8195692300796509,
      "learning_rate": 8.969384799176541e-05,
      "loss": 3.4003,
      "step": 48380
    },
    {
      "epoch": 0.6342569533646158,
      "grad_norm": 0.708659291267395,
      "learning_rate": 8.963692564100393e-05,
      "loss": 3.1676,
      "step": 48390
    },
    {
      "epoch": 0.6343880252706635,
      "grad_norm": 0.6275567412376404,
      "learning_rate": 8.958001366078402e-05,
      "loss": 3.2392,
      "step": 48400
    },
    {
      "epoch": 0.6345190971767112,
      "grad_norm": 0.7094308733940125,
      "learning_rate": 8.952311206088324e-05,
      "loss": 3.2386,
      "step": 48410
    },
    {
      "epoch": 0.6346501690827588,
      "grad_norm": 0.6498938798904419,
      "learning_rate": 8.946622085107746e-05,
      "loss": 3.186,
      "step": 48420
    },
    {
      "epoch": 0.6347812409888065,
      "grad_norm": 0.5871189832687378,
      "learning_rate": 8.940934004114078e-05,
      "loss": 3.232,
      "step": 48430
    },
    {
      "epoch": 0.6349123128948542,
      "grad_norm": 0.6324211359024048,
      "learning_rate": 8.935246964084551e-05,
      "loss": 3.2528,
      "step": 48440
    },
    {
      "epoch": 0.6350433848009017,
      "grad_norm": 0.6860491633415222,
      "learning_rate": 8.929560965996203e-05,
      "loss": 3.2225,
      "step": 48450
    },
    {
      "epoch": 0.6351744567069494,
      "grad_norm": 0.6484309434890747,
      "learning_rate": 8.923876010825913e-05,
      "loss": 3.2235,
      "step": 48460
    },
    {
      "epoch": 0.635305528612997,
      "grad_norm": 0.9210099577903748,
      "learning_rate": 8.918192099550367e-05,
      "loss": 3.2229,
      "step": 48470
    },
    {
      "epoch": 0.6354366005190447,
      "grad_norm": 0.6456104516983032,
      "learning_rate": 8.912509233146085e-05,
      "loss": 3.2355,
      "step": 48480
    },
    {
      "epoch": 0.6355676724250924,
      "grad_norm": 0.6721792817115784,
      "learning_rate": 8.906827412589387e-05,
      "loss": 3.1688,
      "step": 48490
    },
    {
      "epoch": 0.63569874433114,
      "grad_norm": 0.6328034400939941,
      "learning_rate": 8.901146638856429e-05,
      "loss": 3.2429,
      "step": 48500
    },
    {
      "epoch": 0.63569874433114,
      "eval_accuracy": 0.40005522971652,
      "eval_loss": 3.223724126815796,
      "eval_runtime": 62.7514,
      "eval_samples_per_second": 31.872,
      "eval_steps_per_second": 0.51,
      "step": 48500
    },
    {
      "epoch": 0.6358298162371877,
      "grad_norm": 0.6821357011795044,
      "learning_rate": 8.895466912923194e-05,
      "loss": 3.266,
      "step": 48510
    },
    {
      "epoch": 0.6359608881432354,
      "grad_norm": 0.6612139940261841,
      "learning_rate": 8.889788235765458e-05,
      "loss": 3.2355,
      "step": 48520
    },
    {
      "epoch": 0.636091960049283,
      "grad_norm": 0.6270757913589478,
      "learning_rate": 8.884110608358843e-05,
      "loss": 3.2726,
      "step": 48530
    },
    {
      "epoch": 0.6362230319553307,
      "grad_norm": 0.6388996243476868,
      "learning_rate": 8.878434031678782e-05,
      "loss": 3.2156,
      "step": 48540
    },
    {
      "epoch": 0.6363541038613784,
      "grad_norm": 0.6656998991966248,
      "learning_rate": 8.872758506700524e-05,
      "loss": 3.236,
      "step": 48550
    },
    {
      "epoch": 0.636485175767426,
      "grad_norm": 0.6329711675643921,
      "learning_rate": 8.867084034399139e-05,
      "loss": 3.2314,
      "step": 48560
    },
    {
      "epoch": 0.6366162476734737,
      "grad_norm": 0.6939536929130554,
      "learning_rate": 8.861410615749515e-05,
      "loss": 3.2058,
      "step": 48570
    },
    {
      "epoch": 0.6367473195795214,
      "grad_norm": 0.7360872626304626,
      "learning_rate": 8.855738251726365e-05,
      "loss": 3.2279,
      "step": 48580
    },
    {
      "epoch": 0.636878391485569,
      "grad_norm": 0.7883435487747192,
      "learning_rate": 8.850066943304217e-05,
      "loss": 3.2219,
      "step": 48590
    },
    {
      "epoch": 0.6370094633916167,
      "grad_norm": 0.6447556018829346,
      "learning_rate": 8.84439669145741e-05,
      "loss": 3.25,
      "step": 48600
    },
    {
      "epoch": 0.6371405352976643,
      "grad_norm": 0.6142321228981018,
      "learning_rate": 8.838727497160116e-05,
      "loss": 3.2359,
      "step": 48610
    },
    {
      "epoch": 0.637271607203712,
      "grad_norm": 0.6500118970870972,
      "learning_rate": 8.83305936138632e-05,
      "loss": 3.2235,
      "step": 48620
    },
    {
      "epoch": 0.6374026791097596,
      "grad_norm": 0.7137872576713562,
      "learning_rate": 8.827392285109815e-05,
      "loss": 3.2184,
      "step": 48630
    },
    {
      "epoch": 0.6375337510158072,
      "grad_norm": 0.6164005398750305,
      "learning_rate": 8.821726269304227e-05,
      "loss": 3.2147,
      "step": 48640
    },
    {
      "epoch": 0.6376648229218549,
      "grad_norm": 0.6556917428970337,
      "learning_rate": 8.816061314942987e-05,
      "loss": 3.2311,
      "step": 48650
    },
    {
      "epoch": 0.6377958948279026,
      "grad_norm": 0.7349936366081238,
      "learning_rate": 8.810397422999361e-05,
      "loss": 3.2457,
      "step": 48660
    },
    {
      "epoch": 0.6379269667339502,
      "grad_norm": 0.6263675093650818,
      "learning_rate": 8.804734594446408e-05,
      "loss": 3.2332,
      "step": 48670
    },
    {
      "epoch": 0.6380580386399979,
      "grad_norm": 0.6568204760551453,
      "learning_rate": 8.799072830257019e-05,
      "loss": 3.2081,
      "step": 48680
    },
    {
      "epoch": 0.6381891105460455,
      "grad_norm": 0.6226121783256531,
      "learning_rate": 8.793412131403907e-05,
      "loss": 3.2536,
      "step": 48690
    },
    {
      "epoch": 0.6383201824520932,
      "grad_norm": 0.6527653336524963,
      "learning_rate": 8.7877524988596e-05,
      "loss": 3.1668,
      "step": 48700
    },
    {
      "epoch": 0.6384512543581409,
      "grad_norm": 0.6220067143440247,
      "learning_rate": 8.782093933596426e-05,
      "loss": 3.2269,
      "step": 48710
    },
    {
      "epoch": 0.6385823262641885,
      "grad_norm": 0.6496166586875916,
      "learning_rate": 8.776436436586548e-05,
      "loss": 3.1956,
      "step": 48720
    },
    {
      "epoch": 0.6387133981702362,
      "grad_norm": 0.6590812802314758,
      "learning_rate": 8.770780008801938e-05,
      "loss": 3.1934,
      "step": 48730
    },
    {
      "epoch": 0.6388444700762839,
      "grad_norm": 0.5929205417633057,
      "learning_rate": 8.76512465121439e-05,
      "loss": 3.1913,
      "step": 48740
    },
    {
      "epoch": 0.6389755419823315,
      "grad_norm": 0.7077128291130066,
      "learning_rate": 8.759470364795504e-05,
      "loss": 3.2173,
      "step": 48750
    },
    {
      "epoch": 0.6391066138883792,
      "grad_norm": 0.6505268812179565,
      "learning_rate": 8.753817150516703e-05,
      "loss": 3.2057,
      "step": 48760
    },
    {
      "epoch": 0.6392376857944269,
      "grad_norm": 0.6044105887413025,
      "learning_rate": 8.748165009349233e-05,
      "loss": 3.2765,
      "step": 48770
    },
    {
      "epoch": 0.6393687577004745,
      "grad_norm": 0.6355426907539368,
      "learning_rate": 8.742513942264132e-05,
      "loss": 3.1995,
      "step": 48780
    },
    {
      "epoch": 0.6394998296065222,
      "grad_norm": 0.677350640296936,
      "learning_rate": 8.736863950232281e-05,
      "loss": 3.2441,
      "step": 48790
    },
    {
      "epoch": 0.6396309015125698,
      "grad_norm": 0.6349044442176819,
      "learning_rate": 8.731215034224357e-05,
      "loss": 3.2377,
      "step": 48800
    },
    {
      "epoch": 0.6397619734186174,
      "grad_norm": 0.6088805198669434,
      "learning_rate": 8.725567195210873e-05,
      "loss": 3.3147,
      "step": 48810
    },
    {
      "epoch": 0.6398930453246651,
      "grad_norm": 0.6461312174797058,
      "learning_rate": 8.719920434162124e-05,
      "loss": 3.1875,
      "step": 48820
    },
    {
      "epoch": 0.6400241172307127,
      "grad_norm": 0.6237476468086243,
      "learning_rate": 8.714274752048248e-05,
      "loss": 3.2489,
      "step": 48830
    },
    {
      "epoch": 0.6401551891367604,
      "grad_norm": 0.7434705495834351,
      "learning_rate": 8.708630149839195e-05,
      "loss": 3.2673,
      "step": 48840
    },
    {
      "epoch": 0.6402862610428081,
      "grad_norm": 0.6502662301063538,
      "learning_rate": 8.702986628504712e-05,
      "loss": 3.2355,
      "step": 48850
    },
    {
      "epoch": 0.6404173329488557,
      "grad_norm": 0.6460471153259277,
      "learning_rate": 8.697344189014382e-05,
      "loss": 3.2408,
      "step": 48860
    },
    {
      "epoch": 0.6405484048549034,
      "grad_norm": 0.7010906934738159,
      "learning_rate": 8.69170283233758e-05,
      "loss": 3.2306,
      "step": 48870
    },
    {
      "epoch": 0.640679476760951,
      "grad_norm": 0.7609075903892517,
      "learning_rate": 8.686062559443517e-05,
      "loss": 3.2796,
      "step": 48880
    },
    {
      "epoch": 0.6408105486669987,
      "grad_norm": 0.6554475426673889,
      "learning_rate": 8.680423371301196e-05,
      "loss": 3.1965,
      "step": 48890
    },
    {
      "epoch": 0.6409416205730464,
      "grad_norm": 0.6742631196975708,
      "learning_rate": 8.674785268879455e-05,
      "loss": 3.298,
      "step": 48900
    },
    {
      "epoch": 0.641072692479094,
      "grad_norm": 0.6487531661987305,
      "learning_rate": 8.669148253146929e-05,
      "loss": 3.2117,
      "step": 48910
    },
    {
      "epoch": 0.6412037643851417,
      "grad_norm": 0.6006836891174316,
      "learning_rate": 8.663512325072083e-05,
      "loss": 3.2564,
      "step": 48920
    },
    {
      "epoch": 0.6413348362911894,
      "grad_norm": 0.6650765538215637,
      "learning_rate": 8.657877485623172e-05,
      "loss": 3.2229,
      "step": 48930
    },
    {
      "epoch": 0.641465908197237,
      "grad_norm": 0.6072572469711304,
      "learning_rate": 8.652243735768281e-05,
      "loss": 3.2212,
      "step": 48940
    },
    {
      "epoch": 0.6415969801032847,
      "grad_norm": 0.6461341977119446,
      "learning_rate": 8.646611076475305e-05,
      "loss": 3.2028,
      "step": 48950
    },
    {
      "epoch": 0.6417280520093324,
      "grad_norm": 0.671175479888916,
      "learning_rate": 8.640979508711959e-05,
      "loss": 3.2133,
      "step": 48960
    },
    {
      "epoch": 0.64185912391538,
      "grad_norm": 0.6019599437713623,
      "learning_rate": 8.635349033445746e-05,
      "loss": 3.188,
      "step": 48970
    },
    {
      "epoch": 0.6419901958214277,
      "grad_norm": 0.606680691242218,
      "learning_rate": 8.629719651644003e-05,
      "loss": 3.2211,
      "step": 48980
    },
    {
      "epoch": 0.6421212677274752,
      "grad_norm": 0.6697722673416138,
      "learning_rate": 8.624091364273875e-05,
      "loss": 3.2238,
      "step": 48990
    },
    {
      "epoch": 0.6422523396335229,
      "grad_norm": 0.6737632751464844,
      "learning_rate": 8.618464172302324e-05,
      "loss": 3.2089,
      "step": 49000
    },
    {
      "epoch": 0.6422523396335229,
      "eval_accuracy": 0.4004916911045943,
      "eval_loss": 3.2214460372924805,
      "eval_runtime": 62.8424,
      "eval_samples_per_second": 31.826,
      "eval_steps_per_second": 0.509,
      "step": 49000
    },
    {
      "epoch": 5e-09,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00010745841306752502,
      "loss": 3.1778,
      "step": 49010
    },
    {
      "epoch": 1e-08,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00010734454578537477,
      "loss": 3.1465,
      "step": 49020
    },
    {
      "epoch": 1.5e-08,
      "grad_norm": 0.609375,
      "learning_rate": 0.00010723072864920421,
      "loss": 3.1177,
      "step": 49030
    },
    {
      "epoch": 2e-08,
      "grad_norm": 0.625,
      "learning_rate": 0.00010711696175800602,
      "loss": 3.1645,
      "step": 49040
    },
    {
      "epoch": 2.5e-08,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0001070032452107293,
      "loss": 3.1674,
      "step": 49050
    },
    {
      "epoch": 3e-08,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0001068895791062792,
      "loss": 3.1582,
      "step": 49060
    },
    {
      "epoch": 3.5e-08,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00010677596354351713,
      "loss": 3.1465,
      "step": 49070
    },
    {
      "epoch": 4e-08,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00010666239862126061,
      "loss": 3.1316,
      "step": 49080
    },
    {
      "epoch": 4.5e-08,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00010654888443828301,
      "loss": 3.1577,
      "step": 49090
    },
    {
      "epoch": 5e-08,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00010643542109331357,
      "loss": 3.1044,
      "step": 49100
    },
    {
      "epoch": 5.5e-08,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00010632200868503724,
      "loss": 3.1189,
      "step": 49110
    },
    {
      "epoch": 6e-08,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00010620864731209483,
      "loss": 3.0972,
      "step": 49120
    },
    {
      "epoch": 6.5e-08,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00010609533707308268,
      "loss": 3.1177,
      "step": 49130
    },
    {
      "epoch": 7e-08,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00010598207806655264,
      "loss": 3.1101,
      "step": 49140
    },
    {
      "epoch": 7.5e-08,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00010586887039101204,
      "loss": 3.1387,
      "step": 49150
    },
    {
      "epoch": 8e-08,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00010575571414492357,
      "loss": 3.1128,
      "step": 49160
    },
    {
      "epoch": 8.5e-08,
      "grad_norm": 0.65625,
      "learning_rate": 0.0001056426094267051,
      "loss": 3.1292,
      "step": 49170
    },
    {
      "epoch": 9e-08,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0001055295563347298,
      "loss": 3.1134,
      "step": 49180
    },
    {
      "epoch": 9.5e-08,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0001054165549673259,
      "loss": 3.0817,
      "step": 49190
    },
    {
      "epoch": 1e-07,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00010530360542277661,
      "loss": 3.1199,
      "step": 49200
    },
    {
      "epoch": 1.05e-07,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00010519070779932005,
      "loss": 3.1137,
      "step": 49210
    },
    {
      "epoch": 1.1e-07,
      "grad_norm": 0.76171875,
      "learning_rate": 0.00010507786219514923,
      "loss": 3.0613,
      "step": 49220
    },
    {
      "epoch": 1.15e-07,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00010496506870841195,
      "loss": 3.1172,
      "step": 49230
    },
    {
      "epoch": 1.2e-07,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00010485232743721069,
      "loss": 3.0776,
      "step": 49240
    },
    {
      "epoch": 1.25e-07,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0001047396384796023,
      "loss": 3.1018,
      "step": 49250
    },
    {
      "epoch": 1.3e-07,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00010462700193359841,
      "loss": 3.0794,
      "step": 49260
    },
    {
      "epoch": 1.35e-07,
      "grad_norm": 0.578125,
      "learning_rate": 0.00010451441789716487,
      "loss": 3.1154,
      "step": 49270
    },
    {
      "epoch": 1.4e-07,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00010440188646822201,
      "loss": 3.1138,
      "step": 49280
    },
    {
      "epoch": 1.45e-07,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00010428940774464427,
      "loss": 3.1024,
      "step": 49290
    },
    {
      "epoch": 1.5e-07,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00010417698182426035,
      "loss": 3.0942,
      "step": 49300
    },
    {
      "epoch": 1.55e-07,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00010406460880485295,
      "loss": 3.0977,
      "step": 49310
    },
    {
      "epoch": 1.6e-07,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00010395228878415881,
      "loss": 3.0713,
      "step": 49320
    },
    {
      "epoch": 1.65e-07,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00010384002185986863,
      "loss": 3.1061,
      "step": 49330
    },
    {
      "epoch": 1.7e-07,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00010372780812962673,
      "loss": 3.0984,
      "step": 49340
    },
    {
      "epoch": 1.75e-07,
      "grad_norm": 0.640625,
      "learning_rate": 0.00010361564769103133,
      "loss": 3.1019,
      "step": 49350
    },
    {
      "epoch": 1.8e-07,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00010350354064163419,
      "loss": 3.0661,
      "step": 49360
    },
    {
      "epoch": 1.85e-07,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00010339148707894089,
      "loss": 3.1016,
      "step": 49370
    },
    {
      "epoch": 1.9e-07,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00010327948710041024,
      "loss": 3.0893,
      "step": 49380
    },
    {
      "epoch": 1.95e-07,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00010316754080345442,
      "loss": 3.1009,
      "step": 49390
    },
    {
      "epoch": 2e-07,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00010305564828543907,
      "loss": 3.1064,
      "step": 49400
    },
    {
      "epoch": 2.05e-07,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00010294380964368298,
      "loss": 3.065,
      "step": 49410
    },
    {
      "epoch": 2.1e-07,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00010283202497545811,
      "loss": 3.0552,
      "step": 49420
    },
    {
      "epoch": 2.15e-07,
      "grad_norm": 0.671875,
      "learning_rate": 0.00010272029437798943,
      "loss": 3.0983,
      "step": 49430
    },
    {
      "epoch": 2.2e-07,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00010260861794845497,
      "loss": 3.1077,
      "step": 49440
    },
    {
      "epoch": 2.25e-07,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00010249699578398557,
      "loss": 3.0824,
      "step": 49450
    },
    {
      "epoch": 2.3e-07,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00010238542798166487,
      "loss": 3.0985,
      "step": 49460
    },
    {
      "epoch": 2.35e-07,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0001022739146385293,
      "loss": 3.1138,
      "step": 49470
    },
    {
      "epoch": 2.4e-07,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00010216245585156784,
      "loss": 3.1316,
      "step": 49480
    },
    {
      "epoch": 2.45e-07,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00010205105171772204,
      "loss": 3.0916,
      "step": 49490
    },
    {
      "epoch": 2.5e-07,
      "grad_norm": 0.703125,
      "learning_rate": 0.00010193970233388597,
      "loss": 3.1142,
      "step": 49500
    },
    {
      "epoch": 2.5e-07,
      "eval_loss": 3.164301872253418,
      "eval_runtime": 0.2006,
      "eval_samples_per_second": 159.558,
      "eval_steps_per_second": 4.986,
      "step": 49500
    },
    {
      "epoch": 2.55e-07,
      "grad_norm": 0.7578125,
      "learning_rate": 0.00010182840779690595,
      "loss": 3.0552,
      "step": 49510
    },
    {
      "epoch": 2.6e-07,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0001017171682035809,
      "loss": 3.0932,
      "step": 49520
    },
    {
      "epoch": 2.65e-07,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00010160598365066155,
      "loss": 3.1036,
      "step": 49530
    },
    {
      "epoch": 2.7e-07,
      "grad_norm": 0.6796875,
      "learning_rate": 0.000101494854234851,
      "loss": 3.0786,
      "step": 49540
    },
    {
      "epoch": 2.75e-07,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00010138378005280439,
      "loss": 3.0714,
      "step": 49550
    },
    {
      "epoch": 2.8e-07,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00010127276120112878,
      "loss": 3.0573,
      "step": 49560
    },
    {
      "epoch": 2.85e-07,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00010116179777638298,
      "loss": 3.0492,
      "step": 49570
    },
    {
      "epoch": 2.9e-07,
      "grad_norm": 0.765625,
      "learning_rate": 0.00010105088987507782,
      "loss": 3.1214,
      "step": 49580
    },
    {
      "epoch": 2.95e-07,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00010094003759367574,
      "loss": 3.0896,
      "step": 49590
    },
    {
      "epoch": 3e-07,
      "grad_norm": 0.59375,
      "learning_rate": 0.00010082924102859073,
      "loss": 3.1125,
      "step": 49600
    },
    {
      "epoch": 3.05e-07,
      "grad_norm": 0.578125,
      "learning_rate": 0.00010071850027618852,
      "loss": 3.0446,
      "step": 49610
    },
    {
      "epoch": 3.1e-07,
      "grad_norm": 0.6328125,
      "learning_rate": 0.000100607815432786,
      "loss": 3.1128,
      "step": 49620
    },
    {
      "epoch": 3.15e-07,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00010049718659465167,
      "loss": 3.1234,
      "step": 49630
    },
    {
      "epoch": 3.2e-07,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00010038661385800525,
      "loss": 3.0634,
      "step": 49640
    },
    {
      "epoch": 3.25e-07,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00010027609731901759,
      "loss": 3.0568,
      "step": 49650
    },
    {
      "epoch": 3.3e-07,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00010016563707381093,
      "loss": 3.0853,
      "step": 49660
    },
    {
      "epoch": 3.35e-07,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00010005523321845814,
      "loss": 3.1028,
      "step": 49670
    },
    {
      "epoch": 3.4e-07,
      "grad_norm": 0.59375,
      "learning_rate": 9.994488584898336e-05,
      "loss": 3.0982,
      "step": 49680
    },
    {
      "epoch": 3.45e-07,
      "grad_norm": 0.87890625,
      "learning_rate": 9.98345950613615e-05,
      "loss": 3.1122,
      "step": 49690
    },
    {
      "epoch": 3.5e-07,
      "grad_norm": 0.64453125,
      "learning_rate": 9.972436095151826e-05,
      "loss": 3.1003,
      "step": 49700
    },
    {
      "epoch": 3.55e-07,
      "grad_norm": 0.6328125,
      "learning_rate": 9.961418361532994e-05,
      "loss": 3.1298,
      "step": 49710
    },
    {
      "epoch": 3.6e-07,
      "grad_norm": 0.63671875,
      "learning_rate": 9.950406314862365e-05,
      "loss": 3.0694,
      "step": 49720
    },
    {
      "epoch": 3.65e-07,
      "grad_norm": 0.6875,
      "learning_rate": 9.939399964717698e-05,
      "loss": 3.08,
      "step": 49730
    },
    {
      "epoch": 3.7e-07,
      "grad_norm": 0.71875,
      "learning_rate": 9.928399320671787e-05,
      "loss": 3.0896,
      "step": 49740
    },
    {
      "epoch": 3.75e-07,
      "grad_norm": 0.62109375,
      "learning_rate": 9.91740439229248e-05,
      "loss": 3.0944,
      "step": 49750
    },
    {
      "epoch": 3.8e-07,
      "grad_norm": 0.59375,
      "learning_rate": 9.906415189142632e-05,
      "loss": 3.0935,
      "step": 49760
    },
    {
      "epoch": 3.85e-07,
      "grad_norm": 0.63671875,
      "learning_rate": 9.895431720780137e-05,
      "loss": 3.0593,
      "step": 49770
    },
    {
      "epoch": 3.9e-07,
      "grad_norm": 0.61328125,
      "learning_rate": 9.884453996757894e-05,
      "loss": 3.0615,
      "step": 49780
    },
    {
      "epoch": 3.95e-07,
      "grad_norm": 0.66796875,
      "learning_rate": 9.873482026623804e-05,
      "loss": 3.066,
      "step": 49790
    },
    {
      "epoch": 4e-07,
      "grad_norm": 0.65625,
      "learning_rate": 9.862515819920776e-05,
      "loss": 3.0987,
      "step": 49800
    },
    {
      "epoch": 4.05e-07,
      "grad_norm": 0.67578125,
      "learning_rate": 9.851555386186687e-05,
      "loss": 3.0693,
      "step": 49810
    },
    {
      "epoch": 4.1e-07,
      "grad_norm": 0.66015625,
      "learning_rate": 9.840600734954405e-05,
      "loss": 3.1125,
      "step": 49820
    },
    {
      "epoch": 4.15e-07,
      "grad_norm": 0.609375,
      "learning_rate": 9.829651875751764e-05,
      "loss": 3.0705,
      "step": 49830
    },
    {
      "epoch": 4.2e-07,
      "grad_norm": 0.67578125,
      "learning_rate": 9.818708818101574e-05,
      "loss": 3.0935,
      "step": 49840
    },
    {
      "epoch": 4.25e-07,
      "grad_norm": 0.6328125,
      "learning_rate": 9.807771571521565e-05,
      "loss": 3.0583,
      "step": 49850
    },
    {
      "epoch": 4.3e-07,
      "grad_norm": 0.62109375,
      "learning_rate": 9.796840145524456e-05,
      "loss": 3.0835,
      "step": 49860
    },
    {
      "epoch": 4.35e-07,
      "grad_norm": 0.60546875,
      "learning_rate": 9.785914549617873e-05,
      "loss": 3.0431,
      "step": 49870
    },
    {
      "epoch": 4.4e-07,
      "grad_norm": 0.65234375,
      "learning_rate": 9.774994793304384e-05,
      "loss": 3.0697,
      "step": 49880
    },
    {
      "epoch": 4.45e-07,
      "grad_norm": 0.7265625,
      "learning_rate": 9.764080886081484e-05,
      "loss": 3.0997,
      "step": 49890
    },
    {
      "epoch": 4.5e-07,
      "grad_norm": 0.7578125,
      "learning_rate": 9.753172837441554e-05,
      "loss": 3.0633,
      "step": 49900
    },
    {
      "epoch": 4.55e-07,
      "grad_norm": 0.6484375,
      "learning_rate": 9.742270656871906e-05,
      "loss": 3.0752,
      "step": 49910
    },
    {
      "epoch": 4.6e-07,
      "grad_norm": 0.69140625,
      "learning_rate": 9.731374353854742e-05,
      "loss": 3.0803,
      "step": 49920
    },
    {
      "epoch": 4.65e-07,
      "grad_norm": 0.5859375,
      "learning_rate": 9.720483937867147e-05,
      "loss": 3.073,
      "step": 49930
    },
    {
      "epoch": 4.7e-07,
      "grad_norm": 0.61328125,
      "learning_rate": 9.709599418381087e-05,
      "loss": 3.0693,
      "step": 49940
    },
    {
      "epoch": 4.75e-07,
      "grad_norm": 0.640625,
      "learning_rate": 9.698720804863402e-05,
      "loss": 3.124,
      "step": 49950
    },
    {
      "epoch": 4.8e-07,
      "grad_norm": 0.60546875,
      "learning_rate": 9.687848106775797e-05,
      "loss": 3.09,
      "step": 49960
    },
    {
      "epoch": 4.85e-07,
      "grad_norm": 0.68359375,
      "learning_rate": 9.676981333574823e-05,
      "loss": 3.0619,
      "step": 49970
    },
    {
      "epoch": 4.9e-07,
      "grad_norm": 0.625,
      "learning_rate": 9.666120494711896e-05,
      "loss": 3.0917,
      "step": 49980
    },
    {
      "epoch": 4.95e-07,
      "grad_norm": 0.64453125,
      "learning_rate": 9.655265599633246e-05,
      "loss": 3.0923,
      "step": 49990
    },
    {
      "epoch": 5e-07,
      "grad_norm": 0.61328125,
      "learning_rate": 9.644416657779942e-05,
      "loss": 3.1021,
      "step": 50000
    },
    {
      "epoch": 5e-07,
      "eval_loss": 3.1472666263580322,
      "eval_runtime": 0.0603,
      "eval_samples_per_second": 530.541,
      "eval_steps_per_second": 16.579,
      "step": 50000
    },
    {
      "epoch": 5e-09,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00010246413961327904,
      "loss": 3.077,
      "step": 50010
    },
    {
      "epoch": 1e-08,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00010237117601769932,
      "loss": 3.0892,
      "step": 50020
    },
    {
      "epoch": 1.5e-08,
      "grad_norm": 0.58203125,
      "learning_rate": 0.000102278250249607,
      "loss": 3.0528,
      "step": 50030
    },
    {
      "epoch": 2e-08,
      "grad_norm": 0.478515625,
      "learning_rate": 0.000102185362365129,
      "loss": 3.0804,
      "step": 50040
    },
    {
      "epoch": 2.5e-08,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00010209251242036886,
      "loss": 3.0685,
      "step": 50050
    },
    {
      "epoch": 3e-08,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00010199970047140765,
      "loss": 3.0533,
      "step": 50060
    },
    {
      "epoch": 3.5e-08,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00010190692657430326,
      "loss": 3.0581,
      "step": 50070
    },
    {
      "epoch": 4e-08,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00010181419078509062,
      "loss": 3.0981,
      "step": 50080
    },
    {
      "epoch": 4.5e-08,
      "grad_norm": 0.46875,
      "learning_rate": 0.0001017214931597817,
      "loss": 3.0777,
      "step": 50090
    },
    {
      "epoch": 5e-08,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00010162883375436538,
      "loss": 3.0681,
      "step": 50100
    },
    {
      "epoch": 5.5e-08,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0001015362126248074,
      "loss": 3.0786,
      "step": 50110
    },
    {
      "epoch": 6e-08,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0001014436298270506,
      "loss": 3.0386,
      "step": 50120
    },
    {
      "epoch": 6.5e-08,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00010135108541701434,
      "loss": 3.1031,
      "step": 50130
    },
    {
      "epoch": 7e-08,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00010125857945059509,
      "loss": 3.0772,
      "step": 50140
    },
    {
      "epoch": 7.5e-08,
      "grad_norm": 0.44921875,
      "learning_rate": 0.00010116611198366592,
      "loss": 3.0655,
      "step": 50150
    },
    {
      "epoch": 8e-08,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00010107368307207674,
      "loss": 3.0504,
      "step": 50160
    },
    {
      "epoch": 8.5e-08,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00010098129277165412,
      "loss": 3.0748,
      "step": 50170
    },
    {
      "epoch": 9e-08,
      "grad_norm": 0.443359375,
      "learning_rate": 0.0001008889411382013,
      "loss": 3.0916,
      "step": 50180
    },
    {
      "epoch": 9.5e-08,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00010079662822749818,
      "loss": 3.0565,
      "step": 50190
    },
    {
      "epoch": 1e-07,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00010070435409530144,
      "loss": 3.0814,
      "step": 50200
    },
    {
      "epoch": 1.05e-07,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00010061211879734392,
      "loss": 3.0656,
      "step": 50210
    },
    {
      "epoch": 1.1e-07,
      "grad_norm": 0.451171875,
      "learning_rate": 0.00010051992238933552,
      "loss": 3.1134,
      "step": 50220
    },
    {
      "epoch": 1.15e-07,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00010042776492696218,
      "loss": 3.097,
      "step": 50230
    },
    {
      "epoch": 1.2e-07,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00010033564646588669,
      "loss": 3.0581,
      "step": 50240
    },
    {
      "epoch": 1.25e-07,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00010024356706174806,
      "loss": 3.0447,
      "step": 50250
    },
    {
      "epoch": 1.3e-07,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00010015152677016179,
      "loss": 3.0488,
      "step": 50260
    },
    {
      "epoch": 1.35e-07,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00010005952564671973,
      "loss": 3.076,
      "step": 50270
    },
    {
      "epoch": 1.4e-07,
      "grad_norm": 0.52734375,
      "learning_rate": 9.99675637469901e-05,
      "loss": 3.0726,
      "step": 50280
    },
    {
      "epoch": 1.45e-07,
      "grad_norm": 0.5,
      "learning_rate": 9.987564112651737e-05,
      "loss": 3.0732,
      "step": 50290
    },
    {
      "epoch": 1.5e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 9.978375784082248e-05,
      "loss": 3.0587,
      "step": 50300
    },
    {
      "epoch": 1.55e-07,
      "grad_norm": 0.4921875,
      "learning_rate": 9.969191394540223e-05,
      "loss": 3.0667,
      "step": 50310
    },
    {
      "epoch": 1.6e-07,
      "grad_norm": 0.50390625,
      "learning_rate": 9.960010949573002e-05,
      "loss": 3.0913,
      "step": 50320
    },
    {
      "epoch": 1.65e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 9.950834454725524e-05,
      "loss": 3.0945,
      "step": 50330
    },
    {
      "epoch": 1.7e-07,
      "grad_norm": 0.5234375,
      "learning_rate": 9.941661915540343e-05,
      "loss": 3.0876,
      "step": 50340
    },
    {
      "epoch": 1.75e-07,
      "grad_norm": 0.470703125,
      "learning_rate": 9.932493337557626e-05,
      "loss": 3.0706,
      "step": 50350
    },
    {
      "epoch": 1.8e-07,
      "grad_norm": 0.46875,
      "learning_rate": 9.923328726315147e-05,
      "loss": 3.0699,
      "step": 50360
    },
    {
      "epoch": 1.85e-07,
      "grad_norm": 0.484375,
      "learning_rate": 9.914168087348284e-05,
      "loss": 3.0659,
      "step": 50370
    },
    {
      "epoch": 1.9e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 9.905011426190015e-05,
      "loss": 3.1036,
      "step": 50380
    },
    {
      "epoch": 1.95e-07,
      "grad_norm": 0.451171875,
      "learning_rate": 9.895858748370918e-05,
      "loss": 3.0681,
      "step": 50390
    },
    {
      "epoch": 2e-07,
      "grad_norm": 0.5390625,
      "learning_rate": 9.886710059419155e-05,
      "loss": 3.0689,
      "step": 50400
    },
    {
      "epoch": 2.05e-07,
      "grad_norm": 0.46484375,
      "learning_rate": 9.877565364860505e-05,
      "loss": 3.0513,
      "step": 50410
    },
    {
      "epoch": 2.1e-07,
      "grad_norm": 0.4921875,
      "learning_rate": 9.868424670218297e-05,
      "loss": 3.061,
      "step": 50420
    },
    {
      "epoch": 2.15e-07,
      "grad_norm": 0.486328125,
      "learning_rate": 9.859287981013483e-05,
      "loss": 3.0643,
      "step": 50430
    },
    {
      "epoch": 2.2e-07,
      "grad_norm": 0.70703125,
      "learning_rate": 9.850155302764554e-05,
      "loss": 3.0509,
      "step": 50440
    },
    {
      "epoch": 2.25e-07,
      "grad_norm": 0.490234375,
      "learning_rate": 9.841026640987615e-05,
      "loss": 3.076,
      "step": 50450
    },
    {
      "epoch": 2.3e-07,
      "grad_norm": 0.474609375,
      "learning_rate": 9.83190200119633e-05,
      "loss": 3.0528,
      "step": 50460
    },
    {
      "epoch": 2.35e-07,
      "grad_norm": 0.52734375,
      "learning_rate": 9.822781388901932e-05,
      "loss": 3.0495,
      "step": 50470
    },
    {
      "epoch": 2.4e-07,
      "grad_norm": 0.52734375,
      "learning_rate": 9.813664809613225e-05,
      "loss": 3.0791,
      "step": 50480
    },
    {
      "epoch": 2.45e-07,
      "grad_norm": 0.58984375,
      "learning_rate": 9.804552268836576e-05,
      "loss": 3.0763,
      "step": 50490
    },
    {
      "epoch": 2.5e-07,
      "grad_norm": 0.494140625,
      "learning_rate": 9.795443772075904e-05,
      "loss": 3.0677,
      "step": 50500
    },
    {
      "epoch": 2.5e-07,
      "eval_loss": 3.129530191421509,
      "eval_runtime": 0.2149,
      "eval_samples_per_second": 148.921,
      "eval_steps_per_second": 4.654,
      "step": 50500
    },
    {
      "epoch": 2.55e-07,
      "grad_norm": 0.453125,
      "learning_rate": 9.786339324832714e-05,
      "loss": 3.0699,
      "step": 50510
    },
    {
      "epoch": 2.6e-07,
      "grad_norm": 0.455078125,
      "learning_rate": 9.777238932606025e-05,
      "loss": 3.053,
      "step": 50520
    },
    {
      "epoch": 2.65e-07,
      "grad_norm": 0.4765625,
      "learning_rate": 9.768142600892437e-05,
      "loss": 3.0643,
      "step": 50530
    },
    {
      "epoch": 2.7e-07,
      "grad_norm": 0.515625,
      "learning_rate": 9.759050335186092e-05,
      "loss": 3.0705,
      "step": 50540
    },
    {
      "epoch": 2.75e-07,
      "grad_norm": 0.462890625,
      "learning_rate": 9.749962140978659e-05,
      "loss": 3.0879,
      "step": 50550
    },
    {
      "epoch": 2.8e-07,
      "grad_norm": 0.47265625,
      "learning_rate": 9.74087802375937e-05,
      "loss": 3.0484,
      "step": 50560
    },
    {
      "epoch": 2.85e-07,
      "grad_norm": 0.48046875,
      "learning_rate": 9.731797989014985e-05,
      "loss": 3.0952,
      "step": 50570
    },
    {
      "epoch": 2.9e-07,
      "grad_norm": 0.53125,
      "learning_rate": 9.722722042229788e-05,
      "loss": 3.0392,
      "step": 50580
    },
    {
      "epoch": 2.95e-07,
      "grad_norm": 0.52734375,
      "learning_rate": 9.713650188885624e-05,
      "loss": 3.0659,
      "step": 50590
    },
    {
      "epoch": 3e-07,
      "grad_norm": 0.474609375,
      "learning_rate": 9.704582434461823e-05,
      "loss": 3.05,
      "step": 50600
    },
    {
      "epoch": 3.05e-07,
      "grad_norm": 0.5390625,
      "learning_rate": 9.695518784435277e-05,
      "loss": 3.0352,
      "step": 50610
    },
    {
      "epoch": 3.1e-07,
      "grad_norm": 0.5078125,
      "learning_rate": 9.686459244280379e-05,
      "loss": 3.0766,
      "step": 50620
    },
    {
      "epoch": 3.15e-07,
      "grad_norm": 0.447265625,
      "learning_rate": 9.677403819469046e-05,
      "loss": 3.087,
      "step": 50630
    },
    {
      "epoch": 3.2e-07,
      "grad_norm": 0.48828125,
      "learning_rate": 9.668352515470711e-05,
      "loss": 3.0798,
      "step": 50640
    },
    {
      "epoch": 3.25e-07,
      "grad_norm": 0.466796875,
      "learning_rate": 9.659305337752307e-05,
      "loss": 3.1194,
      "step": 50650
    },
    {
      "epoch": 3.3e-07,
      "grad_norm": 0.48828125,
      "learning_rate": 9.650262291778282e-05,
      "loss": 3.0624,
      "step": 50660
    },
    {
      "epoch": 3.35e-07,
      "grad_norm": 0.486328125,
      "learning_rate": 9.641223383010607e-05,
      "loss": 3.0865,
      "step": 50670
    },
    {
      "epoch": 3.4e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 9.632188616908713e-05,
      "loss": 3.0837,
      "step": 50680
    },
    {
      "epoch": 3.45e-07,
      "grad_norm": 0.52734375,
      "learning_rate": 9.623157998929572e-05,
      "loss": 3.0708,
      "step": 50690
    },
    {
      "epoch": 3.5e-07,
      "grad_norm": 0.4453125,
      "learning_rate": 9.614131534527612e-05,
      "loss": 3.0575,
      "step": 50700
    },
    {
      "epoch": 3.55e-07,
      "grad_norm": 0.474609375,
      "learning_rate": 9.605109229154786e-05,
      "loss": 3.0563,
      "step": 50710
    },
    {
      "epoch": 3.6e-07,
      "grad_norm": 0.5078125,
      "learning_rate": 9.596091088260518e-05,
      "loss": 3.043,
      "step": 50720
    },
    {
      "epoch": 3.65e-07,
      "grad_norm": 0.484375,
      "learning_rate": 9.587077117291711e-05,
      "loss": 3.0689,
      "step": 50730
    },
    {
      "epoch": 3.7e-07,
      "grad_norm": 0.54296875,
      "learning_rate": 9.578067321692766e-05,
      "loss": 3.0457,
      "step": 50740
    },
    {
      "epoch": 3.75e-07,
      "grad_norm": 0.52734375,
      "learning_rate": 9.569061706905547e-05,
      "loss": 3.0884,
      "step": 50750
    },
    {
      "epoch": 3.8e-07,
      "grad_norm": 0.47265625,
      "learning_rate": 9.560060278369402e-05,
      "loss": 3.0726,
      "step": 50760
    },
    {
      "epoch": 3.85e-07,
      "grad_norm": 0.4609375,
      "learning_rate": 9.551063041521143e-05,
      "loss": 3.0716,
      "step": 50770
    },
    {
      "epoch": 3.9e-07,
      "grad_norm": 0.52734375,
      "learning_rate": 9.542070001795062e-05,
      "loss": 3.0654,
      "step": 50780
    },
    {
      "epoch": 3.95e-07,
      "grad_norm": 0.470703125,
      "learning_rate": 9.533081164622898e-05,
      "loss": 3.0563,
      "step": 50790
    },
    {
      "epoch": 4e-07,
      "grad_norm": 0.462890625,
      "learning_rate": 9.524096535433879e-05,
      "loss": 3.0715,
      "step": 50800
    },
    {
      "epoch": 4.05e-07,
      "grad_norm": 0.466796875,
      "learning_rate": 9.515116119654659e-05,
      "loss": 3.0468,
      "step": 50810
    },
    {
      "epoch": 4.1e-07,
      "grad_norm": 0.5234375,
      "learning_rate": 9.50613992270937e-05,
      "loss": 3.0638,
      "step": 50820
    },
    {
      "epoch": 4.15e-07,
      "grad_norm": 0.484375,
      "learning_rate": 9.49716795001959e-05,
      "loss": 3.0668,
      "step": 50830
    },
    {
      "epoch": 4.2e-07,
      "grad_norm": 0.453125,
      "learning_rate": 9.488200207004341e-05,
      "loss": 3.0756,
      "step": 50840
    },
    {
      "epoch": 4.25e-07,
      "grad_norm": 0.478515625,
      "learning_rate": 9.479236699080096e-05,
      "loss": 3.0522,
      "step": 50850
    },
    {
      "epoch": 4.3e-07,
      "grad_norm": 0.52734375,
      "learning_rate": 9.470277431660766e-05,
      "loss": 3.0698,
      "step": 50860
    },
    {
      "epoch": 4.35e-07,
      "grad_norm": 0.45703125,
      "learning_rate": 9.4613224101577e-05,
      "loss": 3.0707,
      "step": 50870
    },
    {
      "epoch": 4.4e-07,
      "grad_norm": 0.439453125,
      "learning_rate": 9.452371639979695e-05,
      "loss": 3.0788,
      "step": 50880
    },
    {
      "epoch": 4.45e-07,
      "grad_norm": 0.490234375,
      "learning_rate": 9.443425126532956e-05,
      "loss": 3.0516,
      "step": 50890
    },
    {
      "epoch": 4.5e-07,
      "grad_norm": 0.486328125,
      "learning_rate": 9.434482875221145e-05,
      "loss": 3.075,
      "step": 50900
    },
    {
      "epoch": 4.55e-07,
      "grad_norm": 0.494140625,
      "learning_rate": 9.42554489144532e-05,
      "loss": 3.0823,
      "step": 50910
    },
    {
      "epoch": 4.6e-07,
      "grad_norm": 0.4609375,
      "learning_rate": 9.416611180603991e-05,
      "loss": 3.0312,
      "step": 50920
    },
    {
      "epoch": 4.65e-07,
      "grad_norm": 0.486328125,
      "learning_rate": 9.407681748093071e-05,
      "loss": 3.0439,
      "step": 50930
    },
    {
      "epoch": 4.7e-07,
      "grad_norm": 0.5234375,
      "learning_rate": 9.398756599305886e-05,
      "loss": 3.0656,
      "step": 50940
    },
    {
      "epoch": 4.75e-07,
      "grad_norm": 0.486328125,
      "learning_rate": 9.389835739633186e-05,
      "loss": 3.0561,
      "step": 50950
    },
    {
      "epoch": 4.8e-07,
      "grad_norm": 0.51171875,
      "learning_rate": 9.380919174463118e-05,
      "loss": 3.0647,
      "step": 50960
    },
    {
      "epoch": 4.85e-07,
      "grad_norm": 0.5625,
      "learning_rate": 9.37200690918124e-05,
      "loss": 3.0431,
      "step": 50970
    },
    {
      "epoch": 4.9e-07,
      "grad_norm": 0.50390625,
      "learning_rate": 9.363098949170531e-05,
      "loss": 3.0374,
      "step": 50980
    },
    {
      "epoch": 4.95e-07,
      "grad_norm": 0.498046875,
      "learning_rate": 9.354195299811329e-05,
      "loss": 3.0627,
      "step": 50990
    },
    {
      "epoch": 5e-07,
      "grad_norm": 0.484375,
      "learning_rate": 9.345295966481413e-05,
      "loss": 3.0979,
      "step": 51000
    },
    {
      "epoch": 5e-07,
      "eval_loss": 3.1213490962982178,
      "eval_runtime": 0.0603,
      "eval_samples_per_second": 530.996,
      "eval_steps_per_second": 16.594,
      "step": 51000
    },
    {
      "epoch": 5.05e-07,
      "grad_norm": 0.5,
      "learning_rate": 9.336400954555923e-05,
      "loss": 3.0963,
      "step": 51010
    },
    {
      "epoch": 5.1e-07,
      "grad_norm": 0.50390625,
      "learning_rate": 9.327510269407403e-05,
      "loss": 3.0656,
      "step": 51020
    },
    {
      "epoch": 5.15e-07,
      "grad_norm": 0.453125,
      "learning_rate": 9.318623916405782e-05,
      "loss": 3.0855,
      "step": 51030
    },
    {
      "epoch": 5.2e-07,
      "grad_norm": 0.4765625,
      "learning_rate": 9.309741900918372e-05,
      "loss": 3.0654,
      "step": 51040
    },
    {
      "epoch": 5.25e-07,
      "grad_norm": 0.4921875,
      "learning_rate": 9.30086422830986e-05,
      "loss": 3.0488,
      "step": 51050
    },
    {
      "epoch": 5.3e-07,
      "grad_norm": 0.50390625,
      "learning_rate": 9.291990903942324e-05,
      "loss": 3.0746,
      "step": 51060
    },
    {
      "epoch": 5.35e-07,
      "grad_norm": 0.50390625,
      "learning_rate": 9.283121933175194e-05,
      "loss": 3.0698,
      "step": 51070
    },
    {
      "epoch": 5.4e-07,
      "grad_norm": 0.498046875,
      "learning_rate": 9.274257321365295e-05,
      "loss": 3.0538,
      "step": 51080
    },
    {
      "epoch": 5.45e-07,
      "grad_norm": 0.466796875,
      "learning_rate": 9.265397073866799e-05,
      "loss": 3.0553,
      "step": 51090
    },
    {
      "epoch": 5.5e-07,
      "grad_norm": 2.03125,
      "learning_rate": 9.256541196031252e-05,
      "loss": 3.0775,
      "step": 51100
    },
    {
      "epoch": 5.55e-07,
      "grad_norm": 0.5703125,
      "learning_rate": 9.247689693207562e-05,
      "loss": 3.0664,
      "step": 51110
    },
    {
      "epoch": 5.6e-07,
      "grad_norm": 0.490234375,
      "learning_rate": 9.238842570741985e-05,
      "loss": 3.0337,
      "step": 51120
    },
    {
      "epoch": 5.65e-07,
      "grad_norm": 0.498046875,
      "learning_rate": 9.22999983397814e-05,
      "loss": 3.0708,
      "step": 51130
    },
    {
      "epoch": 5.7e-07,
      "grad_norm": 0.546875,
      "learning_rate": 9.221161488256997e-05,
      "loss": 3.0573,
      "step": 51140
    },
    {
      "epoch": 5.75e-07,
      "grad_norm": 0.47265625,
      "learning_rate": 9.212327538916862e-05,
      "loss": 3.0555,
      "step": 51150
    },
    {
      "epoch": 5.8e-07,
      "grad_norm": 0.453125,
      "learning_rate": 9.203497991293414e-05,
      "loss": 3.0541,
      "step": 51160
    },
    {
      "epoch": 5.85e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 9.194672850719638e-05,
      "loss": 3.0425,
      "step": 51170
    },
    {
      "epoch": 5.9e-07,
      "grad_norm": 0.5078125,
      "learning_rate": 9.185852122525873e-05,
      "loss": 3.0713,
      "step": 51180
    },
    {
      "epoch": 5.95e-07,
      "grad_norm": 0.5,
      "learning_rate": 9.177035812039808e-05,
      "loss": 3.0829,
      "step": 51190
    },
    {
      "epoch": 6e-07,
      "grad_norm": 0.45703125,
      "learning_rate": 9.168223924586428e-05,
      "loss": 3.0628,
      "step": 51200
    },
    {
      "epoch": 6.05e-07,
      "grad_norm": 0.53515625,
      "learning_rate": 9.159416465488085e-05,
      "loss": 3.0838,
      "step": 51210
    },
    {
      "epoch": 6.1e-07,
      "grad_norm": 0.48046875,
      "learning_rate": 9.150613440064436e-05,
      "loss": 3.0512,
      "step": 51220
    },
    {
      "epoch": 6.15e-07,
      "grad_norm": 0.53125,
      "learning_rate": 9.141814853632458e-05,
      "loss": 3.0805,
      "step": 51230
    },
    {
      "epoch": 6.2e-07,
      "grad_norm": 0.4765625,
      "learning_rate": 9.133020711506455e-05,
      "loss": 3.036,
      "step": 51240
    },
    {
      "epoch": 6.25e-07,
      "grad_norm": 0.439453125,
      "learning_rate": 9.124231018998044e-05,
      "loss": 3.0639,
      "step": 51250
    },
    {
      "epoch": 6.3e-07,
      "grad_norm": 0.453125,
      "learning_rate": 9.115445781416147e-05,
      "loss": 3.0884,
      "step": 51260
    },
    {
      "epoch": 6.35e-07,
      "grad_norm": 0.52734375,
      "learning_rate": 9.106665004067018e-05,
      "loss": 3.0888,
      "step": 51270
    },
    {
      "epoch": 6.4e-07,
      "grad_norm": 0.515625,
      "learning_rate": 9.097888692254183e-05,
      "loss": 3.0709,
      "step": 51280
    },
    {
      "epoch": 6.45e-07,
      "grad_norm": 0.49609375,
      "learning_rate": 9.0891168512785e-05,
      "loss": 3.062,
      "step": 51290
    },
    {
      "epoch": 6.5e-07,
      "grad_norm": 0.462890625,
      "learning_rate": 9.080349486438118e-05,
      "loss": 3.0852,
      "step": 51300
    },
    {
      "epoch": 6.55e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 9.071586603028473e-05,
      "loss": 3.0751,
      "step": 51310
    },
    {
      "epoch": 6.6e-07,
      "grad_norm": 0.466796875,
      "learning_rate": 9.062828206342305e-05,
      "loss": 3.0617,
      "step": 51320
    },
    {
      "epoch": 6.65e-07,
      "grad_norm": 0.498046875,
      "learning_rate": 9.054074301669638e-05,
      "loss": 3.0574,
      "step": 51330
    },
    {
      "epoch": 6.7e-07,
      "grad_norm": 0.48046875,
      "learning_rate": 9.045324894297791e-05,
      "loss": 3.0712,
      "step": 51340
    },
    {
      "epoch": 6.75e-07,
      "grad_norm": 0.4921875,
      "learning_rate": 9.036579989511356e-05,
      "loss": 3.0725,
      "step": 51350
    },
    {
      "epoch": 6.8e-07,
      "grad_norm": 0.51171875,
      "learning_rate": 9.027839592592207e-05,
      "loss": 3.072,
      "step": 51360
    },
    {
      "epoch": 6.85e-07,
      "grad_norm": 0.484375,
      "learning_rate": 9.019103708819514e-05,
      "loss": 3.0584,
      "step": 51370
    },
    {
      "epoch": 6.9e-07,
      "grad_norm": 0.484375,
      "learning_rate": 9.010372343469686e-05,
      "loss": 3.0186,
      "step": 51380
    },
    {
      "epoch": 6.95e-07,
      "grad_norm": 0.478515625,
      "learning_rate": 9.001645501816436e-05,
      "loss": 3.0659,
      "step": 51390
    },
    {
      "epoch": 7e-07,
      "grad_norm": 0.5390625,
      "learning_rate": 8.992923189130728e-05,
      "loss": 3.04,
      "step": 51400
    },
    {
      "epoch": 7.05e-07,
      "grad_norm": 0.462890625,
      "learning_rate": 8.984205410680794e-05,
      "loss": 3.0649,
      "step": 51410
    },
    {
      "epoch": 7.1e-07,
      "grad_norm": 0.48828125,
      "learning_rate": 8.975492171732129e-05,
      "loss": 3.034,
      "step": 51420
    },
    {
      "epoch": 7.15e-07,
      "grad_norm": 0.5,
      "learning_rate": 8.96678347754748e-05,
      "loss": 3.0646,
      "step": 51430
    },
    {
      "epoch": 7.2e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 8.958079333386852e-05,
      "loss": 3.0774,
      "step": 51440
    },
    {
      "epoch": 7.25e-07,
      "grad_norm": 0.515625,
      "learning_rate": 8.949379744507516e-05,
      "loss": 3.0537,
      "step": 51450
    },
    {
      "epoch": 7.3e-07,
      "grad_norm": 0.45703125,
      "learning_rate": 8.940684716163963e-05,
      "loss": 3.0421,
      "step": 51460
    },
    {
      "epoch": 7.35e-07,
      "grad_norm": 0.5078125,
      "learning_rate": 8.931994253607957e-05,
      "loss": 3.054,
      "step": 51470
    },
    {
      "epoch": 7.4e-07,
      "grad_norm": 0.45703125,
      "learning_rate": 8.923308362088488e-05,
      "loss": 3.0506,
      "step": 51480
    },
    {
      "epoch": 7.45e-07,
      "grad_norm": 0.5,
      "learning_rate": 8.914627046851789e-05,
      "loss": 3.0629,
      "step": 51490
    },
    {
      "epoch": 7.5e-07,
      "grad_norm": 0.515625,
      "learning_rate": 8.905950313141332e-05,
      "loss": 3.0423,
      "step": 51500
    },
    {
      "epoch": 7.5e-07,
      "eval_loss": 3.111537218093872,
      "eval_runtime": 0.063,
      "eval_samples_per_second": 507.937,
      "eval_steps_per_second": 15.873,
      "step": 51500
    },
    {
      "epoch": 7.55e-07,
      "grad_norm": 0.515625,
      "learning_rate": 8.897278166197817e-05,
      "loss": 3.0612,
      "step": 51510
    },
    {
      "epoch": 7.6e-07,
      "grad_norm": 0.4765625,
      "learning_rate": 8.888610611259177e-05,
      "loss": 3.0569,
      "step": 51520
    },
    {
      "epoch": 7.65e-07,
      "grad_norm": 0.515625,
      "learning_rate": 8.879947653560578e-05,
      "loss": 3.0693,
      "step": 51530
    },
    {
      "epoch": 7.7e-07,
      "grad_norm": 0.5078125,
      "learning_rate": 8.871289298334383e-05,
      "loss": 3.0439,
      "step": 51540
    },
    {
      "epoch": 7.75e-07,
      "grad_norm": 0.474609375,
      "learning_rate": 8.862635550810219e-05,
      "loss": 3.0643,
      "step": 51550
    },
    {
      "epoch": 7.8e-07,
      "grad_norm": 0.48828125,
      "learning_rate": 8.853986416214884e-05,
      "loss": 3.0453,
      "step": 51560
    },
    {
      "epoch": 7.85e-07,
      "grad_norm": 0.515625,
      "learning_rate": 8.84534189977242e-05,
      "loss": 3.0771,
      "step": 51570
    },
    {
      "epoch": 7.9e-07,
      "grad_norm": 0.51953125,
      "learning_rate": 8.836702006704075e-05,
      "loss": 3.0559,
      "step": 51580
    },
    {
      "epoch": 7.95e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 8.828066742228283e-05,
      "loss": 3.064,
      "step": 51590
    },
    {
      "epoch": 8e-07,
      "grad_norm": 0.458984375,
      "learning_rate": 8.819436111560719e-05,
      "loss": 3.0784,
      "step": 51600
    },
    {
      "epoch": 8.05e-07,
      "grad_norm": 0.50390625,
      "learning_rate": 8.81081011991423e-05,
      "loss": 3.067,
      "step": 51610
    },
    {
      "epoch": 8.1e-07,
      "grad_norm": 0.470703125,
      "learning_rate": 8.802188772498871e-05,
      "loss": 3.0762,
      "step": 51620
    },
    {
      "epoch": 8.15e-07,
      "grad_norm": 0.55078125,
      "learning_rate": 8.793572074521892e-05,
      "loss": 3.096,
      "step": 51630
    },
    {
      "epoch": 8.2e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 8.784960031187736e-05,
      "loss": 3.0663,
      "step": 51640
    },
    {
      "epoch": 8.25e-07,
      "grad_norm": 0.46875,
      "learning_rate": 8.776352647698025e-05,
      "loss": 3.0662,
      "step": 51650
    },
    {
      "epoch": 8.3e-07,
      "grad_norm": 0.51171875,
      "learning_rate": 8.767749929251591e-05,
      "loss": 3.0433,
      "step": 51660
    },
    {
      "epoch": 8.35e-07,
      "grad_norm": 0.53515625,
      "learning_rate": 8.75915188104441e-05,
      "loss": 3.0467,
      "step": 51670
    },
    {
      "epoch": 8.4e-07,
      "grad_norm": 0.48046875,
      "learning_rate": 8.750558508269677e-05,
      "loss": 3.0472,
      "step": 51680
    },
    {
      "epoch": 8.45e-07,
      "grad_norm": 0.5,
      "learning_rate": 8.741969816117739e-05,
      "loss": 3.0681,
      "step": 51690
    },
    {
      "epoch": 8.5e-07,
      "grad_norm": 0.46484375,
      "learning_rate": 8.73338580977612e-05,
      "loss": 3.0703,
      "step": 51700
    },
    {
      "epoch": 8.55e-07,
      "grad_norm": 0.515625,
      "learning_rate": 8.724806494429514e-05,
      "loss": 3.0697,
      "step": 51710
    },
    {
      "epoch": 8.6e-07,
      "grad_norm": 0.498046875,
      "learning_rate": 8.71623187525979e-05,
      "loss": 3.049,
      "step": 51720
    },
    {
      "epoch": 8.65e-07,
      "grad_norm": 0.53125,
      "learning_rate": 8.707661957445964e-05,
      "loss": 3.0461,
      "step": 51730
    },
    {
      "epoch": 8.7e-07,
      "grad_norm": 0.5078125,
      "learning_rate": 8.699096746164237e-05,
      "loss": 3.0484,
      "step": 51740
    },
    {
      "epoch": 8.75e-07,
      "grad_norm": 0.4765625,
      "learning_rate": 8.690536246587934e-05,
      "loss": 3.0563,
      "step": 51750
    },
    {
      "epoch": 8.8e-07,
      "grad_norm": 0.5390625,
      "learning_rate": 8.681980463887575e-05,
      "loss": 3.0285,
      "step": 51760
    },
    {
      "epoch": 8.85e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 8.673429403230783e-05,
      "loss": 3.0502,
      "step": 51770
    },
    {
      "epoch": 8.9e-07,
      "grad_norm": 0.462890625,
      "learning_rate": 8.664883069782372e-05,
      "loss": 3.0344,
      "step": 51780
    },
    {
      "epoch": 8.95e-07,
      "grad_norm": 0.498046875,
      "learning_rate": 8.656341468704282e-05,
      "loss": 3.0759,
      "step": 51790
    },
    {
      "epoch": 9e-07,
      "grad_norm": 0.494140625,
      "learning_rate": 8.647804605155591e-05,
      "loss": 3.0505,
      "step": 51800
    },
    {
      "epoch": 9.05e-07,
      "grad_norm": 0.515625,
      "learning_rate": 8.639272484292526e-05,
      "loss": 3.0457,
      "step": 51810
    },
    {
      "epoch": 9.1e-07,
      "grad_norm": 0.478515625,
      "learning_rate": 8.63074511126844e-05,
      "loss": 3.0269,
      "step": 51820
    },
    {
      "epoch": 9.15e-07,
      "grad_norm": 0.482421875,
      "learning_rate": 8.622222491233819e-05,
      "loss": 3.0507,
      "step": 51830
    },
    {
      "epoch": 9.2e-07,
      "grad_norm": 0.4765625,
      "learning_rate": 8.613704629336296e-05,
      "loss": 3.0477,
      "step": 51840
    },
    {
      "epoch": 9.25e-07,
      "grad_norm": 0.490234375,
      "learning_rate": 8.605191530720594e-05,
      "loss": 3.0362,
      "step": 51850
    },
    {
      "epoch": 9.3e-07,
      "grad_norm": 0.4765625,
      "learning_rate": 8.5966832005286e-05,
      "loss": 3.0317,
      "step": 51860
    },
    {
      "epoch": 9.35e-07,
      "grad_norm": 0.486328125,
      "learning_rate": 8.588179643899294e-05,
      "loss": 3.0339,
      "step": 51870
    },
    {
      "epoch": 9.4e-07,
      "grad_norm": 0.45703125,
      "learning_rate": 8.579680865968783e-05,
      "loss": 3.0329,
      "step": 51880
    },
    {
      "epoch": 9.45e-07,
      "grad_norm": 0.46875,
      "learning_rate": 8.571186871870279e-05,
      "loss": 3.0419,
      "step": 51890
    },
    {
      "epoch": 9.5e-07,
      "grad_norm": 0.54296875,
      "learning_rate": 8.562697666734117e-05,
      "loss": 3.0322,
      "step": 51900
    },
    {
      "epoch": 9.55e-07,
      "grad_norm": 0.48828125,
      "learning_rate": 8.554213255687725e-05,
      "loss": 3.0449,
      "step": 51910
    },
    {
      "epoch": 9.6e-07,
      "grad_norm": 0.46875,
      "learning_rate": 8.54573364385566e-05,
      "loss": 3.0549,
      "step": 51920
    },
    {
      "epoch": 9.65e-07,
      "grad_norm": 0.486328125,
      "learning_rate": 8.537258836359544e-05,
      "loss": 3.056,
      "step": 51930
    },
    {
      "epoch": 9.7e-07,
      "grad_norm": 0.4609375,
      "learning_rate": 8.528788838318132e-05,
      "loss": 3.037,
      "step": 51940
    },
    {
      "epoch": 9.75e-07,
      "grad_norm": 0.54296875,
      "learning_rate": 8.520323654847253e-05,
      "loss": 3.0611,
      "step": 51950
    },
    {
      "epoch": 9.8e-07,
      "grad_norm": 0.51171875,
      "learning_rate": 8.511863291059835e-05,
      "loss": 3.0313,
      "step": 51960
    },
    {
      "epoch": 9.85e-07,
      "grad_norm": 0.5390625,
      "learning_rate": 8.503407752065897e-05,
      "loss": 3.0129,
      "step": 51970
    },
    {
      "epoch": 9.9e-07,
      "grad_norm": 0.55859375,
      "learning_rate": 8.494957042972535e-05,
      "loss": 3.0485,
      "step": 51980
    },
    {
      "epoch": 9.95e-07,
      "grad_norm": 0.478515625,
      "learning_rate": 8.48651116888394e-05,
      "loss": 3.0369,
      "step": 51990
    },
    {
      "epoch": 1e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 8.478070134901371e-05,
      "loss": 3.0345,
      "step": 52000
    },
    {
      "epoch": 1e-06,
      "eval_loss": 3.106865882873535,
      "eval_runtime": 0.064,
      "eval_samples_per_second": 500.325,
      "eval_steps_per_second": 15.635,
      "step": 52000
    },
    {
      "epoch": 1.005e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 8.469633946123175e-05,
      "loss": 3.0313,
      "step": 52010
    },
    {
      "epoch": 1.01e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 8.461202607644761e-05,
      "loss": 3.0478,
      "step": 52020
    },
    {
      "epoch": 1.015e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 8.452776124558618e-05,
      "loss": 3.0471,
      "step": 52030
    },
    {
      "epoch": 1.02e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 8.444354501954292e-05,
      "loss": 3.053,
      "step": 52040
    },
    {
      "epoch": 1.025e-06,
      "grad_norm": 0.484375,
      "learning_rate": 8.435937744918413e-05,
      "loss": 3.0378,
      "step": 52050
    },
    {
      "epoch": 1.03e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 8.427525858534644e-05,
      "loss": 3.0725,
      "step": 52060
    },
    {
      "epoch": 1.035e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 8.419118847883731e-05,
      "loss": 3.0044,
      "step": 52070
    },
    {
      "epoch": 1.04e-06,
      "grad_norm": 0.515625,
      "learning_rate": 8.410716718043462e-05,
      "loss": 3.0213,
      "step": 52080
    },
    {
      "epoch": 1.045e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 8.40231947408868e-05,
      "loss": 3.0209,
      "step": 52090
    },
    {
      "epoch": 1.05e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 8.393927121091276e-05,
      "loss": 3.0581,
      "step": 52100
    },
    {
      "epoch": 1.055e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 8.38553966412019e-05,
      "loss": 3.0104,
      "step": 52110
    },
    {
      "epoch": 1.06e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 8.377157108241399e-05,
      "loss": 3.0266,
      "step": 52120
    },
    {
      "epoch": 1.065e-06,
      "grad_norm": 0.54296875,
      "learning_rate": 8.368779458517935e-05,
      "loss": 3.0537,
      "step": 52130
    },
    {
      "epoch": 1.07e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 8.360406720009839e-05,
      "loss": 3.047,
      "step": 52140
    },
    {
      "epoch": 1.075e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 8.352038897774213e-05,
      "loss": 3.0502,
      "step": 52150
    },
    {
      "epoch": 1.08e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 8.343675996865178e-05,
      "loss": 3.028,
      "step": 52160
    },
    {
      "epoch": 1.085e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 8.335318022333878e-05,
      "loss": 3.0312,
      "step": 52170
    },
    {
      "epoch": 1.09e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 8.326964979228493e-05,
      "loss": 3.0443,
      "step": 52180
    },
    {
      "epoch": 1.095e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 8.318616872594211e-05,
      "loss": 3.0671,
      "step": 52190
    },
    {
      "epoch": 1.1e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 8.310273707473246e-05,
      "loss": 3.0629,
      "step": 52200
    },
    {
      "epoch": 1.105e-06,
      "grad_norm": 0.5390625,
      "learning_rate": 8.30193548890484e-05,
      "loss": 3.0152,
      "step": 52210
    },
    {
      "epoch": 1.11e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 8.293602221925213e-05,
      "loss": 3.019,
      "step": 52220
    },
    {
      "epoch": 1.115e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 8.285273911567638e-05,
      "loss": 3.0314,
      "step": 52230
    },
    {
      "epoch": 1.12e-06,
      "grad_norm": 0.484375,
      "learning_rate": 8.27695056286235e-05,
      "loss": 3.0169,
      "step": 52240
    },
    {
      "epoch": 1.125e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 8.268632180836621e-05,
      "loss": 3.0317,
      "step": 52250
    },
    {
      "epoch": 1.13e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 8.26031877051471e-05,
      "loss": 3.0676,
      "step": 52260
    },
    {
      "epoch": 1.135e-06,
      "grad_norm": 0.5390625,
      "learning_rate": 8.252010336917873e-05,
      "loss": 3.0462,
      "step": 52270
    },
    {
      "epoch": 1.14e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 8.243706885064361e-05,
      "loss": 3.0421,
      "step": 52280
    },
    {
      "epoch": 1.145e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 8.235408419969416e-05,
      "loss": 3.0569,
      "step": 52290
    },
    {
      "epoch": 1.15e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 8.227114946645264e-05,
      "loss": 3.0182,
      "step": 52300
    },
    {
      "epoch": 1.155e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 8.218826470101133e-05,
      "loss": 3.0491,
      "step": 52310
    },
    {
      "epoch": 1.16e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 8.210542995343203e-05,
      "loss": 3.0268,
      "step": 52320
    },
    {
      "epoch": 1.165e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 8.202264527374663e-05,
      "loss": 3.0356,
      "step": 52330
    },
    {
      "epoch": 1.17e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 8.193991071195662e-05,
      "loss": 3.036,
      "step": 52340
    },
    {
      "epoch": 1.175e-06,
      "grad_norm": 0.484375,
      "learning_rate": 8.185722631803323e-05,
      "loss": 3.0329,
      "step": 52350
    },
    {
      "epoch": 1.18e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 8.177459214191744e-05,
      "loss": 3.0284,
      "step": 52360
    },
    {
      "epoch": 1.185e-06,
      "grad_norm": 0.55078125,
      "learning_rate": 8.16920082335198e-05,
      "loss": 3.0187,
      "step": 52370
    },
    {
      "epoch": 1.19e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 8.160947464272062e-05,
      "loss": 3.0333,
      "step": 52380
    },
    {
      "epoch": 1.195e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 8.152699141936981e-05,
      "loss": 3.0419,
      "step": 52390
    },
    {
      "epoch": 1.2e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 8.14445586132867e-05,
      "loss": 3.0308,
      "step": 52400
    },
    {
      "epoch": 1.205e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 8.136217627426032e-05,
      "loss": 3.0559,
      "step": 52410
    },
    {
      "epoch": 1.21e-06,
      "grad_norm": 0.546875,
      "learning_rate": 8.127984445204928e-05,
      "loss": 3.0059,
      "step": 52420
    },
    {
      "epoch": 1.215e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 8.119756319638142e-05,
      "loss": 3.0506,
      "step": 52430
    },
    {
      "epoch": 1.22e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 8.111533255695434e-05,
      "loss": 3.0467,
      "step": 52440
    },
    {
      "epoch": 1.225e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 8.103315258343477e-05,
      "loss": 3.0548,
      "step": 52450
    },
    {
      "epoch": 1.23e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 8.095102332545911e-05,
      "loss": 3.0425,
      "step": 52460
    },
    {
      "epoch": 1.235e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 8.0868944832633e-05,
      "loss": 3.0459,
      "step": 52470
    },
    {
      "epoch": 1.24e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 8.078691715453138e-05,
      "loss": 3.0413,
      "step": 52480
    },
    {
      "epoch": 1.245e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 8.070494034069854e-05,
      "loss": 3.0205,
      "step": 52490
    },
    {
      "epoch": 1.25e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 8.062301444064812e-05,
      "loss": 3.0107,
      "step": 52500
    },
    {
      "epoch": 1.25e-06,
      "eval_loss": 3.10275936126709,
      "eval_runtime": 0.0609,
      "eval_samples_per_second": 525.678,
      "eval_steps_per_second": 16.427,
      "step": 52500
    },
    {
      "epoch": 1.255e-06,
      "grad_norm": 0.453125,
      "learning_rate": 8.054113950386285e-05,
      "loss": 3.0403,
      "step": 52510
    },
    {
      "epoch": 1.26e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 8.045931557979492e-05,
      "loss": 3.0551,
      "step": 52520
    },
    {
      "epoch": 1.265e-06,
      "grad_norm": 0.484375,
      "learning_rate": 8.037754271786537e-05,
      "loss": 3.0274,
      "step": 52530
    },
    {
      "epoch": 1.27e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 8.029582096746468e-05,
      "loss": 3.0469,
      "step": 52540
    },
    {
      "epoch": 1.275e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 8.021415037795241e-05,
      "loss": 3.0215,
      "step": 52550
    },
    {
      "epoch": 1.28e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 8.013253099865715e-05,
      "loss": 3.0286,
      "step": 52560
    },
    {
      "epoch": 1.285e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 8.005096287887653e-05,
      "loss": 3.0483,
      "step": 52570
    },
    {
      "epoch": 1.29e-06,
      "grad_norm": 0.484375,
      "learning_rate": 7.996944606787734e-05,
      "loss": 3.0445,
      "step": 52580
    },
    {
      "epoch": 1.295e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 7.988798061489522e-05,
      "loss": 3.037,
      "step": 52590
    },
    {
      "epoch": 1.3e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 7.980656656913505e-05,
      "loss": 3.0108,
      "step": 52600
    },
    {
      "epoch": 1.305e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 7.97252039797703e-05,
      "loss": 3.0436,
      "step": 52610
    },
    {
      "epoch": 1.31e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 7.964389289594372e-05,
      "loss": 3.0363,
      "step": 52620
    },
    {
      "epoch": 1.315e-06,
      "grad_norm": 0.515625,
      "learning_rate": 7.956263336676669e-05,
      "loss": 3.0206,
      "step": 52630
    },
    {
      "epoch": 1.32e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 7.948142544131959e-05,
      "loss": 3.0669,
      "step": 52640
    },
    {
      "epoch": 1.325e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 7.940026916865158e-05,
      "loss": 3.0378,
      "step": 52650
    },
    {
      "epoch": 1.33e-06,
      "grad_norm": 0.58984375,
      "learning_rate": 7.931916459778061e-05,
      "loss": 3.0371,
      "step": 52660
    },
    {
      "epoch": 1.335e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 7.923811177769344e-05,
      "loss": 3.0373,
      "step": 52670
    },
    {
      "epoch": 1.34e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 7.915711075734557e-05,
      "loss": 3.0057,
      "step": 52680
    },
    {
      "epoch": 1.345e-06,
      "grad_norm": 0.484375,
      "learning_rate": 7.907616158566114e-05,
      "loss": 3.0255,
      "step": 52690
    },
    {
      "epoch": 1.35e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 7.899526431153318e-05,
      "loss": 3.0482,
      "step": 52700
    },
    {
      "epoch": 1.355e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 7.891441898382303e-05,
      "loss": 3.0293,
      "step": 52710
    },
    {
      "epoch": 1.36e-06,
      "grad_norm": 0.46875,
      "learning_rate": 7.883362565136098e-05,
      "loss": 3.021,
      "step": 52720
    },
    {
      "epoch": 1.365e-06,
      "grad_norm": 0.5,
      "learning_rate": 7.875288436294577e-05,
      "loss": 3.0617,
      "step": 52730
    },
    {
      "epoch": 1.37e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 7.867219516734468e-05,
      "loss": 3.0296,
      "step": 52740
    },
    {
      "epoch": 1.375e-06,
      "grad_norm": 0.5546875,
      "learning_rate": 7.859155811329361e-05,
      "loss": 3.0677,
      "step": 52750
    },
    {
      "epoch": 1.38e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.851097324949694e-05,
      "loss": 3.0229,
      "step": 52760
    },
    {
      "epoch": 1.385e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.84304406246274e-05,
      "loss": 3.0346,
      "step": 52770
    },
    {
      "epoch": 1.39e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 7.834996028732651e-05,
      "loss": 3.0553,
      "step": 52780
    },
    {
      "epoch": 1.395e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 7.826953228620371e-05,
      "loss": 3.0365,
      "step": 52790
    },
    {
      "epoch": 1.4e-06,
      "grad_norm": 0.56640625,
      "learning_rate": 7.818915666983727e-05,
      "loss": 3.0388,
      "step": 52800
    },
    {
      "epoch": 1.405e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 7.810883348677366e-05,
      "loss": 3.0197,
      "step": 52810
    },
    {
      "epoch": 1.41e-06,
      "grad_norm": 0.515625,
      "learning_rate": 7.802856278552752e-05,
      "loss": 3.0387,
      "step": 52820
    },
    {
      "epoch": 1.415e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 7.794834461458209e-05,
      "loss": 3.0399,
      "step": 52830
    },
    {
      "epoch": 1.42e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 7.786817902238868e-05,
      "loss": 3.042,
      "step": 52840
    },
    {
      "epoch": 1.425e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 7.778806605736686e-05,
      "loss": 3.0575,
      "step": 52850
    },
    {
      "epoch": 1.43e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 7.770800576790451e-05,
      "loss": 3.0402,
      "step": 52860
    },
    {
      "epoch": 1.435e-06,
      "grad_norm": 0.55859375,
      "learning_rate": 7.762799820235763e-05,
      "loss": 3.0296,
      "step": 52870
    },
    {
      "epoch": 1.44e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 7.754804340905029e-05,
      "loss": 3.0476,
      "step": 52880
    },
    {
      "epoch": 1.445e-06,
      "grad_norm": 0.53125,
      "learning_rate": 7.746814143627487e-05,
      "loss": 3.0532,
      "step": 52890
    },
    {
      "epoch": 1.45e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 7.738829233229167e-05,
      "loss": 3.0281,
      "step": 52900
    },
    {
      "epoch": 1.455e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 7.730849614532927e-05,
      "loss": 3.0554,
      "step": 52910
    },
    {
      "epoch": 1.46e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.7228752923584e-05,
      "loss": 3.0659,
      "step": 52920
    },
    {
      "epoch": 1.465e-06,
      "grad_norm": 0.5,
      "learning_rate": 7.714906271522044e-05,
      "loss": 2.9975,
      "step": 52930
    },
    {
      "epoch": 1.47e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 7.706942556837109e-05,
      "loss": 3.0531,
      "step": 52940
    },
    {
      "epoch": 1.475e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 7.698984153113631e-05,
      "loss": 3.0351,
      "step": 52950
    },
    {
      "epoch": 1.48e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 7.69103106515845e-05,
      "loss": 3.0433,
      "step": 52960
    },
    {
      "epoch": 1.485e-06,
      "grad_norm": 0.515625,
      "learning_rate": 7.683083297775186e-05,
      "loss": 3.0431,
      "step": 52970
    },
    {
      "epoch": 1.49e-06,
      "grad_norm": 0.53515625,
      "learning_rate": 7.67514085576425e-05,
      "loss": 3.055,
      "step": 52980
    },
    {
      "epoch": 1.495e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 7.667203743922844e-05,
      "loss": 3.0372,
      "step": 52990
    },
    {
      "epoch": 1.5e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 7.659271967044929e-05,
      "loss": 3.0184,
      "step": 53000
    },
    {
      "epoch": 1.5e-06,
      "eval_loss": 3.1008410453796387,
      "eval_runtime": 0.0647,
      "eval_samples_per_second": 494.323,
      "eval_steps_per_second": 15.448,
      "step": 53000
    },
    {
      "epoch": 1.505e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 7.65134552992127e-05,
      "loss": 3.0403,
      "step": 53010
    },
    {
      "epoch": 1.51e-06,
      "grad_norm": 0.46875,
      "learning_rate": 7.643424437339388e-05,
      "loss": 3.027,
      "step": 53020
    },
    {
      "epoch": 1.515e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 7.635508694083585e-05,
      "loss": 3.0603,
      "step": 53030
    },
    {
      "epoch": 1.52e-06,
      "grad_norm": 0.484375,
      "learning_rate": 7.627598304934924e-05,
      "loss": 3.0123,
      "step": 53040
    },
    {
      "epoch": 1.525e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 7.619693274671247e-05,
      "loss": 3.0107,
      "step": 53050
    },
    {
      "epoch": 1.53e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 7.61179360806714e-05,
      "loss": 3.0384,
      "step": 53060
    },
    {
      "epoch": 1.535e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 7.60389930989398e-05,
      "loss": 3.0456,
      "step": 53070
    },
    {
      "epoch": 1.54e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 7.59601038491986e-05,
      "loss": 3.0511,
      "step": 53080
    },
    {
      "epoch": 1.545e-06,
      "grad_norm": 0.53515625,
      "learning_rate": 7.588126837909671e-05,
      "loss": 3.0622,
      "step": 53090
    },
    {
      "epoch": 1.55e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 7.580248673625017e-05,
      "loss": 3.0289,
      "step": 53100
    },
    {
      "epoch": 1.555e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 7.57237589682428e-05,
      "loss": 3.0143,
      "step": 53110
    },
    {
      "epoch": 1.56e-06,
      "grad_norm": 0.46875,
      "learning_rate": 7.564508512262574e-05,
      "loss": 3.0677,
      "step": 53120
    },
    {
      "epoch": 1.565e-06,
      "grad_norm": 0.515625,
      "learning_rate": 7.556646524691758e-05,
      "loss": 3.05,
      "step": 53130
    },
    {
      "epoch": 1.57e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 7.548789938860431e-05,
      "loss": 3.0401,
      "step": 53140
    },
    {
      "epoch": 1.575e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 7.540938759513933e-05,
      "loss": 3.0375,
      "step": 53150
    },
    {
      "epoch": 1.58e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 7.53309299139433e-05,
      "loss": 3.0354,
      "step": 53160
    },
    {
      "epoch": 1.585e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 7.52525263924044e-05,
      "loss": 3.0318,
      "step": 53170
    },
    {
      "epoch": 1.59e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 7.517417707787776e-05,
      "loss": 3.0517,
      "step": 53180
    },
    {
      "epoch": 1.595e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 7.509588201768609e-05,
      "loss": 3.0316,
      "step": 53190
    },
    {
      "epoch": 1.6e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 7.501764125911917e-05,
      "loss": 3.0634,
      "step": 53200
    },
    {
      "epoch": 1.605e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 7.493945484943403e-05,
      "loss": 3.0587,
      "step": 53210
    },
    {
      "epoch": 1.61e-06,
      "grad_norm": 0.46875,
      "learning_rate": 7.486132283585485e-05,
      "loss": 3.0265,
      "step": 53220
    },
    {
      "epoch": 1.615e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 7.478324526557295e-05,
      "loss": 3.0531,
      "step": 53230
    },
    {
      "epoch": 1.62e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 7.470522218574678e-05,
      "loss": 3.0433,
      "step": 53240
    },
    {
      "epoch": 1.625e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 7.462725364350193e-05,
      "loss": 3.0541,
      "step": 53250
    },
    {
      "epoch": 1.63e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 7.454933968593092e-05,
      "loss": 3.0492,
      "step": 53260
    },
    {
      "epoch": 1.635e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 7.447148036009339e-05,
      "loss": 3.0513,
      "step": 53270
    },
    {
      "epoch": 1.64e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 7.439367571301609e-05,
      "loss": 3.0375,
      "step": 53280
    },
    {
      "epoch": 1.645e-06,
      "grad_norm": 0.515625,
      "learning_rate": 7.431592579169245e-05,
      "loss": 3.0224,
      "step": 53290
    },
    {
      "epoch": 1.65e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 7.42382306430832e-05,
      "loss": 3.0463,
      "step": 53300
    },
    {
      "epoch": 1.655e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 7.416059031411562e-05,
      "loss": 3.0174,
      "step": 53310
    },
    {
      "epoch": 1.66e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 7.408300485168421e-05,
      "loss": 3.0391,
      "step": 53320
    },
    {
      "epoch": 1.665e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.400547430265015e-05,
      "loss": 3.0377,
      "step": 53330
    },
    {
      "epoch": 1.67e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 7.392799871384151e-05,
      "loss": 3.0276,
      "step": 53340
    },
    {
      "epoch": 1.675e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 7.385057813205311e-05,
      "loss": 3.0441,
      "step": 53350
    },
    {
      "epoch": 1.68e-06,
      "grad_norm": 0.46875,
      "learning_rate": 7.377321260404662e-05,
      "loss": 3.0333,
      "step": 53360
    },
    {
      "epoch": 1.685e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.369590217655036e-05,
      "loss": 3.0265,
      "step": 53370
    },
    {
      "epoch": 1.69e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 7.361864689625955e-05,
      "loss": 3.0515,
      "step": 53380
    },
    {
      "epoch": 1.695e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 7.354144680983583e-05,
      "loss": 3.04,
      "step": 53390
    },
    {
      "epoch": 1.7e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 7.346430196390775e-05,
      "loss": 3.0304,
      "step": 53400
    },
    {
      "epoch": 1.705e-06,
      "grad_norm": 0.484375,
      "learning_rate": 7.338721240507039e-05,
      "loss": 3.0461,
      "step": 53410
    },
    {
      "epoch": 1.71e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 7.331017817988545e-05,
      "loss": 3.0125,
      "step": 53420
    },
    {
      "epoch": 1.715e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 7.323319933488117e-05,
      "loss": 3.0283,
      "step": 53430
    },
    {
      "epoch": 1.72e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 7.315627591655244e-05,
      "loss": 3.0337,
      "step": 53440
    },
    {
      "epoch": 1.725e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 7.307940797136048e-05,
      "loss": 3.0678,
      "step": 53450
    },
    {
      "epoch": 1.73e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 7.300259554573334e-05,
      "loss": 3.0282,
      "step": 53460
    },
    {
      "epoch": 1.735e-06,
      "grad_norm": 0.46875,
      "learning_rate": 7.29258386860651e-05,
      "loss": 3.0362,
      "step": 53470
    },
    {
      "epoch": 1.74e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 7.284913743871668e-05,
      "loss": 3.0499,
      "step": 53480
    },
    {
      "epoch": 1.745e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 7.277249185001517e-05,
      "loss": 3.0561,
      "step": 53490
    },
    {
      "epoch": 1.75e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 7.269590196625413e-05,
      "loss": 3.0498,
      "step": 53500
    },
    {
      "epoch": 1.75e-06,
      "eval_loss": 3.098611354827881,
      "eval_runtime": 0.0757,
      "eval_samples_per_second": 422.565,
      "eval_steps_per_second": 13.205,
      "step": 53500
    },
    {
      "epoch": 1.755e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 7.261936783369345e-05,
      "loss": 3.0227,
      "step": 53510
    },
    {
      "epoch": 1.76e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 7.254288949855933e-05,
      "loss": 3.0587,
      "step": 53520
    },
    {
      "epoch": 1.765e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.246646700704433e-05,
      "loss": 3.0528,
      "step": 53530
    },
    {
      "epoch": 1.77e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 7.23901004053072e-05,
      "loss": 3.0421,
      "step": 53540
    },
    {
      "epoch": 1.775e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 7.231378973947297e-05,
      "loss": 3.0374,
      "step": 53550
    },
    {
      "epoch": 1.78e-06,
      "grad_norm": 0.546875,
      "learning_rate": 7.2237535055633e-05,
      "loss": 3.0521,
      "step": 53560
    },
    {
      "epoch": 1.785e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.216133639984455e-05,
      "loss": 3.0285,
      "step": 53570
    },
    {
      "epoch": 1.79e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 7.208519381813136e-05,
      "loss": 3.0473,
      "step": 53580
    },
    {
      "epoch": 1.795e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 7.200910735648309e-05,
      "loss": 3.0643,
      "step": 53590
    },
    {
      "epoch": 1.8e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 7.193307706085558e-05,
      "loss": 3.0224,
      "step": 53600
    },
    {
      "epoch": 1.805e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 7.185710297717076e-05,
      "loss": 3.0586,
      "step": 53610
    },
    {
      "epoch": 1.81e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 7.178118515131658e-05,
      "loss": 3.033,
      "step": 53620
    },
    {
      "epoch": 1.815e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.170532362914699e-05,
      "loss": 3.0426,
      "step": 53630
    },
    {
      "epoch": 1.82e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.1629518456482e-05,
      "loss": 3.0697,
      "step": 53640
    },
    {
      "epoch": 1.825e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 7.15537696791075e-05,
      "loss": 3.0462,
      "step": 53650
    },
    {
      "epoch": 1.83e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 7.147807734277534e-05,
      "loss": 3.027,
      "step": 53660
    },
    {
      "epoch": 1.835e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 7.140244149320342e-05,
      "loss": 3.0462,
      "step": 53670
    },
    {
      "epoch": 1.84e-06,
      "grad_norm": 0.5,
      "learning_rate": 7.132686217607526e-05,
      "loss": 3.0377,
      "step": 53680
    },
    {
      "epoch": 1.845e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 7.125133943704049e-05,
      "loss": 3.0398,
      "step": 53690
    },
    {
      "epoch": 1.85e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 7.117587332171439e-05,
      "loss": 3.0174,
      "step": 53700
    },
    {
      "epoch": 1.855e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.110046387567815e-05,
      "loss": 3.035,
      "step": 53710
    },
    {
      "epoch": 1.86e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 7.102511114447865e-05,
      "loss": 3.0212,
      "step": 53720
    },
    {
      "epoch": 1.865e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 7.094981517362858e-05,
      "loss": 3.0093,
      "step": 53730
    },
    {
      "epoch": 1.87e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 7.087457600860625e-05,
      "loss": 3.0118,
      "step": 53740
    },
    {
      "epoch": 1.875e-06,
      "grad_norm": 0.515625,
      "learning_rate": 7.079939369485589e-05,
      "loss": 3.0441,
      "step": 53750
    },
    {
      "epoch": 1.88e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 7.072426827778703e-05,
      "loss": 3.039,
      "step": 53760
    },
    {
      "epoch": 1.885e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 7.06491998027752e-05,
      "loss": 3.0517,
      "step": 53770
    },
    {
      "epoch": 1.89e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 7.057418831516123e-05,
      "loss": 3.0201,
      "step": 53780
    },
    {
      "epoch": 1.895e-06,
      "grad_norm": 0.515625,
      "learning_rate": 7.049923386025179e-05,
      "loss": 3.0581,
      "step": 53790
    },
    {
      "epoch": 1.9e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 7.042433648331894e-05,
      "loss": 3.0318,
      "step": 53800
    },
    {
      "epoch": 1.905e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 7.034949622960033e-05,
      "loss": 3.0377,
      "step": 53810
    },
    {
      "epoch": 1.91e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 7.027471314429905e-05,
      "loss": 3.0453,
      "step": 53820
    },
    {
      "epoch": 1.915e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 7.019998727258372e-05,
      "loss": 3.0372,
      "step": 53830
    },
    {
      "epoch": 1.92e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 7.012531865958834e-05,
      "loss": 3.0412,
      "step": 53840
    },
    {
      "epoch": 1.925e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 7.005070735041249e-05,
      "loss": 3.0232,
      "step": 53850
    },
    {
      "epoch": 1.93e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 6.997615339012086e-05,
      "loss": 3.0362,
      "step": 53860
    },
    {
      "epoch": 1.935e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 6.990165682374377e-05,
      "loss": 3.0121,
      "step": 53870
    },
    {
      "epoch": 1.94e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 6.982721769627673e-05,
      "loss": 3.0317,
      "step": 53880
    },
    {
      "epoch": 1.945e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 6.975283605268058e-05,
      "loss": 3.0382,
      "step": 53890
    },
    {
      "epoch": 1.95e-06,
      "grad_norm": 0.6484375,
      "learning_rate": 6.967851193788148e-05,
      "loss": 3.0309,
      "step": 53900
    },
    {
      "epoch": 1.955e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 6.960424539677075e-05,
      "loss": 3.0375,
      "step": 53910
    },
    {
      "epoch": 1.96e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 6.953003647420502e-05,
      "loss": 3.0521,
      "step": 53920
    },
    {
      "epoch": 1.965e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 6.945588521500623e-05,
      "loss": 3.0623,
      "step": 53930
    },
    {
      "epoch": 1.97e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 6.938179166396113e-05,
      "loss": 3.0164,
      "step": 53940
    },
    {
      "epoch": 1.975e-06,
      "grad_norm": 0.55078125,
      "learning_rate": 6.930775586582199e-05,
      "loss": 3.0519,
      "step": 53950
    },
    {
      "epoch": 1.98e-06,
      "grad_norm": 0.53125,
      "learning_rate": 6.923377786530605e-05,
      "loss": 3.0225,
      "step": 53960
    },
    {
      "epoch": 1.985e-06,
      "grad_norm": 0.5390625,
      "learning_rate": 6.91598577070956e-05,
      "loss": 3.0781,
      "step": 53970
    },
    {
      "epoch": 1.99e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 6.908599543583804e-05,
      "loss": 3.0287,
      "step": 53980
    },
    {
      "epoch": 1.995e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 6.901219109614582e-05,
      "loss": 3.0296,
      "step": 53990
    },
    {
      "epoch": 2e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 6.893844473259635e-05,
      "loss": 3.0228,
      "step": 54000
    },
    {
      "epoch": 2e-06,
      "eval_loss": 3.0967648029327393,
      "eval_runtime": 0.0604,
      "eval_samples_per_second": 530.111,
      "eval_steps_per_second": 16.566,
      "step": 54000
    },
    {
      "epoch": 2.005e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 6.886475638973211e-05,
      "loss": 3.0489,
      "step": 54010
    },
    {
      "epoch": 2.01e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 6.879112611206041e-05,
      "loss": 3.0541,
      "step": 54020
    },
    {
      "epoch": 2.015e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 6.87175539440536e-05,
      "loss": 3.0685,
      "step": 54030
    },
    {
      "epoch": 2.02e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 6.864403993014891e-05,
      "loss": 3.0675,
      "step": 54040
    },
    {
      "epoch": 2.025e-06,
      "grad_norm": 0.46875,
      "learning_rate": 6.857058411474834e-05,
      "loss": 3.0433,
      "step": 54050
    },
    {
      "epoch": 2.03e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 6.849718654221897e-05,
      "loss": 3.0465,
      "step": 54060
    },
    {
      "epoch": 2.035e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 6.842384725689239e-05,
      "loss": 3.0283,
      "step": 54070
    },
    {
      "epoch": 2.04e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 6.835056630306528e-05,
      "loss": 3.0432,
      "step": 54080
    },
    {
      "epoch": 2.045e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 6.82773437249989e-05,
      "loss": 3.0407,
      "step": 54090
    },
    {
      "epoch": 2.05e-06,
      "grad_norm": 0.53515625,
      "learning_rate": 6.820417956691934e-05,
      "loss": 3.0267,
      "step": 54100
    },
    {
      "epoch": 2.055e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 6.813107387301735e-05,
      "loss": 3.0482,
      "step": 54110
    },
    {
      "epoch": 2.06e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 6.80580266874484e-05,
      "loss": 3.0282,
      "step": 54120
    },
    {
      "epoch": 2.065e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 6.798503805433259e-05,
      "loss": 3.0147,
      "step": 54130
    },
    {
      "epoch": 2.07e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 6.791210801775476e-05,
      "loss": 3.0482,
      "step": 54140
    },
    {
      "epoch": 2.075e-06,
      "grad_norm": 0.484375,
      "learning_rate": 6.783923662176415e-05,
      "loss": 3.0196,
      "step": 54150
    },
    {
      "epoch": 2.08e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 6.776642391037481e-05,
      "loss": 3.0388,
      "step": 54160
    },
    {
      "epoch": 2.085e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 6.769366992756522e-05,
      "loss": 3.0555,
      "step": 54170
    },
    {
      "epoch": 2.09e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 6.76209747172784e-05,
      "loss": 3.0372,
      "step": 54180
    },
    {
      "epoch": 2.095e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 6.75483383234219e-05,
      "loss": 3.0579,
      "step": 54190
    },
    {
      "epoch": 2.1e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 6.747576078986771e-05,
      "loss": 3.0731,
      "step": 54200
    },
    {
      "epoch": 2.105e-06,
      "grad_norm": 0.5390625,
      "learning_rate": 6.740324216045229e-05,
      "loss": 3.0065,
      "step": 54210
    },
    {
      "epoch": 2.11e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 6.733078247897652e-05,
      "loss": 3.054,
      "step": 54220
    },
    {
      "epoch": 2.115e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 6.725838178920564e-05,
      "loss": 3.0552,
      "step": 54230
    },
    {
      "epoch": 2.12e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 6.71860401348694e-05,
      "loss": 3.0346,
      "step": 54240
    },
    {
      "epoch": 2.125e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 6.711375755966163e-05,
      "loss": 3.0433,
      "step": 54250
    },
    {
      "epoch": 2.13e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 6.704153410724079e-05,
      "loss": 3.0616,
      "step": 54260
    },
    {
      "epoch": 2.135e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 6.696936982122936e-05,
      "loss": 3.0663,
      "step": 54270
    },
    {
      "epoch": 2.14e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 6.689726474521428e-05,
      "loss": 3.0636,
      "step": 54280
    },
    {
      "epoch": 2.145e-06,
      "grad_norm": 0.5390625,
      "learning_rate": 6.682521892274659e-05,
      "loss": 3.0658,
      "step": 54290
    },
    {
      "epoch": 2.15e-06,
      "grad_norm": 0.515625,
      "learning_rate": 6.675323239734161e-05,
      "loss": 3.0156,
      "step": 54300
    },
    {
      "epoch": 2.155e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 6.66813052124788e-05,
      "loss": 3.0307,
      "step": 54310
    },
    {
      "epoch": 2.16e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 6.660943741160193e-05,
      "loss": 3.0451,
      "step": 54320
    },
    {
      "epoch": 2.165e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 6.653762903811861e-05,
      "loss": 3.0127,
      "step": 54330
    },
    {
      "epoch": 2.17e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 6.646588013540086e-05,
      "loss": 3.046,
      "step": 54340
    },
    {
      "epoch": 2.175e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 6.63941907467846e-05,
      "loss": 3.0223,
      "step": 54350
    },
    {
      "epoch": 2.18e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 6.632256091556984e-05,
      "loss": 3.0418,
      "step": 54360
    },
    {
      "epoch": 2.185e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 6.625099068502064e-05,
      "loss": 3.0318,
      "step": 54370
    },
    {
      "epoch": 2.19e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 6.617948009836508e-05,
      "loss": 3.0232,
      "step": 54380
    },
    {
      "epoch": 2.195e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 6.610802919879509e-05,
      "loss": 3.0383,
      "step": 54390
    },
    {
      "epoch": 2.2e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 6.603663802946681e-05,
      "loss": 3.05,
      "step": 54400
    },
    {
      "epoch": 2.205e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 6.596530663349999e-05,
      "loss": 3.0407,
      "step": 54410
    },
    {
      "epoch": 2.21e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 6.589403505397852e-05,
      "loss": 3.0343,
      "step": 54420
    },
    {
      "epoch": 2.215e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 6.582282333395e-05,
      "loss": 3.0227,
      "step": 54430
    },
    {
      "epoch": 2.22e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 6.575167151642595e-05,
      "loss": 3.0278,
      "step": 54440
    },
    {
      "epoch": 2.225e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 6.56805796443818e-05,
      "loss": 3.0517,
      "step": 54450
    },
    {
      "epoch": 2.23e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 6.560954776075651e-05,
      "loss": 3.0489,
      "step": 54460
    },
    {
      "epoch": 2.235e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 6.55385759084531e-05,
      "loss": 3.0179,
      "step": 54470
    },
    {
      "epoch": 2.24e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 6.546766413033817e-05,
      "loss": 3.0276,
      "step": 54480
    },
    {
      "epoch": 2.245e-06,
      "grad_norm": 0.484375,
      "learning_rate": 6.539681246924207e-05,
      "loss": 3.0384,
      "step": 54490
    },
    {
      "epoch": 2.25e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 6.532602096795882e-05,
      "loss": 3.0365,
      "step": 54500
    },
    {
      "epoch": 2.25e-06,
      "eval_loss": 3.095414161682129,
      "eval_runtime": 0.0627,
      "eval_samples_per_second": 510.473,
      "eval_steps_per_second": 15.952,
      "step": 54500
    },
    {
      "epoch": 2.255e-06,
      "grad_norm": 0.53125,
      "learning_rate": 6.525528966924612e-05,
      "loss": 3.0353,
      "step": 54510
    },
    {
      "epoch": 2.26e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 6.51846186158253e-05,
      "loss": 3.0247,
      "step": 54520
    },
    {
      "epoch": 2.265e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 6.511400785038144e-05,
      "loss": 3.0141,
      "step": 54530
    },
    {
      "epoch": 2.27e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 6.504345741556287e-05,
      "loss": 3.0336,
      "step": 54540
    },
    {
      "epoch": 2.275e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 6.497296735398184e-05,
      "loss": 3.0334,
      "step": 54550
    },
    {
      "epoch": 2.28e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 6.490253770821396e-05,
      "loss": 3.0185,
      "step": 54560
    },
    {
      "epoch": 2.285e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 6.483216852079834e-05,
      "loss": 3.0674,
      "step": 54570
    },
    {
      "epoch": 2.29e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 6.476185983423762e-05,
      "loss": 3.0396,
      "step": 54580
    },
    {
      "epoch": 2.295e-06,
      "grad_norm": 0.4375,
      "learning_rate": 6.46916116909979e-05,
      "loss": 3.0127,
      "step": 54590
    },
    {
      "epoch": 2.3e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 6.462142413350863e-05,
      "loss": 3.0699,
      "step": 54600
    },
    {
      "epoch": 2.305e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 6.455129720416286e-05,
      "loss": 3.0278,
      "step": 54610
    },
    {
      "epoch": 2.31e-06,
      "grad_norm": 0.484375,
      "learning_rate": 6.448123094531677e-05,
      "loss": 3.0478,
      "step": 54620
    },
    {
      "epoch": 2.315e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 6.441122539929014e-05,
      "loss": 3.0328,
      "step": 54630
    },
    {
      "epoch": 2.32e-06,
      "grad_norm": 0.46875,
      "learning_rate": 6.434128060836581e-05,
      "loss": 3.0378,
      "step": 54640
    },
    {
      "epoch": 2.325e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 6.42713966147902e-05,
      "loss": 3.0619,
      "step": 54650
    },
    {
      "epoch": 2.33e-06,
      "grad_norm": 0.5,
      "learning_rate": 6.420157346077282e-05,
      "loss": 3.0234,
      "step": 54660
    },
    {
      "epoch": 2.335e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 6.413181118848653e-05,
      "loss": 3.0304,
      "step": 54670
    },
    {
      "epoch": 2.34e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 6.406210984006738e-05,
      "loss": 3.0297,
      "step": 54680
    },
    {
      "epoch": 2.345e-06,
      "grad_norm": 0.453125,
      "learning_rate": 6.399246945761461e-05,
      "loss": 3.0217,
      "step": 54690
    },
    {
      "epoch": 2.35e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 6.39228900831906e-05,
      "loss": 3.0211,
      "step": 54700
    },
    {
      "epoch": 2.355e-06,
      "grad_norm": 0.5,
      "learning_rate": 6.385337175882107e-05,
      "loss": 3.0612,
      "step": 54710
    },
    {
      "epoch": 2.36e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 6.378391452649458e-05,
      "loss": 3.065,
      "step": 54720
    },
    {
      "epoch": 2.365e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 6.371451842816306e-05,
      "loss": 3.0446,
      "step": 54730
    },
    {
      "epoch": 2.37e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 6.364518350574134e-05,
      "loss": 3.0203,
      "step": 54740
    },
    {
      "epoch": 2.375e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 6.357590980110738e-05,
      "loss": 3.0139,
      "step": 54750
    },
    {
      "epoch": 2.38e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 6.350669735610212e-05,
      "loss": 3.0238,
      "step": 54760
    },
    {
      "epoch": 2.385e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 6.343754621252951e-05,
      "loss": 3.0593,
      "step": 54770
    },
    {
      "epoch": 2.39e-06,
      "grad_norm": 0.55859375,
      "learning_rate": 6.336845641215646e-05,
      "loss": 3.056,
      "step": 54780
    },
    {
      "epoch": 2.395e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 6.3299427996713e-05,
      "loss": 3.0556,
      "step": 54790
    },
    {
      "epoch": 2.4e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 6.323046100789174e-05,
      "loss": 3.0396,
      "step": 54800
    },
    {
      "epoch": 2.405e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 6.31615554873485e-05,
      "loss": 3.0291,
      "step": 54810
    },
    {
      "epoch": 2.41e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 6.309271147670189e-05,
      "loss": 3.0521,
      "step": 54820
    },
    {
      "epoch": 2.415e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 6.302392901753325e-05,
      "loss": 3.0434,
      "step": 54830
    },
    {
      "epoch": 2.42e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 6.295520815138691e-05,
      "loss": 3.0223,
      "step": 54840
    },
    {
      "epoch": 2.425e-06,
      "grad_norm": 0.5546875,
      "learning_rate": 6.288654891976982e-05,
      "loss": 3.0391,
      "step": 54850
    },
    {
      "epoch": 2.43e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 6.281795136415189e-05,
      "loss": 3.0312,
      "step": 54860
    },
    {
      "epoch": 2.435e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 6.274941552596566e-05,
      "loss": 3.0331,
      "step": 54870
    },
    {
      "epoch": 2.44e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 6.268094144660644e-05,
      "loss": 3.0288,
      "step": 54880
    },
    {
      "epoch": 2.445e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 6.26125291674322e-05,
      "loss": 3.0574,
      "step": 54890
    },
    {
      "epoch": 2.45e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 6.254417872976362e-05,
      "loss": 3.0114,
      "step": 54900
    },
    {
      "epoch": 2.455e-06,
      "grad_norm": 0.5390625,
      "learning_rate": 6.247589017488395e-05,
      "loss": 3.0439,
      "step": 54910
    },
    {
      "epoch": 2.46e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 6.240766354403925e-05,
      "loss": 3.038,
      "step": 54920
    },
    {
      "epoch": 2.465e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 6.233949887843794e-05,
      "loss": 3.0588,
      "step": 54930
    },
    {
      "epoch": 2.47e-06,
      "grad_norm": 0.5,
      "learning_rate": 6.22713962192512e-05,
      "loss": 3.0388,
      "step": 54940
    },
    {
      "epoch": 2.475e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 6.220335560761267e-05,
      "loss": 3.0408,
      "step": 54950
    },
    {
      "epoch": 2.48e-06,
      "grad_norm": 0.515625,
      "learning_rate": 6.21353770846185e-05,
      "loss": 3.0442,
      "step": 54960
    },
    {
      "epoch": 2.485e-06,
      "grad_norm": 0.484375,
      "learning_rate": 6.206746069132737e-05,
      "loss": 3.0322,
      "step": 54970
    },
    {
      "epoch": 2.49e-06,
      "grad_norm": 0.515625,
      "learning_rate": 6.199960646876048e-05,
      "loss": 3.0314,
      "step": 54980
    },
    {
      "epoch": 2.495e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 6.193181445790135e-05,
      "loss": 3.0443,
      "step": 54990
    },
    {
      "epoch": 2.5e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 6.186408469969614e-05,
      "loss": 3.0292,
      "step": 55000
    },
    {
      "epoch": 2.5e-06,
      "eval_loss": 3.0910887718200684,
      "eval_runtime": 0.0601,
      "eval_samples_per_second": 532.665,
      "eval_steps_per_second": 16.646,
      "step": 55000
    },
    {
      "epoch": 2.505e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 6.17964172350531e-05,
      "loss": 3.0389,
      "step": 55010
    },
    {
      "epoch": 2.51e-06,
      "grad_norm": 0.54296875,
      "learning_rate": 6.172881210484317e-05,
      "loss": 3.0451,
      "step": 55020
    },
    {
      "epoch": 2.515e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 6.166126934989944e-05,
      "loss": 3.017,
      "step": 55030
    },
    {
      "epoch": 2.52e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 6.159378901101741e-05,
      "loss": 3.0316,
      "step": 55040
    },
    {
      "epoch": 2.525e-06,
      "grad_norm": 0.515625,
      "learning_rate": 6.152637112895486e-05,
      "loss": 3.0433,
      "step": 55050
    },
    {
      "epoch": 2.53e-06,
      "grad_norm": 0.5859375,
      "learning_rate": 6.14590157444318e-05,
      "loss": 3.061,
      "step": 55060
    },
    {
      "epoch": 2.535e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 6.139172289813058e-05,
      "loss": 3.0261,
      "step": 55070
    },
    {
      "epoch": 2.54e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 6.132449263069571e-05,
      "loss": 3.0451,
      "step": 55080
    },
    {
      "epoch": 2.545e-06,
      "grad_norm": 0.453125,
      "learning_rate": 6.125732498273392e-05,
      "loss": 3.0331,
      "step": 55090
    },
    {
      "epoch": 2.55e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 6.11902199948142e-05,
      "loss": 3.0517,
      "step": 55100
    },
    {
      "epoch": 2.555e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 6.112317770746748e-05,
      "loss": 3.0527,
      "step": 55110
    },
    {
      "epoch": 2.56e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 6.105619816118706e-05,
      "loss": 3.0257,
      "step": 55120
    },
    {
      "epoch": 2.565e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 6.098928139642824e-05,
      "loss": 3.0497,
      "step": 55130
    },
    {
      "epoch": 2.57e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 6.0922427453608354e-05,
      "loss": 3.0282,
      "step": 55140
    },
    {
      "epoch": 2.575e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 6.085563637310687e-05,
      "loss": 3.0345,
      "step": 55150
    },
    {
      "epoch": 2.58e-06,
      "grad_norm": 0.5,
      "learning_rate": 6.078890819526526e-05,
      "loss": 3.0353,
      "step": 55160
    },
    {
      "epoch": 2.585e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 6.072224296038694e-05,
      "loss": 3.043,
      "step": 55170
    },
    {
      "epoch": 2.59e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 6.065564070873749e-05,
      "loss": 3.0395,
      "step": 55180
    },
    {
      "epoch": 2.595e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 6.0589101480544175e-05,
      "loss": 3.0346,
      "step": 55190
    },
    {
      "epoch": 2.6e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 6.0522625315996473e-05,
      "loss": 3.0296,
      "step": 55200
    },
    {
      "epoch": 2.605e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 6.0456212255245584e-05,
      "loss": 3.0591,
      "step": 55210
    },
    {
      "epoch": 2.61e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 6.03898623384047e-05,
      "loss": 3.0443,
      "step": 55220
    },
    {
      "epoch": 2.615e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 6.032357560554876e-05,
      "loss": 3.0232,
      "step": 55230
    },
    {
      "epoch": 2.62e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 6.025735209671468e-05,
      "loss": 3.0433,
      "step": 55240
    },
    {
      "epoch": 2.625e-06,
      "grad_norm": 0.546875,
      "learning_rate": 6.0191191851901065e-05,
      "loss": 3.0297,
      "step": 55250
    },
    {
      "epoch": 2.63e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 6.01250949110684e-05,
      "loss": 3.0405,
      "step": 55260
    },
    {
      "epoch": 2.635e-06,
      "grad_norm": 0.4375,
      "learning_rate": 6.0059061314138895e-05,
      "loss": 3.0623,
      "step": 55270
    },
    {
      "epoch": 2.64e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 5.999309110099643e-05,
      "loss": 3.0405,
      "step": 55280
    },
    {
      "epoch": 2.645e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 5.9927184311486806e-05,
      "loss": 3.0186,
      "step": 55290
    },
    {
      "epoch": 2.65e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 5.986134098541726e-05,
      "loss": 3.0294,
      "step": 55300
    },
    {
      "epoch": 2.655e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 5.9795561162556946e-05,
      "loss": 3.0448,
      "step": 55310
    },
    {
      "epoch": 2.66e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 5.9729844882636363e-05,
      "loss": 3.0305,
      "step": 55320
    },
    {
      "epoch": 2.665e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 5.9664192185347975e-05,
      "loss": 3.0102,
      "step": 55330
    },
    {
      "epoch": 2.67e-06,
      "grad_norm": 0.484375,
      "learning_rate": 5.959860311034561e-05,
      "loss": 3.0084,
      "step": 55340
    },
    {
      "epoch": 2.675e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 5.953307769724473e-05,
      "loss": 3.0436,
      "step": 55350
    },
    {
      "epoch": 2.68e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 5.946761598562233e-05,
      "loss": 3.0366,
      "step": 55360
    },
    {
      "epoch": 2.685e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 5.9402218015017e-05,
      "loss": 3.053,
      "step": 55370
    },
    {
      "epoch": 2.69e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 5.933688382492868e-05,
      "loss": 3.0248,
      "step": 55380
    },
    {
      "epoch": 2.695e-06,
      "grad_norm": 0.5,
      "learning_rate": 5.927161345481903e-05,
      "loss": 3.0176,
      "step": 55390
    },
    {
      "epoch": 2.7e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 5.920640694411085e-05,
      "loss": 3.0404,
      "step": 55400
    },
    {
      "epoch": 2.705e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 5.914126433218867e-05,
      "loss": 3.0305,
      "step": 55410
    },
    {
      "epoch": 2.71e-06,
      "grad_norm": 0.55078125,
      "learning_rate": 5.9076185658398264e-05,
      "loss": 3.0283,
      "step": 55420
    },
    {
      "epoch": 2.715e-06,
      "grad_norm": 0.46875,
      "learning_rate": 5.901117096204678e-05,
      "loss": 3.0162,
      "step": 55430
    },
    {
      "epoch": 2.72e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 5.894622028240277e-05,
      "loss": 3.0423,
      "step": 55440
    },
    {
      "epoch": 2.725e-06,
      "grad_norm": 0.46875,
      "learning_rate": 5.8881333658696125e-05,
      "loss": 3.0468,
      "step": 55450
    },
    {
      "epoch": 2.73e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 5.8816511130118e-05,
      "loss": 3.0161,
      "step": 55460
    },
    {
      "epoch": 2.735e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 5.875175273582095e-05,
      "loss": 3.0312,
      "step": 55470
    },
    {
      "epoch": 2.74e-06,
      "grad_norm": 0.484375,
      "learning_rate": 5.86870585149186e-05,
      "loss": 3.0566,
      "step": 55480
    },
    {
      "epoch": 2.745e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 5.8622428506486e-05,
      "loss": 3.0267,
      "step": 55490
    },
    {
      "epoch": 2.75e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 5.855786274955933e-05,
      "loss": 3.0238,
      "step": 55500
    },
    {
      "epoch": 2.75e-06,
      "eval_loss": 3.0897293090820312,
      "eval_runtime": 0.0628,
      "eval_samples_per_second": 509.512,
      "eval_steps_per_second": 15.922,
      "step": 55500
    },
    {
      "epoch": 2.755e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 5.849336128313597e-05,
      "loss": 3.0123,
      "step": 55510
    },
    {
      "epoch": 2.76e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 5.842892414617449e-05,
      "loss": 3.0462,
      "step": 55520
    },
    {
      "epoch": 2.765e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 5.836455137759457e-05,
      "loss": 3.0373,
      "step": 55530
    },
    {
      "epoch": 2.77e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 5.830024301627702e-05,
      "loss": 3.0388,
      "step": 55540
    },
    {
      "epoch": 2.775e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 5.823599910106379e-05,
      "loss": 3.0458,
      "step": 55550
    },
    {
      "epoch": 2.78e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 5.817181967075779e-05,
      "loss": 3.0684,
      "step": 55560
    },
    {
      "epoch": 2.785e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 5.8107704764123245e-05,
      "loss": 3.0172,
      "step": 55570
    },
    {
      "epoch": 2.79e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 5.8043654419885035e-05,
      "loss": 3.041,
      "step": 55580
    },
    {
      "epoch": 2.795e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 5.7979668676729346e-05,
      "loss": 3.0268,
      "step": 55590
    },
    {
      "epoch": 2.8e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 5.7915747573303205e-05,
      "loss": 3.059,
      "step": 55600
    },
    {
      "epoch": 2.805e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 5.785189114821466e-05,
      "loss": 3.0387,
      "step": 55610
    },
    {
      "epoch": 2.81e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 5.778809944003262e-05,
      "loss": 3.048,
      "step": 55620
    },
    {
      "epoch": 2.815e-06,
      "grad_norm": 0.453125,
      "learning_rate": 5.772437248728697e-05,
      "loss": 3.0239,
      "step": 55630
    },
    {
      "epoch": 2.82e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 5.7660710328468436e-05,
      "loss": 3.039,
      "step": 55640
    },
    {
      "epoch": 2.825e-06,
      "grad_norm": 0.46875,
      "learning_rate": 5.759711300202874e-05,
      "loss": 3.0157,
      "step": 55650
    },
    {
      "epoch": 2.83e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 5.753358054638021e-05,
      "loss": 3.0459,
      "step": 55660
    },
    {
      "epoch": 2.835e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 5.747011299989613e-05,
      "loss": 3.0517,
      "step": 55670
    },
    {
      "epoch": 2.84e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 5.74067104009107e-05,
      "loss": 3.0487,
      "step": 55680
    },
    {
      "epoch": 2.845e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 5.734337278771859e-05,
      "loss": 3.0544,
      "step": 55690
    },
    {
      "epoch": 2.85e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 5.728010019857552e-05,
      "loss": 3.0328,
      "step": 55700
    },
    {
      "epoch": 2.855e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 5.7216892671697773e-05,
      "loss": 3.034,
      "step": 55710
    },
    {
      "epoch": 2.86e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 5.7153750245262366e-05,
      "loss": 3.0147,
      "step": 55720
    },
    {
      "epoch": 2.865e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 5.7090672957407e-05,
      "loss": 3.0206,
      "step": 55730
    },
    {
      "epoch": 2.87e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 5.7027660846230014e-05,
      "loss": 3.0128,
      "step": 55740
    },
    {
      "epoch": 2.875e-06,
      "grad_norm": 0.53125,
      "learning_rate": 5.696471394979042e-05,
      "loss": 3.0288,
      "step": 55750
    },
    {
      "epoch": 2.88e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 5.690183230610782e-05,
      "loss": 3.0382,
      "step": 55760
    },
    {
      "epoch": 2.885e-06,
      "grad_norm": 0.46875,
      "learning_rate": 5.6839015953162364e-05,
      "loss": 3.0401,
      "step": 55770
    },
    {
      "epoch": 2.89e-06,
      "grad_norm": 0.453125,
      "learning_rate": 5.6776264928894916e-05,
      "loss": 3.047,
      "step": 55780
    },
    {
      "epoch": 2.895e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 5.6713579271206615e-05,
      "loss": 3.0568,
      "step": 55790
    },
    {
      "epoch": 2.9e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 5.6650959017959414e-05,
      "loss": 3.0512,
      "step": 55800
    },
    {
      "epoch": 2.905e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 5.6588404206975536e-05,
      "loss": 3.0493,
      "step": 55810
    },
    {
      "epoch": 2.91e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 5.652591487603783e-05,
      "loss": 3.028,
      "step": 55820
    },
    {
      "epoch": 2.915e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 5.6463491062889473e-05,
      "loss": 3.06,
      "step": 55830
    },
    {
      "epoch": 2.92e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 5.6401132805234154e-05,
      "loss": 3.0395,
      "step": 55840
    },
    {
      "epoch": 2.925e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 5.63388401407359e-05,
      "loss": 3.0233,
      "step": 55850
    },
    {
      "epoch": 2.93e-06,
      "grad_norm": 0.55859375,
      "learning_rate": 5.627661310701927e-05,
      "loss": 3.0621,
      "step": 55860
    },
    {
      "epoch": 2.935e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 5.621445174166893e-05,
      "loss": 3.0299,
      "step": 55870
    },
    {
      "epoch": 2.94e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 5.615235608223012e-05,
      "loss": 3.0212,
      "step": 55880
    },
    {
      "epoch": 2.945e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 5.609032616620827e-05,
      "loss": 3.0173,
      "step": 55890
    },
    {
      "epoch": 2.95e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 5.6028362031069145e-05,
      "loss": 3.0407,
      "step": 55900
    },
    {
      "epoch": 2.955e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 5.5966463714238746e-05,
      "loss": 3.0125,
      "step": 55910
    },
    {
      "epoch": 2.96e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 5.590463125310335e-05,
      "loss": 3.0375,
      "step": 55920
    },
    {
      "epoch": 2.965e-06,
      "grad_norm": 0.484375,
      "learning_rate": 5.58428646850094e-05,
      "loss": 3.0448,
      "step": 55930
    },
    {
      "epoch": 2.97e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 5.5781164047263695e-05,
      "loss": 3.0199,
      "step": 55940
    },
    {
      "epoch": 2.975e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 5.571952937713294e-05,
      "loss": 3.0354,
      "step": 55950
    },
    {
      "epoch": 2.98e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 5.565796071184433e-05,
      "loss": 3.0509,
      "step": 55960
    },
    {
      "epoch": 2.985e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 5.559645808858489e-05,
      "loss": 3.0148,
      "step": 55970
    },
    {
      "epoch": 2.99e-06,
      "grad_norm": 0.46875,
      "learning_rate": 5.5535021544501964e-05,
      "loss": 3.0353,
      "step": 55980
    },
    {
      "epoch": 2.995e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 5.547365111670289e-05,
      "loss": 3.0273,
      "step": 55990
    },
    {
      "epoch": 3e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 5.541234684225508e-05,
      "loss": 3.0392,
      "step": 56000
    },
    {
      "epoch": 3e-06,
      "eval_loss": 3.0882744789123535,
      "eval_runtime": 0.0616,
      "eval_samples_per_second": 519.813,
      "eval_steps_per_second": 16.244,
      "step": 56000
    },
    {
      "epoch": 3.005e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 5.535110875818601e-05,
      "loss": 3.0393,
      "step": 56010
    },
    {
      "epoch": 3.01e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 5.5289936901483165e-05,
      "loss": 3.0311,
      "step": 56020
    },
    {
      "epoch": 3.015e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 5.522883130909399e-05,
      "loss": 3.0387,
      "step": 56030
    },
    {
      "epoch": 3.02e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 5.516779201792605e-05,
      "loss": 3.0418,
      "step": 56040
    },
    {
      "epoch": 3.025e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 5.510681906484666e-05,
      "loss": 3.0303,
      "step": 56050
    },
    {
      "epoch": 3.03e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 5.504591248668322e-05,
      "loss": 3.0254,
      "step": 56060
    },
    {
      "epoch": 3.035e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 5.4985072320223035e-05,
      "loss": 3.0246,
      "step": 56070
    },
    {
      "epoch": 3.04e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 5.4924298602213124e-05,
      "loss": 3.0501,
      "step": 56080
    },
    {
      "epoch": 3.045e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 5.48635913693606e-05,
      "loss": 3.03,
      "step": 56090
    },
    {
      "epoch": 3.05e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 5.480295065833231e-05,
      "loss": 3.0406,
      "step": 56100
    },
    {
      "epoch": 3.055e-06,
      "grad_norm": 0.53125,
      "learning_rate": 5.4742376505754896e-05,
      "loss": 3.0288,
      "step": 56110
    },
    {
      "epoch": 3.06e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 5.4681868948214876e-05,
      "loss": 3.0415,
      "step": 56120
    },
    {
      "epoch": 3.065e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 5.4621428022258476e-05,
      "loss": 3.0557,
      "step": 56130
    },
    {
      "epoch": 3.07e-06,
      "grad_norm": 0.453125,
      "learning_rate": 5.456105376439167e-05,
      "loss": 3.049,
      "step": 56140
    },
    {
      "epoch": 3.075e-06,
      "grad_norm": 0.55078125,
      "learning_rate": 5.450074621108031e-05,
      "loss": 3.0445,
      "step": 56150
    },
    {
      "epoch": 3.08e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 5.444050539874971e-05,
      "loss": 3.0246,
      "step": 56160
    },
    {
      "epoch": 3.085e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 5.438033136378516e-05,
      "loss": 3.0429,
      "step": 56170
    },
    {
      "epoch": 3.09e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 5.43202241425313e-05,
      "loss": 3.0454,
      "step": 56180
    },
    {
      "epoch": 3.095e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 5.4260183771292704e-05,
      "loss": 3.0261,
      "step": 56190
    },
    {
      "epoch": 3.1e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 5.420021028633342e-05,
      "loss": 3.039,
      "step": 56200
    },
    {
      "epoch": 3.105e-06,
      "grad_norm": 0.453125,
      "learning_rate": 5.414030372387711e-05,
      "loss": 3.0048,
      "step": 56210
    },
    {
      "epoch": 3.11e-06,
      "grad_norm": 0.53125,
      "learning_rate": 5.408046412010703e-05,
      "loss": 3.0254,
      "step": 56220
    },
    {
      "epoch": 3.115e-06,
      "grad_norm": 0.5625,
      "learning_rate": 5.402069151116601e-05,
      "loss": 3.0646,
      "step": 56230
    },
    {
      "epoch": 3.12e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 5.396098593315635e-05,
      "loss": 3.0374,
      "step": 56240
    },
    {
      "epoch": 3.125e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 5.390134742214002e-05,
      "loss": 3.0636,
      "step": 56250
    },
    {
      "epoch": 3.13e-06,
      "grad_norm": 0.5,
      "learning_rate": 5.384177601413823e-05,
      "loss": 3.0126,
      "step": 56260
    },
    {
      "epoch": 3.135e-06,
      "grad_norm": 0.53125,
      "learning_rate": 5.378227174513191e-05,
      "loss": 3.022,
      "step": 56270
    },
    {
      "epoch": 3.14e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 5.3722834651061304e-05,
      "loss": 3.0478,
      "step": 56280
    },
    {
      "epoch": 3.145e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 5.366346476782611e-05,
      "loss": 3.0544,
      "step": 56290
    },
    {
      "epoch": 3.15e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 5.3604162131285404e-05,
      "loss": 3.0231,
      "step": 56300
    },
    {
      "epoch": 3.155e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 5.354492677725772e-05,
      "loss": 3.0486,
      "step": 56310
    },
    {
      "epoch": 3.16e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 5.3485758741520845e-05,
      "loss": 3.0163,
      "step": 56320
    },
    {
      "epoch": 3.165e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 5.342665805981207e-05,
      "loss": 3.029,
      "step": 56330
    },
    {
      "epoch": 3.17e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 5.3367624767827783e-05,
      "loss": 3.0407,
      "step": 56340
    },
    {
      "epoch": 3.175e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 5.3308658901223864e-05,
      "loss": 3.0329,
      "step": 56350
    },
    {
      "epoch": 3.18e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 5.3249760495615386e-05,
      "loss": 3.0116,
      "step": 56360
    },
    {
      "epoch": 3.185e-06,
      "grad_norm": 0.53125,
      "learning_rate": 5.319092958657668e-05,
      "loss": 3.0383,
      "step": 56370
    },
    {
      "epoch": 3.19e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 5.3132166209641284e-05,
      "loss": 3.0493,
      "step": 56380
    },
    {
      "epoch": 3.195e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 5.307347040030201e-05,
      "loss": 2.9958,
      "step": 56390
    },
    {
      "epoch": 3.2e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 5.301484219401081e-05,
      "loss": 3.0409,
      "step": 56400
    },
    {
      "epoch": 3.205e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 5.295628162617881e-05,
      "loss": 3.0473,
      "step": 56410
    },
    {
      "epoch": 3.21e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 5.289778873217628e-05,
      "loss": 3.0385,
      "step": 56420
    },
    {
      "epoch": 3.215e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 5.283936354733272e-05,
      "loss": 3.0433,
      "step": 56430
    },
    {
      "epoch": 3.22e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 5.2781006106936485e-05,
      "loss": 3.0615,
      "step": 56440
    },
    {
      "epoch": 3.225e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 5.2722716446235286e-05,
      "loss": 3.0456,
      "step": 56450
    },
    {
      "epoch": 3.23e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 5.26644946004358e-05,
      "loss": 3.0323,
      "step": 56460
    },
    {
      "epoch": 3.235e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 5.260634060470358e-05,
      "loss": 3.0368,
      "step": 56470
    },
    {
      "epoch": 3.24e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 5.254825449416348e-05,
      "loss": 3.0267,
      "step": 56480
    },
    {
      "epoch": 3.245e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 5.249023630389916e-05,
      "loss": 2.9978,
      "step": 56490
    },
    {
      "epoch": 3.25e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 5.243228606895329e-05,
      "loss": 3.0386,
      "step": 56500
    },
    {
      "epoch": 3.25e-06,
      "eval_loss": 3.087893486022949,
      "eval_runtime": 0.0618,
      "eval_samples_per_second": 517.741,
      "eval_steps_per_second": 16.179,
      "step": 56500
    },
    {
      "epoch": 3.255e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 5.2374403824327544e-05,
      "loss": 3.0251,
      "step": 56510
    },
    {
      "epoch": 3.26e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 5.231658960498247e-05,
      "loss": 3.0412,
      "step": 56520
    },
    {
      "epoch": 3.265e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 5.225884344583758e-05,
      "loss": 3.0285,
      "step": 56530
    },
    {
      "epoch": 3.27e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 5.220116538177129e-05,
      "loss": 3.0349,
      "step": 56540
    },
    {
      "epoch": 3.275e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 5.2143555447620765e-05,
      "loss": 3.0378,
      "step": 56550
    },
    {
      "epoch": 3.28e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 5.20860136781822e-05,
      "loss": 3.0472,
      "step": 56560
    },
    {
      "epoch": 3.285e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 5.2028540108210486e-05,
      "loss": 3.0287,
      "step": 56570
    },
    {
      "epoch": 3.29e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 5.197113477241935e-05,
      "loss": 3.0381,
      "step": 56580
    },
    {
      "epoch": 3.295e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 5.1913797705481364e-05,
      "loss": 3.0267,
      "step": 56590
    },
    {
      "epoch": 3.3e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 5.185652894202779e-05,
      "loss": 3.0344,
      "step": 56600
    },
    {
      "epoch": 3.305e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 5.179932851664866e-05,
      "loss": 3.0122,
      "step": 56610
    },
    {
      "epoch": 3.31e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 5.174219646389274e-05,
      "loss": 3.0292,
      "step": 56620
    },
    {
      "epoch": 3.315e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 5.16851328182675e-05,
      "loss": 3.0353,
      "step": 56630
    },
    {
      "epoch": 3.32e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 5.162813761423913e-05,
      "loss": 3.0516,
      "step": 56640
    },
    {
      "epoch": 3.325e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 5.157121088623235e-05,
      "loss": 3.0283,
      "step": 56650
    },
    {
      "epoch": 3.33e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 5.151435266863069e-05,
      "loss": 3.0489,
      "step": 56660
    },
    {
      "epoch": 3.335e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 5.145756299577619e-05,
      "loss": 3.0307,
      "step": 56670
    },
    {
      "epoch": 3.34e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 5.1400841901969535e-05,
      "loss": 3.0096,
      "step": 56680
    },
    {
      "epoch": 3.345e-06,
      "grad_norm": 0.515625,
      "learning_rate": 5.1344189421469966e-05,
      "loss": 3.0348,
      "step": 56690
    },
    {
      "epoch": 3.35e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 5.128760558849527e-05,
      "loss": 3.0201,
      "step": 56700
    },
    {
      "epoch": 3.355e-06,
      "grad_norm": 0.431640625,
      "learning_rate": 5.12310904372218e-05,
      "loss": 3.0465,
      "step": 56710
    },
    {
      "epoch": 3.36e-06,
      "grad_norm": 0.515625,
      "learning_rate": 5.1174644001784465e-05,
      "loss": 3.0555,
      "step": 56720
    },
    {
      "epoch": 3.365e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 5.111826631627653e-05,
      "loss": 3.0203,
      "step": 56730
    },
    {
      "epoch": 3.37e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 5.106195741474989e-05,
      "loss": 3.0128,
      "step": 56740
    },
    {
      "epoch": 3.375e-06,
      "grad_norm": 0.609375,
      "learning_rate": 5.100571733121484e-05,
      "loss": 3.0344,
      "step": 56750
    },
    {
      "epoch": 3.38e-06,
      "grad_norm": 0.46875,
      "learning_rate": 5.094954609964008e-05,
      "loss": 3.0237,
      "step": 56760
    },
    {
      "epoch": 3.385e-06,
      "grad_norm": 0.5703125,
      "learning_rate": 5.089344375395274e-05,
      "loss": 3.0247,
      "step": 56770
    },
    {
      "epoch": 3.39e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 5.0837410328038354e-05,
      "loss": 3.038,
      "step": 56780
    },
    {
      "epoch": 3.395e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 5.078144585574079e-05,
      "loss": 3.0339,
      "step": 56790
    },
    {
      "epoch": 3.4e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 5.07255503708624e-05,
      "loss": 3.0661,
      "step": 56800
    },
    {
      "epoch": 3.405e-06,
      "grad_norm": 0.5,
      "learning_rate": 5.066972390716365e-05,
      "loss": 3.0529,
      "step": 56810
    },
    {
      "epoch": 3.41e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 5.061396649836354e-05,
      "loss": 3.0566,
      "step": 56820
    },
    {
      "epoch": 3.415e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 5.055827817813915e-05,
      "loss": 3.0318,
      "step": 56830
    },
    {
      "epoch": 3.42e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 5.0502658980126033e-05,
      "loss": 3.0514,
      "step": 56840
    },
    {
      "epoch": 3.425e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 5.044710893791786e-05,
      "loss": 3.032,
      "step": 56850
    },
    {
      "epoch": 3.43e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 5.0391628085066574e-05,
      "loss": 3.0452,
      "step": 56860
    },
    {
      "epoch": 3.435e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 5.033621645508233e-05,
      "loss": 3.0373,
      "step": 56870
    },
    {
      "epoch": 3.44e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 5.0280874081433447e-05,
      "loss": 3.0321,
      "step": 56880
    },
    {
      "epoch": 3.445e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 5.0225600997546456e-05,
      "loss": 3.0581,
      "step": 56890
    },
    {
      "epoch": 3.45e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 5.017039723680601e-05,
      "loss": 3.0514,
      "step": 56900
    },
    {
      "epoch": 3.455e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 5.01152628325549e-05,
      "loss": 3.0138,
      "step": 56910
    },
    {
      "epoch": 3.46e-06,
      "grad_norm": 0.46875,
      "learning_rate": 5.0060197818093976e-05,
      "loss": 3.0428,
      "step": 56920
    },
    {
      "epoch": 3.465e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 5.000520222668233e-05,
      "loss": 3.0427,
      "step": 56930
    },
    {
      "epoch": 3.47e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 4.995027609153691e-05,
      "loss": 3.0222,
      "step": 56940
    },
    {
      "epoch": 3.475e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 4.989541944583286e-05,
      "loss": 3.0479,
      "step": 56950
    },
    {
      "epoch": 3.48e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 4.9840632322703316e-05,
      "loss": 3.0479,
      "step": 56960
    },
    {
      "epoch": 3.485e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 4.978591475523943e-05,
      "loss": 3.0424,
      "step": 56970
    },
    {
      "epoch": 3.49e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 4.9731266776490305e-05,
      "loss": 3.0347,
      "step": 56980
    },
    {
      "epoch": 3.495e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 4.967668841946305e-05,
      "loss": 3.0167,
      "step": 56990
    },
    {
      "epoch": 3.5e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 4.962217971712269e-05,
      "loss": 3.0236,
      "step": 57000
    },
    {
      "epoch": 3.5e-06,
      "eval_loss": 3.0873687267303467,
      "eval_runtime": 0.0616,
      "eval_samples_per_second": 519.109,
      "eval_steps_per_second": 16.222,
      "step": 57000
    },
    {
      "epoch": 3.505e-06,
      "grad_norm": 0.484375,
      "learning_rate": 4.956774070239227e-05,
      "loss": 3.0128,
      "step": 57010
    },
    {
      "epoch": 3.51e-06,
      "grad_norm": 0.453125,
      "learning_rate": 4.951337140815256e-05,
      "loss": 3.0061,
      "step": 57020
    },
    {
      "epoch": 3.515e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 4.945907186724246e-05,
      "loss": 3.0391,
      "step": 57030
    },
    {
      "epoch": 3.52e-06,
      "grad_norm": 0.484375,
      "learning_rate": 4.9404842112458525e-05,
      "loss": 3.0407,
      "step": 57040
    },
    {
      "epoch": 3.525e-06,
      "grad_norm": 0.53515625,
      "learning_rate": 4.93506821765553e-05,
      "loss": 3.0193,
      "step": 57050
    },
    {
      "epoch": 3.53e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 4.929659209224508e-05,
      "loss": 3.0432,
      "step": 57060
    },
    {
      "epoch": 3.535e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 4.924257189219799e-05,
      "loss": 2.9957,
      "step": 57070
    },
    {
      "epoch": 3.54e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 4.918862160904198e-05,
      "loss": 3.0288,
      "step": 57080
    },
    {
      "epoch": 3.545e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 4.9134741275362735e-05,
      "loss": 3.0417,
      "step": 57090
    },
    {
      "epoch": 3.55e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 4.908093092370364e-05,
      "loss": 3.0475,
      "step": 57100
    },
    {
      "epoch": 3.555e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 4.9027190586566e-05,
      "loss": 3.0488,
      "step": 57110
    },
    {
      "epoch": 3.56e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 4.897352029640856e-05,
      "loss": 3.056,
      "step": 57120
    },
    {
      "epoch": 3.565e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 4.8919920085648e-05,
      "loss": 3.0423,
      "step": 57130
    },
    {
      "epoch": 3.57e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 4.886638998665855e-05,
      "loss": 3.0458,
      "step": 57140
    },
    {
      "epoch": 3.575e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 4.8812930031772074e-05,
      "loss": 3.049,
      "step": 57150
    },
    {
      "epoch": 3.58e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 4.875954025327816e-05,
      "loss": 3.0045,
      "step": 57160
    },
    {
      "epoch": 3.585e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 4.870622068342391e-05,
      "loss": 3.0388,
      "step": 57170
    },
    {
      "epoch": 3.59e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 4.8652971354414086e-05,
      "loss": 3.0488,
      "step": 57180
    },
    {
      "epoch": 3.595e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 4.8599792298411076e-05,
      "loss": 3.0424,
      "step": 57190
    },
    {
      "epoch": 3.6e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 4.8546683547534654e-05,
      "loss": 3.0623,
      "step": 57200
    },
    {
      "epoch": 3.605e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 4.8493645133862284e-05,
      "loss": 3.0353,
      "step": 57210
    },
    {
      "epoch": 3.61e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 4.84406770894289e-05,
      "loss": 3.0128,
      "step": 57220
    },
    {
      "epoch": 3.615e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 4.838777944622691e-05,
      "loss": 3.0337,
      "step": 57230
    },
    {
      "epoch": 3.62e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 4.833495223620624e-05,
      "loss": 3.0205,
      "step": 57240
    },
    {
      "epoch": 3.625e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 4.828219549127422e-05,
      "loss": 3.0281,
      "step": 57250
    },
    {
      "epoch": 3.63e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 4.822950924329565e-05,
      "loss": 3.0471,
      "step": 57260
    },
    {
      "epoch": 3.635e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 4.8176893524092844e-05,
      "loss": 3.0169,
      "step": 57270
    },
    {
      "epoch": 3.64e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 4.8124348365445325e-05,
      "loss": 3.0423,
      "step": 57280
    },
    {
      "epoch": 3.645e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 4.807187379909011e-05,
      "loss": 3.0256,
      "step": 57290
    },
    {
      "epoch": 3.65e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 4.801946985672161e-05,
      "loss": 3.0222,
      "step": 57300
    },
    {
      "epoch": 3.655e-06,
      "grad_norm": 0.484375,
      "learning_rate": 4.796713656999146e-05,
      "loss": 3.0559,
      "step": 57310
    },
    {
      "epoch": 3.66e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 4.791487397050882e-05,
      "loss": 3.0485,
      "step": 57320
    },
    {
      "epoch": 3.665e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 4.7862682089839895e-05,
      "loss": 3.0434,
      "step": 57330
    },
    {
      "epoch": 3.67e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 4.781056095950841e-05,
      "loss": 3.0444,
      "step": 57340
    },
    {
      "epoch": 3.675e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 4.7758510610995244e-05,
      "loss": 3.0176,
      "step": 57350
    },
    {
      "epoch": 3.68e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 4.770653107573851e-05,
      "loss": 3.0333,
      "step": 57360
    },
    {
      "epoch": 3.685e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 4.765462238513358e-05,
      "loss": 3.0264,
      "step": 57370
    },
    {
      "epoch": 3.69e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 4.760278457053306e-05,
      "loss": 3.0358,
      "step": 57380
    },
    {
      "epoch": 3.695e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 4.7551017663246685e-05,
      "loss": 3.0042,
      "step": 57390
    },
    {
      "epoch": 3.7e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 4.7499321694541475e-05,
      "loss": 3.0172,
      "step": 57400
    },
    {
      "epoch": 3.705e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 4.744769669564144e-05,
      "loss": 3.0453,
      "step": 57410
    },
    {
      "epoch": 3.71e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 4.739614269772787e-05,
      "loss": 3.0138,
      "step": 57420
    },
    {
      "epoch": 3.715e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 4.7344659731939115e-05,
      "loss": 3.0527,
      "step": 57430
    },
    {
      "epoch": 3.72e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 4.729324782937059e-05,
      "loss": 3.0299,
      "step": 57440
    },
    {
      "epoch": 3.725e-06,
      "grad_norm": 0.5390625,
      "learning_rate": 4.724190702107485e-05,
      "loss": 3.0201,
      "step": 57450
    },
    {
      "epoch": 3.73e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 4.7190637338061464e-05,
      "loss": 3.039,
      "step": 57460
    },
    {
      "epoch": 3.735e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 4.713943881129701e-05,
      "loss": 3.0292,
      "step": 57470
    },
    {
      "epoch": 3.74e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 4.708831147170527e-05,
      "loss": 3.0647,
      "step": 57480
    },
    {
      "epoch": 3.745e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 4.7037255350166735e-05,
      "loss": 3.0109,
      "step": 57490
    },
    {
      "epoch": 3.75e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 4.6986270477519167e-05,
      "loss": 3.0336,
      "step": 57500
    },
    {
      "epoch": 3.75e-06,
      "eval_loss": 3.0866641998291016,
      "eval_runtime": 0.0672,
      "eval_samples_per_second": 476.356,
      "eval_steps_per_second": 14.886,
      "step": 57500
    },
    {
      "epoch": 3.755e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 4.693535688455706e-05,
      "loss": 3.0596,
      "step": 57510
    },
    {
      "epoch": 3.76e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 4.688451460203204e-05,
      "loss": 3.0226,
      "step": 57520
    },
    {
      "epoch": 3.765e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 4.683374366065256e-05,
      "loss": 3.0367,
      "step": 57530
    },
    {
      "epoch": 3.77e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 4.678304409108399e-05,
      "loss": 3.0649,
      "step": 57540
    },
    {
      "epoch": 3.775e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 4.673241592394864e-05,
      "loss": 3.0397,
      "step": 57550
    },
    {
      "epoch": 3.78e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 4.6681859189825646e-05,
      "loss": 3.0128,
      "step": 57560
    },
    {
      "epoch": 3.785e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 4.663137391925099e-05,
      "loss": 3.048,
      "step": 57570
    },
    {
      "epoch": 3.79e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 4.65809601427176e-05,
      "loss": 3.0438,
      "step": 57580
    },
    {
      "epoch": 3.795e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 4.653061789067502e-05,
      "loss": 3.0507,
      "step": 57590
    },
    {
      "epoch": 3.8e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 4.64803471935298e-05,
      "loss": 3.0454,
      "step": 57600
    },
    {
      "epoch": 3.805e-06,
      "grad_norm": 0.5,
      "learning_rate": 4.643014808164516e-05,
      "loss": 3.0269,
      "step": 57610
    },
    {
      "epoch": 3.81e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 4.638002058534108e-05,
      "loss": 3.0587,
      "step": 57620
    },
    {
      "epoch": 3.815e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 4.632996473489434e-05,
      "loss": 3.0413,
      "step": 57630
    },
    {
      "epoch": 3.82e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 4.627998056053841e-05,
      "loss": 3.0521,
      "step": 57640
    },
    {
      "epoch": 3.825e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 4.6230068092463427e-05,
      "loss": 3.0244,
      "step": 57650
    },
    {
      "epoch": 3.83e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 4.6180227360816373e-05,
      "loss": 3.0377,
      "step": 57660
    },
    {
      "epoch": 3.835e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 4.613045839570068e-05,
      "loss": 3.031,
      "step": 57670
    },
    {
      "epoch": 3.84e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 4.60807612271766e-05,
      "loss": 3.0441,
      "step": 57680
    },
    {
      "epoch": 3.845e-06,
      "grad_norm": 0.484375,
      "learning_rate": 4.6031135885261004e-05,
      "loss": 3.0159,
      "step": 57690
    },
    {
      "epoch": 3.85e-06,
      "grad_norm": 0.484375,
      "learning_rate": 4.5981582399927254e-05,
      "loss": 3.0318,
      "step": 57700
    },
    {
      "epoch": 3.855e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 4.593210080110549e-05,
      "loss": 3.0609,
      "step": 57710
    },
    {
      "epoch": 3.86e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 4.588269111868224e-05,
      "loss": 3.0359,
      "step": 57720
    },
    {
      "epoch": 3.865e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 4.5833353382500786e-05,
      "loss": 3.0017,
      "step": 57730
    },
    {
      "epoch": 3.87e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 4.5784087622360845e-05,
      "loss": 3.0226,
      "step": 57740
    },
    {
      "epoch": 3.875e-06,
      "grad_norm": 0.484375,
      "learning_rate": 4.5734893868018665e-05,
      "loss": 3.051,
      "step": 57750
    },
    {
      "epoch": 3.88e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 4.568577214918706e-05,
      "loss": 3.0253,
      "step": 57760
    },
    {
      "epoch": 3.885e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 4.563672249553525e-05,
      "loss": 3.0389,
      "step": 57770
    },
    {
      "epoch": 3.89e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 4.558774493668896e-05,
      "loss": 3.0268,
      "step": 57780
    },
    {
      "epoch": 3.895e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 4.553883950223047e-05,
      "loss": 3.0334,
      "step": 57790
    },
    {
      "epoch": 3.9e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 4.549000622169831e-05,
      "loss": 3.0053,
      "step": 57800
    },
    {
      "epoch": 3.905e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 4.5441245124587594e-05,
      "loss": 3.0351,
      "step": 57810
    },
    {
      "epoch": 3.91e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 4.539255624034977e-05,
      "loss": 3.0549,
      "step": 57820
    },
    {
      "epoch": 3.915e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 4.5343939598392644e-05,
      "loss": 3.0443,
      "step": 57830
    },
    {
      "epoch": 3.92e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 4.5295395228080426e-05,
      "loss": 3.0512,
      "step": 57840
    },
    {
      "epoch": 3.925e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 4.5246923158733695e-05,
      "loss": 3.0158,
      "step": 57850
    },
    {
      "epoch": 3.93e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 4.5198523419629264e-05,
      "loss": 3.0327,
      "step": 57860
    },
    {
      "epoch": 3.935e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 4.5150196040000436e-05,
      "loss": 3.0406,
      "step": 57870
    },
    {
      "epoch": 3.94e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 4.5101941049036575e-05,
      "loss": 3.0524,
      "step": 57880
    },
    {
      "epoch": 3.945e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 4.505375847588353e-05,
      "loss": 3.0266,
      "step": 57890
    },
    {
      "epoch": 3.95e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 4.500564834964331e-05,
      "loss": 3.0524,
      "step": 57900
    },
    {
      "epoch": 3.955e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 4.495761069937419e-05,
      "loss": 3.0356,
      "step": 57910
    },
    {
      "epoch": 3.96e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 4.490964555409064e-05,
      "loss": 3.038,
      "step": 57920
    },
    {
      "epoch": 3.965e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 4.486175294276337e-05,
      "loss": 3.0396,
      "step": 57930
    },
    {
      "epoch": 3.97e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 4.481393289431929e-05,
      "loss": 3.0425,
      "step": 57940
    },
    {
      "epoch": 3.975e-06,
      "grad_norm": 0.46875,
      "learning_rate": 4.4766185437641437e-05,
      "loss": 3.0307,
      "step": 57950
    },
    {
      "epoch": 3.98e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 4.4718510601569016e-05,
      "loss": 3.023,
      "step": 57960
    },
    {
      "epoch": 3.985e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 4.4670908414897454e-05,
      "loss": 3.0441,
      "step": 57970
    },
    {
      "epoch": 3.99e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 4.462337890637812e-05,
      "loss": 3.0096,
      "step": 57980
    },
    {
      "epoch": 3.995e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 4.457592210471869e-05,
      "loss": 3.0315,
      "step": 57990
    },
    {
      "epoch": 4e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 4.4528538038582794e-05,
      "loss": 3.0336,
      "step": 58000
    },
    {
      "epoch": 4e-06,
      "eval_loss": 3.087162971496582,
      "eval_runtime": 0.0617,
      "eval_samples_per_second": 518.477,
      "eval_steps_per_second": 16.202,
      "step": 58000
    },
    {
      "epoch": 4.005e-06,
      "grad_norm": 0.46875,
      "learning_rate": 4.448122673659015e-05,
      "loss": 3.0309,
      "step": 58010
    },
    {
      "epoch": 4.01e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 4.443398822731655e-05,
      "loss": 3.0302,
      "step": 58020
    },
    {
      "epoch": 4.015e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 4.438682253929382e-05,
      "loss": 3.0298,
      "step": 58030
    },
    {
      "epoch": 4.02e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 4.433972970100975e-05,
      "loss": 3.032,
      "step": 58040
    },
    {
      "epoch": 4.025e-06,
      "grad_norm": 0.46875,
      "learning_rate": 4.429270974090827e-05,
      "loss": 3.0376,
      "step": 58050
    },
    {
      "epoch": 4.03e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 4.4245762687389075e-05,
      "loss": 3.0736,
      "step": 58060
    },
    {
      "epoch": 4.035e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 4.4198888568808017e-05,
      "loss": 3.0349,
      "step": 58070
    },
    {
      "epoch": 4.04e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 4.41520874134768e-05,
      "loss": 3.0325,
      "step": 58080
    },
    {
      "epoch": 4.045e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 4.4105359249663104e-05,
      "loss": 3.0232,
      "step": 58090
    },
    {
      "epoch": 4.05e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 4.405870410559047e-05,
      "loss": 3.0316,
      "step": 58100
    },
    {
      "epoch": 4.055e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 4.4012122009438385e-05,
      "loss": 3.0552,
      "step": 58110
    },
    {
      "epoch": 4.06e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 4.396561298934218e-05,
      "loss": 3.0167,
      "step": 58120
    },
    {
      "epoch": 4.065e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 4.391917707339306e-05,
      "loss": 3.0116,
      "step": 58130
    },
    {
      "epoch": 4.07e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 4.387281428963811e-05,
      "loss": 3.05,
      "step": 58140
    },
    {
      "epoch": 4.075e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 4.38265246660802e-05,
      "loss": 3.029,
      "step": 58150
    },
    {
      "epoch": 4.08e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 4.378030823067799e-05,
      "loss": 3.0376,
      "step": 58160
    },
    {
      "epoch": 4.085e-06,
      "grad_norm": 0.453125,
      "learning_rate": 4.373416501134598e-05,
      "loss": 3.0072,
      "step": 58170
    },
    {
      "epoch": 4.09e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 4.368809503595452e-05,
      "loss": 3.0229,
      "step": 58180
    },
    {
      "epoch": 4.095e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 4.36420983323295e-05,
      "loss": 3.0378,
      "step": 58190
    },
    {
      "epoch": 4.1e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 4.3596174928252805e-05,
      "loss": 3.0396,
      "step": 58200
    },
    {
      "epoch": 4.105e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 4.35503248514619e-05,
      "loss": 3.0374,
      "step": 58210
    },
    {
      "epoch": 4.11e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 4.350454812964998e-05,
      "loss": 3.0261,
      "step": 58220
    },
    {
      "epoch": 4.115e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 4.345884479046595e-05,
      "loss": 3.0536,
      "step": 58230
    },
    {
      "epoch": 4.12e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 4.34132148615144e-05,
      "loss": 3.0226,
      "step": 58240
    },
    {
      "epoch": 4.125e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 4.336765837035556e-05,
      "loss": 3.0173,
      "step": 58250
    },
    {
      "epoch": 4.13e-06,
      "grad_norm": 0.515625,
      "learning_rate": 4.3322175344505384e-05,
      "loss": 3.0156,
      "step": 58260
    },
    {
      "epoch": 4.135e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 4.327676581143527e-05,
      "loss": 3.0516,
      "step": 58270
    },
    {
      "epoch": 4.14e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 4.323142979857244e-05,
      "loss": 3.0318,
      "step": 58280
    },
    {
      "epoch": 4.145e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 4.318616733329957e-05,
      "loss": 3.0311,
      "step": 58290
    },
    {
      "epoch": 4.15e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 4.314097844295496e-05,
      "loss": 3.0353,
      "step": 58300
    },
    {
      "epoch": 4.155e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 4.309586315483245e-05,
      "loss": 3.034,
      "step": 58310
    },
    {
      "epoch": 4.16e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 4.305082149618149e-05,
      "loss": 3.0549,
      "step": 58320
    },
    {
      "epoch": 4.165e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 4.3005853494206956e-05,
      "loss": 3.0392,
      "step": 58330
    },
    {
      "epoch": 4.17e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 4.296095917606936e-05,
      "loss": 3.0359,
      "step": 58340
    },
    {
      "epoch": 4.175e-06,
      "grad_norm": 0.5,
      "learning_rate": 4.291613856888454e-05,
      "loss": 3.0215,
      "step": 58350
    },
    {
      "epoch": 4.18e-06,
      "grad_norm": 0.46875,
      "learning_rate": 4.287139169972402e-05,
      "loss": 3.0506,
      "step": 58360
    },
    {
      "epoch": 4.185e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 4.282671859561457e-05,
      "loss": 3.0381,
      "step": 58370
    },
    {
      "epoch": 4.19e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 4.27821192835386e-05,
      "loss": 3.036,
      "step": 58380
    },
    {
      "epoch": 4.195e-06,
      "grad_norm": 0.5,
      "learning_rate": 4.273759379043383e-05,
      "loss": 3.0144,
      "step": 58390
    },
    {
      "epoch": 4.2e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 4.269314214319344e-05,
      "loss": 3.0005,
      "step": 58400
    },
    {
      "epoch": 4.205e-06,
      "grad_norm": 0.46875,
      "learning_rate": 4.264876436866599e-05,
      "loss": 3.0421,
      "step": 58410
    },
    {
      "epoch": 4.21e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 4.260446049365543e-05,
      "loss": 3.0562,
      "step": 58420
    },
    {
      "epoch": 4.215e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 4.256023054492105e-05,
      "loss": 3.0338,
      "step": 58430
    },
    {
      "epoch": 4.22e-06,
      "grad_norm": 0.46875,
      "learning_rate": 4.251607454917758e-05,
      "loss": 3.0245,
      "step": 58440
    },
    {
      "epoch": 4.225e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 4.247199253309491e-05,
      "loss": 3.0094,
      "step": 58450
    },
    {
      "epoch": 4.23e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 4.2427984523298436e-05,
      "loss": 3.0437,
      "step": 58460
    },
    {
      "epoch": 4.235e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 4.238405054636875e-05,
      "loss": 3.0348,
      "step": 58470
    },
    {
      "epoch": 4.24e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 4.234019062884172e-05,
      "loss": 3.0336,
      "step": 58480
    },
    {
      "epoch": 4.245e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 4.2296404797208546e-05,
      "loss": 3.0658,
      "step": 58490
    },
    {
      "epoch": 4.25e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 4.225269307791562e-05,
      "loss": 3.063,
      "step": 58500
    },
    {
      "epoch": 4.25e-06,
      "eval_loss": 3.0869736671447754,
      "eval_runtime": 0.063,
      "eval_samples_per_second": 508.325,
      "eval_steps_per_second": 15.885,
      "step": 58500
    },
    {
      "epoch": 4.255e-06,
      "grad_norm": 0.6484375,
      "learning_rate": 4.2209055497364606e-05,
      "loss": 3.0559,
      "step": 58510
    },
    {
      "epoch": 4.26e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 4.2165492081912375e-05,
      "loss": 3.034,
      "step": 58520
    },
    {
      "epoch": 4.265e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 4.212200285787102e-05,
      "loss": 3.0549,
      "step": 58530
    },
    {
      "epoch": 4.27e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 4.207858785150775e-05,
      "loss": 3.0166,
      "step": 58540
    },
    {
      "epoch": 4.275e-06,
      "grad_norm": 0.46875,
      "learning_rate": 4.203524708904512e-05,
      "loss": 3.0667,
      "step": 58550
    },
    {
      "epoch": 4.28e-06,
      "grad_norm": 0.484375,
      "learning_rate": 4.1991980596660585e-05,
      "loss": 3.0136,
      "step": 58560
    },
    {
      "epoch": 4.285e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 4.1948788400487004e-05,
      "loss": 3.062,
      "step": 58570
    },
    {
      "epoch": 4.29e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 4.1905670526612186e-05,
      "loss": 3.0321,
      "step": 58580
    },
    {
      "epoch": 4.295e-06,
      "grad_norm": 0.5,
      "learning_rate": 4.186262700107912e-05,
      "loss": 3.0287,
      "step": 58590
    },
    {
      "epoch": 4.3e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 4.181965784988588e-05,
      "loss": 3.0229,
      "step": 58600
    },
    {
      "epoch": 4.305e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 4.177676309898558e-05,
      "loss": 3.0494,
      "step": 58610
    },
    {
      "epoch": 4.31e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 4.173394277428645e-05,
      "loss": 3.029,
      "step": 58620
    },
    {
      "epoch": 4.315e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 4.169119690165176e-05,
      "loss": 3.0398,
      "step": 58630
    },
    {
      "epoch": 4.32e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 4.164852550689974e-05,
      "loss": 3.0387,
      "step": 58640
    },
    {
      "epoch": 4.325e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 4.1605928615803785e-05,
      "loss": 3.0461,
      "step": 58650
    },
    {
      "epoch": 4.33e-06,
      "grad_norm": 0.46875,
      "learning_rate": 4.15634062540921e-05,
      "loss": 3.0283,
      "step": 58660
    },
    {
      "epoch": 4.335e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 4.152095844744806e-05,
      "loss": 3.0601,
      "step": 58670
    },
    {
      "epoch": 4.34e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 4.1478585221509875e-05,
      "loss": 3.0296,
      "step": 58680
    },
    {
      "epoch": 4.345e-06,
      "grad_norm": 0.5,
      "learning_rate": 4.143628660187076e-05,
      "loss": 3.037,
      "step": 58690
    },
    {
      "epoch": 4.35e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 4.139406261407886e-05,
      "loss": 3.03,
      "step": 58700
    },
    {
      "epoch": 4.355e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 4.135191328363726e-05,
      "loss": 3.0511,
      "step": 58710
    },
    {
      "epoch": 4.36e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 4.130983863600391e-05,
      "loss": 3.041,
      "step": 58720
    },
    {
      "epoch": 4.365e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 4.126783869659175e-05,
      "loss": 3.0145,
      "step": 58730
    },
    {
      "epoch": 4.37e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 4.122591349076844e-05,
      "loss": 3.0271,
      "step": 58740
    },
    {
      "epoch": 4.375e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 4.118406304385664e-05,
      "loss": 3.0453,
      "step": 58750
    },
    {
      "epoch": 4.38e-06,
      "grad_norm": 0.453125,
      "learning_rate": 4.114228738113379e-05,
      "loss": 3.0487,
      "step": 58760
    },
    {
      "epoch": 4.385e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 4.1100586527832185e-05,
      "loss": 3.0391,
      "step": 58770
    },
    {
      "epoch": 4.39e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 4.105896050913892e-05,
      "loss": 3.02,
      "step": 58780
    },
    {
      "epoch": 4.395e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 4.1017409350195895e-05,
      "loss": 3.0306,
      "step": 58790
    },
    {
      "epoch": 4.4e-06,
      "grad_norm": 0.515625,
      "learning_rate": 4.0975933076099777e-05,
      "loss": 3.0156,
      "step": 58800
    },
    {
      "epoch": 4.405e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 4.0934531711902085e-05,
      "loss": 3.0304,
      "step": 58810
    },
    {
      "epoch": 4.41e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 4.0893205282608955e-05,
      "loss": 3.0441,
      "step": 58820
    },
    {
      "epoch": 4.415e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 4.0851953813181404e-05,
      "loss": 3.0382,
      "step": 58830
    },
    {
      "epoch": 4.42e-06,
      "grad_norm": 0.5,
      "learning_rate": 4.081077732853501e-05,
      "loss": 3.0369,
      "step": 58840
    },
    {
      "epoch": 4.425e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 4.076967585354027e-05,
      "loss": 3.0373,
      "step": 58850
    },
    {
      "epoch": 4.43e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 4.07286494130222e-05,
      "loss": 3.027,
      "step": 58860
    },
    {
      "epoch": 4.435e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 4.068769803176058e-05,
      "loss": 3.0699,
      "step": 58870
    },
    {
      "epoch": 4.44e-06,
      "grad_norm": 0.46875,
      "learning_rate": 4.064682173448984e-05,
      "loss": 3.027,
      "step": 58880
    },
    {
      "epoch": 4.445e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 4.060602054589901e-05,
      "loss": 3.0186,
      "step": 58890
    },
    {
      "epoch": 4.45e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 4.0565294490631834e-05,
      "loss": 3.0149,
      "step": 58900
    },
    {
      "epoch": 4.455e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 4.0524643593286656e-05,
      "loss": 3.0356,
      "step": 58910
    },
    {
      "epoch": 4.46e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 4.048406787841636e-05,
      "loss": 3.0103,
      "step": 58920
    },
    {
      "epoch": 4.465e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 4.04435673705285e-05,
      "loss": 3.0627,
      "step": 58930
    },
    {
      "epoch": 4.47e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 4.040314209408521e-05,
      "loss": 3.0325,
      "step": 58940
    },
    {
      "epoch": 4.475e-06,
      "grad_norm": 0.484375,
      "learning_rate": 4.036279207350306e-05,
      "loss": 3.0455,
      "step": 58950
    },
    {
      "epoch": 4.48e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 4.0322517333153324e-05,
      "loss": 3.0258,
      "step": 58960
    },
    {
      "epoch": 4.485e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 4.028231789736172e-05,
      "loss": 3.028,
      "step": 58970
    },
    {
      "epoch": 4.49e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 4.0242193790408496e-05,
      "loss": 3.0158,
      "step": 58980
    },
    {
      "epoch": 4.495e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 4.020214503652841e-05,
      "loss": 3.0313,
      "step": 58990
    },
    {
      "epoch": 4.5e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 4.0162171659910705e-05,
      "loss": 3.0079,
      "step": 59000
    },
    {
      "epoch": 4.5e-06,
      "eval_loss": 3.0862205028533936,
      "eval_runtime": 0.0608,
      "eval_samples_per_second": 525.995,
      "eval_steps_per_second": 16.437,
      "step": 59000
    },
    {
      "epoch": 4.505e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 4.0122273684699076e-05,
      "loss": 3.0127,
      "step": 59010
    },
    {
      "epoch": 4.51e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 4.008245113499175e-05,
      "loss": 3.0293,
      "step": 59020
    },
    {
      "epoch": 4.515e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 4.004270403484125e-05,
      "loss": 3.0393,
      "step": 59030
    },
    {
      "epoch": 4.52e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 4.000303240825472e-05,
      "loss": 3.0413,
      "step": 59040
    },
    {
      "epoch": 4.525e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.996343627919353e-05,
      "loss": 3.0523,
      "step": 59050
    },
    {
      "epoch": 4.53e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 3.9923915671573596e-05,
      "loss": 3.0408,
      "step": 59060
    },
    {
      "epoch": 4.535e-06,
      "grad_norm": 0.59375,
      "learning_rate": 3.988447060926515e-05,
      "loss": 3.0606,
      "step": 59070
    },
    {
      "epoch": 4.54e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.984510111609281e-05,
      "loss": 3.031,
      "step": 59080
    },
    {
      "epoch": 4.545e-06,
      "grad_norm": 0.431640625,
      "learning_rate": 3.980580721583553e-05,
      "loss": 3.0423,
      "step": 59090
    },
    {
      "epoch": 4.55e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.976658893222665e-05,
      "loss": 3.0357,
      "step": 59100
    },
    {
      "epoch": 4.555e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.972744628895375e-05,
      "loss": 2.9917,
      "step": 59110
    },
    {
      "epoch": 4.56e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.968837930965888e-05,
      "loss": 3.0213,
      "step": 59120
    },
    {
      "epoch": 4.565e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.96493880179382e-05,
      "loss": 3.0338,
      "step": 59130
    },
    {
      "epoch": 4.57e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.961047243734232e-05,
      "loss": 3.0235,
      "step": 59140
    },
    {
      "epoch": 4.575e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.957163259137602e-05,
      "loss": 3.0082,
      "step": 59150
    },
    {
      "epoch": 4.58e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.953286850349838e-05,
      "loss": 3.0297,
      "step": 59160
    },
    {
      "epoch": 4.585e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.9494180197122704e-05,
      "loss": 3.024,
      "step": 59170
    },
    {
      "epoch": 4.59e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.9455567695616516e-05,
      "loss": 3.024,
      "step": 59180
    },
    {
      "epoch": 4.595e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.9417031022301555e-05,
      "loss": 3.0455,
      "step": 59190
    },
    {
      "epoch": 4.6e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.9378570200453844e-05,
      "loss": 3.0491,
      "step": 59200
    },
    {
      "epoch": 4.605e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.934018525330343e-05,
      "loss": 3.0381,
      "step": 59210
    },
    {
      "epoch": 4.61e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.93018762040347e-05,
      "loss": 3.0388,
      "step": 59220
    },
    {
      "epoch": 4.615e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 3.926364307578608e-05,
      "loss": 3.0372,
      "step": 59230
    },
    {
      "epoch": 4.62e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 3.92254858916502e-05,
      "loss": 3.0297,
      "step": 59240
    },
    {
      "epoch": 4.625e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.918740467467381e-05,
      "loss": 3.0633,
      "step": 59250
    },
    {
      "epoch": 4.63e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.914939944785773e-05,
      "loss": 3.024,
      "step": 59260
    },
    {
      "epoch": 4.635e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.911147023415697e-05,
      "loss": 3.0255,
      "step": 59270
    },
    {
      "epoch": 4.64e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 3.907361705648055e-05,
      "loss": 3.0303,
      "step": 59280
    },
    {
      "epoch": 4.645e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.903583993769158e-05,
      "loss": 3.0449,
      "step": 59290
    },
    {
      "epoch": 4.65e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.8998138900607306e-05,
      "loss": 3.0203,
      "step": 59300
    },
    {
      "epoch": 4.655e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.896051396799891e-05,
      "loss": 3.0362,
      "step": 59310
    },
    {
      "epoch": 4.66e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.892296516259165e-05,
      "loss": 3.0286,
      "step": 59320
    },
    {
      "epoch": 4.665e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.888549250706487e-05,
      "loss": 3.028,
      "step": 59330
    },
    {
      "epoch": 4.67e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.8848096024051786e-05,
      "loss": 3.0514,
      "step": 59340
    },
    {
      "epoch": 4.675e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.8810775736139735e-05,
      "loss": 3.0105,
      "step": 59350
    },
    {
      "epoch": 4.68e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 3.8773531665869956e-05,
      "loss": 3.0374,
      "step": 59360
    },
    {
      "epoch": 4.685e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.873636383573768e-05,
      "loss": 3.0357,
      "step": 59370
    },
    {
      "epoch": 4.69e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 3.8699272268192106e-05,
      "loss": 3.0355,
      "step": 59380
    },
    {
      "epoch": 4.695e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.866225698563631e-05,
      "loss": 3.049,
      "step": 59390
    },
    {
      "epoch": 4.7e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.862531801042735e-05,
      "loss": 3.0383,
      "step": 59400
    },
    {
      "epoch": 4.705e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 3.85884553648762e-05,
      "loss": 3.0607,
      "step": 59410
    },
    {
      "epoch": 4.71e-06,
      "grad_norm": 0.4375,
      "learning_rate": 3.855166907124766e-05,
      "loss": 3.0615,
      "step": 59420
    },
    {
      "epoch": 4.715e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.851495915176051e-05,
      "loss": 3.0758,
      "step": 59430
    },
    {
      "epoch": 4.72e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 3.8478325628587333e-05,
      "loss": 3.0454,
      "step": 59440
    },
    {
      "epoch": 4.725e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.844176852385459e-05,
      "loss": 3.0781,
      "step": 59450
    },
    {
      "epoch": 4.73e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.840528785964258e-05,
      "loss": 3.0399,
      "step": 59460
    },
    {
      "epoch": 4.735e-06,
      "grad_norm": 0.5,
      "learning_rate": 3.836888365798544e-05,
      "loss": 3.0473,
      "step": 59470
    },
    {
      "epoch": 4.74e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.8332555940871116e-05,
      "loss": 3.0281,
      "step": 59480
    },
    {
      "epoch": 4.745e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.8296304730241373e-05,
      "loss": 3.0279,
      "step": 59490
    },
    {
      "epoch": 4.75e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 3.826013004799171e-05,
      "loss": 3.0488,
      "step": 59500
    },
    {
      "epoch": 4.75e-06,
      "eval_loss": 3.0857510566711426,
      "eval_runtime": 0.0597,
      "eval_samples_per_second": 536.461,
      "eval_steps_per_second": 16.764,
      "step": 59500
    },
    {
      "epoch": 4.755e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.822403191597155e-05,
      "loss": 3.0382,
      "step": 59510
    },
    {
      "epoch": 4.76e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.8188010355983867e-05,
      "loss": 3.0729,
      "step": 59520
    },
    {
      "epoch": 4.765e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.815206538978556e-05,
      "loss": 3.0307,
      "step": 59530
    },
    {
      "epoch": 4.77e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.8116197039087214e-05,
      "loss": 3.0631,
      "step": 59540
    },
    {
      "epoch": 4.775e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.8080405325553084e-05,
      "loss": 3.0335,
      "step": 59550
    },
    {
      "epoch": 4.78e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.804469027080122e-05,
      "loss": 3.0385,
      "step": 59560
    },
    {
      "epoch": 4.785e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.8009051896403314e-05,
      "loss": 3.0451,
      "step": 59570
    },
    {
      "epoch": 4.79e-06,
      "grad_norm": 0.515625,
      "learning_rate": 3.7973490223884724e-05,
      "loss": 3.0262,
      "step": 59580
    },
    {
      "epoch": 4.795e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.793800527472461e-05,
      "loss": 3.0147,
      "step": 59590
    },
    {
      "epoch": 4.8e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 3.790259707035559e-05,
      "loss": 3.0267,
      "step": 59600
    },
    {
      "epoch": 4.805e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 3.78672656321641e-05,
      "loss": 3.0362,
      "step": 59610
    },
    {
      "epoch": 4.81e-06,
      "grad_norm": 0.4296875,
      "learning_rate": 3.7832010981490165e-05,
      "loss": 3.0154,
      "step": 59620
    },
    {
      "epoch": 4.815e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.7796833139627365e-05,
      "loss": 3.028,
      "step": 59630
    },
    {
      "epoch": 4.82e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.776173212782296e-05,
      "loss": 3.0346,
      "step": 59640
    },
    {
      "epoch": 4.825e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.772670796727777e-05,
      "loss": 3.0404,
      "step": 59650
    },
    {
      "epoch": 4.83e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.769176067914621e-05,
      "loss": 3.0259,
      "step": 59660
    },
    {
      "epoch": 4.835e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.7656890284536285e-05,
      "loss": 3.0127,
      "step": 59670
    },
    {
      "epoch": 4.84e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 3.762209680450949e-05,
      "loss": 3.0477,
      "step": 59680
    },
    {
      "epoch": 4.845e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.7587380260080955e-05,
      "loss": 3.0418,
      "step": 59690
    },
    {
      "epoch": 4.85e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.755274067221925e-05,
      "loss": 3.067,
      "step": 59700
    },
    {
      "epoch": 4.855e-06,
      "grad_norm": 0.5390625,
      "learning_rate": 3.751817806184654e-05,
      "loss": 3.0363,
      "step": 59710
    },
    {
      "epoch": 4.86e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.7483692449838476e-05,
      "loss": 3.0685,
      "step": 59720
    },
    {
      "epoch": 4.865e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 3.744928385702413e-05,
      "loss": 3.0503,
      "step": 59730
    },
    {
      "epoch": 4.87e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.741495230418616e-05,
      "loss": 3.0299,
      "step": 59740
    },
    {
      "epoch": 4.875e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.7380697812060656e-05,
      "loss": 3.0336,
      "step": 59750
    },
    {
      "epoch": 4.88e-06,
      "grad_norm": 0.53125,
      "learning_rate": 3.734652040133713e-05,
      "loss": 3.039,
      "step": 59760
    },
    {
      "epoch": 4.885e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.7312420092658574e-05,
      "loss": 3.0528,
      "step": 59770
    },
    {
      "epoch": 4.89e-06,
      "grad_norm": 0.5,
      "learning_rate": 3.7278396906621405e-05,
      "loss": 3.0421,
      "step": 59780
    },
    {
      "epoch": 4.895e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.724445086377541e-05,
      "loss": 3.0449,
      "step": 59790
    },
    {
      "epoch": 4.9e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.7210581984623895e-05,
      "loss": 3.0506,
      "step": 59800
    },
    {
      "epoch": 4.905e-06,
      "grad_norm": 0.4375,
      "learning_rate": 3.7176790289623433e-05,
      "loss": 3.0147,
      "step": 59810
    },
    {
      "epoch": 4.91e-06,
      "grad_norm": 0.5,
      "learning_rate": 3.714307579918404e-05,
      "loss": 3.0537,
      "step": 59820
    },
    {
      "epoch": 4.915e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 3.710943853366914e-05,
      "loss": 3.0433,
      "step": 59830
    },
    {
      "epoch": 4.92e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.707587851339541e-05,
      "loss": 3.0292,
      "step": 59840
    },
    {
      "epoch": 4.925e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.7042395758632966e-05,
      "loss": 3.0574,
      "step": 59850
    },
    {
      "epoch": 4.93e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.700899028960522e-05,
      "loss": 3.0283,
      "step": 59860
    },
    {
      "epoch": 4.935e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.697566212648886e-05,
      "loss": 3.0513,
      "step": 59870
    },
    {
      "epoch": 4.94e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.694241128941399e-05,
      "loss": 3.0761,
      "step": 59880
    },
    {
      "epoch": 4.945e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.6909237798463885e-05,
      "loss": 3.0443,
      "step": 59890
    },
    {
      "epoch": 4.95e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.687614167367523e-05,
      "loss": 3.0081,
      "step": 59900
    },
    {
      "epoch": 4.955e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.684312293503782e-05,
      "loss": 3.0219,
      "step": 59910
    },
    {
      "epoch": 4.96e-06,
      "grad_norm": 0.4375,
      "learning_rate": 3.681018160249489e-05,
      "loss": 3.0233,
      "step": 59920
    },
    {
      "epoch": 4.965e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.6777317695942806e-05,
      "loss": 3.0505,
      "step": 59930
    },
    {
      "epoch": 4.97e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.674453123523118e-05,
      "loss": 3.0284,
      "step": 59940
    },
    {
      "epoch": 4.975e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.6711822240162894e-05,
      "loss": 3.0427,
      "step": 59950
    },
    {
      "epoch": 4.98e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.667919073049399e-05,
      "loss": 3.0312,
      "step": 59960
    },
    {
      "epoch": 4.985e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.664663672593372e-05,
      "loss": 3.05,
      "step": 59970
    },
    {
      "epoch": 4.99e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.661416024614461e-05,
      "loss": 3.0506,
      "step": 59980
    },
    {
      "epoch": 4.995e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.658176131074218e-05,
      "loss": 3.0483,
      "step": 59990
    },
    {
      "epoch": 5e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 3.65494399392953e-05,
      "loss": 3.0289,
      "step": 60000
    },
    {
      "epoch": 5e-06,
      "eval_loss": 3.0852837562561035,
      "eval_runtime": 0.0593,
      "eval_samples_per_second": 539.367,
      "eval_steps_per_second": 16.855,
      "step": 60000
    },
    {
      "epoch": 5.005e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.651719615132589e-05,
      "loss": 3.0406,
      "step": 60010
    },
    {
      "epoch": 5.01e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.648502996630902e-05,
      "loss": 3.0424,
      "step": 60020
    },
    {
      "epoch": 5.015e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 3.64529414036729e-05,
      "loss": 3.0241,
      "step": 60030
    },
    {
      "epoch": 5.02e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.6420930482798884e-05,
      "loss": 3.0499,
      "step": 60040
    },
    {
      "epoch": 5.025e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.638899722302136e-05,
      "loss": 3.0608,
      "step": 60050
    },
    {
      "epoch": 5.03e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.6357141643627916e-05,
      "loss": 3.0505,
      "step": 60060
    },
    {
      "epoch": 5.035e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 3.6325363763859064e-05,
      "loss": 3.022,
      "step": 60070
    },
    {
      "epoch": 5.04e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.629366360290857e-05,
      "loss": 3.0508,
      "step": 60080
    },
    {
      "epoch": 5.045e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.626204117992313e-05,
      "loss": 3.0232,
      "step": 60090
    },
    {
      "epoch": 5.05e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 3.623049651400253e-05,
      "loss": 3.0551,
      "step": 60100
    },
    {
      "epoch": 5.055e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.619902962419958e-05,
      "loss": 3.0218,
      "step": 60110
    },
    {
      "epoch": 5.06e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 3.6167640529520125e-05,
      "loss": 3.041,
      "step": 60120
    },
    {
      "epoch": 5.065e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.6136329248923014e-05,
      "loss": 3.0214,
      "step": 60130
    },
    {
      "epoch": 5.07e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.6105095801320104e-05,
      "loss": 3.0296,
      "step": 60140
    },
    {
      "epoch": 5.075e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 3.607394020557623e-05,
      "loss": 3.034,
      "step": 60150
    },
    {
      "epoch": 5.08e-06,
      "grad_norm": 0.5625,
      "learning_rate": 3.6042862480509204e-05,
      "loss": 3.0418,
      "step": 60160
    },
    {
      "epoch": 5.085e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.601186264488983e-05,
      "loss": 3.0388,
      "step": 60170
    },
    {
      "epoch": 5.09e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.598094071744183e-05,
      "loss": 3.0308,
      "step": 60180
    },
    {
      "epoch": 5.095e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.5950096716841946e-05,
      "loss": 3.0488,
      "step": 60190
    },
    {
      "epoch": 5.1e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.5919330661719705e-05,
      "loss": 3.0319,
      "step": 60200
    },
    {
      "epoch": 5.105e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.58886425706577e-05,
      "loss": 3.0307,
      "step": 60210
    },
    {
      "epoch": 5.11e-06,
      "grad_norm": 0.484375,
      "learning_rate": 3.585803246219138e-05,
      "loss": 3.0401,
      "step": 60220
    },
    {
      "epoch": 5.115e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 3.5827500354809087e-05,
      "loss": 3.0296,
      "step": 60230
    },
    {
      "epoch": 5.12e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.579704626695206e-05,
      "loss": 3.0509,
      "step": 60240
    },
    {
      "epoch": 5.125e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.576667021701439e-05,
      "loss": 3.0433,
      "step": 60250
    },
    {
      "epoch": 5.13e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.573637222334306e-05,
      "loss": 3.0174,
      "step": 60260
    },
    {
      "epoch": 5.135e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.570615230423795e-05,
      "loss": 3.0384,
      "step": 60270
    },
    {
      "epoch": 5.14e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.567601047795166e-05,
      "loss": 3.0233,
      "step": 60280
    },
    {
      "epoch": 5.145e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.564594676268976e-05,
      "loss": 3.058,
      "step": 60290
    },
    {
      "epoch": 5.15e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.5615961176610556e-05,
      "loss": 3.0106,
      "step": 60300
    },
    {
      "epoch": 5.155e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 3.5586053737825185e-05,
      "loss": 3.0274,
      "step": 60310
    },
    {
      "epoch": 5.16e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 3.555622446439759e-05,
      "loss": 3.0462,
      "step": 60320
    },
    {
      "epoch": 5.165e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.552647337434451e-05,
      "loss": 3.0171,
      "step": 60330
    },
    {
      "epoch": 5.17e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.5496800485635405e-05,
      "loss": 3.0594,
      "step": 60340
    },
    {
      "epoch": 5.175e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.5467205816192625e-05,
      "loss": 3.0224,
      "step": 60350
    },
    {
      "epoch": 5.18e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.543768938389111e-05,
      "loss": 3.0154,
      "step": 60360
    },
    {
      "epoch": 5.185e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.540825120655871e-05,
      "loss": 3.0207,
      "step": 60370
    },
    {
      "epoch": 5.19e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 3.537889130197588e-05,
      "loss": 3.0181,
      "step": 60380
    },
    {
      "epoch": 5.195e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 3.534960968787588e-05,
      "loss": 3.0459,
      "step": 60390
    },
    {
      "epoch": 5.2e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 3.532040638194464e-05,
      "loss": 3.0218,
      "step": 60400
    },
    {
      "epoch": 5.205e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.5291281401820816e-05,
      "loss": 3.0116,
      "step": 60410
    },
    {
      "epoch": 5.21e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 3.526223476509574e-05,
      "loss": 3.0385,
      "step": 60420
    },
    {
      "epoch": 5.215e-06,
      "grad_norm": 0.515625,
      "learning_rate": 3.523326648931343e-05,
      "loss": 3.0438,
      "step": 60430
    },
    {
      "epoch": 5.22e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 3.520437659197057e-05,
      "loss": 3.0284,
      "step": 60440
    },
    {
      "epoch": 5.225e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 3.5175565090516556e-05,
      "loss": 3.0723,
      "step": 60450
    },
    {
      "epoch": 5.23e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.51468320023533e-05,
      "loss": 3.0537,
      "step": 60460
    },
    {
      "epoch": 5.235e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.5118177344835525e-05,
      "loss": 3.0178,
      "step": 60470
    },
    {
      "epoch": 5.24e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 3.508960113527044e-05,
      "loss": 3.0416,
      "step": 60480
    },
    {
      "epoch": 5.245e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.5061103390917954e-05,
      "loss": 3.052,
      "step": 60490
    },
    {
      "epoch": 5.25e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 3.5032684128990545e-05,
      "loss": 3.0286,
      "step": 60500
    },
    {
      "epoch": 5.25e-06,
      "eval_loss": 3.0846943855285645,
      "eval_runtime": 0.0599,
      "eval_samples_per_second": 534.611,
      "eval_steps_per_second": 16.707,
      "step": 60500
    },
    {
      "epoch": 5.255e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.500434336665332e-05,
      "loss": 3.0132,
      "step": 60510
    },
    {
      "epoch": 5.26e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.497608112102392e-05,
      "loss": 3.0262,
      "step": 60520
    },
    {
      "epoch": 5.265e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.494789740917265e-05,
      "loss": 3.0234,
      "step": 60530
    },
    {
      "epoch": 5.27e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.491979224812228e-05,
      "loss": 3.0236,
      "step": 60540
    },
    {
      "epoch": 5.275e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.489176565484818e-05,
      "loss": 3.0386,
      "step": 60550
    },
    {
      "epoch": 5.28e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.4863817646278315e-05,
      "loss": 3.0422,
      "step": 60560
    },
    {
      "epoch": 5.285e-06,
      "grad_norm": 0.5,
      "learning_rate": 3.483594823929309e-05,
      "loss": 3.0328,
      "step": 60570
    },
    {
      "epoch": 5.29e-06,
      "grad_norm": 0.4296875,
      "learning_rate": 3.480815745072552e-05,
      "loss": 3.0231,
      "step": 60580
    },
    {
      "epoch": 5.295e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 3.478044529736105e-05,
      "loss": 3.0393,
      "step": 60590
    },
    {
      "epoch": 5.3e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.475281179593773e-05,
      "loss": 3.0386,
      "step": 60600
    },
    {
      "epoch": 5.305e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.472525696314604e-05,
      "loss": 3.0583,
      "step": 60610
    },
    {
      "epoch": 5.31e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.4697780815628933e-05,
      "loss": 3.0432,
      "step": 60620
    },
    {
      "epoch": 5.315e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.467038336998188e-05,
      "loss": 3.0293,
      "step": 60630
    },
    {
      "epoch": 5.32e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.464306464275281e-05,
      "loss": 3.0556,
      "step": 60640
    },
    {
      "epoch": 5.325e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.461582465044205e-05,
      "loss": 3.0158,
      "step": 60650
    },
    {
      "epoch": 5.33e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 3.4588663409502464e-05,
      "loss": 3.0368,
      "step": 60660
    },
    {
      "epoch": 5.335e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 3.456158093633926e-05,
      "loss": 3.0525,
      "step": 60670
    },
    {
      "epoch": 5.34e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 3.453457724731015e-05,
      "loss": 2.9989,
      "step": 60680
    },
    {
      "epoch": 5.345e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.45076523587252e-05,
      "loss": 3.0361,
      "step": 60690
    },
    {
      "epoch": 5.35e-06,
      "grad_norm": 0.5,
      "learning_rate": 3.4480806286846915e-05,
      "loss": 3.0227,
      "step": 60700
    },
    {
      "epoch": 5.355e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.445403904789017e-05,
      "loss": 3.0443,
      "step": 60710
    },
    {
      "epoch": 5.36e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 3.4427350658022266e-05,
      "loss": 3.0639,
      "step": 60720
    },
    {
      "epoch": 5.365e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 3.4400741133362806e-05,
      "loss": 3.0332,
      "step": 60730
    },
    {
      "epoch": 5.37e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.437421048998389e-05,
      "loss": 3.0399,
      "step": 60740
    },
    {
      "epoch": 5.375e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 3.434775874390978e-05,
      "loss": 3.0268,
      "step": 60750
    },
    {
      "epoch": 5.38e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.4321385911117284e-05,
      "loss": 3.0133,
      "step": 60760
    },
    {
      "epoch": 5.385e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.429509200753542e-05,
      "loss": 3.0256,
      "step": 60770
    },
    {
      "epoch": 5.39e-06,
      "grad_norm": 0.57421875,
      "learning_rate": 3.426887704904557e-05,
      "loss": 3.0545,
      "step": 60780
    },
    {
      "epoch": 5.395e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.424274105148143e-05,
      "loss": 3.0549,
      "step": 60790
    },
    {
      "epoch": 5.4e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.4216684030629027e-05,
      "loss": 3.0207,
      "step": 60800
    },
    {
      "epoch": 5.405e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 3.4190706002226674e-05,
      "loss": 3.0224,
      "step": 60810
    },
    {
      "epoch": 5.41e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.416480698196493e-05,
      "loss": 3.0319,
      "step": 60820
    },
    {
      "epoch": 5.415e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.413898698548668e-05,
      "loss": 3.0149,
      "step": 60830
    },
    {
      "epoch": 5.42e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 3.4113246028387106e-05,
      "loss": 3.0471,
      "step": 60840
    },
    {
      "epoch": 5.425e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.4087584126213566e-05,
      "loss": 3.0249,
      "step": 60850
    },
    {
      "epoch": 5.43e-06,
      "grad_norm": 0.4296875,
      "learning_rate": 3.406200129446574e-05,
      "loss": 3.0524,
      "step": 60860
    },
    {
      "epoch": 5.435e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.403649754859554e-05,
      "loss": 3.0163,
      "step": 60870
    },
    {
      "epoch": 5.44e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.401107290400708e-05,
      "loss": 3.021,
      "step": 60880
    },
    {
      "epoch": 5.445e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.3985727376056705e-05,
      "loss": 3.0482,
      "step": 60890
    },
    {
      "epoch": 5.45e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.3960460980053005e-05,
      "loss": 3.0195,
      "step": 60900
    },
    {
      "epoch": 5.455e-06,
      "grad_norm": 0.42578125,
      "learning_rate": 3.393527373125673e-05,
      "loss": 3.0108,
      "step": 60910
    },
    {
      "epoch": 5.46e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.391016564488088e-05,
      "loss": 3.0381,
      "step": 60920
    },
    {
      "epoch": 5.465e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.388513673609057e-05,
      "loss": 3.0389,
      "step": 60930
    },
    {
      "epoch": 5.47e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.386018702000317e-05,
      "loss": 3.038,
      "step": 60940
    },
    {
      "epoch": 5.475e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.3835316511688167e-05,
      "loss": 3.0332,
      "step": 60950
    },
    {
      "epoch": 5.48e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.38105252261672e-05,
      "loss": 3.0181,
      "step": 60960
    },
    {
      "epoch": 5.485e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.3785813178414114e-05,
      "loss": 3.0446,
      "step": 60970
    },
    {
      "epoch": 5.49e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.3761180383354816e-05,
      "loss": 3.0425,
      "step": 60980
    },
    {
      "epoch": 5.495e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.3736626855867436e-05,
      "loss": 3.0675,
      "step": 60990
    },
    {
      "epoch": 5.5e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.371215261078215e-05,
      "loss": 3.0325,
      "step": 61000
    },
    {
      "epoch": 5.5e-06,
      "eval_loss": 3.08465576171875,
      "eval_runtime": 0.0622,
      "eval_samples_per_second": 514.745,
      "eval_steps_per_second": 16.086,
      "step": 61000
    },
    {
      "epoch": 5.505e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.3687757662881264e-05,
      "loss": 3.0267,
      "step": 61010
    },
    {
      "epoch": 5.51e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.366344202689924e-05,
      "loss": 3.0438,
      "step": 61020
    },
    {
      "epoch": 5.515e-06,
      "grad_norm": 0.4296875,
      "learning_rate": 3.363920571752256e-05,
      "loss": 3.0282,
      "step": 61030
    },
    {
      "epoch": 5.52e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.361504874938982e-05,
      "loss": 3.0085,
      "step": 61040
    },
    {
      "epoch": 5.525e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.359097113709177e-05,
      "loss": 3.0684,
      "step": 61050
    },
    {
      "epoch": 5.53e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.35669728951711e-05,
      "loss": 3.0342,
      "step": 61060
    },
    {
      "epoch": 5.535e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 3.3543054038122644e-05,
      "loss": 3.0305,
      "step": 61070
    },
    {
      "epoch": 5.54e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.35192145803933e-05,
      "loss": 3.0409,
      "step": 61080
    },
    {
      "epoch": 5.545e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.349545453638192e-05,
      "loss": 3.0163,
      "step": 61090
    },
    {
      "epoch": 5.55e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.3471773920439485e-05,
      "loss": 3.0266,
      "step": 61100
    },
    {
      "epoch": 5.555e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.344817274686897e-05,
      "loss": 3.044,
      "step": 61110
    },
    {
      "epoch": 5.56e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.342465102992532e-05,
      "loss": 3.0506,
      "step": 61120
    },
    {
      "epoch": 5.565e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.34012087838156e-05,
      "loss": 3.0199,
      "step": 61130
    },
    {
      "epoch": 5.57e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.3377846022698745e-05,
      "loss": 3.0224,
      "step": 61140
    },
    {
      "epoch": 5.575e-06,
      "grad_norm": 0.484375,
      "learning_rate": 3.335456276068577e-05,
      "loss": 3.0378,
      "step": 61150
    },
    {
      "epoch": 5.58e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.333135901183964e-05,
      "loss": 3.0493,
      "step": 61160
    },
    {
      "epoch": 5.585e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 3.33082347901753e-05,
      "loss": 3.0362,
      "step": 61170
    },
    {
      "epoch": 5.59e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.3285190109659674e-05,
      "loss": 3.0236,
      "step": 61180
    },
    {
      "epoch": 5.595e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.3262224984211594e-05,
      "loss": 3.0328,
      "step": 61190
    },
    {
      "epoch": 5.6e-06,
      "grad_norm": 0.55078125,
      "learning_rate": 3.323933942770191e-05,
      "loss": 3.0391,
      "step": 61200
    },
    {
      "epoch": 5.605e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.3216533453953396e-05,
      "loss": 3.0319,
      "step": 61210
    },
    {
      "epoch": 5.61e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.319380707674069e-05,
      "loss": 3.0491,
      "step": 61220
    },
    {
      "epoch": 5.615e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.317116030979045e-05,
      "loss": 3.0147,
      "step": 61230
    },
    {
      "epoch": 5.62e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.3148593166781184e-05,
      "loss": 3.013,
      "step": 61240
    },
    {
      "epoch": 5.625e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.312610566134334e-05,
      "loss": 3.066,
      "step": 61250
    },
    {
      "epoch": 5.63e-06,
      "grad_norm": 0.58203125,
      "learning_rate": 3.310369780705925e-05,
      "loss": 3.0378,
      "step": 61260
    },
    {
      "epoch": 5.635e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.308136961746316e-05,
      "loss": 3.0435,
      "step": 61270
    },
    {
      "epoch": 5.64e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.305912110604117e-05,
      "loss": 3.0087,
      "step": 61280
    },
    {
      "epoch": 5.645e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.303695228623127e-05,
      "loss": 3.0136,
      "step": 61290
    },
    {
      "epoch": 5.65e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 3.3014863171423267e-05,
      "loss": 3.0532,
      "step": 61300
    },
    {
      "epoch": 5.655e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.2992853774958967e-05,
      "loss": 3.0455,
      "step": 61310
    },
    {
      "epoch": 5.66e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.297092411013184e-05,
      "loss": 3.0331,
      "step": 61320
    },
    {
      "epoch": 5.665e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.2949074190187345e-05,
      "loss": 3.0396,
      "step": 61330
    },
    {
      "epoch": 5.67e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 3.2927304028322705e-05,
      "loss": 3.0622,
      "step": 61340
    },
    {
      "epoch": 5.675e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.2905613637686994e-05,
      "loss": 3.0437,
      "step": 61350
    },
    {
      "epoch": 5.68e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 3.288400303138109e-05,
      "loss": 3.0543,
      "step": 61360
    },
    {
      "epoch": 5.685e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.286247222245767e-05,
      "loss": 3.0124,
      "step": 61370
    },
    {
      "epoch": 5.69e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 3.284102122392127e-05,
      "loss": 3.0396,
      "step": 61380
    },
    {
      "epoch": 5.695e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.281965004872814e-05,
      "loss": 3.0256,
      "step": 61390
    },
    {
      "epoch": 5.7e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.279835870978641e-05,
      "loss": 3.0482,
      "step": 61400
    },
    {
      "epoch": 5.705e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.277714721995588e-05,
      "loss": 3.0389,
      "step": 61410
    },
    {
      "epoch": 5.71e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 3.2756015592048245e-05,
      "loss": 3.0119,
      "step": 61420
    },
    {
      "epoch": 5.715e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 3.273496383882683e-05,
      "loss": 3.0374,
      "step": 61430
    },
    {
      "epoch": 5.72e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.2713991973006846e-05,
      "loss": 3.0338,
      "step": 61440
    },
    {
      "epoch": 5.725e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.2693100007255156e-05,
      "loss": 3.0476,
      "step": 61450
    },
    {
      "epoch": 5.73e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.2672287954190415e-05,
      "loss": 3.0564,
      "step": 61460
    },
    {
      "epoch": 5.735e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.2651555826383003e-05,
      "loss": 3.0471,
      "step": 61470
    },
    {
      "epoch": 5.74e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 3.2630903636355004e-05,
      "loss": 3.0278,
      "step": 61480
    },
    {
      "epoch": 5.745e-06,
      "grad_norm": 0.53515625,
      "learning_rate": 3.2610331396580236e-05,
      "loss": 3.0591,
      "step": 61490
    },
    {
      "epoch": 5.75e-06,
      "grad_norm": 0.484375,
      "learning_rate": 3.258983911948423e-05,
      "loss": 3.0557,
      "step": 61500
    },
    {
      "epoch": 5.75e-06,
      "eval_loss": 3.0839545726776123,
      "eval_runtime": 0.0957,
      "eval_samples_per_second": 334.488,
      "eval_steps_per_second": 10.453,
      "step": 61500
    },
    {
      "epoch": 5.755e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 3.25694268174442e-05,
      "loss": 3.0437,
      "step": 61510
    },
    {
      "epoch": 5.76e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.2549094502789115e-05,
      "loss": 3.0163,
      "step": 61520
    },
    {
      "epoch": 5.765e-06,
      "grad_norm": 0.578125,
      "learning_rate": 3.2528842187799525e-05,
      "loss": 3.0244,
      "step": 61530
    },
    {
      "epoch": 5.77e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.250866988470778e-05,
      "loss": 3.0336,
      "step": 61540
    },
    {
      "epoch": 5.775e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.248857760569783e-05,
      "loss": 3.0395,
      "step": 61550
    },
    {
      "epoch": 5.78e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.2468565362905304e-05,
      "loss": 3.062,
      "step": 61560
    },
    {
      "epoch": 5.785e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 3.244863316841748e-05,
      "loss": 3.0343,
      "step": 61570
    },
    {
      "epoch": 5.79e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.242878103427332e-05,
      "loss": 3.0232,
      "step": 61580
    },
    {
      "epoch": 5.795e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.2409008972463405e-05,
      "loss": 3.0498,
      "step": 61590
    },
    {
      "epoch": 5.8e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.238931699492998e-05,
      "loss": 3.0453,
      "step": 61600
    },
    {
      "epoch": 5.805e-06,
      "grad_norm": 0.515625,
      "learning_rate": 3.236970511356685e-05,
      "loss": 3.0243,
      "step": 61610
    },
    {
      "epoch": 5.81e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.235017334021953e-05,
      "loss": 3.0525,
      "step": 61620
    },
    {
      "epoch": 5.815e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 3.233072168668509e-05,
      "loss": 3.045,
      "step": 61630
    },
    {
      "epoch": 5.82e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 3.2311350164712245e-05,
      "loss": 3.0315,
      "step": 61640
    },
    {
      "epoch": 5.825e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.229205878600129e-05,
      "loss": 3.0189,
      "step": 61650
    },
    {
      "epoch": 5.83e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.227284756220411e-05,
      "loss": 3.0462,
      "step": 61660
    },
    {
      "epoch": 5.835e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 3.2253716504924176e-05,
      "loss": 3.0236,
      "step": 61670
    },
    {
      "epoch": 5.84e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.223466562571659e-05,
      "loss": 3.0436,
      "step": 61680
    },
    {
      "epoch": 5.845e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.221569493608794e-05,
      "loss": 3.005,
      "step": 61690
    },
    {
      "epoch": 5.85e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 3.219680444749646e-05,
      "loss": 3.034,
      "step": 61700
    },
    {
      "epoch": 5.855e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.217799417135187e-05,
      "loss": 3.0469,
      "step": 61710
    },
    {
      "epoch": 5.86e-06,
      "grad_norm": 0.515625,
      "learning_rate": 3.2159264119015505e-05,
      "loss": 3.0802,
      "step": 61720
    },
    {
      "epoch": 5.865e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.214061430180023e-05,
      "loss": 3.0502,
      "step": 61730
    },
    {
      "epoch": 5.87e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.212204473097042e-05,
      "loss": 3.0568,
      "step": 61740
    },
    {
      "epoch": 5.875e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.2103555417742005e-05,
      "loss": 3.0341,
      "step": 61750
    },
    {
      "epoch": 5.88e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.208514637328242e-05,
      "loss": 3.0293,
      "step": 61760
    },
    {
      "epoch": 5.885e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.206681760871066e-05,
      "loss": 3.0316,
      "step": 61770
    },
    {
      "epoch": 5.89e-06,
      "grad_norm": 0.5,
      "learning_rate": 3.20485691350972e-05,
      "loss": 3.0193,
      "step": 61780
    },
    {
      "epoch": 5.895e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.203040096346401e-05,
      "loss": 3.0133,
      "step": 61790
    },
    {
      "epoch": 5.9e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.2012313104784576e-05,
      "loss": 3.0352,
      "step": 61800
    },
    {
      "epoch": 5.905e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 3.199430556998388e-05,
      "loss": 3.0357,
      "step": 61810
    },
    {
      "epoch": 5.91e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.197637836993836e-05,
      "loss": 3.0398,
      "step": 61820
    },
    {
      "epoch": 5.915e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.195853151547599e-05,
      "loss": 3.0161,
      "step": 61830
    },
    {
      "epoch": 5.92e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.1940765017376144e-05,
      "loss": 3.0337,
      "step": 61840
    },
    {
      "epoch": 5.925e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.192307888636971e-05,
      "loss": 3.0339,
      "step": 61850
    },
    {
      "epoch": 5.93e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.1905473133139004e-05,
      "loss": 3.0503,
      "step": 61860
    },
    {
      "epoch": 5.935e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 3.188794776831785e-05,
      "loss": 3.0456,
      "step": 61870
    },
    {
      "epoch": 5.94e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.1870502802491426e-05,
      "loss": 3.0702,
      "step": 61880
    },
    {
      "epoch": 5.945e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 3.185313824619645e-05,
      "loss": 3.066,
      "step": 61890
    },
    {
      "epoch": 5.95e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 3.183585410992099e-05,
      "loss": 3.0132,
      "step": 61900
    },
    {
      "epoch": 5.955e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.181865040410459e-05,
      "loss": 3.0068,
      "step": 61910
    },
    {
      "epoch": 5.96e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.180152713913819e-05,
      "loss": 3.0361,
      "step": 61920
    },
    {
      "epoch": 5.965e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.178448432536417e-05,
      "loss": 3.0255,
      "step": 61930
    },
    {
      "epoch": 5.97e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.176752197307629e-05,
      "loss": 3.0247,
      "step": 61940
    },
    {
      "epoch": 5.975e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.175064009251973e-05,
      "loss": 3.0466,
      "step": 61950
    },
    {
      "epoch": 5.98e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.173383869389106e-05,
      "loss": 3.0208,
      "step": 61960
    },
    {
      "epoch": 5.985e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 3.1717117787338214e-05,
      "loss": 3.0271,
      "step": 61970
    },
    {
      "epoch": 5.99e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.170047738296057e-05,
      "loss": 3.0562,
      "step": 61980
    },
    {
      "epoch": 5.995e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 3.168391749080883e-05,
      "loss": 3.0333,
      "step": 61990
    },
    {
      "epoch": 6e-06,
      "grad_norm": 0.5,
      "learning_rate": 3.166743812088508e-05,
      "loss": 3.0383,
      "step": 62000
    },
    {
      "epoch": 6e-06,
      "eval_loss": 3.0848124027252197,
      "eval_runtime": 0.0611,
      "eval_samples_per_second": 523.795,
      "eval_steps_per_second": 16.369,
      "step": 62000
    },
    {
      "epoch": 6.005e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 3.1651039283142785e-05,
      "loss": 3.0705,
      "step": 62010
    },
    {
      "epoch": 6.01e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.163472098748674e-05,
      "loss": 3.018,
      "step": 62020
    },
    {
      "epoch": 6.015e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.161848324377314e-05,
      "loss": 3.0347,
      "step": 62030
    },
    {
      "epoch": 6.02e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.1602326061809477e-05,
      "loss": 3.034,
      "step": 62040
    },
    {
      "epoch": 6.025e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.1586249451354604e-05,
      "loss": 3.0129,
      "step": 62050
    },
    {
      "epoch": 6.03e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.157025342211872e-05,
      "loss": 3.0371,
      "step": 62060
    },
    {
      "epoch": 6.035e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.1554337983763355e-05,
      "loss": 2.9919,
      "step": 62070
    },
    {
      "epoch": 6.04e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 3.153850314590133e-05,
      "loss": 3.043,
      "step": 62080
    },
    {
      "epoch": 6.045e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 3.1522748918096816e-05,
      "loss": 3.0101,
      "step": 62090
    },
    {
      "epoch": 6.05e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.150707530986528e-05,
      "loss": 3.0136,
      "step": 62100
    },
    {
      "epoch": 6.055e-06,
      "grad_norm": 0.59765625,
      "learning_rate": 3.149148233067352e-05,
      "loss": 3.0558,
      "step": 62110
    },
    {
      "epoch": 6.06e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.147596998993959e-05,
      "loss": 3.0124,
      "step": 62120
    },
    {
      "epoch": 6.065e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.146053829703288e-05,
      "loss": 3.0001,
      "step": 62130
    },
    {
      "epoch": 6.07e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.144518726127404e-05,
      "loss": 3.0428,
      "step": 62140
    },
    {
      "epoch": 6.075e-06,
      "grad_norm": 0.427734375,
      "learning_rate": 3.142991689193503e-05,
      "loss": 3.0201,
      "step": 62150
    },
    {
      "epoch": 6.08e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.1414727198239066e-05,
      "loss": 3.0272,
      "step": 62160
    },
    {
      "epoch": 6.085e-06,
      "grad_norm": 0.53125,
      "learning_rate": 3.139961818936066e-05,
      "loss": 3.0511,
      "step": 62170
    },
    {
      "epoch": 6.09e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.1384589874425546e-05,
      "loss": 3.0169,
      "step": 62180
    },
    {
      "epoch": 6.095e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.136964226251078e-05,
      "loss": 3.0654,
      "step": 62190
    },
    {
      "epoch": 6.1e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.135477536264462e-05,
      "loss": 3.0352,
      "step": 62200
    },
    {
      "epoch": 6.105e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.133998918380661e-05,
      "loss": 3.0438,
      "step": 62210
    },
    {
      "epoch": 6.11e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 3.132528373492751e-05,
      "loss": 3.0602,
      "step": 62220
    },
    {
      "epoch": 6.115e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 3.131065902488937e-05,
      "loss": 3.0239,
      "step": 62230
    },
    {
      "epoch": 6.12e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.12961150625254e-05,
      "loss": 3.0573,
      "step": 62240
    },
    {
      "epoch": 6.125e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.12816518566201e-05,
      "loss": 3.0118,
      "step": 62250
    },
    {
      "epoch": 6.13e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.126726941590917e-05,
      "loss": 3.0207,
      "step": 62260
    },
    {
      "epoch": 6.135e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.125296774907952e-05,
      "loss": 3.0375,
      "step": 62270
    },
    {
      "epoch": 6.14e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 3.1238746864769304e-05,
      "loss": 3.033,
      "step": 62280
    },
    {
      "epoch": 6.145e-06,
      "grad_norm": 0.54296875,
      "learning_rate": 3.122460677156784e-05,
      "loss": 3.0401,
      "step": 62290
    },
    {
      "epoch": 6.15e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 3.1210547478015696e-05,
      "loss": 3.0282,
      "step": 62300
    },
    {
      "epoch": 6.155e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.119656899260458e-05,
      "loss": 3.0394,
      "step": 62310
    },
    {
      "epoch": 6.16e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 3.118267132377745e-05,
      "loss": 3.0414,
      "step": 62320
    },
    {
      "epoch": 6.165e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.116885447992843e-05,
      "loss": 3.0479,
      "step": 62330
    },
    {
      "epoch": 6.17e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.11551184694028e-05,
      "loss": 3.0433,
      "step": 62340
    },
    {
      "epoch": 6.175e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.1141463300497066e-05,
      "loss": 3.0408,
      "step": 62350
    },
    {
      "epoch": 6.18e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.112788898145886e-05,
      "loss": 3.0659,
      "step": 62360
    },
    {
      "epoch": 6.185e-06,
      "grad_norm": 0.484375,
      "learning_rate": 3.1114395520486995e-05,
      "loss": 3.0591,
      "step": 62370
    },
    {
      "epoch": 6.19e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 3.1100982925731486e-05,
      "loss": 3.039,
      "step": 62380
    },
    {
      "epoch": 6.195e-06,
      "grad_norm": 0.4375,
      "learning_rate": 3.1087651205293434e-05,
      "loss": 3.0464,
      "step": 62390
    },
    {
      "epoch": 6.2e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.1074400367225146e-05,
      "loss": 3.0489,
      "step": 62400
    },
    {
      "epoch": 6.205e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.106123041953006e-05,
      "loss": 3.046,
      "step": 62410
    },
    {
      "epoch": 6.21e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 3.104814137016276e-05,
      "loss": 3.0524,
      "step": 62420
    },
    {
      "epoch": 6.215e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.1035133227028954e-05,
      "loss": 3.0438,
      "step": 62430
    },
    {
      "epoch": 6.22e-06,
      "grad_norm": 0.431640625,
      "learning_rate": 3.1022205997985504e-05,
      "loss": 3.0222,
      "step": 62440
    },
    {
      "epoch": 6.225e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.1009359690840386e-05,
      "loss": 3.0416,
      "step": 62450
    },
    {
      "epoch": 6.23e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.09965943133527e-05,
      "loss": 3.0592,
      "step": 62460
    },
    {
      "epoch": 6.235e-06,
      "grad_norm": 0.546875,
      "learning_rate": 3.098390987323268e-05,
      "loss": 3.0322,
      "step": 62470
    },
    {
      "epoch": 6.24e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.0971306378141646e-05,
      "loss": 3.031,
      "step": 62480
    },
    {
      "epoch": 6.245e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 3.0958783835692056e-05,
      "loss": 3.0101,
      "step": 62490
    },
    {
      "epoch": 6.25e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.094634225344745e-05,
      "loss": 3.027,
      "step": 62500
    },
    {
      "epoch": 6.25e-06,
      "eval_loss": 3.0835986137390137,
      "eval_runtime": 0.0595,
      "eval_samples_per_second": 538.024,
      "eval_steps_per_second": 16.813,
      "step": 62500
    },
    {
      "epoch": 6.255e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.0933981638922495e-05,
      "loss": 3.0348,
      "step": 62510
    },
    {
      "epoch": 6.26e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.0921701999582926e-05,
      "loss": 3.0738,
      "step": 62520
    },
    {
      "epoch": 6.265e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.090950334284558e-05,
      "loss": 3.0406,
      "step": 62530
    },
    {
      "epoch": 6.27e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 3.08973856760784e-05,
      "loss": 3.0498,
      "step": 62540
    },
    {
      "epoch": 6.275e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.088534900660038e-05,
      "loss": 3.0416,
      "step": 62550
    },
    {
      "epoch": 6.28e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 3.087339334168162e-05,
      "loss": 3.0103,
      "step": 62560
    },
    {
      "epoch": 6.285e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.086151868854326e-05,
      "loss": 3.0516,
      "step": 62570
    },
    {
      "epoch": 6.29e-06,
      "grad_norm": 0.484375,
      "learning_rate": 3.084972505435754e-05,
      "loss": 3.0404,
      "step": 62580
    },
    {
      "epoch": 6.295e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.0838012446247766e-05,
      "loss": 3.0473,
      "step": 62590
    },
    {
      "epoch": 6.3e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.0826380871288256e-05,
      "loss": 3.03,
      "step": 62600
    },
    {
      "epoch": 6.305e-06,
      "grad_norm": 0.54296875,
      "learning_rate": 3.081483033650446e-05,
      "loss": 3.059,
      "step": 62610
    },
    {
      "epoch": 6.31e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.080336084887284e-05,
      "loss": 3.0237,
      "step": 62620
    },
    {
      "epoch": 6.315e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.079197241532088e-05,
      "loss": 3.0566,
      "step": 62630
    },
    {
      "epoch": 6.32e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.078066504272716e-05,
      "loss": 3.0253,
      "step": 62640
    },
    {
      "epoch": 6.325e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.0769438737921254e-05,
      "loss": 3.0261,
      "step": 62650
    },
    {
      "epoch": 6.33e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 3.075829350768381e-05,
      "loss": 3.0297,
      "step": 62660
    },
    {
      "epoch": 6.335e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.07472293587465e-05,
      "loss": 3.0335,
      "step": 62670
    },
    {
      "epoch": 6.34e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 3.073624629779197e-05,
      "loss": 3.0238,
      "step": 62680
    },
    {
      "epoch": 6.345e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.072534433145398e-05,
      "loss": 3.0209,
      "step": 62690
    },
    {
      "epoch": 6.35e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.071452346631723e-05,
      "loss": 3.0414,
      "step": 62700
    },
    {
      "epoch": 6.355e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.0703783708917485e-05,
      "loss": 3.0279,
      "step": 62710
    },
    {
      "epoch": 6.36e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.069312506574149e-05,
      "loss": 3.053,
      "step": 62720
    },
    {
      "epoch": 6.365e-06,
      "grad_norm": 0.423828125,
      "learning_rate": 3.068254754322702e-05,
      "loss": 3.003,
      "step": 62730
    },
    {
      "epoch": 6.37e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.0672051147762844e-05,
      "loss": 3.0495,
      "step": 62740
    },
    {
      "epoch": 6.375e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.066163588568872e-05,
      "loss": 3.0056,
      "step": 62750
    },
    {
      "epoch": 6.38e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.065130176329542e-05,
      "loss": 3.032,
      "step": 62760
    },
    {
      "epoch": 6.385e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.0641048786824725e-05,
      "loss": 3.0657,
      "step": 62770
    },
    {
      "epoch": 6.39e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.063087696246933e-05,
      "loss": 3.0405,
      "step": 62780
    },
    {
      "epoch": 6.395e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.0620786296372984e-05,
      "loss": 3.0232,
      "step": 62790
    },
    {
      "epoch": 6.4e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.061077679463042e-05,
      "loss": 3.0455,
      "step": 62800
    },
    {
      "epoch": 6.405e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.0600848463287305e-05,
      "loss": 3.0434,
      "step": 62810
    },
    {
      "epoch": 6.41e-06,
      "grad_norm": 0.5703125,
      "learning_rate": 3.059100130834031e-05,
      "loss": 3.0374,
      "step": 62820
    },
    {
      "epoch": 6.415e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.058123533573707e-05,
      "loss": 3.0423,
      "step": 62830
    },
    {
      "epoch": 6.42e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.057155055137615e-05,
      "loss": 3.0335,
      "step": 62840
    },
    {
      "epoch": 6.425e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.0561946961107145e-05,
      "loss": 3.0528,
      "step": 62850
    },
    {
      "epoch": 6.43e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.0552424570730564e-05,
      "loss": 2.9976,
      "step": 62860
    },
    {
      "epoch": 6.435e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.0542983385997883e-05,
      "loss": 3.0213,
      "step": 62870
    },
    {
      "epoch": 6.44e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.053362341261152e-05,
      "loss": 3.0585,
      "step": 62880
    },
    {
      "epoch": 6.445e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.0524344656224846e-05,
      "loss": 3.0266,
      "step": 62890
    },
    {
      "epoch": 6.45e-06,
      "grad_norm": 0.4375,
      "learning_rate": 3.0515147122442208e-05,
      "loss": 3.0345,
      "step": 62900
    },
    {
      "epoch": 6.455e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.0506030816818843e-05,
      "loss": 3.0525,
      "step": 62910
    },
    {
      "epoch": 6.46e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 3.0496995744860966e-05,
      "loss": 3.0328,
      "step": 62920
    },
    {
      "epoch": 6.465e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 3.0488041912025727e-05,
      "loss": 3.0449,
      "step": 62930
    },
    {
      "epoch": 6.47e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.0479169323721162e-05,
      "loss": 3.0285,
      "step": 62940
    },
    {
      "epoch": 6.475e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.0470377985306294e-05,
      "loss": 3.0222,
      "step": 62950
    },
    {
      "epoch": 6.48e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.046166790209103e-05,
      "loss": 3.0411,
      "step": 62960
    },
    {
      "epoch": 6.485e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.0453039079336234e-05,
      "loss": 3.0731,
      "step": 62970
    },
    {
      "epoch": 6.49e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.0444491522253663e-05,
      "loss": 3.0319,
      "step": 62980
    },
    {
      "epoch": 6.495e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.0436025236005974e-05,
      "loss": 3.0117,
      "step": 62990
    },
    {
      "epoch": 6.5e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.0427640225706802e-05,
      "loss": 3.048,
      "step": 63000
    },
    {
      "epoch": 6.5e-06,
      "eval_loss": 3.0833935737609863,
      "eval_runtime": 0.0612,
      "eval_samples_per_second": 523.001,
      "eval_steps_per_second": 16.344,
      "step": 63000
    },
    {
      "epoch": 6.505e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.041933649642062e-05,
      "loss": 3.0379,
      "step": 63010
    },
    {
      "epoch": 6.51e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.041111405316285e-05,
      "loss": 3.0309,
      "step": 63020
    },
    {
      "epoch": 6.515e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.0402972900899798e-05,
      "loss": 3.0227,
      "step": 63030
    },
    {
      "epoch": 6.52e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.039491304454869e-05,
      "loss": 3.0294,
      "step": 63040
    },
    {
      "epoch": 6.525e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.038693448897763e-05,
      "loss": 3.0564,
      "step": 63050
    },
    {
      "epoch": 6.53e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.0379037239005623e-05,
      "loss": 3.0354,
      "step": 63060
    },
    {
      "epoch": 6.535e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.037122129940256e-05,
      "loss": 3.0398,
      "step": 63070
    },
    {
      "epoch": 6.54e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 3.0363486674889242e-05,
      "loss": 3.039,
      "step": 63080
    },
    {
      "epoch": 6.545e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.0355833370137347e-05,
      "loss": 3.0539,
      "step": 63090
    },
    {
      "epoch": 6.55e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.034826138976941e-05,
      "loss": 3.0362,
      "step": 63100
    },
    {
      "epoch": 6.555e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 3.0340770738358892e-05,
      "loss": 3.0117,
      "step": 63110
    },
    {
      "epoch": 6.56e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.0333361420430092e-05,
      "loss": 3.0176,
      "step": 63120
    },
    {
      "epoch": 6.565e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.0326033440458197e-05,
      "loss": 3.0335,
      "step": 63130
    },
    {
      "epoch": 6.57e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 3.0318786802869297e-05,
      "loss": 3.0305,
      "step": 63140
    },
    {
      "epoch": 6.575e-06,
      "grad_norm": 0.5,
      "learning_rate": 3.0311621512040287e-05,
      "loss": 3.038,
      "step": 63150
    },
    {
      "epoch": 6.58e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.030453757229899e-05,
      "loss": 3.0364,
      "step": 63160
    },
    {
      "epoch": 6.585e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 3.0297534987924056e-05,
      "loss": 3.0531,
      "step": 63170
    },
    {
      "epoch": 6.59e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.0290613763145023e-05,
      "loss": 3.0425,
      "step": 63180
    },
    {
      "epoch": 6.595e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.0283773902142265e-05,
      "loss": 3.0619,
      "step": 63190
    },
    {
      "epoch": 6.6e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.027701540904702e-05,
      "loss": 3.0254,
      "step": 63200
    },
    {
      "epoch": 6.605e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.0270338287941388e-05,
      "loss": 3.0546,
      "step": 63210
    },
    {
      "epoch": 6.61e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.026374254285832e-05,
      "loss": 3.0334,
      "step": 63220
    },
    {
      "epoch": 6.615e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 3.0257228177781606e-05,
      "loss": 3.0591,
      "step": 63230
    },
    {
      "epoch": 6.62e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.0250795196645893e-05,
      "loss": 3.0298,
      "step": 63240
    },
    {
      "epoch": 6.625e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.0244443603336668e-05,
      "loss": 3.0474,
      "step": 63250
    },
    {
      "epoch": 6.63e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.023817340169026e-05,
      "loss": 3.0313,
      "step": 63260
    },
    {
      "epoch": 6.635e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.0231984595493846e-05,
      "loss": 3.0408,
      "step": 63270
    },
    {
      "epoch": 6.64e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.0225877188485416e-05,
      "loss": 3.0301,
      "step": 63280
    },
    {
      "epoch": 6.645e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.0219851184353815e-05,
      "loss": 3.0367,
      "step": 63290
    },
    {
      "epoch": 6.65e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.0213906586738742e-05,
      "loss": 3.0448,
      "step": 63300
    },
    {
      "epoch": 6.655e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.020804339923067e-05,
      "loss": 3.0388,
      "step": 63310
    },
    {
      "epoch": 6.66e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.020226162537096e-05,
      "loss": 3.0594,
      "step": 63320
    },
    {
      "epoch": 6.665e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.0196561268651768e-05,
      "loss": 3.026,
      "step": 63330
    },
    {
      "epoch": 6.67e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.0190942332516066e-05,
      "loss": 3.0518,
      "step": 63340
    },
    {
      "epoch": 6.675e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.0185404820357683e-05,
      "loss": 3.0423,
      "step": 63350
    },
    {
      "epoch": 6.68e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 3.0179948735521233e-05,
      "loss": 3.0296,
      "step": 63360
    },
    {
      "epoch": 6.685e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.0174574081302176e-05,
      "loss": 3.0488,
      "step": 63370
    },
    {
      "epoch": 6.69e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.0169280860946752e-05,
      "loss": 3.038,
      "step": 63380
    },
    {
      "epoch": 6.695e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 3.0164069077652067e-05,
      "loss": 3.0292,
      "step": 63390
    },
    {
      "epoch": 6.7e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.0158938734565994e-05,
      "loss": 3.0347,
      "step": 63400
    },
    {
      "epoch": 6.705e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.015388983478725e-05,
      "loss": 3.0286,
      "step": 63410
    },
    {
      "epoch": 6.71e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.014892238136531e-05,
      "loss": 3.0315,
      "step": 63420
    },
    {
      "epoch": 6.715e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 3.0144036377300536e-05,
      "loss": 3.0368,
      "step": 63430
    },
    {
      "epoch": 6.72e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 3.0139231825544012e-05,
      "loss": 3.027,
      "step": 63440
    },
    {
      "epoch": 6.725e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.013450872899769e-05,
      "loss": 3.0308,
      "step": 63450
    },
    {
      "epoch": 6.73e-06,
      "grad_norm": 0.484375,
      "learning_rate": 3.0129867090514267e-05,
      "loss": 3.0274,
      "step": 63460
    },
    {
      "epoch": 6.735e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.01253069128973e-05,
      "loss": 3.0306,
      "step": 63470
    },
    {
      "epoch": 6.74e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 3.0120828198901088e-05,
      "loss": 3.0237,
      "step": 63480
    },
    {
      "epoch": 6.745e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.0116430951230763e-05,
      "loss": 3.0481,
      "step": 63490
    },
    {
      "epoch": 6.75e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.011211517254223e-05,
      "loss": 3.0555,
      "step": 63500
    },
    {
      "epoch": 6.75e-06,
      "eval_loss": 3.08327317237854,
      "eval_runtime": 0.0603,
      "eval_samples_per_second": 530.402,
      "eval_steps_per_second": 16.575,
      "step": 63500
    },
    {
      "epoch": 6.755e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 3.010788086544223e-05,
      "loss": 3.0309,
      "step": 63510
    },
    {
      "epoch": 6.76e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 3.010372803248822e-05,
      "loss": 3.0307,
      "step": 63520
    },
    {
      "epoch": 6.765e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.0099656676188508e-05,
      "loss": 3.0408,
      "step": 63530
    },
    {
      "epoch": 6.77e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 3.0095666799002165e-05,
      "loss": 3.0413,
      "step": 63540
    },
    {
      "epoch": 6.775e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 3.009175840333908e-05,
      "loss": 3.0354,
      "step": 63550
    },
    {
      "epoch": 6.78e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 3.008793149155988e-05,
      "loss": 3.0388,
      "step": 63560
    },
    {
      "epoch": 6.785e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.0084186065976e-05,
      "loss": 3.0422,
      "step": 63570
    },
    {
      "epoch": 6.79e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.0080522128849674e-05,
      "loss": 3.0437,
      "step": 63580
    },
    {
      "epoch": 6.795e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.0076939682393893e-05,
      "loss": 3.0458,
      "step": 63590
    },
    {
      "epoch": 6.8e-06,
      "grad_norm": 0.484375,
      "learning_rate": 3.007343872877242e-05,
      "loss": 3.0292,
      "step": 63600
    },
    {
      "epoch": 6.805e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.0070019270099847e-05,
      "loss": 3.0531,
      "step": 63610
    },
    {
      "epoch": 6.81e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.0066681308441487e-05,
      "loss": 3.0607,
      "step": 63620
    },
    {
      "epoch": 6.815e-06,
      "grad_norm": 0.53515625,
      "learning_rate": 3.006342484581344e-05,
      "loss": 3.0163,
      "step": 63630
    },
    {
      "epoch": 6.82e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 3.006024988418262e-05,
      "loss": 3.0359,
      "step": 63640
    },
    {
      "epoch": 6.825e-06,
      "grad_norm": 0.427734375,
      "learning_rate": 3.0057156425466686e-05,
      "loss": 3.039,
      "step": 63650
    },
    {
      "epoch": 6.83e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 3.0054144471534037e-05,
      "loss": 3.0405,
      "step": 63660
    },
    {
      "epoch": 6.835e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.0051214024203922e-05,
      "loss": 3.0413,
      "step": 63670
    },
    {
      "epoch": 6.84e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.004836508524629e-05,
      "loss": 3.0417,
      "step": 63680
    },
    {
      "epoch": 6.845e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.0045597656381878e-05,
      "loss": 3.0409,
      "step": 63690
    },
    {
      "epoch": 6.85e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 3.0042911739282228e-05,
      "loss": 3.064,
      "step": 63700
    },
    {
      "epoch": 6.855e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.0040307335569605e-05,
      "loss": 3.0331,
      "step": 63710
    },
    {
      "epoch": 6.86e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 3.0037784446817037e-05,
      "loss": 3.0454,
      "step": 63720
    },
    {
      "epoch": 6.865e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 3.003534307454837e-05,
      "loss": 3.0351,
      "step": 63730
    },
    {
      "epoch": 6.87e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.0032983220238157e-05,
      "loss": 3.034,
      "step": 63740
    },
    {
      "epoch": 6.875e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.003070488531174e-05,
      "loss": 3.036,
      "step": 63750
    },
    {
      "epoch": 6.88e-06,
      "grad_norm": 0.46875,
      "learning_rate": 3.0028508071145236e-05,
      "loss": 3.0274,
      "step": 63760
    },
    {
      "epoch": 6.885e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 3.002639277906549e-05,
      "loss": 3.0161,
      "step": 63770
    },
    {
      "epoch": 6.89e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.0024359010350137e-05,
      "loss": 3.0502,
      "step": 63780
    },
    {
      "epoch": 6.895e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.0022406766227565e-05,
      "loss": 3.0289,
      "step": 63790
    },
    {
      "epoch": 6.9e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 3.0020536047876918e-05,
      "loss": 3.0495,
      "step": 63800
    },
    {
      "epoch": 6.905e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.0018746856428098e-05,
      "loss": 3.0528,
      "step": 63810
    },
    {
      "epoch": 6.91e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.0017039192961777e-05,
      "loss": 3.032,
      "step": 63820
    },
    {
      "epoch": 6.915e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.001541305850935e-05,
      "loss": 3.0394,
      "step": 63830
    },
    {
      "epoch": 6.92e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 3.0013868454053028e-05,
      "loss": 3.0593,
      "step": 63840
    },
    {
      "epoch": 6.925e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 3.001240538052572e-05,
      "loss": 3.0386,
      "step": 63850
    },
    {
      "epoch": 6.93e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 3.0011023838811114e-05,
      "loss": 3.0161,
      "step": 63860
    },
    {
      "epoch": 6.935e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 3.000972382974367e-05,
      "loss": 3.0411,
      "step": 63870
    },
    {
      "epoch": 6.94e-06,
      "grad_norm": 0.453125,
      "learning_rate": 3.0008505354108577e-05,
      "loss": 3.0213,
      "step": 63880
    },
    {
      "epoch": 6.945e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 3.0007368412641783e-05,
      "loss": 3.0314,
      "step": 63890
    },
    {
      "epoch": 6.95e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 3.0006313006030014e-05,
      "loss": 3.0518,
      "step": 63900
    },
    {
      "epoch": 6.955e-06,
      "grad_norm": 0.484375,
      "learning_rate": 3.0005339134910707e-05,
      "loss": 3.0246,
      "step": 63910
    },
    {
      "epoch": 6.96e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.0004446799872097e-05,
      "loss": 3.0388,
      "step": 63920
    },
    {
      "epoch": 6.965e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 3.0003636001453123e-05,
      "loss": 3.0642,
      "step": 63930
    },
    {
      "epoch": 6.97e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 3.0002906740143532e-05,
      "loss": 3.0401,
      "step": 63940
    },
    {
      "epoch": 6.975e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.0002259016383762e-05,
      "loss": 3.0406,
      "step": 63950
    },
    {
      "epoch": 6.98e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.0001692830565078e-05,
      "loss": 3.0625,
      "step": 63960
    },
    {
      "epoch": 6.985e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 3.0001208183029415e-05,
      "loss": 3.0375,
      "step": 63970
    },
    {
      "epoch": 6.99e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 3.0000805074069504e-05,
      "loss": 3.0482,
      "step": 63980
    },
    {
      "epoch": 6.995e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.000048350392884e-05,
      "loss": 3.0307,
      "step": 63990
    },
    {
      "epoch": 7e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 3.0000243472801637e-05,
      "loss": 3.0391,
      "step": 64000
    },
    {
      "epoch": 7e-06,
      "eval_loss": 3.0830166339874268,
      "eval_runtime": 0.0603,
      "eval_samples_per_second": 530.398,
      "eval_steps_per_second": 16.575,
      "step": 64000
    },
    {
      "epoch": 7.005e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.000008498083287e-05,
      "loss": 3.0437,
      "step": 64010
    },
    {
      "epoch": 7.01e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 3.0000008028118274e-05,
      "loss": 3.0368,
      "step": 64020
    },
    {
      "epoch": 7.015e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "step": 64030
    },
    {
      "epoch": 7.02e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0165,
      "step": 64040
    },
    {
      "epoch": 7.025e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "step": 64050
    },
    {
      "epoch": 7.03e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0088,
      "step": 64060
    },
    {
      "epoch": 7.035e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "step": 64070
    },
    {
      "epoch": 7.04e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "step": 64080
    },
    {
      "epoch": 7.045e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0778,
      "step": 64090
    },
    {
      "epoch": 7.05e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0224,
      "step": 64100
    },
    {
      "epoch": 7.055e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "step": 64110
    },
    {
      "epoch": 7.06e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.029,
      "step": 64120
    },
    {
      "epoch": 7.065e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "step": 64130
    },
    {
      "epoch": 7.07e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "step": 64140
    },
    {
      "epoch": 7.075e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0511,
      "step": 64150
    },
    {
      "epoch": 7.08e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "step": 64160
    },
    {
      "epoch": 7.085e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "step": 64170
    },
    {
      "epoch": 7.09e-06,
      "grad_norm": 0.423828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0244,
      "step": 64180
    },
    {
      "epoch": 7.095e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0216,
      "step": 64190
    },
    {
      "epoch": 7.1e-06,
      "grad_norm": 0.55078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "step": 64200
    },
    {
      "epoch": 7.105e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "step": 64210
    },
    {
      "epoch": 7.11e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0405,
      "step": 64220
    },
    {
      "epoch": 7.115e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0301,
      "step": 64230
    },
    {
      "epoch": 7.12e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0335,
      "step": 64240
    },
    {
      "epoch": 7.125e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0602,
      "step": 64250
    },
    {
      "epoch": 7.13e-06,
      "grad_norm": 0.5703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0227,
      "step": 64260
    },
    {
      "epoch": 7.135e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "step": 64270
    },
    {
      "epoch": 7.14e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "step": 64280
    },
    {
      "epoch": 7.145e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0238,
      "step": 64290
    },
    {
      "epoch": 7.15e-06,
      "grad_norm": 0.53515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0313,
      "step": 64300
    },
    {
      "epoch": 7.155e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0083,
      "step": 64310
    },
    {
      "epoch": 7.16e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "step": 64320
    },
    {
      "epoch": 7.165e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.042,
      "step": 64330
    },
    {
      "epoch": 7.17e-06,
      "grad_norm": 0.53125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "step": 64340
    },
    {
      "epoch": 7.175e-06,
      "grad_norm": 0.5625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "step": 64350
    },
    {
      "epoch": 7.18e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "step": 64360
    },
    {
      "epoch": 7.185e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "step": 64370
    },
    {
      "epoch": 7.19e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0525,
      "step": 64380
    },
    {
      "epoch": 7.195e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "step": 64390
    },
    {
      "epoch": 7.2e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0737,
      "step": 64400
    },
    {
      "epoch": 7.205e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0361,
      "step": 64410
    },
    {
      "epoch": 7.21e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0793,
      "step": 64420
    },
    {
      "epoch": 7.215e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0429,
      "step": 64430
    },
    {
      "epoch": 7.22e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0193,
      "step": 64440
    },
    {
      "epoch": 7.225e-06,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "step": 64450
    },
    {
      "epoch": 7.23e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0617,
      "step": 64460
    },
    {
      "epoch": 7.235e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "step": 64470
    },
    {
      "epoch": 7.24e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "step": 64480
    },
    {
      "epoch": 7.245e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "step": 64490
    },
    {
      "epoch": 7.25e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0243,
      "step": 64500
    },
    {
      "epoch": 7.25e-06,
      "eval_loss": 3.0821197032928467,
      "eval_runtime": 0.0629,
      "eval_samples_per_second": 508.925,
      "eval_steps_per_second": 15.904,
      "step": 64500
    },
    {
      "epoch": 7.255e-06,
      "grad_norm": 0.4296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0193,
      "step": 64510
    },
    {
      "epoch": 7.26e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0261,
      "step": 64520
    },
    {
      "epoch": 7.265e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0667,
      "step": 64530
    },
    {
      "epoch": 7.27e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0655,
      "step": 64540
    },
    {
      "epoch": 7.275e-06,
      "grad_norm": 0.431640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "step": 64550
    },
    {
      "epoch": 7.28e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0604,
      "step": 64560
    },
    {
      "epoch": 7.285e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9941,
      "step": 64570
    },
    {
      "epoch": 7.29e-06,
      "grad_norm": 2.75,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "step": 64580
    },
    {
      "epoch": 7.295e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0296,
      "step": 64590
    },
    {
      "epoch": 7.3e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "step": 64600
    },
    {
      "epoch": 7.305e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0215,
      "step": 64610
    },
    {
      "epoch": 7.31e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0489,
      "step": 64620
    },
    {
      "epoch": 7.315e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0275,
      "step": 64630
    },
    {
      "epoch": 7.32e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0333,
      "step": 64640
    },
    {
      "epoch": 7.325e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0416,
      "step": 64650
    },
    {
      "epoch": 7.33e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "step": 64660
    },
    {
      "epoch": 7.335e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0211,
      "step": 64670
    },
    {
      "epoch": 7.34e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0275,
      "step": 64680
    },
    {
      "epoch": 7.345e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0303,
      "step": 64690
    },
    {
      "epoch": 7.35e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "step": 64700
    },
    {
      "epoch": 7.355e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "step": 64710
    },
    {
      "epoch": 7.36e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0178,
      "step": 64720
    },
    {
      "epoch": 7.365e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0349,
      "step": 64730
    },
    {
      "epoch": 7.37e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "step": 64740
    },
    {
      "epoch": 7.375e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "step": 64750
    },
    {
      "epoch": 7.38e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "step": 64760
    },
    {
      "epoch": 7.385e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "step": 64770
    },
    {
      "epoch": 7.39e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.039,
      "step": 64780
    },
    {
      "epoch": 7.395e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0355,
      "step": 64790
    },
    {
      "epoch": 7.4e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0192,
      "step": 64800
    },
    {
      "epoch": 7.405e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0429,
      "step": 64810
    },
    {
      "epoch": 7.41e-06,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "step": 64820
    },
    {
      "epoch": 7.415e-06,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "step": 64830
    },
    {
      "epoch": 7.42e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0329,
      "step": 64840
    },
    {
      "epoch": 7.425e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0322,
      "step": 64850
    },
    {
      "epoch": 7.43e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "step": 64860
    },
    {
      "epoch": 7.435e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "step": 64870
    },
    {
      "epoch": 7.44e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "step": 64880
    },
    {
      "epoch": 7.445e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0665,
      "step": 64890
    },
    {
      "epoch": 7.45e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0242,
      "step": 64900
    },
    {
      "epoch": 7.455e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "step": 64910
    },
    {
      "epoch": 7.46e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0366,
      "step": 64920
    },
    {
      "epoch": 7.465e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "step": 64930
    },
    {
      "epoch": 7.47e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "step": 64940
    },
    {
      "epoch": 7.475e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.023,
      "step": 64950
    },
    {
      "epoch": 7.48e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0094,
      "step": 64960
    },
    {
      "epoch": 7.485e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "step": 64970
    },
    {
      "epoch": 7.49e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "step": 64980
    },
    {
      "epoch": 7.495e-06,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "step": 64990
    },
    {
      "epoch": 7.5e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0288,
      "step": 65000
    },
    {
      "epoch": 7.5e-06,
      "eval_loss": 3.0829381942749023,
      "eval_runtime": 0.058,
      "eval_samples_per_second": 552.189,
      "eval_steps_per_second": 17.256,
      "step": 65000
    },
    {
      "epoch": 7.505e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0558,
      "step": 65010
    },
    {
      "epoch": 7.51e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0339,
      "step": 65020
    },
    {
      "epoch": 7.515e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "step": 65030
    },
    {
      "epoch": 7.52e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.031,
      "step": 65040
    },
    {
      "epoch": 7.525e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0271,
      "step": 65050
    },
    {
      "epoch": 7.53e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "step": 65060
    },
    {
      "epoch": 7.535e-06,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0578,
      "step": 65070
    },
    {
      "epoch": 7.54e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "step": 65080
    },
    {
      "epoch": 7.545e-06,
      "grad_norm": 0.4296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0399,
      "step": 65090
    },
    {
      "epoch": 7.55e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "step": 65100
    },
    {
      "epoch": 7.555e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "step": 65110
    },
    {
      "epoch": 7.56e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "step": 65120
    },
    {
      "epoch": 7.565e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "step": 65130
    },
    {
      "epoch": 7.57e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0498,
      "step": 65140
    },
    {
      "epoch": 7.575e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0648,
      "step": 65150
    },
    {
      "epoch": 7.58e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "step": 65160
    },
    {
      "epoch": 7.585e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0346,
      "step": 65170
    },
    {
      "epoch": 7.59e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0284,
      "step": 65180
    },
    {
      "epoch": 7.595e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0189,
      "step": 65190
    },
    {
      "epoch": 7.6e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 65200
    },
    {
      "epoch": 7.605e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0153,
      "step": 65210
    },
    {
      "epoch": 7.61e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "step": 65220
    },
    {
      "epoch": 7.615e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "step": 65230
    },
    {
      "epoch": 7.62e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0308,
      "step": 65240
    },
    {
      "epoch": 7.625e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0261,
      "step": 65250
    },
    {
      "epoch": 7.63e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0353,
      "step": 65260
    },
    {
      "epoch": 7.635e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "step": 65270
    },
    {
      "epoch": 7.64e-06,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0374,
      "step": 65280
    },
    {
      "epoch": 7.645e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0164,
      "step": 65290
    },
    {
      "epoch": 7.65e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "step": 65300
    },
    {
      "epoch": 7.655e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0523,
      "step": 65310
    },
    {
      "epoch": 7.66e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0023,
      "step": 65320
    },
    {
      "epoch": 7.665e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0346,
      "step": 65330
    },
    {
      "epoch": 7.67e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.003,
      "step": 65340
    },
    {
      "epoch": 7.675e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "step": 65350
    },
    {
      "epoch": 7.68e-06,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0157,
      "step": 65360
    },
    {
      "epoch": 7.685e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0073,
      "step": 65370
    },
    {
      "epoch": 7.69e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0498,
      "step": 65380
    },
    {
      "epoch": 7.695e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0566,
      "step": 65390
    },
    {
      "epoch": 7.7e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "step": 65400
    },
    {
      "epoch": 7.705e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0206,
      "step": 65410
    },
    {
      "epoch": 7.71e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "step": 65420
    },
    {
      "epoch": 7.715e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "step": 65430
    },
    {
      "epoch": 7.72e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "step": 65440
    },
    {
      "epoch": 7.725e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.007,
      "step": 65450
    },
    {
      "epoch": 7.73e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "step": 65460
    },
    {
      "epoch": 7.735e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "step": 65470
    },
    {
      "epoch": 7.74e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "step": 65480
    },
    {
      "epoch": 7.745e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0401,
      "step": 65490
    },
    {
      "epoch": 7.75e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "step": 65500
    },
    {
      "epoch": 7.75e-06,
      "eval_loss": 3.0822300910949707,
      "eval_runtime": 0.0604,
      "eval_samples_per_second": 529.693,
      "eval_steps_per_second": 16.553,
      "step": 65500
    },
    {
      "epoch": 7.755e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "step": 65510
    },
    {
      "epoch": 7.76e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0039,
      "step": 65520
    },
    {
      "epoch": 7.765e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "step": 65530
    },
    {
      "epoch": 7.77e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "step": 65540
    },
    {
      "epoch": 7.775e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0669,
      "step": 65550
    },
    {
      "epoch": 7.78e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0288,
      "step": 65560
    },
    {
      "epoch": 7.785e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0224,
      "step": 65570
    },
    {
      "epoch": 7.79e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "step": 65580
    },
    {
      "epoch": 7.795e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.048,
      "step": 65590
    },
    {
      "epoch": 7.8e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0177,
      "step": 65600
    },
    {
      "epoch": 7.805e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0639,
      "step": 65610
    },
    {
      "epoch": 7.81e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0379,
      "step": 65620
    },
    {
      "epoch": 7.815e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "step": 65630
    },
    {
      "epoch": 7.82e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "step": 65640
    },
    {
      "epoch": 7.825e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0333,
      "step": 65650
    },
    {
      "epoch": 7.83e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0143,
      "step": 65660
    },
    {
      "epoch": 7.835e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "step": 65670
    },
    {
      "epoch": 7.84e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "step": 65680
    },
    {
      "epoch": 7.845e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "step": 65690
    },
    {
      "epoch": 7.85e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0391,
      "step": 65700
    },
    {
      "epoch": 7.855e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "step": 65710
    },
    {
      "epoch": 7.86e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0289,
      "step": 65720
    },
    {
      "epoch": 7.865e-06,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0328,
      "step": 65730
    },
    {
      "epoch": 7.87e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0323,
      "step": 65740
    },
    {
      "epoch": 7.875e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "step": 65750
    },
    {
      "epoch": 7.88e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "step": 65760
    },
    {
      "epoch": 7.885e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "step": 65770
    },
    {
      "epoch": 7.89e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0184,
      "step": 65780
    },
    {
      "epoch": 7.895e-06,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "step": 65790
    },
    {
      "epoch": 7.9e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0382,
      "step": 65800
    },
    {
      "epoch": 7.905e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "step": 65810
    },
    {
      "epoch": 7.91e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "step": 65820
    },
    {
      "epoch": 7.915e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "step": 65830
    },
    {
      "epoch": 7.92e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0354,
      "step": 65840
    },
    {
      "epoch": 7.925e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "step": 65850
    },
    {
      "epoch": 7.93e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "step": 65860
    },
    {
      "epoch": 7.935e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0441,
      "step": 65870
    },
    {
      "epoch": 7.94e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "step": 65880
    },
    {
      "epoch": 7.945e-06,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "step": 65890
    },
    {
      "epoch": 7.95e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "step": 65900
    },
    {
      "epoch": 7.955e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0314,
      "step": 65910
    },
    {
      "epoch": 7.96e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0306,
      "step": 65920
    },
    {
      "epoch": 7.965e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "step": 65930
    },
    {
      "epoch": 7.97e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "step": 65940
    },
    {
      "epoch": 7.975e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "step": 65950
    },
    {
      "epoch": 7.98e-06,
      "grad_norm": 0.87109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "step": 65960
    },
    {
      "epoch": 7.985e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0345,
      "step": 65970
    },
    {
      "epoch": 7.99e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "step": 65980
    },
    {
      "epoch": 7.995e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0289,
      "step": 65990
    },
    {
      "epoch": 8e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "step": 66000
    },
    {
      "epoch": 8e-06,
      "eval_loss": 3.0830793380737305,
      "eval_runtime": 0.0582,
      "eval_samples_per_second": 550.129,
      "eval_steps_per_second": 17.192,
      "step": 66000
    },
    {
      "epoch": 8.005e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "step": 66010
    },
    {
      "epoch": 8.01e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "step": 66020
    },
    {
      "epoch": 8.015e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "step": 66030
    },
    {
      "epoch": 8.02e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0215,
      "step": 66040
    },
    {
      "epoch": 8.025e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9968,
      "step": 66050
    },
    {
      "epoch": 8.03e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0261,
      "step": 66060
    },
    {
      "epoch": 8.035e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "step": 66070
    },
    {
      "epoch": 8.04e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0226,
      "step": 66080
    },
    {
      "epoch": 8.045e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0362,
      "step": 66090
    },
    {
      "epoch": 8.05e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0285,
      "step": 66100
    },
    {
      "epoch": 8.055e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0267,
      "step": 66110
    },
    {
      "epoch": 8.06e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "step": 66120
    },
    {
      "epoch": 8.065e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0164,
      "step": 66130
    },
    {
      "epoch": 8.07e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0595,
      "step": 66140
    },
    {
      "epoch": 8.075e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "step": 66150
    },
    {
      "epoch": 8.08e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0647,
      "step": 66160
    },
    {
      "epoch": 8.085e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "step": 66170
    },
    {
      "epoch": 8.09e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "step": 66180
    },
    {
      "epoch": 8.095e-06,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "step": 66190
    },
    {
      "epoch": 8.1e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0318,
      "step": 66200
    },
    {
      "epoch": 8.105e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "step": 66210
    },
    {
      "epoch": 8.11e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "step": 66220
    },
    {
      "epoch": 8.115e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0354,
      "step": 66230
    },
    {
      "epoch": 8.12e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "step": 66240
    },
    {
      "epoch": 8.125e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "step": 66250
    },
    {
      "epoch": 8.13e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0317,
      "step": 66260
    },
    {
      "epoch": 8.135e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0437,
      "step": 66270
    },
    {
      "epoch": 8.14e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0361,
      "step": 66280
    },
    {
      "epoch": 8.145e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "step": 66290
    },
    {
      "epoch": 8.15e-06,
      "grad_norm": 0.53125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "step": 66300
    },
    {
      "epoch": 8.155e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "step": 66310
    },
    {
      "epoch": 8.16e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0546,
      "step": 66320
    },
    {
      "epoch": 8.165e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0254,
      "step": 66330
    },
    {
      "epoch": 8.17e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "step": 66340
    },
    {
      "epoch": 8.175e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "step": 66350
    },
    {
      "epoch": 8.18e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0448,
      "step": 66360
    },
    {
      "epoch": 8.185e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0486,
      "step": 66370
    },
    {
      "epoch": 8.19e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0656,
      "step": 66380
    },
    {
      "epoch": 8.195e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "step": 66390
    },
    {
      "epoch": 8.2e-06,
      "grad_norm": 0.55078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0541,
      "step": 66400
    },
    {
      "epoch": 8.205e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "step": 66410
    },
    {
      "epoch": 8.21e-06,
      "grad_norm": 0.53125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0438,
      "step": 66420
    },
    {
      "epoch": 8.215e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0318,
      "step": 66430
    },
    {
      "epoch": 8.22e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0375,
      "step": 66440
    },
    {
      "epoch": 8.225e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0149,
      "step": 66450
    },
    {
      "epoch": 8.23e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "step": 66460
    },
    {
      "epoch": 8.235e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0313,
      "step": 66470
    },
    {
      "epoch": 8.24e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "step": 66480
    },
    {
      "epoch": 8.245e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0351,
      "step": 66490
    },
    {
      "epoch": 8.25e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.091,
      "step": 66500
    },
    {
      "epoch": 8.25e-06,
      "eval_loss": 3.0829520225524902,
      "eval_runtime": 0.0582,
      "eval_samples_per_second": 549.386,
      "eval_steps_per_second": 17.168,
      "step": 66500
    },
    {
      "epoch": 8.255e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0257,
      "step": 66510
    },
    {
      "epoch": 8.26e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "step": 66520
    },
    {
      "epoch": 8.265e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0337,
      "step": 66530
    },
    {
      "epoch": 8.27e-06,
      "grad_norm": 0.578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0175,
      "step": 66540
    },
    {
      "epoch": 8.275e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "step": 66550
    },
    {
      "epoch": 8.28e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "step": 66560
    },
    {
      "epoch": 8.285e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0479,
      "step": 66570
    },
    {
      "epoch": 8.29e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0206,
      "step": 66580
    },
    {
      "epoch": 8.295e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "step": 66590
    },
    {
      "epoch": 8.3e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "step": 66600
    },
    {
      "epoch": 8.305e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0513,
      "step": 66610
    },
    {
      "epoch": 8.31e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0018,
      "step": 66620
    },
    {
      "epoch": 8.315e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0578,
      "step": 66630
    },
    {
      "epoch": 8.32e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "step": 66640
    },
    {
      "epoch": 8.325e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "step": 66650
    },
    {
      "epoch": 8.33e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "step": 66660
    },
    {
      "epoch": 8.335e-06,
      "grad_norm": 0.6015625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.018,
      "step": 66670
    },
    {
      "epoch": 8.34e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0815,
      "step": 66680
    },
    {
      "epoch": 8.345e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0623,
      "step": 66690
    },
    {
      "epoch": 8.35e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0369,
      "step": 66700
    },
    {
      "epoch": 8.355e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0238,
      "step": 66710
    },
    {
      "epoch": 8.36e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0206,
      "step": 66720
    },
    {
      "epoch": 8.365e-06,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "step": 66730
    },
    {
      "epoch": 8.37e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0436,
      "step": 66740
    },
    {
      "epoch": 8.375e-06,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0285,
      "step": 66750
    },
    {
      "epoch": 8.38e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "step": 66760
    },
    {
      "epoch": 8.385e-06,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "step": 66770
    },
    {
      "epoch": 8.39e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0163,
      "step": 66780
    },
    {
      "epoch": 8.395e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "step": 66790
    },
    {
      "epoch": 8.4e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "step": 66800
    },
    {
      "epoch": 8.405e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "step": 66810
    },
    {
      "epoch": 8.41e-06,
      "grad_norm": 0.54296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.028,
      "step": 66820
    },
    {
      "epoch": 8.415e-06,
      "grad_norm": 0.546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "step": 66830
    },
    {
      "epoch": 8.42e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "step": 66840
    },
    {
      "epoch": 8.425e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "step": 66850
    },
    {
      "epoch": 8.43e-06,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0303,
      "step": 66860
    },
    {
      "epoch": 8.435e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0315,
      "step": 66870
    },
    {
      "epoch": 8.44e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0364,
      "step": 66880
    },
    {
      "epoch": 8.445e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "step": 66890
    },
    {
      "epoch": 8.45e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0219,
      "step": 66900
    },
    {
      "epoch": 8.455e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0278,
      "step": 66910
    },
    {
      "epoch": 8.46e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0317,
      "step": 66920
    },
    {
      "epoch": 8.465e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "step": 66930
    },
    {
      "epoch": 8.47e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "step": 66940
    },
    {
      "epoch": 8.475e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "step": 66950
    },
    {
      "epoch": 8.48e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "step": 66960
    },
    {
      "epoch": 8.485e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0645,
      "step": 66970
    },
    {
      "epoch": 8.49e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0365,
      "step": 66980
    },
    {
      "epoch": 8.495e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "step": 66990
    },
    {
      "epoch": 8.5e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "step": 67000
    },
    {
      "epoch": 8.5e-06,
      "eval_loss": 3.0823678970336914,
      "eval_runtime": 0.0605,
      "eval_samples_per_second": 528.633,
      "eval_steps_per_second": 16.52,
      "step": 67000
    },
    {
      "epoch": 8.505e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "step": 67010
    },
    {
      "epoch": 8.51e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0319,
      "step": 67020
    },
    {
      "epoch": 8.515e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0308,
      "step": 67030
    },
    {
      "epoch": 8.52e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0466,
      "step": 67040
    },
    {
      "epoch": 8.525e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0416,
      "step": 67050
    },
    {
      "epoch": 8.53e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0737,
      "step": 67060
    },
    {
      "epoch": 8.535e-06,
      "grad_norm": 0.5625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "step": 67070
    },
    {
      "epoch": 8.54e-06,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "step": 67080
    },
    {
      "epoch": 8.545e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "step": 67090
    },
    {
      "epoch": 8.55e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "step": 67100
    },
    {
      "epoch": 8.555e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0331,
      "step": 67110
    },
    {
      "epoch": 8.56e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0373,
      "step": 67120
    },
    {
      "epoch": 8.565e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "step": 67130
    },
    {
      "epoch": 8.57e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "step": 67140
    },
    {
      "epoch": 8.575e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "step": 67150
    },
    {
      "epoch": 8.58e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "step": 67160
    },
    {
      "epoch": 8.585e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "step": 67170
    },
    {
      "epoch": 8.59e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0203,
      "step": 67180
    },
    {
      "epoch": 8.595e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.025,
      "step": 67190
    },
    {
      "epoch": 8.6e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0311,
      "step": 67200
    },
    {
      "epoch": 8.605e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "step": 67210
    },
    {
      "epoch": 8.61e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0429,
      "step": 67220
    },
    {
      "epoch": 8.615e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0174,
      "step": 67230
    },
    {
      "epoch": 8.62e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "step": 67240
    },
    {
      "epoch": 8.625e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "step": 67250
    },
    {
      "epoch": 8.63e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "step": 67260
    },
    {
      "epoch": 8.635e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0345,
      "step": 67270
    },
    {
      "epoch": 8.64e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "step": 67280
    },
    {
      "epoch": 8.645e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0441,
      "step": 67290
    },
    {
      "epoch": 8.65e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0313,
      "step": 67300
    },
    {
      "epoch": 8.655e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "step": 67310
    },
    {
      "epoch": 8.66e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0681,
      "step": 67320
    },
    {
      "epoch": 8.665e-06,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0329,
      "step": 67330
    },
    {
      "epoch": 8.67e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "step": 67340
    },
    {
      "epoch": 8.675e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0357,
      "step": 67350
    },
    {
      "epoch": 8.68e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.048,
      "step": 67360
    },
    {
      "epoch": 8.685e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "step": 67370
    },
    {
      "epoch": 8.69e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0405,
      "step": 67380
    },
    {
      "epoch": 8.695e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "step": 67390
    },
    {
      "epoch": 8.7e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0219,
      "step": 67400
    },
    {
      "epoch": 8.705e-06,
      "grad_norm": 0.53125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "step": 67410
    },
    {
      "epoch": 8.71e-06,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0391,
      "step": 67420
    },
    {
      "epoch": 8.715e-06,
      "grad_norm": 0.5390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0182,
      "step": 67430
    },
    {
      "epoch": 8.72e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "step": 67440
    },
    {
      "epoch": 8.725e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "step": 67450
    },
    {
      "epoch": 8.73e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "step": 67460
    },
    {
      "epoch": 8.735e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "step": 67470
    },
    {
      "epoch": 8.74e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0491,
      "step": 67480
    },
    {
      "epoch": 8.745e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "step": 67490
    },
    {
      "epoch": 8.75e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "step": 67500
    },
    {
      "epoch": 8.75e-06,
      "eval_loss": 3.081859588623047,
      "eval_runtime": 0.0608,
      "eval_samples_per_second": 526.317,
      "eval_steps_per_second": 16.447,
      "step": 67500
    },
    {
      "epoch": 8.755e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "step": 67510
    },
    {
      "epoch": 8.76e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "step": 67520
    },
    {
      "epoch": 8.765e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "step": 67530
    },
    {
      "epoch": 8.77e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "step": 67540
    },
    {
      "epoch": 8.775e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0293,
      "step": 67550
    },
    {
      "epoch": 8.78e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0321,
      "step": 67560
    },
    {
      "epoch": 8.785e-06,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0391,
      "step": 67570
    },
    {
      "epoch": 8.79e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "step": 67580
    },
    {
      "epoch": 8.795e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "step": 67590
    },
    {
      "epoch": 8.8e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "step": 67600
    },
    {
      "epoch": 8.805e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0248,
      "step": 67610
    },
    {
      "epoch": 8.81e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0079,
      "step": 67620
    },
    {
      "epoch": 8.815e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "step": 67630
    },
    {
      "epoch": 8.82e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0348,
      "step": 67640
    },
    {
      "epoch": 8.825e-06,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.013,
      "step": 67650
    },
    {
      "epoch": 8.83e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "step": 67660
    },
    {
      "epoch": 8.835e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0416,
      "step": 67670
    },
    {
      "epoch": 8.84e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0634,
      "step": 67680
    },
    {
      "epoch": 8.845e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "step": 67690
    },
    {
      "epoch": 8.85e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.026,
      "step": 67700
    },
    {
      "epoch": 8.855e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.03,
      "step": 67710
    },
    {
      "epoch": 8.86e-06,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "step": 67720
    },
    {
      "epoch": 8.865e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0614,
      "step": 67730
    },
    {
      "epoch": 8.87e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0263,
      "step": 67740
    },
    {
      "epoch": 8.875e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0236,
      "step": 67750
    },
    {
      "epoch": 8.88e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "step": 67760
    },
    {
      "epoch": 8.885e-06,
      "grad_norm": 0.52734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "step": 67770
    },
    {
      "epoch": 8.89e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0455,
      "step": 67780
    },
    {
      "epoch": 8.895e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "step": 67790
    },
    {
      "epoch": 8.9e-06,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0155,
      "step": 67800
    },
    {
      "epoch": 8.905e-06,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0315,
      "step": 67810
    },
    {
      "epoch": 8.91e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "step": 67820
    },
    {
      "epoch": 8.915e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "step": 67830
    },
    {
      "epoch": 8.92e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "step": 67840
    },
    {
      "epoch": 8.925e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "step": 67850
    },
    {
      "epoch": 8.93e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0103,
      "step": 67860
    },
    {
      "epoch": 8.935e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0529,
      "step": 67870
    },
    {
      "epoch": 8.94e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "step": 67880
    },
    {
      "epoch": 8.945e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.027,
      "step": 67890
    },
    {
      "epoch": 8.95e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "step": 67900
    },
    {
      "epoch": 8.955e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "step": 67910
    },
    {
      "epoch": 8.96e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0308,
      "step": 67920
    },
    {
      "epoch": 8.965e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0558,
      "step": 67930
    },
    {
      "epoch": 8.97e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "step": 67940
    },
    {
      "epoch": 8.975e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "step": 67950
    },
    {
      "epoch": 8.98e-06,
      "grad_norm": 0.58203125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "step": 67960
    },
    {
      "epoch": 8.985e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0355,
      "step": 67970
    },
    {
      "epoch": 8.99e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "step": 67980
    },
    {
      "epoch": 8.995e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "step": 67990
    },
    {
      "epoch": 9e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "step": 68000
    },
    {
      "epoch": 9e-06,
      "eval_loss": 3.0828726291656494,
      "eval_runtime": 0.0615,
      "eval_samples_per_second": 520.008,
      "eval_steps_per_second": 16.25,
      "step": 68000
    },
    {
      "epoch": 9.005e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "step": 68010
    },
    {
      "epoch": 9.01e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "step": 68020
    },
    {
      "epoch": 9.015e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "step": 68030
    },
    {
      "epoch": 9.02e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0288,
      "step": 68040
    },
    {
      "epoch": 9.025e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "step": 68050
    },
    {
      "epoch": 9.03e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "step": 68060
    },
    {
      "epoch": 9.035e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "step": 68070
    },
    {
      "epoch": 9.04e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "step": 68080
    },
    {
      "epoch": 9.045e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0259,
      "step": 68090
    },
    {
      "epoch": 9.05e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "step": 68100
    },
    {
      "epoch": 9.055e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.031,
      "step": 68110
    },
    {
      "epoch": 9.06e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "step": 68120
    },
    {
      "epoch": 9.065e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "step": 68130
    },
    {
      "epoch": 9.07e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "step": 68140
    },
    {
      "epoch": 9.075e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.039,
      "step": 68150
    },
    {
      "epoch": 9.08e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "step": 68160
    },
    {
      "epoch": 9.085e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0182,
      "step": 68170
    },
    {
      "epoch": 9.09e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0604,
      "step": 68180
    },
    {
      "epoch": 9.095e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0318,
      "step": 68190
    },
    {
      "epoch": 9.1e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "step": 68200
    },
    {
      "epoch": 9.105e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "step": 68210
    },
    {
      "epoch": 9.11e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "step": 68220
    },
    {
      "epoch": 9.115e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "step": 68230
    },
    {
      "epoch": 9.12e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "step": 68240
    },
    {
      "epoch": 9.125e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.059,
      "step": 68250
    },
    {
      "epoch": 9.13e-06,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0156,
      "step": 68260
    },
    {
      "epoch": 9.135e-06,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "step": 68270
    },
    {
      "epoch": 9.14e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0382,
      "step": 68280
    },
    {
      "epoch": 9.145e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0143,
      "step": 68290
    },
    {
      "epoch": 9.15e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0337,
      "step": 68300
    },
    {
      "epoch": 9.155e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0244,
      "step": 68310
    },
    {
      "epoch": 9.16e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0256,
      "step": 68320
    },
    {
      "epoch": 9.165e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0293,
      "step": 68330
    },
    {
      "epoch": 9.17e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0486,
      "step": 68340
    },
    {
      "epoch": 9.175e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0348,
      "step": 68350
    },
    {
      "epoch": 9.18e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "step": 68360
    },
    {
      "epoch": 9.185e-06,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "step": 68370
    },
    {
      "epoch": 9.19e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "step": 68380
    },
    {
      "epoch": 9.195e-06,
      "grad_norm": 0.55078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0605,
      "step": 68390
    },
    {
      "epoch": 9.2e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.06,
      "step": 68400
    },
    {
      "epoch": 9.205e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "step": 68410
    },
    {
      "epoch": 9.21e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0176,
      "step": 68420
    },
    {
      "epoch": 9.215e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0582,
      "step": 68430
    },
    {
      "epoch": 9.22e-06,
      "grad_norm": 0.59375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0499,
      "step": 68440
    },
    {
      "epoch": 9.225e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "step": 68450
    },
    {
      "epoch": 9.23e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 68460
    },
    {
      "epoch": 9.235e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0428,
      "step": 68470
    },
    {
      "epoch": 9.24e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0526,
      "step": 68480
    },
    {
      "epoch": 9.245e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0117,
      "step": 68490
    },
    {
      "epoch": 9.25e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0232,
      "step": 68500
    },
    {
      "epoch": 9.25e-06,
      "eval_loss": 3.0828802585601807,
      "eval_runtime": 0.0653,
      "eval_samples_per_second": 490.316,
      "eval_steps_per_second": 15.322,
      "step": 68500
    },
    {
      "epoch": 9.255e-06,
      "grad_norm": 0.671875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0604,
      "step": 68510
    },
    {
      "epoch": 9.26e-06,
      "grad_norm": 0.546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0111,
      "step": 68520
    },
    {
      "epoch": 9.265e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "step": 68530
    },
    {
      "epoch": 9.27e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "step": 68540
    },
    {
      "epoch": 9.275e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.028,
      "step": 68550
    },
    {
      "epoch": 9.28e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "step": 68560
    },
    {
      "epoch": 9.285e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "step": 68570
    },
    {
      "epoch": 9.29e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "step": 68580
    },
    {
      "epoch": 9.295e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0312,
      "step": 68590
    },
    {
      "epoch": 9.3e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0238,
      "step": 68600
    },
    {
      "epoch": 9.305e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "step": 68610
    },
    {
      "epoch": 9.31e-06,
      "grad_norm": 0.5703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0625,
      "step": 68620
    },
    {
      "epoch": 9.315e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "step": 68630
    },
    {
      "epoch": 9.32e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0311,
      "step": 68640
    },
    {
      "epoch": 9.325e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "step": 68650
    },
    {
      "epoch": 9.33e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0264,
      "step": 68660
    },
    {
      "epoch": 9.335e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0273,
      "step": 68670
    },
    {
      "epoch": 9.34e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "step": 68680
    },
    {
      "epoch": 9.345e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0353,
      "step": 68690
    },
    {
      "epoch": 9.35e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0108,
      "step": 68700
    },
    {
      "epoch": 9.355e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0369,
      "step": 68710
    },
    {
      "epoch": 9.36e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "step": 68720
    },
    {
      "epoch": 9.365e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "step": 68730
    },
    {
      "epoch": 9.37e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "step": 68740
    },
    {
      "epoch": 9.375e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "step": 68750
    },
    {
      "epoch": 9.38e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "step": 68760
    },
    {
      "epoch": 9.385e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0276,
      "step": 68770
    },
    {
      "epoch": 9.39e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "step": 68780
    },
    {
      "epoch": 9.395e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "step": 68790
    },
    {
      "epoch": 9.4e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "step": 68800
    },
    {
      "epoch": 9.405e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0275,
      "step": 68810
    },
    {
      "epoch": 9.41e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "step": 68820
    },
    {
      "epoch": 9.415e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0314,
      "step": 68830
    },
    {
      "epoch": 9.42e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0273,
      "step": 68840
    },
    {
      "epoch": 9.425e-06,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.018,
      "step": 68850
    },
    {
      "epoch": 9.43e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0344,
      "step": 68860
    },
    {
      "epoch": 9.435e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0649,
      "step": 68870
    },
    {
      "epoch": 9.44e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "step": 68880
    },
    {
      "epoch": 9.445e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "step": 68890
    },
    {
      "epoch": 9.45e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "step": 68900
    },
    {
      "epoch": 9.455e-06,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0686,
      "step": 68910
    },
    {
      "epoch": 9.46e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0588,
      "step": 68920
    },
    {
      "epoch": 9.465e-06,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "step": 68930
    },
    {
      "epoch": 9.47e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "step": 68940
    },
    {
      "epoch": 9.475e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0368,
      "step": 68950
    },
    {
      "epoch": 9.48e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0272,
      "step": 68960
    },
    {
      "epoch": 9.485e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "step": 68970
    },
    {
      "epoch": 9.49e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0642,
      "step": 68980
    },
    {
      "epoch": 9.495e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "step": 68990
    },
    {
      "epoch": 9.5e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0262,
      "step": 69000
    },
    {
      "epoch": 9.5e-06,
      "eval_loss": 3.0823965072631836,
      "eval_runtime": 0.0619,
      "eval_samples_per_second": 517.085,
      "eval_steps_per_second": 16.159,
      "step": 69000
    },
    {
      "epoch": 9.505e-06,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "step": 69010
    },
    {
      "epoch": 9.51e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0714,
      "step": 69020
    },
    {
      "epoch": 9.515e-06,
      "grad_norm": 0.427734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9955,
      "step": 69030
    },
    {
      "epoch": 9.52e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0481,
      "step": 69040
    },
    {
      "epoch": 9.525e-06,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0602,
      "step": 69050
    },
    {
      "epoch": 9.53e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0226,
      "step": 69060
    },
    {
      "epoch": 9.535e-06,
      "grad_norm": 0.5546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "step": 69070
    },
    {
      "epoch": 9.54e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0322,
      "step": 69080
    },
    {
      "epoch": 9.545e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "step": 69090
    },
    {
      "epoch": 9.55e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0512,
      "step": 69100
    },
    {
      "epoch": 9.555e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.049,
      "step": 69110
    },
    {
      "epoch": 9.56e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0239,
      "step": 69120
    },
    {
      "epoch": 9.565e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "step": 69130
    },
    {
      "epoch": 9.57e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0255,
      "step": 69140
    },
    {
      "epoch": 9.575e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0192,
      "step": 69150
    },
    {
      "epoch": 9.58e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "step": 69160
    },
    {
      "epoch": 9.585e-06,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "step": 69170
    },
    {
      "epoch": 9.59e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "step": 69180
    },
    {
      "epoch": 9.595e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "step": 69190
    },
    {
      "epoch": 9.6e-06,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0503,
      "step": 69200
    },
    {
      "epoch": 9.605e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "step": 69210
    },
    {
      "epoch": 9.61e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "step": 69220
    },
    {
      "epoch": 9.615e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0598,
      "step": 69230
    },
    {
      "epoch": 9.62e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "step": 69240
    },
    {
      "epoch": 9.625e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "step": 69250
    },
    {
      "epoch": 9.63e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "step": 69260
    },
    {
      "epoch": 9.635e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0351,
      "step": 69270
    },
    {
      "epoch": 9.64e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0561,
      "step": 69280
    },
    {
      "epoch": 9.645e-06,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0493,
      "step": 69290
    },
    {
      "epoch": 9.65e-06,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.026,
      "step": 69300
    },
    {
      "epoch": 9.655e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0658,
      "step": 69310
    },
    {
      "epoch": 9.66e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0109,
      "step": 69320
    },
    {
      "epoch": 9.665e-06,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0594,
      "step": 69330
    },
    {
      "epoch": 9.67e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0377,
      "step": 69340
    },
    {
      "epoch": 9.675e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "step": 69350
    },
    {
      "epoch": 9.68e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.035,
      "step": 69360
    },
    {
      "epoch": 9.685e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "step": 69370
    },
    {
      "epoch": 9.69e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "step": 69380
    },
    {
      "epoch": 9.695e-06,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0558,
      "step": 69390
    },
    {
      "epoch": 9.7e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "step": 69400
    },
    {
      "epoch": 9.705e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "step": 69410
    },
    {
      "epoch": 9.71e-06,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0493,
      "step": 69420
    },
    {
      "epoch": 9.715e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "step": 69430
    },
    {
      "epoch": 9.72e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.039,
      "step": 69440
    },
    {
      "epoch": 9.725e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0187,
      "step": 69450
    },
    {
      "epoch": 9.73e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0177,
      "step": 69460
    },
    {
      "epoch": 9.735e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0063,
      "step": 69470
    },
    {
      "epoch": 9.74e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "step": 69480
    },
    {
      "epoch": 9.745e-06,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "step": 69490
    },
    {
      "epoch": 9.75e-06,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0777,
      "step": 69500
    },
    {
      "epoch": 9.75e-06,
      "eval_loss": 3.082871913909912,
      "eval_runtime": 0.0636,
      "eval_samples_per_second": 503.384,
      "eval_steps_per_second": 15.731,
      "step": 69500
    },
    {
      "epoch": 9.755e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "step": 69510
    },
    {
      "epoch": 9.76e-06,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0286,
      "step": 69520
    },
    {
      "epoch": 9.765e-06,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0301,
      "step": 69530
    },
    {
      "epoch": 9.77e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "step": 69540
    },
    {
      "epoch": 9.775e-06,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0646,
      "step": 69550
    },
    {
      "epoch": 9.78e-06,
      "grad_norm": 0.546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0599,
      "step": 69560
    },
    {
      "epoch": 9.785e-06,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "step": 69570
    },
    {
      "epoch": 9.79e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0307,
      "step": 69580
    },
    {
      "epoch": 9.795e-06,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0187,
      "step": 69590
    },
    {
      "epoch": 9.8e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0622,
      "step": 69600
    },
    {
      "epoch": 9.805e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0186,
      "step": 69610
    },
    {
      "epoch": 9.81e-06,
      "grad_norm": 0.51953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "step": 69620
    },
    {
      "epoch": 9.815e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0355,
      "step": 69630
    },
    {
      "epoch": 9.82e-06,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0386,
      "step": 69640
    },
    {
      "epoch": 9.825e-06,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.048,
      "step": 69650
    },
    {
      "epoch": 9.83e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0209,
      "step": 69660
    },
    {
      "epoch": 9.835e-06,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0096,
      "step": 69670
    },
    {
      "epoch": 9.84e-06,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0265,
      "step": 69680
    },
    {
      "epoch": 9.845e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0422,
      "step": 69690
    },
    {
      "epoch": 9.85e-06,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0192,
      "step": 69700
    },
    {
      "epoch": 9.855e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0285,
      "step": 69710
    },
    {
      "epoch": 9.86e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "step": 69720
    },
    {
      "epoch": 9.865e-06,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0217,
      "step": 69730
    },
    {
      "epoch": 9.87e-06,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0344,
      "step": 69740
    },
    {
      "epoch": 9.875e-06,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0269,
      "step": 69750
    },
    {
      "epoch": 9.88e-06,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0206,
      "step": 69760
    },
    {
      "epoch": 9.885e-06,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.035,
      "step": 69770
    },
    {
      "epoch": 9.89e-06,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.027,
      "step": 69780
    },
    {
      "epoch": 9.895e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0153,
      "step": 69790
    },
    {
      "epoch": 9.9e-06,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0281,
      "step": 69800
    },
    {
      "epoch": 9.905e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0556,
      "step": 69810
    },
    {
      "epoch": 9.91e-06,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "step": 69820
    },
    {
      "epoch": 9.915e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0744,
      "step": 69830
    },
    {
      "epoch": 9.92e-06,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0246,
      "step": 69840
    },
    {
      "epoch": 9.925e-06,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0199,
      "step": 69850
    },
    {
      "epoch": 9.93e-06,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "step": 69860
    },
    {
      "epoch": 9.935e-06,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0239,
      "step": 69870
    },
    {
      "epoch": 9.94e-06,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "step": 69880
    },
    {
      "epoch": 9.945e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0183,
      "step": 69890
    },
    {
      "epoch": 9.95e-06,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0489,
      "step": 69900
    },
    {
      "epoch": 9.955e-06,
      "grad_norm": 0.5546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0386,
      "step": 69910
    },
    {
      "epoch": 9.96e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "step": 69920
    },
    {
      "epoch": 9.965e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0745,
      "step": 69930
    },
    {
      "epoch": 9.97e-06,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0528,
      "step": 69940
    },
    {
      "epoch": 9.975e-06,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.063,
      "step": 69950
    },
    {
      "epoch": 9.98e-06,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "step": 69960
    },
    {
      "epoch": 9.985e-06,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "step": 69970
    },
    {
      "epoch": 9.99e-06,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "step": 69980
    },
    {
      "epoch": 9.995e-06,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.017,
      "step": 69990
    },
    {
      "epoch": 1e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "step": 70000
    },
    {
      "epoch": 1e-05,
      "eval_loss": 3.082289218902588,
      "eval_runtime": 0.0664,
      "eval_samples_per_second": 481.893,
      "eval_steps_per_second": 15.059,
      "step": 70000
    },
    {
      "epoch": 1.0005e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "step": 70010
    },
    {
      "epoch": 1.001e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.052,
      "step": 70020
    },
    {
      "epoch": 1.0015e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0338,
      "step": 70030
    },
    {
      "epoch": 1.002e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "step": 70040
    },
    {
      "epoch": 1.0025e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0211,
      "step": 70050
    },
    {
      "epoch": 1.003e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0553,
      "step": 70060
    },
    {
      "epoch": 1.0035e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "step": 70070
    },
    {
      "epoch": 1.004e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "step": 70080
    },
    {
      "epoch": 1.0045e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "step": 70090
    },
    {
      "epoch": 1.005e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "step": 70100
    },
    {
      "epoch": 1.0055e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "step": 70110
    },
    {
      "epoch": 1.006e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0717,
      "step": 70120
    },
    {
      "epoch": 1.0065e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0667,
      "step": 70130
    },
    {
      "epoch": 1.007e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.024,
      "step": 70140
    },
    {
      "epoch": 1.0075e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "step": 70150
    },
    {
      "epoch": 1.008e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "step": 70160
    },
    {
      "epoch": 1.0085e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "step": 70170
    },
    {
      "epoch": 1.009e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 70180
    },
    {
      "epoch": 1.0095e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0725,
      "step": 70190
    },
    {
      "epoch": 1.01e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "step": 70200
    },
    {
      "epoch": 1.0105e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0362,
      "step": 70210
    },
    {
      "epoch": 1.011e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0684,
      "step": 70220
    },
    {
      "epoch": 1.0115e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0529,
      "step": 70230
    },
    {
      "epoch": 1.012e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "step": 70240
    },
    {
      "epoch": 1.0125e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "step": 70250
    },
    {
      "epoch": 1.013e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0127,
      "step": 70260
    },
    {
      "epoch": 1.0135e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0529,
      "step": 70270
    },
    {
      "epoch": 1.014e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.029,
      "step": 70280
    },
    {
      "epoch": 1.0145e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0437,
      "step": 70290
    },
    {
      "epoch": 1.015e-05,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "step": 70300
    },
    {
      "epoch": 1.0155e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0292,
      "step": 70310
    },
    {
      "epoch": 1.016e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0249,
      "step": 70320
    },
    {
      "epoch": 1.0165e-05,
      "grad_norm": 0.53125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0257,
      "step": 70330
    },
    {
      "epoch": 1.017e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0175,
      "step": 70340
    },
    {
      "epoch": 1.0175e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.014,
      "step": 70350
    },
    {
      "epoch": 1.018e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0289,
      "step": 70360
    },
    {
      "epoch": 1.0185e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "step": 70370
    },
    {
      "epoch": 1.019e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0222,
      "step": 70380
    },
    {
      "epoch": 1.0195e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0429,
      "step": 70390
    },
    {
      "epoch": 1.02e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0319,
      "step": 70400
    },
    {
      "epoch": 1.0205e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0483,
      "step": 70410
    },
    {
      "epoch": 1.021e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "step": 70420
    },
    {
      "epoch": 1.0215e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "step": 70430
    },
    {
      "epoch": 1.022e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0232,
      "step": 70440
    },
    {
      "epoch": 1.0225e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0286,
      "step": 70450
    },
    {
      "epoch": 1.023e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0179,
      "step": 70460
    },
    {
      "epoch": 1.0235e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0266,
      "step": 70470
    },
    {
      "epoch": 1.024e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "step": 70480
    },
    {
      "epoch": 1.0245e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "step": 70490
    },
    {
      "epoch": 1.025e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0737,
      "step": 70500
    },
    {
      "epoch": 1.025e-05,
      "eval_loss": 3.0824360847473145,
      "eval_runtime": 0.0628,
      "eval_samples_per_second": 509.541,
      "eval_steps_per_second": 15.923,
      "step": 70500
    },
    {
      "epoch": 1.0255e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "step": 70510
    },
    {
      "epoch": 1.026e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "step": 70520
    },
    {
      "epoch": 1.0265e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0538,
      "step": 70530
    },
    {
      "epoch": 1.027e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "step": 70540
    },
    {
      "epoch": 1.0275e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "step": 70550
    },
    {
      "epoch": 1.028e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "step": 70560
    },
    {
      "epoch": 1.0285e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "step": 70570
    },
    {
      "epoch": 1.029e-05,
      "grad_norm": 0.52734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0278,
      "step": 70580
    },
    {
      "epoch": 1.0295e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0416,
      "step": 70590
    },
    {
      "epoch": 1.03e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "step": 70600
    },
    {
      "epoch": 1.0305e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0511,
      "step": 70610
    },
    {
      "epoch": 1.031e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0314,
      "step": 70620
    },
    {
      "epoch": 1.0315e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0173,
      "step": 70630
    },
    {
      "epoch": 1.032e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "step": 70640
    },
    {
      "epoch": 1.0325e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0591,
      "step": 70650
    },
    {
      "epoch": 1.033e-05,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "step": 70660
    },
    {
      "epoch": 1.0335e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0187,
      "step": 70670
    },
    {
      "epoch": 1.034e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "step": 70680
    },
    {
      "epoch": 1.0345e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "step": 70690
    },
    {
      "epoch": 1.035e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "step": 70700
    },
    {
      "epoch": 1.0355e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0513,
      "step": 70710
    },
    {
      "epoch": 1.036e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "step": 70720
    },
    {
      "epoch": 1.0365e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0247,
      "step": 70730
    },
    {
      "epoch": 1.037e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "step": 70740
    },
    {
      "epoch": 1.0375e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "step": 70750
    },
    {
      "epoch": 1.038e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "step": 70760
    },
    {
      "epoch": 1.0385e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "step": 70770
    },
    {
      "epoch": 1.039e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "step": 70780
    },
    {
      "epoch": 1.0395e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0362,
      "step": 70790
    },
    {
      "epoch": 1.04e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "step": 70800
    },
    {
      "epoch": 1.0405e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "step": 70810
    },
    {
      "epoch": 1.041e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "step": 70820
    },
    {
      "epoch": 1.0415e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0644,
      "step": 70830
    },
    {
      "epoch": 1.042e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "step": 70840
    },
    {
      "epoch": 1.0425e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "step": 70850
    },
    {
      "epoch": 1.043e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "step": 70860
    },
    {
      "epoch": 1.0435e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0261,
      "step": 70870
    },
    {
      "epoch": 1.044e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0362,
      "step": 70880
    },
    {
      "epoch": 1.0445e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0479,
      "step": 70890
    },
    {
      "epoch": 1.045e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0676,
      "step": 70900
    },
    {
      "epoch": 1.0455e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "step": 70910
    },
    {
      "epoch": 1.046e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0273,
      "step": 70920
    },
    {
      "epoch": 1.0465e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0262,
      "step": 70930
    },
    {
      "epoch": 1.047e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "step": 70940
    },
    {
      "epoch": 1.0475e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0276,
      "step": 70950
    },
    {
      "epoch": 1.048e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0472,
      "step": 70960
    },
    {
      "epoch": 1.0485e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0555,
      "step": 70970
    },
    {
      "epoch": 1.049e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "step": 70980
    },
    {
      "epoch": 1.0495e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.031,
      "step": 70990
    },
    {
      "epoch": 1.05e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0346,
      "step": 71000
    },
    {
      "epoch": 1.05e-05,
      "eval_loss": 3.0824408531188965,
      "eval_runtime": 0.0609,
      "eval_samples_per_second": 525.734,
      "eval_steps_per_second": 16.429,
      "step": 71000
    },
    {
      "epoch": 1.0505e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "step": 71010
    },
    {
      "epoch": 1.051e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0217,
      "step": 71020
    },
    {
      "epoch": 1.0515e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0226,
      "step": 71030
    },
    {
      "epoch": 1.052e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0276,
      "step": 71040
    },
    {
      "epoch": 1.0525e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0422,
      "step": 71050
    },
    {
      "epoch": 1.053e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "step": 71060
    },
    {
      "epoch": 1.0535e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0086,
      "step": 71070
    },
    {
      "epoch": 1.054e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "step": 71080
    },
    {
      "epoch": 1.0545e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0437,
      "step": 71090
    },
    {
      "epoch": 1.055e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0296,
      "step": 71100
    },
    {
      "epoch": 1.0555e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0241,
      "step": 71110
    },
    {
      "epoch": 1.056e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "step": 71120
    },
    {
      "epoch": 1.0565e-05,
      "grad_norm": 0.578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0946,
      "step": 71130
    },
    {
      "epoch": 1.057e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0643,
      "step": 71140
    },
    {
      "epoch": 1.0575e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0263,
      "step": 71150
    },
    {
      "epoch": 1.058e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "step": 71160
    },
    {
      "epoch": 1.0585e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "step": 71170
    },
    {
      "epoch": 1.059e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0729,
      "step": 71180
    },
    {
      "epoch": 1.0595e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0213,
      "step": 71190
    },
    {
      "epoch": 1.06e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "step": 71200
    },
    {
      "epoch": 1.0605e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0654,
      "step": 71210
    },
    {
      "epoch": 1.061e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0321,
      "step": 71220
    },
    {
      "epoch": 1.0615e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "step": 71230
    },
    {
      "epoch": 1.062e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "step": 71240
    },
    {
      "epoch": 1.0625e-05,
      "grad_norm": 1.1171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "step": 71250
    },
    {
      "epoch": 1.063e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "step": 71260
    },
    {
      "epoch": 1.0635e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.048,
      "step": 71270
    },
    {
      "epoch": 1.064e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "step": 71280
    },
    {
      "epoch": 1.0645e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "step": 71290
    },
    {
      "epoch": 1.065e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0349,
      "step": 71300
    },
    {
      "epoch": 1.0655e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0313,
      "step": 71310
    },
    {
      "epoch": 1.066e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0362,
      "step": 71320
    },
    {
      "epoch": 1.0665e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "step": 71330
    },
    {
      "epoch": 1.067e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0313,
      "step": 71340
    },
    {
      "epoch": 1.0675e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0361,
      "step": 71350
    },
    {
      "epoch": 1.068e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0785,
      "step": 71360
    },
    {
      "epoch": 1.0685e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "step": 71370
    },
    {
      "epoch": 1.069e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0329,
      "step": 71380
    },
    {
      "epoch": 1.0695e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "step": 71390
    },
    {
      "epoch": 1.07e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "step": 71400
    },
    {
      "epoch": 1.0705e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "step": 71410
    },
    {
      "epoch": 1.071e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0416,
      "step": 71420
    },
    {
      "epoch": 1.0715e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "step": 71430
    },
    {
      "epoch": 1.072e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "step": 71440
    },
    {
      "epoch": 1.0725e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0412,
      "step": 71450
    },
    {
      "epoch": 1.073e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "step": 71460
    },
    {
      "epoch": 1.0735e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0472,
      "step": 71470
    },
    {
      "epoch": 1.074e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "step": 71480
    },
    {
      "epoch": 1.0745e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0335,
      "step": 71490
    },
    {
      "epoch": 1.075e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0406,
      "step": 71500
    },
    {
      "epoch": 1.075e-05,
      "eval_loss": 3.0823135375976562,
      "eval_runtime": 0.0663,
      "eval_samples_per_second": 482.799,
      "eval_steps_per_second": 15.087,
      "step": 71500
    },
    {
      "epoch": 1.0755e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "step": 71510
    },
    {
      "epoch": 1.076e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "step": 71520
    },
    {
      "epoch": 1.0765e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0479,
      "step": 71530
    },
    {
      "epoch": 1.077e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "step": 71540
    },
    {
      "epoch": 1.0775e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0145,
      "step": 71550
    },
    {
      "epoch": 1.078e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0208,
      "step": 71560
    },
    {
      "epoch": 1.0785e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0868,
      "step": 71570
    },
    {
      "epoch": 1.079e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "step": 71580
    },
    {
      "epoch": 1.0795e-05,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "step": 71590
    },
    {
      "epoch": 1.08e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "step": 71600
    },
    {
      "epoch": 1.0805e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "step": 71610
    },
    {
      "epoch": 1.081e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0499,
      "step": 71620
    },
    {
      "epoch": 1.0815e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0228,
      "step": 71630
    },
    {
      "epoch": 1.082e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "step": 71640
    },
    {
      "epoch": 1.0825e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "step": 71650
    },
    {
      "epoch": 1.083e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0553,
      "step": 71660
    },
    {
      "epoch": 1.0835e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0119,
      "step": 71670
    },
    {
      "epoch": 1.084e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "step": 71680
    },
    {
      "epoch": 1.0845e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0289,
      "step": 71690
    },
    {
      "epoch": 1.085e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "step": 71700
    },
    {
      "epoch": 1.0855e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0226,
      "step": 71710
    },
    {
      "epoch": 1.086e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "step": 71720
    },
    {
      "epoch": 1.0865e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "step": 71730
    },
    {
      "epoch": 1.087e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "step": 71740
    },
    {
      "epoch": 1.0875e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0305,
      "step": 71750
    },
    {
      "epoch": 1.088e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "step": 71760
    },
    {
      "epoch": 1.0885e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "step": 71770
    },
    {
      "epoch": 1.089e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "step": 71780
    },
    {
      "epoch": 1.0895e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "step": 71790
    },
    {
      "epoch": 1.09e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.058,
      "step": 71800
    },
    {
      "epoch": 1.0905e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.042,
      "step": 71810
    },
    {
      "epoch": 1.091e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "step": 71820
    },
    {
      "epoch": 1.0915e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0437,
      "step": 71830
    },
    {
      "epoch": 1.092e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0346,
      "step": 71840
    },
    {
      "epoch": 1.0925e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "step": 71850
    },
    {
      "epoch": 1.093e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.03,
      "step": 71860
    },
    {
      "epoch": 1.0935e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "step": 71870
    },
    {
      "epoch": 1.094e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "step": 71880
    },
    {
      "epoch": 1.0945e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "step": 71890
    },
    {
      "epoch": 1.095e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0673,
      "step": 71900
    },
    {
      "epoch": 1.0955e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0284,
      "step": 71910
    },
    {
      "epoch": 1.096e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "step": 71920
    },
    {
      "epoch": 1.0965e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "step": 71930
    },
    {
      "epoch": 1.097e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "step": 71940
    },
    {
      "epoch": 1.0975e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "step": 71950
    },
    {
      "epoch": 1.098e-05,
      "grad_norm": 0.53125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0259,
      "step": 71960
    },
    {
      "epoch": 1.0985e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0422,
      "step": 71970
    },
    {
      "epoch": 1.099e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0614,
      "step": 71980
    },
    {
      "epoch": 1.0995e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0503,
      "step": 71990
    },
    {
      "epoch": 1.1e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0657,
      "step": 72000
    },
    {
      "epoch": 1.1e-05,
      "eval_loss": 3.0822079181671143,
      "eval_runtime": 0.074,
      "eval_samples_per_second": 432.148,
      "eval_steps_per_second": 13.505,
      "step": 72000
    },
    {
      "epoch": 1.1005e-05,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "step": 72010
    },
    {
      "epoch": 1.101e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0347,
      "step": 72020
    },
    {
      "epoch": 1.1015e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0406,
      "step": 72030
    },
    {
      "epoch": 1.102e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0267,
      "step": 72040
    },
    {
      "epoch": 1.1025e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0198,
      "step": 72050
    },
    {
      "epoch": 1.103e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.018,
      "step": 72060
    },
    {
      "epoch": 1.1035e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "step": 72070
    },
    {
      "epoch": 1.104e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0301,
      "step": 72080
    },
    {
      "epoch": 1.1045e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0216,
      "step": 72090
    },
    {
      "epoch": 1.105e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "step": 72100
    },
    {
      "epoch": 1.1055e-05,
      "grad_norm": 0.431640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0355,
      "step": 72110
    },
    {
      "epoch": 1.106e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0555,
      "step": 72120
    },
    {
      "epoch": 1.1065e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0357,
      "step": 72130
    },
    {
      "epoch": 1.107e-05,
      "grad_norm": 0.4296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "step": 72140
    },
    {
      "epoch": 1.1075e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0448,
      "step": 72150
    },
    {
      "epoch": 1.108e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "step": 72160
    },
    {
      "epoch": 1.1085e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0598,
      "step": 72170
    },
    {
      "epoch": 1.109e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0529,
      "step": 72180
    },
    {
      "epoch": 1.1095e-05,
      "grad_norm": 0.54296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "step": 72190
    },
    {
      "epoch": 1.11e-05,
      "grad_norm": 0.57421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "step": 72200
    },
    {
      "epoch": 1.1105e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "step": 72210
    },
    {
      "epoch": 1.111e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0328,
      "step": 72220
    },
    {
      "epoch": 1.1115e-05,
      "grad_norm": 0.53125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "step": 72230
    },
    {
      "epoch": 1.112e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "step": 72240
    },
    {
      "epoch": 1.1125e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0205,
      "step": 72250
    },
    {
      "epoch": 1.113e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.028,
      "step": 72260
    },
    {
      "epoch": 1.1135e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "step": 72270
    },
    {
      "epoch": 1.114e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "step": 72280
    },
    {
      "epoch": 1.1145e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "step": 72290
    },
    {
      "epoch": 1.115e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0242,
      "step": 72300
    },
    {
      "epoch": 1.1155e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "step": 72310
    },
    {
      "epoch": 1.116e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0383,
      "step": 72320
    },
    {
      "epoch": 1.1165e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0409,
      "step": 72330
    },
    {
      "epoch": 1.117e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "step": 72340
    },
    {
      "epoch": 1.1175e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "step": 72350
    },
    {
      "epoch": 1.118e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "step": 72360
    },
    {
      "epoch": 1.1185e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "step": 72370
    },
    {
      "epoch": 1.119e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0629,
      "step": 72380
    },
    {
      "epoch": 1.1195e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "step": 72390
    },
    {
      "epoch": 1.12e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0252,
      "step": 72400
    },
    {
      "epoch": 1.1205e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0746,
      "step": 72410
    },
    {
      "epoch": 1.121e-05,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "step": 72420
    },
    {
      "epoch": 1.1215e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "step": 72430
    },
    {
      "epoch": 1.122e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0318,
      "step": 72440
    },
    {
      "epoch": 1.1225e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0637,
      "step": 72450
    },
    {
      "epoch": 1.123e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0245,
      "step": 72460
    },
    {
      "epoch": 1.1235e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "step": 72470
    },
    {
      "epoch": 1.124e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "step": 72480
    },
    {
      "epoch": 1.1245e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0226,
      "step": 72490
    },
    {
      "epoch": 1.125e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "step": 72500
    },
    {
      "epoch": 1.125e-05,
      "eval_loss": 3.082770586013794,
      "eval_runtime": 0.0632,
      "eval_samples_per_second": 505.942,
      "eval_steps_per_second": 15.811,
      "step": 72500
    },
    {
      "epoch": 1.1255e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 72510
    },
    {
      "epoch": 1.126e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0365,
      "step": 72520
    },
    {
      "epoch": 1.1265e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "step": 72530
    },
    {
      "epoch": 1.127e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "step": 72540
    },
    {
      "epoch": 1.1275e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9997,
      "step": 72550
    },
    {
      "epoch": 1.128e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "step": 72560
    },
    {
      "epoch": 1.1285e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "step": 72570
    },
    {
      "epoch": 1.129e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "step": 72580
    },
    {
      "epoch": 1.1295e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "step": 72590
    },
    {
      "epoch": 1.13e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9987,
      "step": 72600
    },
    {
      "epoch": 1.1305e-05,
      "grad_norm": 0.5703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0351,
      "step": 72610
    },
    {
      "epoch": 1.131e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.06,
      "step": 72620
    },
    {
      "epoch": 1.1315e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0353,
      "step": 72630
    },
    {
      "epoch": 1.132e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0218,
      "step": 72640
    },
    {
      "epoch": 1.1325e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "step": 72650
    },
    {
      "epoch": 1.133e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0342,
      "step": 72660
    },
    {
      "epoch": 1.1335e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0567,
      "step": 72670
    },
    {
      "epoch": 1.134e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "step": 72680
    },
    {
      "epoch": 1.1345e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0522,
      "step": 72690
    },
    {
      "epoch": 1.135e-05,
      "grad_norm": 0.60546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0825,
      "step": 72700
    },
    {
      "epoch": 1.1355e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0603,
      "step": 72710
    },
    {
      "epoch": 1.136e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0466,
      "step": 72720
    },
    {
      "epoch": 1.1365e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "step": 72730
    },
    {
      "epoch": 1.137e-05,
      "grad_norm": 0.53515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0546,
      "step": 72740
    },
    {
      "epoch": 1.1375e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0365,
      "step": 72750
    },
    {
      "epoch": 1.138e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0241,
      "step": 72760
    },
    {
      "epoch": 1.1385e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "step": 72770
    },
    {
      "epoch": 1.139e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "step": 72780
    },
    {
      "epoch": 1.1395e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "step": 72790
    },
    {
      "epoch": 1.14e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "step": 72800
    },
    {
      "epoch": 1.1405e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "step": 72810
    },
    {
      "epoch": 1.141e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "step": 72820
    },
    {
      "epoch": 1.1415e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "step": 72830
    },
    {
      "epoch": 1.142e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "step": 72840
    },
    {
      "epoch": 1.1425e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0598,
      "step": 72850
    },
    {
      "epoch": 1.143e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "step": 72860
    },
    {
      "epoch": 1.1435e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0264,
      "step": 72870
    },
    {
      "epoch": 1.144e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0398,
      "step": 72880
    },
    {
      "epoch": 1.1445e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0148,
      "step": 72890
    },
    {
      "epoch": 1.145e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0753,
      "step": 72900
    },
    {
      "epoch": 1.1455e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "step": 72910
    },
    {
      "epoch": 1.146e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "step": 72920
    },
    {
      "epoch": 1.1465e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "step": 72930
    },
    {
      "epoch": 1.147e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0483,
      "step": 72940
    },
    {
      "epoch": 1.1475e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0269,
      "step": 72950
    },
    {
      "epoch": 1.148e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0549,
      "step": 72960
    },
    {
      "epoch": 1.1485e-05,
      "grad_norm": 0.52734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0286,
      "step": 72970
    },
    {
      "epoch": 1.149e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "step": 72980
    },
    {
      "epoch": 1.1495e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0124,
      "step": 72990
    },
    {
      "epoch": 1.15e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0497,
      "step": 73000
    },
    {
      "epoch": 1.15e-05,
      "eval_loss": 3.0831046104431152,
      "eval_runtime": 0.0619,
      "eval_samples_per_second": 516.922,
      "eval_steps_per_second": 16.154,
      "step": 73000
    },
    {
      "epoch": 1.1505e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "step": 73010
    },
    {
      "epoch": 1.151e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 73020
    },
    {
      "epoch": 1.1515e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 73030
    },
    {
      "epoch": 1.152e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0298,
      "step": 73040
    },
    {
      "epoch": 1.1525e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0383,
      "step": 73050
    },
    {
      "epoch": 1.153e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0365,
      "step": 73060
    },
    {
      "epoch": 1.1535e-05,
      "grad_norm": 0.5703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "step": 73070
    },
    {
      "epoch": 1.154e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "step": 73080
    },
    {
      "epoch": 1.1545e-05,
      "grad_norm": 0.56640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "step": 73090
    },
    {
      "epoch": 1.155e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0706,
      "step": 73100
    },
    {
      "epoch": 1.1555e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0399,
      "step": 73110
    },
    {
      "epoch": 1.156e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0468,
      "step": 73120
    },
    {
      "epoch": 1.1565e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "step": 73130
    },
    {
      "epoch": 1.157e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "step": 73140
    },
    {
      "epoch": 1.1575e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0303,
      "step": 73150
    },
    {
      "epoch": 1.158e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0603,
      "step": 73160
    },
    {
      "epoch": 1.1585e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0441,
      "step": 73170
    },
    {
      "epoch": 1.159e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0318,
      "step": 73180
    },
    {
      "epoch": 1.1595e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0468,
      "step": 73190
    },
    {
      "epoch": 1.16e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0555,
      "step": 73200
    },
    {
      "epoch": 1.1605e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0602,
      "step": 73210
    },
    {
      "epoch": 1.161e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0416,
      "step": 73220
    },
    {
      "epoch": 1.1615e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "step": 73230
    },
    {
      "epoch": 1.162e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "step": 73240
    },
    {
      "epoch": 1.1625e-05,
      "grad_norm": 0.431640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0556,
      "step": 73250
    },
    {
      "epoch": 1.163e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "step": 73260
    },
    {
      "epoch": 1.1635e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0317,
      "step": 73270
    },
    {
      "epoch": 1.164e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "step": 73280
    },
    {
      "epoch": 1.1645e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0594,
      "step": 73290
    },
    {
      "epoch": 1.165e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "step": 73300
    },
    {
      "epoch": 1.1655e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "step": 73310
    },
    {
      "epoch": 1.166e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "step": 73320
    },
    {
      "epoch": 1.1665e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0528,
      "step": 73330
    },
    {
      "epoch": 1.167e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0405,
      "step": 73340
    },
    {
      "epoch": 1.1675e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0576,
      "step": 73350
    },
    {
      "epoch": 1.168e-05,
      "grad_norm": 0.53125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "step": 73360
    },
    {
      "epoch": 1.1685e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.052,
      "step": 73370
    },
    {
      "epoch": 1.169e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0729,
      "step": 73380
    },
    {
      "epoch": 1.1695e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "step": 73390
    },
    {
      "epoch": 1.17e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0306,
      "step": 73400
    },
    {
      "epoch": 1.1705e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0456,
      "step": 73410
    },
    {
      "epoch": 1.171e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0428,
      "step": 73420
    },
    {
      "epoch": 1.1715e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "step": 73430
    },
    {
      "epoch": 1.172e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0289,
      "step": 73440
    },
    {
      "epoch": 1.1725e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "step": 73450
    },
    {
      "epoch": 1.173e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0034,
      "step": 73460
    },
    {
      "epoch": 1.1735e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "step": 73470
    },
    {
      "epoch": 1.174e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0259,
      "step": 73480
    },
    {
      "epoch": 1.1745e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0218,
      "step": 73490
    },
    {
      "epoch": 1.175e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "step": 73500
    },
    {
      "epoch": 1.175e-05,
      "eval_loss": 3.0830891132354736,
      "eval_runtime": 0.074,
      "eval_samples_per_second": 432.283,
      "eval_steps_per_second": 13.509,
      "step": 73500
    },
    {
      "epoch": 1.1755e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0245,
      "step": 73510
    },
    {
      "epoch": 1.176e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "step": 73520
    },
    {
      "epoch": 1.1765e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "step": 73530
    },
    {
      "epoch": 1.177e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "step": 73540
    },
    {
      "epoch": 1.1775e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "step": 73550
    },
    {
      "epoch": 1.178e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0379,
      "step": 73560
    },
    {
      "epoch": 1.1785e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0336,
      "step": 73570
    },
    {
      "epoch": 1.179e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "step": 73580
    },
    {
      "epoch": 1.1795e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "step": 73590
    },
    {
      "epoch": 1.18e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "step": 73600
    },
    {
      "epoch": 1.1805e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.039,
      "step": 73610
    },
    {
      "epoch": 1.181e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "step": 73620
    },
    {
      "epoch": 1.1815e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "step": 73630
    },
    {
      "epoch": 1.182e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "step": 73640
    },
    {
      "epoch": 1.1825e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "step": 73650
    },
    {
      "epoch": 1.183e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0262,
      "step": 73660
    },
    {
      "epoch": 1.1835e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0499,
      "step": 73670
    },
    {
      "epoch": 1.184e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0151,
      "step": 73680
    },
    {
      "epoch": 1.1845e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0406,
      "step": 73690
    },
    {
      "epoch": 1.185e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "step": 73700
    },
    {
      "epoch": 1.1855e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "step": 73710
    },
    {
      "epoch": 1.186e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0667,
      "step": 73720
    },
    {
      "epoch": 1.1865e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "step": 73730
    },
    {
      "epoch": 1.187e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "step": 73740
    },
    {
      "epoch": 1.1875e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0175,
      "step": 73750
    },
    {
      "epoch": 1.188e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0204,
      "step": 73760
    },
    {
      "epoch": 1.1885e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "step": 73770
    },
    {
      "epoch": 1.189e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "step": 73780
    },
    {
      "epoch": 1.1895e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "step": 73790
    },
    {
      "epoch": 1.19e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "step": 73800
    },
    {
      "epoch": 1.1905e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "step": 73810
    },
    {
      "epoch": 1.191e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "step": 73820
    },
    {
      "epoch": 1.1915e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0549,
      "step": 73830
    },
    {
      "epoch": 1.192e-05,
      "grad_norm": 0.53515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "step": 73840
    },
    {
      "epoch": 1.1925e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "step": 73850
    },
    {
      "epoch": 1.193e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "step": 73860
    },
    {
      "epoch": 1.1935e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "step": 73870
    },
    {
      "epoch": 1.194e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0522,
      "step": 73880
    },
    {
      "epoch": 1.1945e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0643,
      "step": 73890
    },
    {
      "epoch": 1.195e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0367,
      "step": 73900
    },
    {
      "epoch": 1.1955e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0379,
      "step": 73910
    },
    {
      "epoch": 1.196e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.009,
      "step": 73920
    },
    {
      "epoch": 1.1965e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0398,
      "step": 73930
    },
    {
      "epoch": 1.197e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0201,
      "step": 73940
    },
    {
      "epoch": 1.1975e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0081,
      "step": 73950
    },
    {
      "epoch": 1.198e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0277,
      "step": 73960
    },
    {
      "epoch": 1.1985e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0597,
      "step": 73970
    },
    {
      "epoch": 1.199e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0113,
      "step": 73980
    },
    {
      "epoch": 1.1995e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "step": 73990
    },
    {
      "epoch": 1.2e-05,
      "grad_norm": 0.51953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0712,
      "step": 74000
    },
    {
      "epoch": 1.2e-05,
      "eval_loss": 3.08304500579834,
      "eval_runtime": 0.0692,
      "eval_samples_per_second": 462.734,
      "eval_steps_per_second": 14.46,
      "step": 74000
    },
    {
      "epoch": 1.2005e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0497,
      "step": 74010
    },
    {
      "epoch": 1.201e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "step": 74020
    },
    {
      "epoch": 1.2015e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "step": 74030
    },
    {
      "epoch": 1.202e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0168,
      "step": 74040
    },
    {
      "epoch": 1.2025e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0312,
      "step": 74050
    },
    {
      "epoch": 1.203e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "step": 74060
    },
    {
      "epoch": 1.2035e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "step": 74070
    },
    {
      "epoch": 1.204e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0235,
      "step": 74080
    },
    {
      "epoch": 1.2045e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "step": 74090
    },
    {
      "epoch": 1.205e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0619,
      "step": 74100
    },
    {
      "epoch": 1.2055e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "step": 74110
    },
    {
      "epoch": 1.206e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "step": 74120
    },
    {
      "epoch": 1.2065e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0437,
      "step": 74130
    },
    {
      "epoch": 1.207e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "step": 74140
    },
    {
      "epoch": 1.2075e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "step": 74150
    },
    {
      "epoch": 1.208e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "step": 74160
    },
    {
      "epoch": 1.2085e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0437,
      "step": 74170
    },
    {
      "epoch": 1.209e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "step": 74180
    },
    {
      "epoch": 1.2095e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0498,
      "step": 74190
    },
    {
      "epoch": 1.21e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0493,
      "step": 74200
    },
    {
      "epoch": 1.2105e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0361,
      "step": 74210
    },
    {
      "epoch": 1.211e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0513,
      "step": 74220
    },
    {
      "epoch": 1.2115e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0745,
      "step": 74230
    },
    {
      "epoch": 1.212e-05,
      "grad_norm": 0.51953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0383,
      "step": 74240
    },
    {
      "epoch": 1.2125e-05,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0239,
      "step": 74250
    },
    {
      "epoch": 1.213e-05,
      "grad_norm": 0.54296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "step": 74260
    },
    {
      "epoch": 1.2135e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "step": 74270
    },
    {
      "epoch": 1.214e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0298,
      "step": 74280
    },
    {
      "epoch": 1.2145e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "step": 74290
    },
    {
      "epoch": 1.215e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0287,
      "step": 74300
    },
    {
      "epoch": 1.2155e-05,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "step": 74310
    },
    {
      "epoch": 1.216e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0401,
      "step": 74320
    },
    {
      "epoch": 1.2165e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "step": 74330
    },
    {
      "epoch": 1.217e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0167,
      "step": 74340
    },
    {
      "epoch": 1.2175e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0337,
      "step": 74350
    },
    {
      "epoch": 1.218e-05,
      "grad_norm": 0.56640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "step": 74360
    },
    {
      "epoch": 1.2185e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "step": 74370
    },
    {
      "epoch": 1.219e-05,
      "grad_norm": 0.4296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0063,
      "step": 74380
    },
    {
      "epoch": 1.2195e-05,
      "grad_norm": 0.98046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.024,
      "step": 74390
    },
    {
      "epoch": 1.22e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "step": 74400
    },
    {
      "epoch": 1.2205e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.075,
      "step": 74410
    },
    {
      "epoch": 1.221e-05,
      "grad_norm": 0.4296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9934,
      "step": 74420
    },
    {
      "epoch": 1.2215e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0843,
      "step": 74430
    },
    {
      "epoch": 1.222e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "step": 74440
    },
    {
      "epoch": 1.2225e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 74450
    },
    {
      "epoch": 1.223e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "step": 74460
    },
    {
      "epoch": 1.2235e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "step": 74470
    },
    {
      "epoch": 1.224e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0441,
      "step": 74480
    },
    {
      "epoch": 1.2245e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "step": 74490
    },
    {
      "epoch": 1.225e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0217,
      "step": 74500
    },
    {
      "epoch": 1.225e-05,
      "eval_loss": 3.0819525718688965,
      "eval_runtime": 0.0614,
      "eval_samples_per_second": 521.163,
      "eval_steps_per_second": 16.286,
      "step": 74500
    },
    {
      "epoch": 1.2255e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0251,
      "step": 74510
    },
    {
      "epoch": 1.226e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "step": 74520
    },
    {
      "epoch": 1.2265e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0466,
      "step": 74530
    },
    {
      "epoch": 1.227e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "step": 74540
    },
    {
      "epoch": 1.2275e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "step": 74550
    },
    {
      "epoch": 1.228e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0409,
      "step": 74560
    },
    {
      "epoch": 1.2285e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "step": 74570
    },
    {
      "epoch": 1.229e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0276,
      "step": 74580
    },
    {
      "epoch": 1.2295e-05,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "step": 74590
    },
    {
      "epoch": 1.23e-05,
      "grad_norm": 0.431640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "step": 74600
    },
    {
      "epoch": 1.2305e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0612,
      "step": 74610
    },
    {
      "epoch": 1.231e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0593,
      "step": 74620
    },
    {
      "epoch": 1.2315e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "step": 74630
    },
    {
      "epoch": 1.232e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0297,
      "step": 74640
    },
    {
      "epoch": 1.2325e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0521,
      "step": 74650
    },
    {
      "epoch": 1.233e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0364,
      "step": 74660
    },
    {
      "epoch": 1.2335e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.054,
      "step": 74670
    },
    {
      "epoch": 1.234e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "step": 74680
    },
    {
      "epoch": 1.2345e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0608,
      "step": 74690
    },
    {
      "epoch": 1.235e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0593,
      "step": 74700
    },
    {
      "epoch": 1.2355e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0608,
      "step": 74710
    },
    {
      "epoch": 1.236e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "step": 74720
    },
    {
      "epoch": 1.2365e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0582,
      "step": 74730
    },
    {
      "epoch": 1.237e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "step": 74740
    },
    {
      "epoch": 1.2375e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0219,
      "step": 74750
    },
    {
      "epoch": 1.238e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0684,
      "step": 74760
    },
    {
      "epoch": 1.2385e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0235,
      "step": 74770
    },
    {
      "epoch": 1.239e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "step": 74780
    },
    {
      "epoch": 1.2395e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "step": 74790
    },
    {
      "epoch": 1.24e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0697,
      "step": 74800
    },
    {
      "epoch": 1.2405e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0379,
      "step": 74810
    },
    {
      "epoch": 1.241e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0036,
      "step": 74820
    },
    {
      "epoch": 1.2415e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "step": 74830
    },
    {
      "epoch": 1.242e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "step": 74840
    },
    {
      "epoch": 1.2425e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0398,
      "step": 74850
    },
    {
      "epoch": 1.243e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "step": 74860
    },
    {
      "epoch": 1.2435e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "step": 74870
    },
    {
      "epoch": 1.244e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "step": 74880
    },
    {
      "epoch": 1.2445e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "step": 74890
    },
    {
      "epoch": 1.245e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "step": 74900
    },
    {
      "epoch": 1.2455e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "step": 74910
    },
    {
      "epoch": 1.246e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.073,
      "step": 74920
    },
    {
      "epoch": 1.2465e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0252,
      "step": 74930
    },
    {
      "epoch": 1.247e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0285,
      "step": 74940
    },
    {
      "epoch": 1.2475e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0629,
      "step": 74950
    },
    {
      "epoch": 1.248e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0646,
      "step": 74960
    },
    {
      "epoch": 1.2485e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "step": 74970
    },
    {
      "epoch": 1.249e-05,
      "grad_norm": 0.625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0603,
      "step": 74980
    },
    {
      "epoch": 1.2495e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.052,
      "step": 74990
    },
    {
      "epoch": 1.25e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0577,
      "step": 75000
    },
    {
      "epoch": 1.25e-05,
      "eval_loss": 3.082608222961426,
      "eval_runtime": 0.0642,
      "eval_samples_per_second": 498.567,
      "eval_steps_per_second": 15.58,
      "step": 75000
    },
    {
      "epoch": 1.2505e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0583,
      "step": 75010
    },
    {
      "epoch": 1.251e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "step": 75020
    },
    {
      "epoch": 1.2515e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "step": 75030
    },
    {
      "epoch": 1.252e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0556,
      "step": 75040
    },
    {
      "epoch": 1.2525e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "step": 75050
    },
    {
      "epoch": 1.253e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "step": 75060
    },
    {
      "epoch": 1.2535e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0353,
      "step": 75070
    },
    {
      "epoch": 1.254e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0306,
      "step": 75080
    },
    {
      "epoch": 1.2545e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0702,
      "step": 75090
    },
    {
      "epoch": 1.255e-05,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0643,
      "step": 75100
    },
    {
      "epoch": 1.2555e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0678,
      "step": 75110
    },
    {
      "epoch": 1.256e-05,
      "grad_norm": 0.60546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0662,
      "step": 75120
    },
    {
      "epoch": 1.2565e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0483,
      "step": 75130
    },
    {
      "epoch": 1.257e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "step": 75140
    },
    {
      "epoch": 1.2575e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0236,
      "step": 75150
    },
    {
      "epoch": 1.258e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0354,
      "step": 75160
    },
    {
      "epoch": 1.2585e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "step": 75170
    },
    {
      "epoch": 1.259e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "step": 75180
    },
    {
      "epoch": 1.2595e-05,
      "grad_norm": 0.6171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "step": 75190
    },
    {
      "epoch": 1.26e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "step": 75200
    },
    {
      "epoch": 1.2605e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0306,
      "step": 75210
    },
    {
      "epoch": 1.261e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0586,
      "step": 75220
    },
    {
      "epoch": 1.2615e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "step": 75230
    },
    {
      "epoch": 1.262e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0321,
      "step": 75240
    },
    {
      "epoch": 1.2625e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.073,
      "step": 75250
    },
    {
      "epoch": 1.263e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0405,
      "step": 75260
    },
    {
      "epoch": 1.2635e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0529,
      "step": 75270
    },
    {
      "epoch": 1.264e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "step": 75280
    },
    {
      "epoch": 1.2645e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0441,
      "step": 75290
    },
    {
      "epoch": 1.265e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "step": 75300
    },
    {
      "epoch": 1.2655e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "step": 75310
    },
    {
      "epoch": 1.266e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0255,
      "step": 75320
    },
    {
      "epoch": 1.2665e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0401,
      "step": 75330
    },
    {
      "epoch": 1.267e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0551,
      "step": 75340
    },
    {
      "epoch": 1.2675e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "step": 75350
    },
    {
      "epoch": 1.268e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0603,
      "step": 75360
    },
    {
      "epoch": 1.2685e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "step": 75370
    },
    {
      "epoch": 1.269e-05,
      "grad_norm": 0.52734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0257,
      "step": 75380
    },
    {
      "epoch": 1.2695e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "step": 75390
    },
    {
      "epoch": 1.27e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0261,
      "step": 75400
    },
    {
      "epoch": 1.2705e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "step": 75410
    },
    {
      "epoch": 1.271e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0217,
      "step": 75420
    },
    {
      "epoch": 1.2715e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0352,
      "step": 75430
    },
    {
      "epoch": 1.272e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "step": 75440
    },
    {
      "epoch": 1.2725e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "step": 75450
    },
    {
      "epoch": 1.273e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0401,
      "step": 75460
    },
    {
      "epoch": 1.2735e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0332,
      "step": 75470
    },
    {
      "epoch": 1.274e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "step": 75480
    },
    {
      "epoch": 1.2745e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0656,
      "step": 75490
    },
    {
      "epoch": 1.275e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "step": 75500
    },
    {
      "epoch": 1.275e-05,
      "eval_loss": 3.0825040340423584,
      "eval_runtime": 0.0636,
      "eval_samples_per_second": 503.282,
      "eval_steps_per_second": 15.728,
      "step": 75500
    },
    {
      "epoch": 1.2755e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0493,
      "step": 75510
    },
    {
      "epoch": 1.276e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0349,
      "step": 75520
    },
    {
      "epoch": 1.2765e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "step": 75530
    },
    {
      "epoch": 1.277e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "step": 75540
    },
    {
      "epoch": 1.2775e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "step": 75550
    },
    {
      "epoch": 1.278e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0119,
      "step": 75560
    },
    {
      "epoch": 1.2785e-05,
      "grad_norm": 0.55859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0448,
      "step": 75570
    },
    {
      "epoch": 1.279e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "step": 75580
    },
    {
      "epoch": 1.2795e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0612,
      "step": 75590
    },
    {
      "epoch": 1.28e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "step": 75600
    },
    {
      "epoch": 1.2805e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0226,
      "step": 75610
    },
    {
      "epoch": 1.281e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "step": 75620
    },
    {
      "epoch": 1.2815e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "step": 75630
    },
    {
      "epoch": 1.282e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "step": 75640
    },
    {
      "epoch": 1.2825e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "step": 75650
    },
    {
      "epoch": 1.283e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "step": 75660
    },
    {
      "epoch": 1.2835e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0549,
      "step": 75670
    },
    {
      "epoch": 1.284e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0614,
      "step": 75680
    },
    {
      "epoch": 1.2845e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "step": 75690
    },
    {
      "epoch": 1.285e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0362,
      "step": 75700
    },
    {
      "epoch": 1.2855e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0342,
      "step": 75710
    },
    {
      "epoch": 1.286e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0526,
      "step": 75720
    },
    {
      "epoch": 1.2865e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0123,
      "step": 75730
    },
    {
      "epoch": 1.287e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0193,
      "step": 75740
    },
    {
      "epoch": 1.2875e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0564,
      "step": 75750
    },
    {
      "epoch": 1.288e-05,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0246,
      "step": 75760
    },
    {
      "epoch": 1.2885e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "step": 75770
    },
    {
      "epoch": 1.289e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "step": 75780
    },
    {
      "epoch": 1.2895e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "step": 75790
    },
    {
      "epoch": 1.29e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0214,
      "step": 75800
    },
    {
      "epoch": 1.2905e-05,
      "grad_norm": 0.51953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "step": 75810
    },
    {
      "epoch": 1.291e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0645,
      "step": 75820
    },
    {
      "epoch": 1.2915e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0726,
      "step": 75830
    },
    {
      "epoch": 1.292e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "step": 75840
    },
    {
      "epoch": 1.2925e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.03,
      "step": 75850
    },
    {
      "epoch": 1.293e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0607,
      "step": 75860
    },
    {
      "epoch": 1.2935e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0342,
      "step": 75870
    },
    {
      "epoch": 1.294e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0324,
      "step": 75880
    },
    {
      "epoch": 1.2945e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "step": 75890
    },
    {
      "epoch": 1.295e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0216,
      "step": 75900
    },
    {
      "epoch": 1.2955e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0644,
      "step": 75910
    },
    {
      "epoch": 1.296e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "step": 75920
    },
    {
      "epoch": 1.2965e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0139,
      "step": 75930
    },
    {
      "epoch": 1.297e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0269,
      "step": 75940
    },
    {
      "epoch": 1.2975e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0691,
      "step": 75950
    },
    {
      "epoch": 1.298e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0304,
      "step": 75960
    },
    {
      "epoch": 1.2985e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "step": 75970
    },
    {
      "epoch": 1.299e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "step": 75980
    },
    {
      "epoch": 1.2995e-05,
      "grad_norm": 0.53515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.068,
      "step": 75990
    },
    {
      "epoch": 1.3e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0276,
      "step": 76000
    },
    {
      "epoch": 1.3e-05,
      "eval_loss": 3.08209228515625,
      "eval_runtime": 0.0672,
      "eval_samples_per_second": 476.536,
      "eval_steps_per_second": 14.892,
      "step": 76000
    },
    {
      "epoch": 1.3005e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "step": 76010
    },
    {
      "epoch": 1.301e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0688,
      "step": 76020
    },
    {
      "epoch": 1.3015e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "step": 76030
    },
    {
      "epoch": 1.302e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "step": 76040
    },
    {
      "epoch": 1.3025e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0643,
      "step": 76050
    },
    {
      "epoch": 1.303e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0159,
      "step": 76060
    },
    {
      "epoch": 1.3035e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0191,
      "step": 76070
    },
    {
      "epoch": 1.304e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0437,
      "step": 76080
    },
    {
      "epoch": 1.3045e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0617,
      "step": 76090
    },
    {
      "epoch": 1.305e-05,
      "grad_norm": 0.5703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0262,
      "step": 76100
    },
    {
      "epoch": 1.3055e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0523,
      "step": 76110
    },
    {
      "epoch": 1.306e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "step": 76120
    },
    {
      "epoch": 1.3065e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0304,
      "step": 76130
    },
    {
      "epoch": 1.307e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0366,
      "step": 76140
    },
    {
      "epoch": 1.3075e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0129,
      "step": 76150
    },
    {
      "epoch": 1.308e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0232,
      "step": 76160
    },
    {
      "epoch": 1.3085e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0632,
      "step": 76170
    },
    {
      "epoch": 1.309e-05,
      "grad_norm": 0.53515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0336,
      "step": 76180
    },
    {
      "epoch": 1.3095e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0633,
      "step": 76190
    },
    {
      "epoch": 1.31e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0386,
      "step": 76200
    },
    {
      "epoch": 1.3105e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "step": 76210
    },
    {
      "epoch": 1.311e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0448,
      "step": 76220
    },
    {
      "epoch": 1.3115e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0503,
      "step": 76230
    },
    {
      "epoch": 1.312e-05,
      "grad_norm": 0.60546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0525,
      "step": 76240
    },
    {
      "epoch": 1.3125e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "step": 76250
    },
    {
      "epoch": 1.313e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0322,
      "step": 76260
    },
    {
      "epoch": 1.3135e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0351,
      "step": 76270
    },
    {
      "epoch": 1.314e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "step": 76280
    },
    {
      "epoch": 1.3145e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0224,
      "step": 76290
    },
    {
      "epoch": 1.315e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "step": 76300
    },
    {
      "epoch": 1.3155e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0409,
      "step": 76310
    },
    {
      "epoch": 1.316e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0511,
      "step": 76320
    },
    {
      "epoch": 1.3165e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0497,
      "step": 76330
    },
    {
      "epoch": 1.317e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "step": 76340
    },
    {
      "epoch": 1.3175e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0617,
      "step": 76350
    },
    {
      "epoch": 1.318e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "step": 76360
    },
    {
      "epoch": 1.3185e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0558,
      "step": 76370
    },
    {
      "epoch": 1.319e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "step": 76380
    },
    {
      "epoch": 1.3195e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0572,
      "step": 76390
    },
    {
      "epoch": 1.32e-05,
      "grad_norm": 0.53125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "step": 76400
    },
    {
      "epoch": 1.3205e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "step": 76410
    },
    {
      "epoch": 1.321e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "step": 76420
    },
    {
      "epoch": 1.3215e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0373,
      "step": 76430
    },
    {
      "epoch": 1.322e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "step": 76440
    },
    {
      "epoch": 1.3225e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0438,
      "step": 76450
    },
    {
      "epoch": 1.323e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "step": 76460
    },
    {
      "epoch": 1.3235e-05,
      "grad_norm": 0.51953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0631,
      "step": 76470
    },
    {
      "epoch": 1.324e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.031,
      "step": 76480
    },
    {
      "epoch": 1.3245e-05,
      "grad_norm": 0.52734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "step": 76490
    },
    {
      "epoch": 1.325e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0495,
      "step": 76500
    },
    {
      "epoch": 1.325e-05,
      "eval_loss": 3.081979751586914,
      "eval_runtime": 0.0631,
      "eval_samples_per_second": 507.227,
      "eval_steps_per_second": 15.851,
      "step": 76500
    },
    {
      "epoch": 1.3255e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "step": 76510
    },
    {
      "epoch": 1.326e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0531,
      "step": 76520
    },
    {
      "epoch": 1.3265e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "step": 76530
    },
    {
      "epoch": 1.327e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0612,
      "step": 76540
    },
    {
      "epoch": 1.3275e-05,
      "grad_norm": 0.51953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "step": 76550
    },
    {
      "epoch": 1.328e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0334,
      "step": 76560
    },
    {
      "epoch": 1.3285e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "step": 76570
    },
    {
      "epoch": 1.329e-05,
      "grad_norm": 0.51953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0565,
      "step": 76580
    },
    {
      "epoch": 1.3295e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0357,
      "step": 76590
    },
    {
      "epoch": 1.33e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0516,
      "step": 76600
    },
    {
      "epoch": 1.3305e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "step": 76610
    },
    {
      "epoch": 1.331e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "step": 76620
    },
    {
      "epoch": 1.3315e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0383,
      "step": 76630
    },
    {
      "epoch": 1.332e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0279,
      "step": 76640
    },
    {
      "epoch": 1.3325e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "step": 76650
    },
    {
      "epoch": 1.333e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0711,
      "step": 76660
    },
    {
      "epoch": 1.3335e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0298,
      "step": 76670
    },
    {
      "epoch": 1.334e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "step": 76680
    },
    {
      "epoch": 1.3345e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "step": 76690
    },
    {
      "epoch": 1.335e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0369,
      "step": 76700
    },
    {
      "epoch": 1.3355e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "step": 76710
    },
    {
      "epoch": 1.336e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "step": 76720
    },
    {
      "epoch": 1.3365e-05,
      "grad_norm": 0.52734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "step": 76730
    },
    {
      "epoch": 1.337e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "step": 76740
    },
    {
      "epoch": 1.3375e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0324,
      "step": 76750
    },
    {
      "epoch": 1.338e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "step": 76760
    },
    {
      "epoch": 1.3385e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.049,
      "step": 76770
    },
    {
      "epoch": 1.339e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0677,
      "step": 76780
    },
    {
      "epoch": 1.3395e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "step": 76790
    },
    {
      "epoch": 1.34e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0503,
      "step": 76800
    },
    {
      "epoch": 1.3405e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0103,
      "step": 76810
    },
    {
      "epoch": 1.341e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "step": 76820
    },
    {
      "epoch": 1.3415e-05,
      "grad_norm": 0.5703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "step": 76830
    },
    {
      "epoch": 1.342e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0185,
      "step": 76840
    },
    {
      "epoch": 1.3425e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0259,
      "step": 76850
    },
    {
      "epoch": 1.343e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "step": 76860
    },
    {
      "epoch": 1.3435e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "step": 76870
    },
    {
      "epoch": 1.344e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0644,
      "step": 76880
    },
    {
      "epoch": 1.3445e-05,
      "grad_norm": 0.53515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.048,
      "step": 76890
    },
    {
      "epoch": 1.345e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "step": 76900
    },
    {
      "epoch": 1.3455e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "step": 76910
    },
    {
      "epoch": 1.346e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0532,
      "step": 76920
    },
    {
      "epoch": 1.3465e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0164,
      "step": 76930
    },
    {
      "epoch": 1.347e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "step": 76940
    },
    {
      "epoch": 1.3475e-05,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0031,
      "step": 76950
    },
    {
      "epoch": 1.348e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "step": 76960
    },
    {
      "epoch": 1.3485e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0288,
      "step": 76970
    },
    {
      "epoch": 1.349e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0379,
      "step": 76980
    },
    {
      "epoch": 1.3495e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0303,
      "step": 76990
    },
    {
      "epoch": 1.35e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0318,
      "step": 77000
    },
    {
      "epoch": 1.35e-05,
      "eval_loss": 3.082359552383423,
      "eval_runtime": 0.0711,
      "eval_samples_per_second": 449.957,
      "eval_steps_per_second": 14.061,
      "step": 77000
    },
    {
      "epoch": 1.3505e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "step": 77010
    },
    {
      "epoch": 1.351e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "step": 77020
    },
    {
      "epoch": 1.3515e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "step": 77030
    },
    {
      "epoch": 1.352e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0235,
      "step": 77040
    },
    {
      "epoch": 1.3525e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "step": 77050
    },
    {
      "epoch": 1.353e-05,
      "grad_norm": 0.431640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "step": 77060
    },
    {
      "epoch": 1.3535e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "step": 77070
    },
    {
      "epoch": 1.354e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "step": 77080
    },
    {
      "epoch": 1.3545e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0683,
      "step": 77090
    },
    {
      "epoch": 1.355e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "step": 77100
    },
    {
      "epoch": 1.3555e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0766,
      "step": 77110
    },
    {
      "epoch": 1.356e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "step": 77120
    },
    {
      "epoch": 1.3565e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0328,
      "step": 77130
    },
    {
      "epoch": 1.357e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "step": 77140
    },
    {
      "epoch": 1.3575e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "step": 77150
    },
    {
      "epoch": 1.358e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0455,
      "step": 77160
    },
    {
      "epoch": 1.3585e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "step": 77170
    },
    {
      "epoch": 1.359e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "step": 77180
    },
    {
      "epoch": 1.3595e-05,
      "grad_norm": 0.51953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "step": 77190
    },
    {
      "epoch": 1.36e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0572,
      "step": 77200
    },
    {
      "epoch": 1.3605e-05,
      "grad_norm": 0.62890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0391,
      "step": 77210
    },
    {
      "epoch": 1.361e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0312,
      "step": 77220
    },
    {
      "epoch": 1.3615e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0406,
      "step": 77230
    },
    {
      "epoch": 1.362e-05,
      "grad_norm": 0.42578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.017,
      "step": 77240
    },
    {
      "epoch": 1.3625e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "step": 77250
    },
    {
      "epoch": 1.363e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0487,
      "step": 77260
    },
    {
      "epoch": 1.3635e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0664,
      "step": 77270
    },
    {
      "epoch": 1.364e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0532,
      "step": 77280
    },
    {
      "epoch": 1.3645e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "step": 77290
    },
    {
      "epoch": 1.365e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0297,
      "step": 77300
    },
    {
      "epoch": 1.3655e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0349,
      "step": 77310
    },
    {
      "epoch": 1.366e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "step": 77320
    },
    {
      "epoch": 1.3665e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0347,
      "step": 77330
    },
    {
      "epoch": 1.367e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "step": 77340
    },
    {
      "epoch": 1.3675e-05,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "step": 77350
    },
    {
      "epoch": 1.368e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "step": 77360
    },
    {
      "epoch": 1.3685e-05,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0251,
      "step": 77370
    },
    {
      "epoch": 1.369e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0468,
      "step": 77380
    },
    {
      "epoch": 1.3695e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0266,
      "step": 77390
    },
    {
      "epoch": 1.37e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "step": 77400
    },
    {
      "epoch": 1.3705e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "step": 77410
    },
    {
      "epoch": 1.371e-05,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "step": 77420
    },
    {
      "epoch": 1.3715e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "step": 77430
    },
    {
      "epoch": 1.372e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0715,
      "step": 77440
    },
    {
      "epoch": 1.3725e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "step": 77450
    },
    {
      "epoch": 1.373e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0543,
      "step": 77460
    },
    {
      "epoch": 1.3735e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "step": 77470
    },
    {
      "epoch": 1.374e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "step": 77480
    },
    {
      "epoch": 1.3745e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "step": 77490
    },
    {
      "epoch": 1.375e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "step": 77500
    },
    {
      "epoch": 1.375e-05,
      "eval_loss": 3.0821151733398438,
      "eval_runtime": 0.0623,
      "eval_samples_per_second": 513.937,
      "eval_steps_per_second": 16.061,
      "step": 77500
    },
    {
      "epoch": 1.3755e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "step": 77510
    },
    {
      "epoch": 1.376e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0332,
      "step": 77520
    },
    {
      "epoch": 1.3765e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0583,
      "step": 77530
    },
    {
      "epoch": 1.377e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0265,
      "step": 77540
    },
    {
      "epoch": 1.3775e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "step": 77550
    },
    {
      "epoch": 1.378e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0133,
      "step": 77560
    },
    {
      "epoch": 1.3785e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "step": 77570
    },
    {
      "epoch": 1.379e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0314,
      "step": 77580
    },
    {
      "epoch": 1.3795e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0196,
      "step": 77590
    },
    {
      "epoch": 1.38e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "step": 77600
    },
    {
      "epoch": 1.3805e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "step": 77610
    },
    {
      "epoch": 1.381e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0366,
      "step": 77620
    },
    {
      "epoch": 1.3815e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "step": 77630
    },
    {
      "epoch": 1.382e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0266,
      "step": 77640
    },
    {
      "epoch": 1.3825e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0353,
      "step": 77650
    },
    {
      "epoch": 1.383e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0676,
      "step": 77660
    },
    {
      "epoch": 1.3835e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "step": 77670
    },
    {
      "epoch": 1.384e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "step": 77680
    },
    {
      "epoch": 1.3845e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0629,
      "step": 77690
    },
    {
      "epoch": 1.385e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "step": 77700
    },
    {
      "epoch": 1.3855e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.02,
      "step": 77710
    },
    {
      "epoch": 1.386e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0633,
      "step": 77720
    },
    {
      "epoch": 1.3865e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "step": 77730
    },
    {
      "epoch": 1.387e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0533,
      "step": 77740
    },
    {
      "epoch": 1.3875e-05,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0479,
      "step": 77750
    },
    {
      "epoch": 1.388e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "step": 77760
    },
    {
      "epoch": 1.3885e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "step": 77770
    },
    {
      "epoch": 1.389e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0618,
      "step": 77780
    },
    {
      "epoch": 1.3895e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0333,
      "step": 77790
    },
    {
      "epoch": 1.39e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0733,
      "step": 77800
    },
    {
      "epoch": 1.3905e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "step": 77810
    },
    {
      "epoch": 1.391e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "step": 77820
    },
    {
      "epoch": 1.3915e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "step": 77830
    },
    {
      "epoch": 1.392e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0149,
      "step": 77840
    },
    {
      "epoch": 1.3925e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.062,
      "step": 77850
    },
    {
      "epoch": 1.393e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "step": 77860
    },
    {
      "epoch": 1.3935e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "step": 77870
    },
    {
      "epoch": 1.394e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.027,
      "step": 77880
    },
    {
      "epoch": 1.3945e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "step": 77890
    },
    {
      "epoch": 1.395e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0152,
      "step": 77900
    },
    {
      "epoch": 1.3955e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0251,
      "step": 77910
    },
    {
      "epoch": 1.396e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0681,
      "step": 77920
    },
    {
      "epoch": 1.3965e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0696,
      "step": 77930
    },
    {
      "epoch": 1.397e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.029,
      "step": 77940
    },
    {
      "epoch": 1.3975e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "step": 77950
    },
    {
      "epoch": 1.398e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "step": 77960
    },
    {
      "epoch": 1.3985e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.072,
      "step": 77970
    },
    {
      "epoch": 1.399e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "step": 77980
    },
    {
      "epoch": 1.3995e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0235,
      "step": 77990
    },
    {
      "epoch": 1.4e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.035,
      "step": 78000
    },
    {
      "epoch": 1.4e-05,
      "eval_loss": 3.081981658935547,
      "eval_runtime": 0.0629,
      "eval_samples_per_second": 508.967,
      "eval_steps_per_second": 15.905,
      "step": 78000
    },
    {
      "epoch": 1.4005e-05,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0369,
      "step": 78010
    },
    {
      "epoch": 1.401e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0665,
      "step": 78020
    },
    {
      "epoch": 1.4015e-05,
      "grad_norm": 0.4296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0357,
      "step": 78030
    },
    {
      "epoch": 1.402e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0584,
      "step": 78040
    },
    {
      "epoch": 1.4025e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "step": 78050
    },
    {
      "epoch": 1.403e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "step": 78060
    },
    {
      "epoch": 1.4035e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "step": 78070
    },
    {
      "epoch": 1.404e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 78080
    },
    {
      "epoch": 1.4045e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0532,
      "step": 78090
    },
    {
      "epoch": 1.405e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "step": 78100
    },
    {
      "epoch": 1.4055e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "step": 78110
    },
    {
      "epoch": 1.406e-05,
      "grad_norm": 0.52734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "step": 78120
    },
    {
      "epoch": 1.4065e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0288,
      "step": 78130
    },
    {
      "epoch": 1.407e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0322,
      "step": 78140
    },
    {
      "epoch": 1.4075e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "step": 78150
    },
    {
      "epoch": 1.408e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0241,
      "step": 78160
    },
    {
      "epoch": 1.4085e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0672,
      "step": 78170
    },
    {
      "epoch": 1.409e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "step": 78180
    },
    {
      "epoch": 1.4095e-05,
      "grad_norm": 0.55078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 78190
    },
    {
      "epoch": 1.41e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0577,
      "step": 78200
    },
    {
      "epoch": 1.4105e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "step": 78210
    },
    {
      "epoch": 1.411e-05,
      "grad_norm": 0.54296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0566,
      "step": 78220
    },
    {
      "epoch": 1.4115e-05,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0195,
      "step": 78230
    },
    {
      "epoch": 1.412e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0259,
      "step": 78240
    },
    {
      "epoch": 1.4125e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0533,
      "step": 78250
    },
    {
      "epoch": 1.413e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0113,
      "step": 78260
    },
    {
      "epoch": 1.4135e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "step": 78270
    },
    {
      "epoch": 1.414e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0748,
      "step": 78280
    },
    {
      "epoch": 1.4145e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0723,
      "step": 78290
    },
    {
      "epoch": 1.415e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0572,
      "step": 78300
    },
    {
      "epoch": 1.4155e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "step": 78310
    },
    {
      "epoch": 1.416e-05,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "step": 78320
    },
    {
      "epoch": 1.4165e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0718,
      "step": 78330
    },
    {
      "epoch": 1.417e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "step": 78340
    },
    {
      "epoch": 1.4175e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0406,
      "step": 78350
    },
    {
      "epoch": 1.418e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "step": 78360
    },
    {
      "epoch": 1.4185e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0323,
      "step": 78370
    },
    {
      "epoch": 1.419e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0401,
      "step": 78380
    },
    {
      "epoch": 1.4195e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0456,
      "step": 78390
    },
    {
      "epoch": 1.42e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0523,
      "step": 78400
    },
    {
      "epoch": 1.4205e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0257,
      "step": 78410
    },
    {
      "epoch": 1.421e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "step": 78420
    },
    {
      "epoch": 1.4215e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "step": 78430
    },
    {
      "epoch": 1.422e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "step": 78440
    },
    {
      "epoch": 1.4225e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "step": 78450
    },
    {
      "epoch": 1.423e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.07,
      "step": 78460
    },
    {
      "epoch": 1.4235e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.031,
      "step": 78470
    },
    {
      "epoch": 1.424e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "step": 78480
    },
    {
      "epoch": 1.4245e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0673,
      "step": 78490
    },
    {
      "epoch": 1.425e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0365,
      "step": 78500
    },
    {
      "epoch": 1.425e-05,
      "eval_loss": 3.081976890563965,
      "eval_runtime": 0.0646,
      "eval_samples_per_second": 495.49,
      "eval_steps_per_second": 15.484,
      "step": 78500
    },
    {
      "epoch": 1.4255e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "step": 78510
    },
    {
      "epoch": 1.426e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "step": 78520
    },
    {
      "epoch": 1.4265e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "step": 78530
    },
    {
      "epoch": 1.427e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0334,
      "step": 78540
    },
    {
      "epoch": 1.4275e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "step": 78550
    },
    {
      "epoch": 1.428e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0487,
      "step": 78560
    },
    {
      "epoch": 1.4285e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0533,
      "step": 78570
    },
    {
      "epoch": 1.429e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "step": 78580
    },
    {
      "epoch": 1.4295e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "step": 78590
    },
    {
      "epoch": 1.43e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0214,
      "step": 78600
    },
    {
      "epoch": 1.4305e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0489,
      "step": 78610
    },
    {
      "epoch": 1.431e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "step": 78620
    },
    {
      "epoch": 1.4315e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0699,
      "step": 78630
    },
    {
      "epoch": 1.432e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0218,
      "step": 78640
    },
    {
      "epoch": 1.4325e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "step": 78650
    },
    {
      "epoch": 1.433e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0429,
      "step": 78660
    },
    {
      "epoch": 1.4335e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0736,
      "step": 78670
    },
    {
      "epoch": 1.434e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "step": 78680
    },
    {
      "epoch": 1.4345e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0319,
      "step": 78690
    },
    {
      "epoch": 1.435e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "step": 78700
    },
    {
      "epoch": 1.4355e-05,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0634,
      "step": 78710
    },
    {
      "epoch": 1.436e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "step": 78720
    },
    {
      "epoch": 1.4365e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.061,
      "step": 78730
    },
    {
      "epoch": 1.437e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "step": 78740
    },
    {
      "epoch": 1.4375e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "step": 78750
    },
    {
      "epoch": 1.438e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0666,
      "step": 78760
    },
    {
      "epoch": 1.4385e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0248,
      "step": 78770
    },
    {
      "epoch": 1.439e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "step": 78780
    },
    {
      "epoch": 1.4395e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0398,
      "step": 78790
    },
    {
      "epoch": 1.44e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "step": 78800
    },
    {
      "epoch": 1.4405e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "step": 78810
    },
    {
      "epoch": 1.441e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "step": 78820
    },
    {
      "epoch": 1.4415e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0574,
      "step": 78830
    },
    {
      "epoch": 1.442e-05,
      "grad_norm": 0.421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0036,
      "step": 78840
    },
    {
      "epoch": 1.4425e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "step": 78850
    },
    {
      "epoch": 1.443e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0224,
      "step": 78860
    },
    {
      "epoch": 1.4435e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "step": 78870
    },
    {
      "epoch": 1.444e-05,
      "grad_norm": 0.52734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "step": 78880
    },
    {
      "epoch": 1.4445e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0307,
      "step": 78890
    },
    {
      "epoch": 1.445e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "step": 78900
    },
    {
      "epoch": 1.4455e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0324,
      "step": 78910
    },
    {
      "epoch": 1.446e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "step": 78920
    },
    {
      "epoch": 1.4465e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "step": 78930
    },
    {
      "epoch": 1.447e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0321,
      "step": 78940
    },
    {
      "epoch": 1.4475e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "step": 78950
    },
    {
      "epoch": 1.448e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0551,
      "step": 78960
    },
    {
      "epoch": 1.4485e-05,
      "grad_norm": 0.5546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "step": 78970
    },
    {
      "epoch": 1.449e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "step": 78980
    },
    {
      "epoch": 1.4495e-05,
      "grad_norm": 0.5546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0757,
      "step": 78990
    },
    {
      "epoch": 1.45e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "step": 79000
    },
    {
      "epoch": 1.45e-05,
      "eval_loss": 3.0814013481140137,
      "eval_runtime": 0.0614,
      "eval_samples_per_second": 520.93,
      "eval_steps_per_second": 16.279,
      "step": 79000
    },
    {
      "epoch": 1.4505e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0687,
      "step": 79010
    },
    {
      "epoch": 1.451e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "step": 79020
    },
    {
      "epoch": 1.4515e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0422,
      "step": 79030
    },
    {
      "epoch": 1.452e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0516,
      "step": 79040
    },
    {
      "epoch": 1.4525e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "step": 79050
    },
    {
      "epoch": 1.453e-05,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "step": 79060
    },
    {
      "epoch": 1.4535e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0348,
      "step": 79070
    },
    {
      "epoch": 1.454e-05,
      "grad_norm": 0.51953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "step": 79080
    },
    {
      "epoch": 1.4545e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0848,
      "step": 79090
    },
    {
      "epoch": 1.455e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "step": 79100
    },
    {
      "epoch": 1.4555e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0824,
      "step": 79110
    },
    {
      "epoch": 1.456e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0412,
      "step": 79120
    },
    {
      "epoch": 1.4565e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "step": 79130
    },
    {
      "epoch": 1.457e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0797,
      "step": 79140
    },
    {
      "epoch": 1.4575e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0497,
      "step": 79150
    },
    {
      "epoch": 1.458e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "step": 79160
    },
    {
      "epoch": 1.4585e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0369,
      "step": 79170
    },
    {
      "epoch": 1.459e-05,
      "grad_norm": 0.435546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "step": 79180
    },
    {
      "epoch": 1.4595e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "step": 79190
    },
    {
      "epoch": 1.46e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "step": 79200
    },
    {
      "epoch": 1.4605e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0285,
      "step": 79210
    },
    {
      "epoch": 1.461e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "step": 79220
    },
    {
      "epoch": 1.4615e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "step": 79230
    },
    {
      "epoch": 1.462e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0623,
      "step": 79240
    },
    {
      "epoch": 1.4625e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "step": 79250
    },
    {
      "epoch": 1.463e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "step": 79260
    },
    {
      "epoch": 1.4635e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0633,
      "step": 79270
    },
    {
      "epoch": 1.464e-05,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "step": 79280
    },
    {
      "epoch": 1.4645e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0212,
      "step": 79290
    },
    {
      "epoch": 1.465e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "step": 79300
    },
    {
      "epoch": 1.4655e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "step": 79310
    },
    {
      "epoch": 1.466e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.06,
      "step": 79320
    },
    {
      "epoch": 1.4665e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "step": 79330
    },
    {
      "epoch": 1.467e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0617,
      "step": 79340
    },
    {
      "epoch": 1.4675e-05,
      "grad_norm": 0.76171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "step": 79350
    },
    {
      "epoch": 1.468e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0729,
      "step": 79360
    },
    {
      "epoch": 1.4685e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "step": 79370
    },
    {
      "epoch": 1.469e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0538,
      "step": 79380
    },
    {
      "epoch": 1.4695e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0638,
      "step": 79390
    },
    {
      "epoch": 1.47e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0731,
      "step": 79400
    },
    {
      "epoch": 1.4705e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0204,
      "step": 79410
    },
    {
      "epoch": 1.471e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0489,
      "step": 79420
    },
    {
      "epoch": 1.4715e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "step": 79430
    },
    {
      "epoch": 1.472e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0576,
      "step": 79440
    },
    {
      "epoch": 1.4725e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0616,
      "step": 79450
    },
    {
      "epoch": 1.473e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "step": 79460
    },
    {
      "epoch": 1.4735e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "step": 79470
    },
    {
      "epoch": 1.474e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "step": 79480
    },
    {
      "epoch": 1.4745e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "step": 79490
    },
    {
      "epoch": 1.475e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "step": 79500
    },
    {
      "epoch": 1.475e-05,
      "eval_loss": 3.082590341567993,
      "eval_runtime": 0.0598,
      "eval_samples_per_second": 535.301,
      "eval_steps_per_second": 16.728,
      "step": 79500
    },
    {
      "epoch": 1.4755e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0369,
      "step": 79510
    },
    {
      "epoch": 1.476e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0569,
      "step": 79520
    },
    {
      "epoch": 1.4765e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0562,
      "step": 79530
    },
    {
      "epoch": 1.477e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0693,
      "step": 79540
    },
    {
      "epoch": 1.4775e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "step": 79550
    },
    {
      "epoch": 1.478e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "step": 79560
    },
    {
      "epoch": 1.4785e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0675,
      "step": 79570
    },
    {
      "epoch": 1.479e-05,
      "grad_norm": 0.5234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "step": 79580
    },
    {
      "epoch": 1.4795e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "step": 79590
    },
    {
      "epoch": 1.48e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "step": 79600
    },
    {
      "epoch": 1.4805e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "step": 79610
    },
    {
      "epoch": 1.481e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "step": 79620
    },
    {
      "epoch": 1.4815e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "step": 79630
    },
    {
      "epoch": 1.482e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "step": 79640
    },
    {
      "epoch": 1.4825e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "step": 79650
    },
    {
      "epoch": 1.483e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "step": 79660
    },
    {
      "epoch": 1.4835e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.035,
      "step": 79670
    },
    {
      "epoch": 1.484e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0352,
      "step": 79680
    },
    {
      "epoch": 1.4845e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "step": 79690
    },
    {
      "epoch": 1.485e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0355,
      "step": 79700
    },
    {
      "epoch": 1.4855e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "step": 79710
    },
    {
      "epoch": 1.486e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0377,
      "step": 79720
    },
    {
      "epoch": 1.4865e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "step": 79730
    },
    {
      "epoch": 1.487e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0486,
      "step": 79740
    },
    {
      "epoch": 1.4875e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "step": 79750
    },
    {
      "epoch": 1.488e-05,
      "grad_norm": 0.4296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0361,
      "step": 79760
    },
    {
      "epoch": 1.4885e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0438,
      "step": 79770
    },
    {
      "epoch": 1.489e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0422,
      "step": 79780
    },
    {
      "epoch": 1.4895e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "step": 79790
    },
    {
      "epoch": 1.49e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0483,
      "step": 79800
    },
    {
      "epoch": 1.4905e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0306,
      "step": 79810
    },
    {
      "epoch": 1.491e-05,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0345,
      "step": 79820
    },
    {
      "epoch": 1.4915e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "step": 79830
    },
    {
      "epoch": 1.492e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "step": 79840
    },
    {
      "epoch": 1.4925e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0516,
      "step": 79850
    },
    {
      "epoch": 1.493e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0355,
      "step": 79860
    },
    {
      "epoch": 1.4935e-05,
      "grad_norm": 0.515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "step": 79870
    },
    {
      "epoch": 1.494e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "step": 79880
    },
    {
      "epoch": 1.4945e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "step": 79890
    },
    {
      "epoch": 1.495e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "step": 79900
    },
    {
      "epoch": 1.4955e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0498,
      "step": 79910
    },
    {
      "epoch": 1.496e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0241,
      "step": 79920
    },
    {
      "epoch": 1.4965e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0597,
      "step": 79930
    },
    {
      "epoch": 1.497e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "step": 79940
    },
    {
      "epoch": 1.4975e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "step": 79950
    },
    {
      "epoch": 1.498e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0338,
      "step": 79960
    },
    {
      "epoch": 1.4985e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "step": 79970
    },
    {
      "epoch": 1.499e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0493,
      "step": 79980
    },
    {
      "epoch": 1.4995e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0499,
      "step": 79990
    },
    {
      "epoch": 1.5e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0266,
      "step": 80000
    },
    {
      "epoch": 1.5e-05,
      "eval_loss": 3.081814765930176,
      "eval_runtime": 0.0645,
      "eval_samples_per_second": 495.881,
      "eval_steps_per_second": 15.496,
      "step": 80000
    },
    {
      "epoch": 1.5005e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "step": 80010
    },
    {
      "epoch": 1.501e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0549,
      "step": 80020
    },
    {
      "epoch": 1.5015e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "step": 80030
    },
    {
      "epoch": 1.502e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.054,
      "step": 80040
    },
    {
      "epoch": 1.5025e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "step": 80050
    },
    {
      "epoch": 1.503e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0569,
      "step": 80060
    },
    {
      "epoch": 1.5035e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "step": 80070
    },
    {
      "epoch": 1.504e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0278,
      "step": 80080
    },
    {
      "epoch": 1.5045e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0412,
      "step": 80090
    },
    {
      "epoch": 1.505e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0698,
      "step": 80100
    },
    {
      "epoch": 1.5055e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "step": 80110
    },
    {
      "epoch": 1.506e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "step": 80120
    },
    {
      "epoch": 1.5065e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0781,
      "step": 80130
    },
    {
      "epoch": 1.507e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0399,
      "step": 80140
    },
    {
      "epoch": 1.5075e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0611,
      "step": 80150
    },
    {
      "epoch": 1.508e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0288,
      "step": 80160
    },
    {
      "epoch": 1.5085e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "step": 80170
    },
    {
      "epoch": 1.509e-05,
      "grad_norm": 0.43359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0287,
      "step": 80180
    },
    {
      "epoch": 1.5095e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0689,
      "step": 80190
    },
    {
      "epoch": 1.51e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0641,
      "step": 80200
    },
    {
      "epoch": 1.5105e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "step": 80210
    },
    {
      "epoch": 1.511e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "step": 80220
    },
    {
      "epoch": 1.5115e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "step": 80230
    },
    {
      "epoch": 1.512e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0616,
      "step": 80240
    },
    {
      "epoch": 1.5125e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0804,
      "step": 80250
    },
    {
      "epoch": 1.513e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "step": 80260
    },
    {
      "epoch": 1.5135e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0438,
      "step": 80270
    },
    {
      "epoch": 1.514e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "step": 80280
    },
    {
      "epoch": 1.5145e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0362,
      "step": 80290
    },
    {
      "epoch": 1.515e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0667,
      "step": 80300
    },
    {
      "epoch": 1.5155e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "step": 80310
    },
    {
      "epoch": 1.516e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0578,
      "step": 80320
    },
    {
      "epoch": 1.5165e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0382,
      "step": 80330
    },
    {
      "epoch": 1.517e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0284,
      "step": 80340
    },
    {
      "epoch": 1.5175e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "step": 80350
    },
    {
      "epoch": 1.518e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "step": 80360
    },
    {
      "epoch": 1.5185e-05,
      "grad_norm": 0.431640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "step": 80370
    },
    {
      "epoch": 1.519e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "step": 80380
    },
    {
      "epoch": 1.5195e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0599,
      "step": 80390
    },
    {
      "epoch": 1.52e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0489,
      "step": 80400
    },
    {
      "epoch": 1.5205e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 80410
    },
    {
      "epoch": 1.521e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "step": 80420
    },
    {
      "epoch": 1.5215e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "step": 80430
    },
    {
      "epoch": 1.522e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "step": 80440
    },
    {
      "epoch": 1.5225e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0331,
      "step": 80450
    },
    {
      "epoch": 1.523e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0566,
      "step": 80460
    },
    {
      "epoch": 1.5235e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0304,
      "step": 80470
    },
    {
      "epoch": 1.524e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0674,
      "step": 80480
    },
    {
      "epoch": 1.5245e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "step": 80490
    },
    {
      "epoch": 1.525e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0525,
      "step": 80500
    },
    {
      "epoch": 1.525e-05,
      "eval_loss": 3.081955909729004,
      "eval_runtime": 0.1072,
      "eval_samples_per_second": 298.406,
      "eval_steps_per_second": 9.325,
      "step": 80500
    },
    {
      "epoch": 1.5255e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0249,
      "step": 80510
    },
    {
      "epoch": 1.526e-05,
      "grad_norm": 0.49609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "step": 80520
    },
    {
      "epoch": 1.5265e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0263,
      "step": 80530
    },
    {
      "epoch": 1.527e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0789,
      "step": 80540
    },
    {
      "epoch": 1.5275e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0239,
      "step": 80550
    },
    {
      "epoch": 1.528e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "step": 80560
    },
    {
      "epoch": 1.5285e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0214,
      "step": 80570
    },
    {
      "epoch": 1.529e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "step": 80580
    },
    {
      "epoch": 1.5295e-05,
      "grad_norm": 0.5078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "step": 80590
    },
    {
      "epoch": 1.53e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "step": 80600
    },
    {
      "epoch": 1.5305e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "step": 80610
    },
    {
      "epoch": 1.531e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0495,
      "step": 80620
    },
    {
      "epoch": 1.5315e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0269,
      "step": 80630
    },
    {
      "epoch": 1.532e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0213,
      "step": 80640
    },
    {
      "epoch": 1.5325e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "step": 80650
    },
    {
      "epoch": 1.533e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "step": 80660
    },
    {
      "epoch": 1.5335e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0828,
      "step": 80670
    },
    {
      "epoch": 1.534e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "step": 80680
    },
    {
      "epoch": 1.5345e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0317,
      "step": 80690
    },
    {
      "epoch": 1.535e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "step": 80700
    },
    {
      "epoch": 1.5355e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0331,
      "step": 80710
    },
    {
      "epoch": 1.536e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "step": 80720
    },
    {
      "epoch": 1.5365e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0416,
      "step": 80730
    },
    {
      "epoch": 1.537e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0608,
      "step": 80740
    },
    {
      "epoch": 1.5375e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0401,
      "step": 80750
    },
    {
      "epoch": 1.538e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0707,
      "step": 80760
    },
    {
      "epoch": 1.5385e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.065,
      "step": 80770
    },
    {
      "epoch": 1.539e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0312,
      "step": 80780
    },
    {
      "epoch": 1.5395e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "step": 80790
    },
    {
      "epoch": 1.54e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "step": 80800
    },
    {
      "epoch": 1.5405e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "step": 80810
    },
    {
      "epoch": 1.541e-05,
      "grad_norm": 0.44140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0292,
      "step": 80820
    },
    {
      "epoch": 1.5415e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "step": 80830
    },
    {
      "epoch": 1.542e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "step": 80840
    },
    {
      "epoch": 1.5425e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "step": 80850
    },
    {
      "epoch": 1.543e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "step": 80860
    },
    {
      "epoch": 1.5435e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "step": 80870
    },
    {
      "epoch": 1.544e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "step": 80880
    },
    {
      "epoch": 1.5445e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0639,
      "step": 80890
    },
    {
      "epoch": 1.545e-05,
      "grad_norm": 0.478515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0149,
      "step": 80900
    },
    {
      "epoch": 1.5455e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0293,
      "step": 80910
    },
    {
      "epoch": 1.546e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0256,
      "step": 80920
    },
    {
      "epoch": 1.5465e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "step": 80930
    },
    {
      "epoch": 1.547e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0288,
      "step": 80940
    },
    {
      "epoch": 1.5475e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0625,
      "step": 80950
    },
    {
      "epoch": 1.548e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0276,
      "step": 80960
    },
    {
      "epoch": 1.5485e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0368,
      "step": 80970
    },
    {
      "epoch": 1.549e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0546,
      "step": 80980
    },
    {
      "epoch": 1.5495e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0308,
      "step": 80990
    },
    {
      "epoch": 1.55e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "step": 81000
    },
    {
      "epoch": 1.55e-05,
      "eval_loss": 3.081806182861328,
      "eval_runtime": 0.0929,
      "eval_samples_per_second": 344.456,
      "eval_steps_per_second": 10.764,
      "step": 81000
    },
    {
      "epoch": 1.5505e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0595,
      "step": 81010
    },
    {
      "epoch": 1.551e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "step": 81020
    },
    {
      "epoch": 1.5515e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0765,
      "step": 81030
    },
    {
      "epoch": 1.552e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0207,
      "step": 81040
    },
    {
      "epoch": 1.5525e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0416,
      "step": 81050
    },
    {
      "epoch": 1.553e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0555,
      "step": 81060
    },
    {
      "epoch": 1.5535e-05,
      "grad_norm": 0.46484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "step": 81070
    },
    {
      "epoch": 1.554e-05,
      "grad_norm": 0.431640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0541,
      "step": 81080
    },
    {
      "epoch": 1.5545e-05,
      "grad_norm": 0.4375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "step": 81090
    },
    {
      "epoch": 1.555e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0675,
      "step": 81100
    },
    {
      "epoch": 1.5555e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0479,
      "step": 81110
    },
    {
      "epoch": 1.556e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0285,
      "step": 81120
    },
    {
      "epoch": 1.5565e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.025,
      "step": 81130
    },
    {
      "epoch": 1.557e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "step": 81140
    },
    {
      "epoch": 1.5575e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0511,
      "step": 81150
    },
    {
      "epoch": 1.558e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "step": 81160
    },
    {
      "epoch": 1.5585e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0487,
      "step": 81170
    },
    {
      "epoch": 1.559e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.079,
      "step": 81180
    },
    {
      "epoch": 1.5595e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0321,
      "step": 81190
    },
    {
      "epoch": 1.56e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "step": 81200
    },
    {
      "epoch": 1.5605e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0503,
      "step": 81210
    },
    {
      "epoch": 1.561e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "step": 81220
    },
    {
      "epoch": 1.5615e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "step": 81230
    },
    {
      "epoch": 1.562e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0207,
      "step": 81240
    },
    {
      "epoch": 1.5625e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0688,
      "step": 81250
    },
    {
      "epoch": 1.563e-05,
      "grad_norm": 0.5546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "step": 81260
    },
    {
      "epoch": 1.5635e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "step": 81270
    },
    {
      "epoch": 1.564e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "step": 81280
    },
    {
      "epoch": 1.5645e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "step": 81290
    },
    {
      "epoch": 1.565e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "step": 81300
    },
    {
      "epoch": 1.5655e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.067,
      "step": 81310
    },
    {
      "epoch": 1.566e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0438,
      "step": 81320
    },
    {
      "epoch": 1.5665e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "step": 81330
    },
    {
      "epoch": 1.567e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "step": 81340
    },
    {
      "epoch": 1.5675e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.066,
      "step": 81350
    },
    {
      "epoch": 1.568e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0335,
      "step": 81360
    },
    {
      "epoch": 1.5685e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.049,
      "step": 81370
    },
    {
      "epoch": 1.569e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "step": 81380
    },
    {
      "epoch": 1.5695e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0361,
      "step": 81390
    },
    {
      "epoch": 1.57e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0682,
      "step": 81400
    },
    {
      "epoch": 1.5705e-05,
      "grad_norm": 0.77734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "step": 81410
    },
    {
      "epoch": 1.571e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "step": 81420
    },
    {
      "epoch": 1.5715e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "step": 81430
    },
    {
      "epoch": 1.572e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0347,
      "step": 81440
    },
    {
      "epoch": 1.5725e-05,
      "grad_norm": 0.53125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "step": 81450
    },
    {
      "epoch": 1.573e-05,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0659,
      "step": 81460
    },
    {
      "epoch": 1.5735e-05,
      "grad_norm": 0.4921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "step": 81470
    },
    {
      "epoch": 1.574e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "step": 81480
    },
    {
      "epoch": 1.5745e-05,
      "grad_norm": 0.45703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "step": 81490
    },
    {
      "epoch": 1.575e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0503,
      "step": 81500
    },
    {
      "epoch": 1.575e-05,
      "eval_loss": 3.082416296005249,
      "eval_runtime": 0.0918,
      "eval_samples_per_second": 348.489,
      "eval_steps_per_second": 10.89,
      "step": 81500
    },
    {
      "epoch": 1.5755e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0301,
      "step": 81510
    },
    {
      "epoch": 1.576e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0276,
      "step": 81520
    },
    {
      "epoch": 1.5765e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "step": 81530
    },
    {
      "epoch": 1.577e-05,
      "grad_norm": 0.5,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "step": 81540
    },
    {
      "epoch": 1.5775e-05,
      "grad_norm": 0.455078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "step": 81550
    },
    {
      "epoch": 1.578e-05,
      "grad_norm": 0.4453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0567,
      "step": 81560
    },
    {
      "epoch": 1.5785e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "step": 81570
    },
    {
      "epoch": 1.579e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "step": 81580
    },
    {
      "epoch": 1.5795e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "step": 81590
    },
    {
      "epoch": 1.58e-05,
      "grad_norm": 0.447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0296,
      "step": 81600
    },
    {
      "epoch": 1.5805e-05,
      "grad_norm": 0.44921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0313,
      "step": 81610
    },
    {
      "epoch": 1.581e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "step": 81620
    },
    {
      "epoch": 1.5815e-05,
      "grad_norm": 0.4765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "step": 81630
    },
    {
      "epoch": 1.582e-05,
      "grad_norm": 0.50390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "step": 81640
    },
    {
      "epoch": 1.5825e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0503,
      "step": 81650
    },
    {
      "epoch": 1.583e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "step": 81660
    },
    {
      "epoch": 1.5835e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "step": 81670
    },
    {
      "epoch": 1.584e-05,
      "grad_norm": 0.46875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "step": 81680
    },
    {
      "epoch": 1.5845e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "step": 81690
    },
    {
      "epoch": 1.585e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0637,
      "step": 81700
    },
    {
      "epoch": 1.5855e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "step": 81710
    },
    {
      "epoch": 1.586e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0525,
      "step": 81720
    },
    {
      "epoch": 1.5865e-05,
      "grad_norm": 0.462890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0543,
      "step": 81730
    },
    {
      "epoch": 1.587e-05,
      "grad_norm": 0.48046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "step": 81740
    },
    {
      "epoch": 1.5875e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0227,
      "step": 81750
    },
    {
      "epoch": 1.588e-05,
      "grad_norm": 0.466796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0787,
      "step": 81760
    },
    {
      "epoch": 1.5885e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "step": 81770
    },
    {
      "epoch": 1.589e-05,
      "grad_norm": 0.458984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0526,
      "step": 81780
    },
    {
      "epoch": 1.5895e-05,
      "grad_norm": 0.451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0401,
      "step": 81790
    },
    {
      "epoch": 1.59e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "step": 81800
    },
    {
      "epoch": 1.5905e-05,
      "grad_norm": 0.47265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "step": 81810
    },
    {
      "epoch": 1.591e-05,
      "grad_norm": 0.470703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0334,
      "step": 81820
    },
    {
      "epoch": 1.5915e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "step": 81830
    },
    {
      "epoch": 1.592e-05,
      "grad_norm": 0.490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "step": 81840
    },
    {
      "epoch": 1.5925e-05,
      "grad_norm": 0.498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "step": 81850
    },
    {
      "epoch": 1.593e-05,
      "grad_norm": 0.55078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0321,
      "step": 81860
    },
    {
      "epoch": 1.5935e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "step": 81870
    },
    {
      "epoch": 1.594e-05,
      "grad_norm": 0.486328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "step": 81880
    },
    {
      "epoch": 1.5945e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0231,
      "step": 81890
    },
    {
      "epoch": 1.595e-05,
      "grad_norm": 0.443359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0368,
      "step": 81900
    },
    {
      "epoch": 1.5955e-05,
      "grad_norm": 0.4609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0512,
      "step": 81910
    },
    {
      "epoch": 1.596e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0596,
      "step": 81920
    },
    {
      "epoch": 1.5965e-05,
      "grad_norm": 0.439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0357,
      "step": 81930
    },
    {
      "epoch": 1.597e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0331,
      "step": 81940
    },
    {
      "epoch": 1.5975e-05,
      "grad_norm": 0.484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.027,
      "step": 81950
    },
    {
      "epoch": 1.598e-05,
      "grad_norm": 0.48828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0579,
      "step": 81960
    },
    {
      "epoch": 1.5985e-05,
      "grad_norm": 0.453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0632,
      "step": 81970
    },
    {
      "epoch": 1.599e-05,
      "grad_norm": 0.494140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0653,
      "step": 81980
    },
    {
      "epoch": 1.5995e-05,
      "grad_norm": 0.474609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "step": 81990
    },
    {
      "epoch": 1.6e-05,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0667,
      "step": 82000
    },
    {
      "epoch": 1.6e-05,
      "eval_loss": 3.082991600036621,
      "eval_runtime": 0.0895,
      "eval_samples_per_second": 357.382,
      "eval_steps_per_second": 11.168,
      "step": 82000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4203776,
      "curriculum_late_phase": 1.0,
      "epoch": 5e-09,
      "grad_norm": 0.419921875,
      "host_rss_gib": 14.485172271728516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0002666666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86347453124931,
      "rewrite6_current_pass_fraction": 0.000754610426281864,
      "rewrite6_full_passes": 0.0,
      "step": 82010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4228352,
      "curriculum_late_phase": 1.0,
      "epoch": 1e-08,
      "grad_norm": 0.4296875,
      "host_rss_gib": 14.510940551757812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0597,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0005466666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.321159854522396,
      "rewrite6_current_pass_fraction": 0.001363167221670464,
      "rewrite6_full_passes": 0.0,
      "step": 82020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4252928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5e-08,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.520858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0349,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0009066666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.798897905687166,
      "rewrite6_current_pass_fraction": 0.0018938287472493234,
      "rewrite6_full_passes": 0.0,
      "step": 82030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4277504,
      "curriculum_late_phase": 1.0,
      "epoch": 2e-08,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.527912139892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0012266666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.002482911725185488,
      "rewrite6_full_passes": 0.0,
      "step": 82040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.430208,
      "curriculum_late_phase": 1.0,
      "epoch": 2.5e-08,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.532970428466797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0311,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0014266666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.720012071656146,
      "rewrite6_current_pass_fraction": 0.003145021518568285,
      "rewrite6_full_passes": 0.0,
      "step": 82050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4326656,
      "curriculum_late_phase": 1.0,
      "epoch": 3e-08,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.536087036132812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0348,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00164,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.796818119887625,
      "rewrite6_current_pass_fraction": 0.003724367587778232,
      "rewrite6_full_passes": 0.0,
      "step": 82060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4351232,
      "curriculum_late_phase": 1.0,
      "epoch": 3.5e-08,
      "grad_norm": 0.44140625,
      "host_rss_gib": 14.538379669189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0019733333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.004279371385172636,
      "rewrite6_full_passes": 0.0,
      "step": 82070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4375808,
      "curriculum_late_phase": 1.0,
      "epoch": 4e-08,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.542137145996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0023866666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.062621991982052,
      "rewrite6_current_pass_fraction": 0.004863585908745692,
      "rewrite6_full_passes": 0.0,
      "step": 82080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4400384,
      "curriculum_late_phase": 1.0,
      "epoch": 4.5e-08,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.545825958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0366,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0027066666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.83428610357238,
      "rewrite6_current_pass_fraction": 0.005486748067223618,
      "rewrite6_full_passes": 0.0,
      "step": 82090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.442496,
      "curriculum_late_phase": 1.0,
      "epoch": 5e-08,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.547012329101562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00312,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.006109910225701544,
      "rewrite6_full_passes": 0.0,
      "step": 82100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4449536,
      "curriculum_late_phase": 1.0,
      "epoch": 5.5e-08,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.549003601074219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0033333333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14704120581873,
      "rewrite6_current_pass_fraction": 0.006635703296917294,
      "rewrite6_full_passes": 0.0,
      "step": 82110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4474112,
      "curriculum_late_phase": 1.0,
      "epoch": 6e-08,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.55020523071289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0224,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0036533333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.540529846395334,
      "rewrite6_current_pass_fraction": 0.0073903137231991585,
      "rewrite6_full_passes": 0.0,
      "step": 82120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4498688,
      "curriculum_late_phase": 1.0,
      "epoch": 6.5e-08,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.554557800292969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.004026666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.221185526912038,
      "rewrite6_current_pass_fraction": 0.007891764522599366,
      "rewrite6_full_passes": 0.0,
      "step": 82130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4523264,
      "curriculum_late_phase": 1.0,
      "epoch": 7e-08,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.555896759033203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0256,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00444,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.606364821685087,
      "rewrite6_current_pass_fraction": 0.00844189986563066,
      "rewrite6_full_passes": 0.0,
      "step": 82140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.454784,
      "curriculum_late_phase": 1.0,
      "epoch": 7.5e-08,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.559032440185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00476,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.936624630107648,
      "rewrite6_current_pass_fraction": 0.009094272750287239,
      "rewrite6_full_passes": 0.0,
      "step": 82150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4572416,
      "curriculum_late_phase": 1.0,
      "epoch": 8e-08,
      "grad_norm": 0.439453125,
      "host_rss_gib": 14.561603546142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.005253333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.229675699044027,
      "rewrite6_current_pass_fraction": 0.009746645634943818,
      "rewrite6_full_passes": 0.0,
      "step": 82160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4596992,
      "curriculum_late_phase": 1.0,
      "epoch": 8.5e-08,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.562095642089844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.035,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.005546666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.800977899475697,
      "rewrite6_current_pass_fraction": 0.010360070884695528,
      "rewrite6_full_passes": 0.0,
      "step": 82170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4621568,
      "curriculum_late_phase": 1.0,
      "epoch": 9e-08,
      "grad_norm": 0.5,
      "host_rss_gib": 14.56243896484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0213,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.005946666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.517947686029533,
      "rewrite6_current_pass_fraction": 0.010890732410274386,
      "rewrite6_full_passes": 0.0,
      "step": 82180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4646144,
      "curriculum_late_phase": 1.0,
      "epoch": 9.5e-08,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.563953399658203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00632,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.011567447566746509,
      "rewrite6_full_passes": 0.0,
      "step": 82190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.467072,
      "curriculum_late_phase": 1.0,
      "epoch": 1e-07,
      "grad_norm": 0.427734375,
      "host_rss_gib": 14.564643859863281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.006706666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.045778632630523,
      "rewrite6_current_pass_fraction": 0.012141925181593348,
      "rewrite6_full_passes": 0.0,
      "step": 82200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4695296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.05e-07,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.568511962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.007026666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.012643375980993554,
      "rewrite6_full_passes": 0.0,
      "step": 82210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4719872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1e-07,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.570701599121094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00744,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34676060372526,
      "rewrite6_current_pass_fraction": 0.013251932776382154,
      "rewrite6_full_passes": 0.0,
      "step": 82220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4744448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.15e-07,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.570735931396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0428,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.007653333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.96385994124339,
      "rewrite6_current_pass_fraction": 0.013845884208681427,
      "rewrite6_full_passes": 0.0,
      "step": 82230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4769024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2e-07,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.57088851928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00796,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.07737098892312,
      "rewrite6_current_pass_fraction": 0.014434967186617593,
      "rewrite6_full_passes": 0.0,
      "step": 82240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.47936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.25e-07,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.572490692138672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.008413333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.014916944168565364,
      "rewrite6_full_passes": 0.0,
      "step": 82250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4818176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.572559356689453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0497,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.008693333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10901076930934,
      "rewrite6_current_pass_fraction": 0.01546221105723355,
      "rewrite6_full_passes": 0.0,
      "step": 82260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4842752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.35e-07,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.575515747070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0588,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00908,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.301979443133135,
      "rewrite6_current_pass_fraction": 0.016124320850616346,
      "rewrite6_full_passes": 0.0,
      "step": 82270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4867328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4e-07,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.575637817382812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0243,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.009453333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.579593952252267,
      "rewrite6_current_pass_fraction": 0.01671827228291562,
      "rewrite6_full_passes": 0.0,
      "step": 82280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4891904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.45e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.57571792602539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.009746666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.989031673191437,
      "rewrite6_current_pass_fraction": 0.017307355260851785,
      "rewrite6_full_passes": 0.0,
      "step": 82290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.491648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5e-07,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.5758056640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01008,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.01788183287569862,
      "rewrite6_full_passes": 0.0,
      "step": 82300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4941056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.55e-07,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.576335906982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0532,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.010493333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18302175066643,
      "rewrite6_current_pass_fraction": 0.018475784307997897,
      "rewrite6_full_passes": 0.0,
      "step": 82310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4965632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6e-07,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.577667236328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0248,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.010946666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.589886322106427,
      "rewrite6_current_pass_fraction": 0.018972366653034995,
      "rewrite6_full_passes": 0.0,
      "step": 82320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.4990208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.65e-07,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.577735900878906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01136,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.014233629348723,
      "rewrite6_current_pass_fraction": 0.01960526572023914,
      "rewrite6_full_passes": 0.0,
      "step": 82330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5014784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7e-07,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.577938079833984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0748,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.011653333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.645552368098528,
      "rewrite6_current_pass_fraction": 0.020131058791454888,
      "rewrite6_full_passes": 0.0,
      "step": 82340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.503936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.75e-07,
      "grad_norm": 0.490234375,
      "host_rss_gib": 14.578758239746094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0884,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.011906666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.94194276666499,
      "rewrite6_current_pass_fraction": 0.020676325680123075,
      "rewrite6_full_passes": 0.0,
      "step": 82350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5063936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8e-07,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.578903198242188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0375,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.012266666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.853045401483406,
      "rewrite6_current_pass_fraction": 0.02127027711242235,
      "rewrite6_full_passes": 0.0,
      "step": 82360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5088512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.85e-07,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.579113006591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0746,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01256,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.641223690507097,
      "rewrite6_current_pass_fraction": 0.02186909699908473,
      "rewrite6_full_passes": 0.0,
      "step": 82370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5113088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.9e-07,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.5494384765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.012773333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.024743373380456,
      "rewrite6_current_pass_fraction": 0.022506864520651985,
      "rewrite6_full_passes": 0.0,
      "step": 82380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5137664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.95e-07,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.82406234741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0479,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.013093333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.071048726013295,
      "rewrite6_current_pass_fraction": 0.023042394500593952,
      "rewrite6_full_passes": 0.0,
      "step": 82390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.516224,
      "curriculum_late_phase": 1.0,
      "epoch": 2e-07,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.82436752319336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01344,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.02365095129598255,
      "rewrite6_full_passes": 0.0,
      "step": 82400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5186816,
      "curriculum_late_phase": 1.0,
      "epoch": 2.05e-07,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.824504852294922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.013746666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.02419134973028763,
      "rewrite6_full_passes": 0.0,
      "step": 82410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5211392,
      "curriculum_late_phase": 1.0,
      "epoch": 2.1e-07,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.108272552490234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0538,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.014026666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.19573537742345,
      "rewrite6_current_pass_fraction": 0.024692800529687836,
      "rewrite6_full_passes": 0.0,
      "step": 82420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5235968,
      "curriculum_late_phase": 1.0,
      "epoch": 2.15e-07,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.408638000488281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.075,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01432,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.649881911512058,
      "rewrite6_current_pass_fraction": 0.025359778777433742,
      "rewrite6_full_passes": 0.0,
      "step": 82430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5260544,
      "curriculum_late_phase": 1.0,
      "epoch": 2.2e-07,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.40890884399414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0146,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.025914782574828144,
      "rewrite6_full_passes": 0.0,
      "step": 82440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.528512,
      "curriculum_late_phase": 1.0,
      "epoch": 2.25e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.409412384033203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0533,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.014853333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18514015876014,
      "rewrite6_current_pass_fraction": 0.026494128644038092,
      "rewrite6_full_passes": 0.0,
      "step": 82450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5309696,
      "curriculum_late_phase": 1.0,
      "epoch": 2.3e-07,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.409656524658203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0429,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0152,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.9659564320603,
      "rewrite6_current_pass_fraction": 0.027127027711242233,
      "rewrite6_full_passes": 0.0,
      "step": 82460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5334272,
      "curriculum_late_phase": 1.0,
      "epoch": 2.35e-07,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.410076141357422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.015493333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.02775505832408327,
      "rewrite6_full_passes": 0.0,
      "step": 82470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5358848,
      "curriculum_late_phase": 1.0,
      "epoch": 2.4e-07,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.410076141357422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.015813333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.028314930575840783,
      "rewrite6_full_passes": 0.0,
      "step": 82480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5383424,
      "curriculum_late_phase": 1.0,
      "epoch": 2.45e-07,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.410602569580078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.048,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.016093333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.073155936244657,
      "rewrite6_current_pass_fraction": 0.028830986738330316,
      "rewrite6_full_passes": 0.0,
      "step": 82490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5408,
      "curriculum_late_phase": 1.0,
      "epoch": 2.5e-07,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.41116714477539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.016453333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.225430188469428,
      "rewrite6_current_pass_fraction": 0.029424938170629588,
      "rewrite6_full_passes": 0.0,
      "step": 82500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 2.5e-07,
      "eval_loss": 3.082862615585327,
      "eval_perplexity": 21.820777573490783,
      "eval_runtime": 0.1047,
      "eval_samples_per_second": 305.565,
      "eval_steps_per_second": 9.549,
      "step": 82500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5432576,
      "curriculum_late_phase": 1.0,
      "epoch": 2.55e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.415607452392578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0428,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.016733333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.96385994124339,
      "rewrite6_current_pass_fraction": 0.02994099433311912,
      "rewrite6_full_passes": 0.0,
      "step": 82510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5457152,
      "curriculum_late_phase": 1.0,
      "epoch": 2.6e-07,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.413032531738281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0793,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.017093333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.743176844084687,
      "rewrite6_current_pass_fraction": 0.030583630309049484,
      "rewrite6_full_passes": 0.0,
      "step": 82520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5481728,
      "curriculum_late_phase": 1.0,
      "epoch": 2.65e-07,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.41341781616211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.052,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.017453333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.157617370242413,
      "rewrite6_current_pass_fraction": 0.031172713286985648,
      "rewrite6_full_passes": 0.0,
      "step": 82530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5506304,
      "curriculum_late_phase": 1.0,
      "epoch": 2.7e-07,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.414131164550781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.017813333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.03177153317364803,
      "rewrite6_full_passes": 0.0,
      "step": 82540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.553088,
      "curriculum_late_phase": 1.0,
      "epoch": 2.75e-07,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.414379119873047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0361,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.018133333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.82387156437236,
      "rewrite6_current_pass_fraction": 0.0323606161515842,
      "rewrite6_full_passes": 0.0,
      "step": 82550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5555456,
      "curriculum_late_phase": 1.0,
      "epoch": 2.8e-07,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.415252685546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.018373333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.64348968071354,
      "rewrite6_current_pass_fraction": 0.032920488403341705,
      "rewrite6_full_passes": 0.0,
      "step": 82560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5580032,
      "curriculum_late_phase": 1.0,
      "epoch": 2.85e-07,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.418777465820312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01872,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.855130810292252,
      "rewrite6_current_pass_fraction": 0.03344628147455746,
      "rewrite6_full_passes": 0.0,
      "step": 82570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5604608,
      "curriculum_late_phase": 1.0,
      "epoch": 2.9e-07,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.418922424316406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0497,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01908,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10901076930934,
      "rewrite6_current_pass_fraction": 0.03414247044848202,
      "rewrite6_full_passes": 0.0,
      "step": 82580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5629184,
      "curriculum_late_phase": 1.0,
      "epoch": 2.95e-07,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.419132232666016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01948,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.928251654962274,
      "rewrite6_current_pass_fraction": 0.034716948063328854,
      "rewrite6_full_passes": 0.0,
      "step": 82590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.565376,
      "curriculum_late_phase": 1.0,
      "epoch": 3e-07,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.41922378540039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0332,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.019813333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.763569816631378,
      "rewrite6_current_pass_fraction": 0.035228135771455275,
      "rewrite6_full_passes": 0.0,
      "step": 82600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5678336,
      "curriculum_late_phase": 1.0,
      "epoch": 3.05e-07,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.419654846191406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0208,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.020106666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.507691276502573,
      "rewrite6_current_pass_fraction": 0.035846429475570096,
      "rewrite6_full_passes": 0.0,
      "step": 82610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5702912,
      "curriculum_late_phase": 1.0,
      "epoch": 3.1e-07,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.419883728027344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.020373333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.905243235092758,
      "rewrite6_current_pass_fraction": 0.036508539268952894,
      "rewrite6_full_passes": 0.0,
      "step": 82620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5727488,
      "curriculum_late_phase": 1.0,
      "epoch": 3.15e-07,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.420059204101562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0558,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02070666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23816926792417,
      "rewrite6_current_pass_fraction": 0.037204728242877454,
      "rewrite6_full_passes": 0.0,
      "step": 82630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5752064,
      "curriculum_late_phase": 1.0,
      "epoch": 3.2e-07,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.420059204101562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0617,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.021026666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.363844844993586,
      "rewrite6_current_pass_fraction": 0.03775973204027185,
      "rewrite6_full_passes": 0.0,
      "step": 82640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.577664,
      "curriculum_late_phase": 1.0,
      "epoch": 3.25e-07,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.4205322265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0633,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.021333333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.398054357057195,
      "rewrite6_current_pass_fraction": 0.03834394656384491,
      "rewrite6_full_passes": 0.0,
      "step": 82650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5801216,
      "curriculum_late_phase": 1.0,
      "epoch": 3.3e-07,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.420791625976562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.021733333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.038928161087417965,
      "rewrite6_full_passes": 0.0,
      "step": 82660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5825792,
      "curriculum_late_phase": 1.0,
      "epoch": 3.35e-07,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.034648895263672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.022053333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.014233629348723,
      "rewrite6_current_pass_fraction": 0.039497770247901694,
      "rewrite6_full_passes": 0.0,
      "step": 82670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5850368,
      "curriculum_late_phase": 1.0,
      "epoch": 3.4e-07,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.05048370361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0436,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0224,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.980637739420835,
      "rewrite6_current_pass_fraction": 0.04013066931510584,
      "rewrite6_full_passes": 0.0,
      "step": 82680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5874944,
      "curriculum_late_phase": 1.0,
      "epoch": 3.45e-07,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.393211364746094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02264,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.913607005029263,
      "rewrite6_current_pass_fraction": 0.040773305291036205,
      "rewrite6_full_passes": 0.0,
      "step": 82690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.589952,
      "curriculum_late_phase": 1.0,
      "epoch": 3.5e-07,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.393413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.022893333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.949190374231975,
      "rewrite6_current_pass_fraction": 0.04136725672333547,
      "rewrite6_full_passes": 0.0,
      "step": 82700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5924096,
      "curriculum_late_phase": 1.0,
      "epoch": 3.55e-07,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.393413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.023146666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.041927128975092987,
      "rewrite6_full_passes": 0.0,
      "step": 82710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5948672,
      "curriculum_late_phase": 1.0,
      "epoch": 3.6e-07,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.393413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0658,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02344,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45161641762865,
      "rewrite6_current_pass_fraction": 0.04249186968121361,
      "rewrite6_full_passes": 0.0,
      "step": 82720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5973248,
      "curriculum_late_phase": 1.0,
      "epoch": 3.65e-07,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.393535614013672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0369,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.023746666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.840537327040092,
      "rewrite6_current_pass_fraction": 0.043100426476602206,
      "rewrite6_full_passes": 0.0,
      "step": 82730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.5997824,
      "curriculum_late_phase": 1.0,
      "epoch": 3.7e-07,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.393539428710938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.023946666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.391635903566247,
      "rewrite6_current_pass_fraction": 0.04366516718272283,
      "rewrite6_full_passes": 0.0,
      "step": 82740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.60224,
      "curriculum_late_phase": 1.0,
      "epoch": 3.75e-07,
      "grad_norm": 0.498046875,
      "host_rss_gib": 16.393646240234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.024333333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.024743373380456,
      "rewrite6_current_pass_fraction": 0.044215302525754126,
      "rewrite6_full_passes": 0.0,
      "step": 82750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6046976,
      "curriculum_late_phase": 1.0,
      "epoch": 3.8e-07,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.393646240234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0284,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.024546666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.664143495580532,
      "rewrite6_current_pass_fraction": 0.04476543786878542,
      "rewrite6_full_passes": 0.0,
      "step": 82760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6071552,
      "curriculum_late_phase": 1.0,
      "epoch": 3.85e-07,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.393722534179688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.024893333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01003120287951,
      "rewrite6_current_pass_fraction": 0.045354520846721584,
      "rewrite6_full_passes": 0.0,
      "step": 82770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6096128,
      "curriculum_late_phase": 1.0,
      "epoch": 3.9e-07,
      "grad_norm": 0.4375,
      "host_rss_gib": 16.393722534179688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.025146666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.04589978773538977,
      "rewrite6_full_passes": 0.0,
      "step": 82780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6120704,
      "curriculum_late_phase": 1.0,
      "epoch": 3.95e-07,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.393844604492188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.025413333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.87808407616535,
      "rewrite6_current_pass_fraction": 0.046508344530778366,
      "rewrite6_full_passes": 0.0,
      "step": 82790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.614528,
      "curriculum_late_phase": 1.0,
      "epoch": 4e-07,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.394309997558594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.068,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.025693333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.498861924749583,
      "rewrite6_current_pass_fraction": 0.047077953691262095,
      "rewrite6_full_passes": 0.0,
      "step": 82800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6169856,
      "curriculum_late_phase": 1.0,
      "epoch": 4.05e-07,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.39432144165039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0337,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02604,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.77395419741855,
      "rewrite6_current_pass_fraction": 0.047744931939008,
      "rewrite6_full_passes": 0.0,
      "step": 82810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6194432,
      "curriculum_late_phase": 1.0,
      "epoch": 4.1e-07,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.394336700439453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.026466666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.04843625245856945,
      "rewrite6_full_passes": 0.0,
      "step": 82820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6219008,
      "curriculum_late_phase": 1.0,
      "epoch": 4.15e-07,
      "grad_norm": 0.42578125,
      "host_rss_gib": 16.39447021484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.026853333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10689997377394,
      "rewrite6_current_pass_fraction": 0.0490691515257736,
      "rewrite6_full_passes": 0.0,
      "step": 82830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6243584,
      "curriculum_late_phase": 1.0,
      "epoch": 4.2e-07,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.39462661743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0523,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.027106666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.16396560764148,
      "rewrite6_current_pass_fraction": 0.04959494459698935,
      "rewrite6_full_passes": 0.0,
      "step": 82840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.626816,
      "curriculum_late_phase": 1.0,
      "epoch": 4.25e-07,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.394882202148438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0285,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.027373333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.666210013254258,
      "rewrite6_current_pass_fraction": 0.05008665848766334,
      "rewrite6_full_passes": 0.0,
      "step": 82850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6292736,
      "curriculum_late_phase": 1.0,
      "epoch": 4.3e-07,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.394969940185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0543,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.027693333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.206335895020718,
      "rewrite6_current_pass_fraction": 0.05059784619578976,
      "rewrite6_full_passes": 0.0,
      "step": 82860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6317312,
      "curriculum_late_phase": 1.0,
      "epoch": 4.35e-07,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.395015716552734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02792,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.936624630107648,
      "rewrite6_current_pass_fraction": 0.0512258768086308,
      "rewrite6_full_passes": 0.0,
      "step": 82870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6341888,
      "curriculum_late_phase": 1.0,
      "epoch": 4.4e-07,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.654796600341797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0367,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.028226666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.836369636357645,
      "rewrite6_current_pass_fraction": 0.05178574906038831,
      "rewrite6_full_passes": 0.0,
      "step": 82880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6366464,
      "curriculum_late_phase": 1.0,
      "epoch": 4.45e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.656875610351562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02856,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.197855056943403,
      "rewrite6_current_pass_fraction": 0.052530622577943956,
      "rewrite6_full_passes": 0.0,
      "step": 82890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.639104,
      "curriculum_late_phase": 1.0,
      "epoch": 4.5e-07,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.657424926757812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.028906666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.05312944246460634,
      "rewrite6_full_passes": 0.0,
      "step": 82900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6415616,
      "curriculum_late_phase": 1.0,
      "epoch": 4.55e-07,
      "grad_norm": 0.44140625,
      "host_rss_gib": 14.658195495605469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.029253333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.053723393896905614,
      "rewrite6_full_passes": 0.0,
      "step": 82910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6440192,
      "curriculum_late_phase": 1.0,
      "epoch": 4.6e-07,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.658321380615234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02968,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.81346223114035,
      "rewrite6_current_pass_fraction": 0.05434168760102043,
      "rewrite6_full_passes": 0.0,
      "step": 82920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6464768,
      "curriculum_late_phase": 1.0,
      "epoch": 4.65e-07,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.658794403076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0604,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03004,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3360898913238,
      "rewrite6_current_pass_fraction": 0.05497458666822457,
      "rewrite6_full_passes": 0.0,
      "step": 82930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6489344,
      "curriculum_late_phase": 1.0,
      "epoch": 4.7e-07,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.65970230102539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.030333333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.821789281331817,
      "rewrite6_current_pass_fraction": 0.05553932737434519,
      "rewrite6_full_passes": 0.0,
      "step": 82940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.651392,
      "curriculum_late_phase": 1.0,
      "epoch": 4.75e-07,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.6605224609375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0486,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03068,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.085803623733216,
      "rewrite6_current_pass_fraction": 0.05602130435629296,
      "rewrite6_full_passes": 0.0,
      "step": 82950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6538496,
      "curriculum_late_phase": 1.0,
      "epoch": 4.8e-07,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.66116714477539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.030933333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13647032815547,
      "rewrite6_current_pass_fraction": 0.05668828260403887,
      "rewrite6_full_passes": 0.0,
      "step": 82960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6563072,
      "curriculum_late_phase": 1.0,
      "epoch": 4.85e-07,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.661293029785156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.031186666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.05729683939942747,
      "rewrite6_full_passes": 0.0,
      "step": 82970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6587648,
      "curriculum_late_phase": 1.0,
      "epoch": 4.9e-07,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.661689758300781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0189,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03144,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.468763656027395,
      "rewrite6_current_pass_fraction": 0.05790539619481607,
      "rewrite6_full_passes": 0.0,
      "step": 82980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6612224,
      "curriculum_late_phase": 1.0,
      "epoch": 4.95e-07,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.662399291992188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03166666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.028948742578034,
      "rewrite6_current_pass_fraction": 0.05853342680765711,
      "rewrite6_full_passes": 0.0,
      "step": 82990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.66368,
      "curriculum_late_phase": 1.0,
      "epoch": 5e-07,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.662487030029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.031933333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.64348968071354,
      "rewrite6_current_pass_fraction": 0.05915172051177192,
      "rewrite6_full_passes": 0.0,
      "step": 83000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 5e-07,
      "eval_loss": 3.0835492610931396,
      "eval_perplexity": 21.83576585761869,
      "eval_runtime": 0.0496,
      "eval_samples_per_second": 644.964,
      "eval_steps_per_second": 20.155,
      "step": 83000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6661376,
      "curriculum_late_phase": 1.0,
      "epoch": 5.05e-07,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.734214782714844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03224,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09845890211131,
      "rewrite6_current_pass_fraction": 0.059799224942065396,
      "rewrite6_full_passes": 0.0,
      "step": 83010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6685952,
      "curriculum_late_phase": 1.0,
      "epoch": 5.1e-07,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.734214782714844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0614,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03264,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.357436652816972,
      "rewrite6_current_pass_fraction": 0.06043699246363265,
      "rewrite6_full_passes": 0.0,
      "step": 83020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6710528,
      "curriculum_late_phase": 1.0,
      "epoch": 5.15e-07,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.734298706054688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0528,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.032946666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.174550236381975,
      "rewrite6_current_pass_fraction": 0.06102607544156881,
      "rewrite6_full_passes": 0.0,
      "step": 83030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6735104,
      "curriculum_late_phase": 1.0,
      "epoch": 5.2e-07,
      "grad_norm": 0.48046875,
      "host_rss_gib": 14.734310150146484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.033253333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.058409888808015,
      "rewrite6_current_pass_fraction": 0.061576210784600106,
      "rewrite6_full_passes": 0.0,
      "step": 83040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.675968,
      "curriculum_late_phase": 1.0,
      "epoch": 5.25e-07,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.734699249267578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0553,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03370666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.22755283761896,
      "rewrite6_current_pass_fraction": 0.06225292594107223,
      "rewrite6_full_passes": 0.0,
      "step": 83050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6784256,
      "curriculum_late_phase": 1.0,
      "epoch": 5.3e-07,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.734725952148438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0633,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03406666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.398054357057195,
      "rewrite6_current_pass_fraction": 0.06281279819282974,
      "rewrite6_full_passes": 0.0,
      "step": 83060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6808832,
      "curriculum_late_phase": 1.0,
      "epoch": 5.35e-07,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.734798431396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03444,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.06329964362914062,
      "rewrite6_full_passes": 0.0,
      "step": 83070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6833408,
      "curriculum_late_phase": 1.0,
      "epoch": 5.4e-07,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.735176086425781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0667,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03473333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.470931562916125,
      "rewrite6_current_pass_fraction": 0.06391306887889232,
      "rewrite6_full_passes": 0.0,
      "step": 83080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6857984,
      "curriculum_late_phase": 1.0,
      "epoch": 5.45e-07,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.735549926757812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0622,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.035,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.37452943834182,
      "rewrite6_current_pass_fraction": 0.06443886195010808,
      "rewrite6_full_passes": 0.0,
      "step": 83090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.688256,
      "curriculum_late_phase": 1.0,
      "epoch": 5.5e-07,
      "grad_norm": 0.515625,
      "host_rss_gib": 14.735771179199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.049,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03544,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.094239632271936,
      "rewrite6_current_pass_fraction": 0.06496952347568694,
      "rewrite6_full_passes": 0.0,
      "step": 83100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6907136,
      "curriculum_late_phase": 1.0,
      "epoch": 5.55e-07,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.736103057861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0612,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03572,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35316559260667,
      "rewrite6_current_pass_fraction": 0.06552452727308135,
      "rewrite6_full_passes": 0.0,
      "step": 83110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6931712,
      "curriculum_late_phase": 1.0,
      "epoch": 5.6e-07,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.73678207397461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03594666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23604555718468,
      "rewrite6_current_pass_fraction": 0.0661087417966544,
      "rewrite6_full_passes": 0.0,
      "step": 83120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6956288,
      "curriculum_late_phase": 1.0,
      "epoch": 5.65e-07,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.736957550048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0437,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03629333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.982735908101457,
      "rewrite6_current_pass_fraction": 0.06667348250277502,
      "rewrite6_full_passes": 0.0,
      "step": 83130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.6980864,
      "curriculum_late_phase": 1.0,
      "epoch": 5.7e-07,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.737251281738281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0779,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.036546666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.71275769487588,
      "rewrite6_current_pass_fraction": 0.06727717084380051,
      "rewrite6_full_passes": 0.0,
      "step": 83140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.700544,
      "curriculum_late_phase": 1.0,
      "epoch": 5.75e-07,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.737556457519531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0398,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03673333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.901062604522732,
      "rewrite6_current_pass_fraction": 0.0679295437284571,
      "rewrite6_full_passes": 0.0,
      "step": 83150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7030016,
      "curriculum_late_phase": 1.0,
      "epoch": 5.8e-07,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.738189697265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03710666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23604555718468,
      "rewrite6_current_pass_fraction": 0.06852836361511948,
      "rewrite6_full_passes": 0.0,
      "step": 83160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7054592,
      "curriculum_late_phase": 1.0,
      "epoch": 5.85e-07,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.738296508789062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0399,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.037373333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.903152815291975,
      "rewrite6_current_pass_fraction": 0.06911744659305563,
      "rewrite6_full_passes": 0.0,
      "step": 83170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7079168,
      "curriculum_late_phase": 1.0,
      "epoch": 5.9e-07,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.738296508789062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0661,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03756,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45805286797322,
      "rewrite6_current_pass_fraction": 0.06971626647971801,
      "rewrite6_full_passes": 0.0,
      "step": 83180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7103744,
      "curriculum_late_phase": 1.0,
      "epoch": 5.95e-07,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.738422393798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03794666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.0703150863663804,
      "rewrite6_full_passes": 0.0,
      "step": 83190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.712832,
      "curriculum_late_phase": 1.0,
      "epoch": 6e-07,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.738643646240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03833333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.07096259079667387,
      "rewrite6_full_passes": 0.0,
      "step": 83200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7152896,
      "curriculum_late_phase": 1.0,
      "epoch": 6.05e-07,
      "grad_norm": 0.6953125,
      "host_rss_gib": 14.739147186279297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0181,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03881333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.452395193360626,
      "rewrite6_current_pass_fraction": 0.07150785768534205,
      "rewrite6_full_passes": 0.0,
      "step": 83210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7177472,
      "curriculum_late_phase": 1.0,
      "epoch": 6.1e-07,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.739933013916016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.039066666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.959667588504388,
      "rewrite6_current_pass_fraction": 0.07214562520690931,
      "rewrite6_full_passes": 0.0,
      "step": 83220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7202048,
      "curriculum_late_phase": 1.0,
      "epoch": 6.15e-07,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.74017333984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03938666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.730374668125208,
      "rewrite6_current_pass_fraction": 0.07265681291503573,
      "rewrite6_full_passes": 0.0,
      "step": 83230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7226624,
      "curriculum_late_phase": 1.0,
      "epoch": 6.2e-07,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.74026107788086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03972,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.898972602764104,
      "rewrite6_current_pass_fraction": 0.07326050125606122,
      "rewrite6_full_passes": 0.0,
      "step": 83240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.72512,
      "curriculum_late_phase": 1.0,
      "epoch": 6.25e-07,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.74026107788086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0648,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.430175523444856,
      "rewrite6_current_pass_fraction": 0.07381063659909252,
      "rewrite6_full_passes": 0.0,
      "step": 83250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7275776,
      "curriculum_late_phase": 1.0,
      "epoch": 6.3e-07,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.644195556640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.040306666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.26367036864598,
      "rewrite6_current_pass_fraction": 0.07434616657903449,
      "rewrite6_full_passes": 0.0,
      "step": 83260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7300352,
      "curriculum_late_phase": 1.0,
      "epoch": 6.35e-07,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.391353607177734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0502,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04070666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11956791376016,
      "rewrite6_current_pass_fraction": 0.07494011801133377,
      "rewrite6_full_passes": 0.0,
      "step": 83270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7324928,
      "curriculum_late_phase": 1.0,
      "epoch": 6.4e-07,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.39138412475586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.041,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.066834937660975,
      "rewrite6_current_pass_fraction": 0.07561196471344278,
      "rewrite6_full_passes": 0.0,
      "step": 83280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7349504,
      "curriculum_late_phase": 1.0,
      "epoch": 6.45e-07,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.39217758178711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.041253333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.155501714289947,
      "rewrite6_current_pass_fraction": 0.0762351268719207,
      "rewrite6_full_passes": 0.0,
      "step": 83290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.737408,
      "curriculum_late_phase": 1.0,
      "epoch": 6.5e-07,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.392425537109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.041626666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.69516296961328,
      "rewrite6_current_pass_fraction": 0.07680473603240444,
      "rewrite6_full_passes": 0.0,
      "step": 83300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7398656,
      "curriculum_late_phase": 1.0,
      "epoch": 6.55e-07,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.393539428710938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.042,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.921974121142934,
      "rewrite6_current_pass_fraction": 0.07740355591906682,
      "rewrite6_full_passes": 0.0,
      "step": 83310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7423232,
      "curriculum_late_phase": 1.0,
      "epoch": 6.6e-07,
      "grad_norm": 0.54296875,
      "host_rss_gib": 15.39404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0572,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.042293333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.267923528021466,
      "rewrite6_current_pass_fraction": 0.07791961208155634,
      "rewrite6_full_passes": 0.0,
      "step": 83320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7447808,
      "curriculum_late_phase": 1.0,
      "epoch": 6.65e-07,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.394119262695312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0677,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04253333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.492413233524204,
      "rewrite6_current_pass_fraction": 0.07844053669840899,
      "rewrite6_full_passes": 0.0,
      "step": 83330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7472384,
      "curriculum_late_phase": 1.0,
      "epoch": 6.7e-07,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.394420623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.042866666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.07896632976962474,
      "rewrite6_full_passes": 0.0,
      "step": 83340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.749696,
      "curriculum_late_phase": 1.0,
      "epoch": 6.75e-07,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.394447326660156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04306666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.07958462347373955,
      "rewrite6_full_passes": 0.0,
      "step": 83350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7521536,
      "curriculum_late_phase": 1.0,
      "epoch": 6.8e-07,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.39461898803711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0795,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.747525914346035,
      "rewrite6_current_pass_fraction": 0.08012502190804463,
      "rewrite6_full_passes": 0.0,
      "step": 83360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7546112,
      "curriculum_late_phase": 1.0,
      "epoch": 6.85e-07,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.394744873046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0315,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.043706666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.728301734306807,
      "rewrite6_current_pass_fraction": 0.08078713170142743,
      "rewrite6_full_passes": 0.0,
      "step": 83370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7570688,
      "curriculum_late_phase": 1.0,
      "epoch": 6.9e-07,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.394996643066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04405333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.191496654234413,
      "rewrite6_current_pass_fraction": 0.08147358376662577,
      "rewrite6_full_passes": 0.0,
      "step": 83380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7595264,
      "curriculum_late_phase": 1.0,
      "epoch": 6.95e-07,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.395980834960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0728,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0444,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.602304525620752,
      "rewrite6_current_pass_fraction": 0.08206753519892504,
      "rewrite6_full_passes": 0.0,
      "step": 83390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.761984,
      "curriculum_late_phase": 1.0,
      "epoch": 7e-07,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.396034240722656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04466666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.08267122353995053,
      "rewrite6_full_passes": 0.0,
      "step": 83400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7644416,
      "curriculum_late_phase": 1.0,
      "epoch": 7.05e-07,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.396251678466797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04494666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86347453124931,
      "rewrite6_current_pass_fraction": 0.08321162197425562,
      "rewrite6_full_passes": 0.0,
      "step": 83410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7668992,
      "curriculum_late_phase": 1.0,
      "epoch": 7.1e-07,
      "grad_norm": 0.5703125,
      "host_rss_gib": 15.396793365478516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0237,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04529333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.567249899467065,
      "rewrite6_current_pass_fraction": 0.08396136394617437,
      "rewrite6_full_passes": 0.0,
      "step": 83420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7693568,
      "curriculum_late_phase": 1.0,
      "epoch": 7.15e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.397117614746094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04568,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.049988209300682,
      "rewrite6_current_pass_fraction": 0.08457965765028919,
      "rewrite6_full_passes": 0.0,
      "step": 83430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7718144,
      "curriculum_late_phase": 1.0,
      "epoch": 7.2e-07,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.397335052490234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04598666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.911515748893304,
      "rewrite6_current_pass_fraction": 0.08515900371949914,
      "rewrite6_full_passes": 0.0,
      "step": 83440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.774272,
      "curriculum_late_phase": 1.0,
      "epoch": 7.25e-07,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.397411346435547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04636,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.197855056943403,
      "rewrite6_current_pass_fraction": 0.08577242896925084,
      "rewrite6_full_passes": 0.0,
      "step": 83450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7767296,
      "curriculum_late_phase": 1.0,
      "epoch": 7.3e-07,
      "grad_norm": 0.4296875,
      "host_rss_gib": 15.397872924804688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0252,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.046586666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.59812392404582,
      "rewrite6_current_pass_fraction": 0.08637124885591323,
      "rewrite6_full_passes": 0.0,
      "step": 83460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7791872,
      "curriculum_late_phase": 1.0,
      "epoch": 7.35e-07,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.398090362548828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0322,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04698666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.742816625139927,
      "rewrite6_current_pass_fraction": 0.08693598956203384,
      "rewrite6_full_passes": 0.0,
      "step": 83470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7816448,
      "curriculum_late_phase": 1.0,
      "epoch": 7.4e-07,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.398792266845703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0555,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04732,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.231798772765842,
      "rewrite6_current_pass_fraction": 0.08756888862923799,
      "rewrite6_full_passes": 0.0,
      "step": 83480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7841024,
      "curriculum_late_phase": 1.0,
      "epoch": 7.45e-07,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.39889144897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04772,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.054198627970376,
      "rewrite6_current_pass_fraction": 0.08822126151389457,
      "rewrite6_full_passes": 0.0,
      "step": 83490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.78656,
      "curriculum_late_phase": 1.0,
      "epoch": 7.5e-07,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.398963928222656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0386,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04805333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.87599637214468,
      "rewrite6_current_pass_fraction": 0.08886389748982493,
      "rewrite6_full_passes": 0.0,
      "step": 83500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 7.5e-07,
      "eval_loss": 3.0831298828125,
      "eval_perplexity": 21.82661033162543,
      "eval_runtime": 0.051,
      "eval_samples_per_second": 628.016,
      "eval_steps_per_second": 19.626,
      "step": 83500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7890176,
      "curriculum_late_phase": 1.0,
      "epoch": 7.55e-07,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.451122283935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04849333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.216941714202072,
      "rewrite6_current_pass_fraction": 0.08947732273957663,
      "rewrite6_full_passes": 0.0,
      "step": 83510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7914752,
      "curriculum_late_phase": 1.0,
      "epoch": 7.6e-07,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.45162582397461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0224,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04884,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.540529846395334,
      "rewrite6_current_pass_fraction": 0.0900615372631497,
      "rewrite6_full_passes": 0.0,
      "step": 83520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7939328,
      "curriculum_late_phase": 1.0,
      "epoch": 7.65e-07,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.451698303222656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.049146666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.882260110570105,
      "rewrite6_current_pass_fraction": 0.0907285155108956,
      "rewrite6_full_passes": 0.0,
      "step": 83530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.7963904,
      "curriculum_late_phase": 1.0,
      "epoch": 7.7e-07,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.451934814453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0821,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.049453333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.80414305210811,
      "rewrite6_current_pass_fraction": 0.09140036221300461,
      "rewrite6_full_passes": 0.0,
      "step": 83540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.798848,
      "curriculum_late_phase": 1.0,
      "epoch": 7.75e-07,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.452442169189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0199,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04985333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.48924265747756,
      "rewrite6_current_pass_fraction": 0.09196023446476213,
      "rewrite6_full_passes": 0.0,
      "step": 83550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8013056,
      "curriculum_late_phase": 1.0,
      "epoch": 7.8e-07,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.452518463134766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0269,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05010666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.63317051587937,
      "rewrite6_current_pass_fraction": 0.09252010671651964,
      "rewrite6_full_passes": 0.0,
      "step": 83560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8037632,
      "curriculum_late_phase": 1.0,
      "epoch": 7.85e-07,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.452762603759766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0715,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.050333333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.574239775777293,
      "rewrite6_current_pass_fraction": 0.09308484742264025,
      "rewrite6_full_passes": 0.0,
      "step": 83570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8062208,
      "curriculum_late_phase": 1.0,
      "epoch": 7.9e-07,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.453357696533203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0638,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.050813333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.40875605943837,
      "rewrite6_current_pass_fraction": 0.09380051021401725,
      "rewrite6_full_passes": 0.0,
      "step": 83580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8086784,
      "curriculum_late_phase": 1.0,
      "epoch": 7.95e-07,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.453487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0588,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.051133333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.301979443133135,
      "rewrite6_current_pass_fraction": 0.09423380265233394,
      "rewrite6_full_passes": 0.0,
      "step": 83590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.811136,
      "curriculum_late_phase": 1.0,
      "epoch": 8e-07,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.453632354736328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05148,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.0947644641779128,
      "rewrite6_full_passes": 0.0,
      "step": 83600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8135936,
      "curriculum_late_phase": 1.0,
      "epoch": 8.05e-07,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.453678131103516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05172,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.74904040364757,
      "rewrite6_current_pass_fraction": 0.09532920488403342,
      "rewrite6_full_passes": 0.0,
      "step": 83610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8160512,
      "curriculum_late_phase": 1.0,
      "epoch": 8.1e-07,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.453754425048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0412,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05198666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.93034458477251,
      "rewrite6_current_pass_fraction": 0.09595236704251134,
      "rewrite6_full_passes": 0.0,
      "step": 83620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8185088,
      "curriculum_late_phase": 1.0,
      "epoch": 8.15e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.454002380371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.052333333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.909424701872513,
      "rewrite6_current_pass_fraction": 0.09661447683589414,
      "rewrite6_full_passes": 0.0,
      "step": 83630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8209664,
      "curriculum_late_phase": 1.0,
      "epoch": 8.2e-07,
      "grad_norm": 0.6171875,
      "host_rss_gib": 15.45461654663086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05278666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.189377610522946,
      "rewrite6_current_pass_fraction": 0.0972522443574614,
      "rewrite6_full_passes": 0.0,
      "step": 83640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.823424,
      "curriculum_late_phase": 1.0,
      "epoch": 8.25e-07,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.45645523071289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05318666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.89061468541791,
      "rewrite6_current_pass_fraction": 0.09784132733539756,
      "rewrite6_full_passes": 0.0,
      "step": 83650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8258816,
      "curriculum_late_phase": 1.0,
      "epoch": 8.3e-07,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.456729888916016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0498,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05352,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.111121775934837,
      "rewrite6_current_pass_fraction": 0.09841093649588128,
      "rewrite6_full_passes": 0.0,
      "step": 83660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8283392,
      "curriculum_late_phase": 1.0,
      "epoch": 8.35e-07,
      "grad_norm": 0.42578125,
      "host_rss_gib": 15.456768035888672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0528,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0538,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.174550236381975,
      "rewrite6_current_pass_fraction": 0.09894159802146014,
      "rewrite6_full_passes": 0.0,
      "step": 83670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8307968,
      "curriculum_late_phase": 1.0,
      "epoch": 8.4e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.456989288330078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05414666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11534442254061,
      "rewrite6_current_pass_fraction": 0.09940896964031859,
      "rewrite6_full_passes": 0.0,
      "step": 83680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8332544,
      "curriculum_late_phase": 1.0,
      "epoch": 8.45e-07,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.457588195800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0522,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05449333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.161849316897023,
      "rewrite6_current_pass_fraction": 0.10006134252497517,
      "rewrite6_full_passes": 0.0,
      "step": 83690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.835712,
      "curriculum_late_phase": 1.0,
      "epoch": 8.5e-07,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.162914276123047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05484,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.031051742600535,
      "rewrite6_current_pass_fraction": 0.10069424159217931,
      "rewrite6_full_passes": 0.0,
      "step": 83700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8381696,
      "curriculum_late_phase": 1.0,
      "epoch": 8.55e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.163021087646484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05508,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.10125411384393683,
      "rewrite6_full_passes": 0.0,
      "step": 83710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8406272,
      "curriculum_late_phase": 1.0,
      "epoch": 8.6e-07,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.163578033447266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0511,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.055466666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.138584080874157,
      "rewrite6_current_pass_fraction": 0.10182372300442055,
      "rewrite6_full_passes": 0.0,
      "step": 83720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8430848,
      "curriculum_late_phase": 1.0,
      "epoch": 8.65e-07,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.16403579711914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0428,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05584,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.96385994124339,
      "rewrite6_current_pass_fraction": 0.10244201670853538,
      "rewrite6_full_passes": 0.0,
      "step": 83730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8455424,
      "curriculum_late_phase": 1.0,
      "epoch": 8.7e-07,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.643260955810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05608,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.197855056943403,
      "rewrite6_current_pass_fraction": 0.10313333722809682,
      "rewrite6_full_passes": 0.0,
      "step": 83740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.848,
      "curriculum_late_phase": 1.0,
      "epoch": 8.75e-07,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.644168853759766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0293,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.056413333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.68274959621593,
      "rewrite6_current_pass_fraction": 0.10368834102549122,
      "rewrite6_full_passes": 0.0,
      "step": 83750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8504576,
      "curriculum_late_phase": 1.0,
      "epoch": 8.8e-07,
      "grad_norm": 0.44140625,
      "host_rss_gib": 14.645015716552734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05674666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.949190374231975,
      "rewrite6_current_pass_fraction": 0.10431150318396916,
      "rewrite6_full_passes": 0.0,
      "step": 83760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8529152,
      "curriculum_late_phase": 1.0,
      "epoch": 8.85e-07,
      "grad_norm": 0.43359375,
      "host_rss_gib": 14.64645767211914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0516,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05706666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14915601567805,
      "rewrite6_current_pass_fraction": 0.10497361297735196,
      "rewrite6_full_passes": 0.0,
      "step": 83770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8553728,
      "curriculum_late_phase": 1.0,
      "epoch": 8.9e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.646736145019531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05748,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91988202833721,
      "rewrite6_current_pass_fraction": 0.10553835368347257,
      "rewrite6_full_passes": 0.0,
      "step": 83780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8578304,
      "curriculum_late_phase": 1.0,
      "epoch": 8.95e-07,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.647270202636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0588,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05785333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.301979443133135,
      "rewrite6_current_pass_fraction": 0.10608362057214075,
      "rewrite6_full_passes": 0.0,
      "step": 83790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.860288,
      "curriculum_late_phase": 1.0,
      "epoch": 9e-07,
      "grad_norm": 0.4375,
      "host_rss_gib": 14.647342681884766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05842666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.970150042693795,
      "rewrite6_current_pass_fraction": 0.10667757200444003,
      "rewrite6_full_passes": 0.0,
      "step": 83800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8627456,
      "curriculum_late_phase": 1.0,
      "epoch": 9.05e-07,
      "grad_norm": 0.49609375,
      "host_rss_gib": 14.649192810058594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0344,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.058706666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.788501056163312,
      "rewrite6_current_pass_fraction": 0.10728126034546552,
      "rewrite6_full_passes": 0.0,
      "step": 83810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8652032,
      "curriculum_late_phase": 1.0,
      "epoch": 9.1e-07,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.649795532226562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.059,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.896882809995212,
      "rewrite6_current_pass_fraction": 0.10795310704757453,
      "rewrite6_full_passes": 0.0,
      "step": 83820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8676608,
      "curriculum_late_phase": 1.0,
      "epoch": 9.15e-07,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.649795532226562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0658,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05932,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45161641762865,
      "rewrite6_current_pass_fraction": 0.10853245311678449,
      "rewrite6_full_passes": 0.0,
      "step": 83830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8701184,
      "curriculum_late_phase": 1.0,
      "epoch": 9.2e-07,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.649940490722656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0348,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05974666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.796818119887625,
      "rewrite6_current_pass_fraction": 0.10904850927927402,
      "rewrite6_full_passes": 0.0,
      "step": 83840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.872576,
      "curriculum_late_phase": 1.0,
      "epoch": 9.25e-07,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.650558471679688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0289,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.060053333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.67447815077682,
      "rewrite6_current_pass_fraction": 0.10953048626122179,
      "rewrite6_full_passes": 0.0,
      "step": 83850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8750336,
      "curriculum_late_phase": 1.0,
      "epoch": 9.3e-07,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.65427017211914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06028,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.17878556994849,
      "rewrite6_current_pass_fraction": 0.11013904305661039,
      "rewrite6_full_passes": 0.0,
      "step": 83860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8774912,
      "curriculum_late_phase": 1.0,
      "epoch": 9.35e-07,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.654739379882812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06052,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.11068917839964168,
      "rewrite6_full_passes": 0.0,
      "step": 83870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8799488,
      "curriculum_late_phase": 1.0,
      "epoch": 9.4e-07,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.654964447021484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06098666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.041569897791508,
      "rewrite6_current_pass_fraction": 0.11124418219703608,
      "rewrite6_full_passes": 0.0,
      "step": 83880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8824064,
      "curriculum_late_phase": 1.0,
      "epoch": 9.45e-07,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.655170440673828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.061386666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.81346223114035,
      "rewrite6_current_pass_fraction": 0.11181865981188292,
      "rewrite6_full_passes": 0.0,
      "step": 83890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.884864,
      "curriculum_late_phase": 1.0,
      "epoch": 9.5e-07,
      "grad_norm": 0.515625,
      "host_rss_gib": 14.655502319335938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.061693333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01003120287951,
      "rewrite6_current_pass_fraction": 0.11248076960526572,
      "rewrite6_full_passes": 0.0,
      "step": 83900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8873216,
      "curriculum_late_phase": 1.0,
      "epoch": 9.55e-07,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.65554428100586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0317,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.062106666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.732447809247343,
      "rewrite6_current_pass_fraction": 0.11308445794629121,
      "rewrite6_full_passes": 0.0,
      "step": 83910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8897792,
      "curriculum_late_phase": 1.0,
      "epoch": 9.6e-07,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.655792236328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0292,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06246666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.680681424666606,
      "rewrite6_current_pass_fraction": 0.11370275165040603,
      "rewrite6_full_passes": 0.0,
      "step": 83920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8922368,
      "curriculum_late_phase": 1.0,
      "epoch": 9.65e-07,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.656177520751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0686,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06276,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.51176511247365,
      "rewrite6_current_pass_fraction": 0.11426749235652665,
      "rewrite6_full_passes": 0.0,
      "step": 83930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8946944,
      "curriculum_late_phase": 1.0,
      "epoch": 9.7e-07,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.65628433227539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06301333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.114910128332457,
      "rewrite6_full_passes": 0.0,
      "step": 83940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.897152,
      "curriculum_late_phase": 1.0,
      "epoch": 9.75e-07,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.656436920166016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0399,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06330666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.903152815291975,
      "rewrite6_current_pass_fraction": 0.11550407976475628,
      "rewrite6_full_passes": 0.0,
      "step": 83950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.8996096,
      "curriculum_late_phase": 1.0,
      "epoch": 9.8e-07,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.656497955322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0332,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0636,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.763569816631378,
      "rewrite6_current_pass_fraction": 0.11609803119705556,
      "rewrite6_full_passes": 0.0,
      "step": 83960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9020672,
      "curriculum_late_phase": 1.0,
      "epoch": 9.85e-07,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.656719207763672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06393333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91988202833721,
      "rewrite6_current_pass_fraction": 0.11674553562734903,
      "rewrite6_full_passes": 0.0,
      "step": 83970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9045248,
      "curriculum_late_phase": 1.0,
      "epoch": 9.9e-07,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.657085418701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0702,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06417333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.54621148640419,
      "rewrite6_current_pass_fraction": 0.11730540787910654,
      "rewrite6_full_passes": 0.0,
      "step": 83980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9069824,
      "curriculum_late_phase": 1.0,
      "epoch": 9.95e-07,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.65740966796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06441333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.909424701872513,
      "rewrite6_current_pass_fraction": 0.11793830694631069,
      "rewrite6_full_passes": 0.0,
      "step": 83990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.90944,
      "curriculum_late_phase": 1.0,
      "epoch": 1e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.657909393310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0566,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06469333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25516660136536,
      "rewrite6_current_pass_fraction": 0.11846410001752644,
      "rewrite6_full_passes": 0.0,
      "step": 84000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1e-06,
      "eval_loss": 3.0836780071258545,
      "eval_perplexity": 21.83857730682173,
      "eval_runtime": 0.0503,
      "eval_samples_per_second": 635.805,
      "eval_steps_per_second": 19.869,
      "step": 84000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9118976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.005e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.812694549560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.065,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.191496654234413,
      "rewrite6_current_pass_fraction": 0.1190580514498257,
      "rewrite6_full_passes": 0.0,
      "step": 84010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9143552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.01e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.813278198242188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0599,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06538666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.325424512944927,
      "rewrite6_current_pass_fraction": 0.11951568615995793,
      "rewrite6_full_passes": 0.0,
      "step": 84020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9168128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.015e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 14.813617706298828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06576,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.014233629348723,
      "rewrite6_current_pass_fraction": 0.12011450604662031,
      "rewrite6_full_passes": 0.0,
      "step": 84030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9192704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.02e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.813766479492188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06593333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.066834937660975,
      "rewrite6_current_pass_fraction": 0.12074253665946136,
      "rewrite6_full_passes": 0.0,
      "step": 84040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.921728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.025e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 14.81463623046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06626666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.949190374231975,
      "rewrite6_current_pass_fraction": 0.12134135654612374,
      "rewrite6_full_passes": 0.0,
      "step": 84050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9241856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.03e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.81500244140625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06661333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.12197425561332788,
      "rewrite6_full_passes": 0.0,
      "step": 84060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9266432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.035e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.81503677368164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06692,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.155501714289947,
      "rewrite6_current_pass_fraction": 0.12254873322817472,
      "rewrite6_full_passes": 0.0,
      "step": 84070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9291008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.04e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.815086364746094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06725333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21482012611182,
      "rewrite6_current_pass_fraction": 0.12319136920410508,
      "rewrite6_full_passes": 0.0,
      "step": 84080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9315584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.045e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.815166473388672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0572,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0676,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.267923528021466,
      "rewrite6_current_pass_fraction": 0.1238047944538568,
      "rewrite6_full_passes": 0.0,
      "step": 84090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.934016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.05e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.49655532836914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.054,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06796,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.199974948441902,
      "rewrite6_current_pass_fraction": 0.12446203579287647,
      "rewrite6_full_passes": 0.0,
      "step": 84100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9364736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.055e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 15.497215270996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0247,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06826666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.58782743642022,
      "rewrite6_current_pass_fraction": 0.1250316449533602,
      "rewrite6_full_passes": 0.0,
      "step": 84110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9389312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.06e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.08563995361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0566,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06864,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25516660136536,
      "rewrite6_current_pass_fraction": 0.12573270238164788,
      "rewrite6_full_passes": 0.0,
      "step": 84120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9413888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.065e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.085731506347656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06898666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.1263509960857627,
      "rewrite6_full_passes": 0.0,
      "step": 84130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9438464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.07e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 16.086380004882812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06922666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.892703851343008,
      "rewrite6_current_pass_fraction": 0.12692060524624643,
      "rewrite6_full_passes": 0.0,
      "step": 84140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.946304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.075e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 16.086429595947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0375,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06945333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.853045401483406,
      "rewrite6_current_pass_fraction": 0.12749508286109326,
      "rewrite6_full_passes": 0.0,
      "step": 84150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9487616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.08e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.08707046508789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0499,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06976,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.113232993671563,
      "rewrite6_current_pass_fraction": 0.1281182450195712,
      "rewrite6_full_passes": 0.0,
      "step": 84160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9512192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.085e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.08722686767578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07008,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.74904040364757,
      "rewrite6_current_pass_fraction": 0.12880956553913264,
      "rewrite6_full_passes": 0.0,
      "step": 84170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9536768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.09e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.087448120117188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07046666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.759417517911764,
      "rewrite6_current_pass_fraction": 0.1293499639734377,
      "rewrite6_full_passes": 0.0,
      "step": 84180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9561344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.095e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.087501525878906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07081333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.12988062549901658,
      "rewrite6_full_passes": 0.0,
      "step": 84190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.958592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.08751678466797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0191,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07124,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.472857818161167,
      "rewrite6_current_pass_fraction": 0.1304453662051372,
      "rewrite6_full_passes": 0.0,
      "step": 84200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9610496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.105e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.087520599365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07154666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.049988209300682,
      "rewrite6_current_pass_fraction": 0.1310587914548889,
      "rewrite6_full_passes": 0.0,
      "step": 84210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9635072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.11e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.087520599365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07186666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.1315066892562949,
      "rewrite6_full_passes": 0.0,
      "step": 84220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9659648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.115e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.088340759277344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07233333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.662077184548252,
      "rewrite6_current_pass_fraction": 0.13208603532550486,
      "rewrite6_full_passes": 0.0,
      "step": 84230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9684224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.12e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.088340759277344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07261333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.13259235457926818,
      "rewrite6_full_passes": 0.0,
      "step": 84240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.97088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.125e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.088340759277344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0017,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07298666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.11971137601164,
      "rewrite6_current_pass_fraction": 0.13319604292029366,
      "rewrite6_full_passes": 0.0,
      "step": 84250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9733376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.13e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.088489532470703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01213231105346,
      "rewrite6_current_pass_fraction": 0.13378025744386673,
      "rewrite6_full_passes": 0.0,
      "step": 84260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9757952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.135e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 16.088489532470703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07378666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.1344423672372495,
      "rewrite6_full_passes": 0.0,
      "step": 84270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9782528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.14e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.088653564453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07402666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.13496329185410216,
      "rewrite6_full_passes": 0.0,
      "step": 84280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9807104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.145e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.088668823242188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07434666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.1355329010145859,
      "rewrite6_full_passes": 0.0,
      "step": 84290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.983168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.15e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.08869171142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07464,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.13602948335962298,
      "rewrite6_full_passes": 0.0,
      "step": 84300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9856256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.155e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.0887451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0679,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07498666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.49671214604783,
      "rewrite6_current_pass_fraction": 0.13665751397246403,
      "rewrite6_full_passes": 0.0,
      "step": 84310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9880832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.16e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.0887451171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07537333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.13725633385912642,
      "rewrite6_full_passes": 0.0,
      "step": 84320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9905408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.165e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 16.089763641357422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07568,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.259418059817307,
      "rewrite6_current_pass_fraction": 0.13787949601760432,
      "rewrite6_full_passes": 0.0,
      "step": 84330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9929984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.17e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.089763641357422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07598666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.1384101575431832,
      "rewrite6_full_passes": 0.0,
      "step": 84340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.995456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.175e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.08984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07638666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.151271037028916,
      "rewrite6_current_pass_fraction": 0.13910147806274464,
      "rewrite6_full_passes": 0.0,
      "step": 84350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 14.9979136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.18e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.08984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07678666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.13971490331249634,
      "rewrite6_full_passes": 0.0,
      "step": 84360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0003712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.185e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.089859008789062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0513,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07704,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.1428122204902,
      "rewrite6_current_pass_fraction": 0.1402893809273432,
      "rewrite6_full_passes": 0.0,
      "step": 84370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0028288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.19e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.089859008789062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07742666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.090021206202152,
      "rewrite6_current_pass_fraction": 0.14083464781601138,
      "rewrite6_full_passes": 0.0,
      "step": 84380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0052864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.195e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.089935302734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.048,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07777333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.073155936244657,
      "rewrite6_current_pass_fraction": 0.1414626784288524,
      "rewrite6_full_passes": 0.0,
      "step": 84390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.007744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.08993911743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0663,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07818666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.462344907736483,
      "rewrite6_current_pass_fraction": 0.14210531440478277,
      "rewrite6_full_passes": 0.0,
      "step": 84400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0102016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.205e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.08993911743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0391,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07856,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.886436980265266,
      "rewrite6_current_pass_fraction": 0.14269926583708203,
      "rewrite6_full_passes": 0.0,
      "step": 84410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0126592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.21e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.08993911743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07890666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.132243456791066,
      "rewrite6_current_pass_fraction": 0.14325426963447643,
      "rewrite6_full_passes": 0.0,
      "step": 84420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0151168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.215e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.090057373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0576,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07921333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.276432399093437,
      "rewrite6_current_pass_fraction": 0.14386769488422815,
      "rewrite6_full_passes": 0.0,
      "step": 84430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0175744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.22e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.090057373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0284,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07952,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.664143495580532,
      "rewrite6_current_pass_fraction": 0.1444373040447119,
      "rewrite6_full_passes": 0.0,
      "step": 84440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.020032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.225e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.090057373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0753,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07981333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.656377850427624,
      "rewrite6_current_pass_fraction": 0.14508967692936847,
      "rewrite6_full_passes": 0.0,
      "step": 84450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0224896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.23e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.090057373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08013333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.959667588504388,
      "rewrite6_current_pass_fraction": 0.1457225759965726,
      "rewrite6_full_passes": 0.0,
      "step": 84460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0249472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.235e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.090057373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0806,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.14640415960740785,
      "rewrite6_full_passes": 0.0,
      "step": 84470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0274048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.24e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.090057373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08093333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08158688469643,
      "rewrite6_current_pass_fraction": 0.14712955930751107,
      "rewrite6_full_passes": 0.0,
      "step": 84480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0298624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.245e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.090057373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0301,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08124,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.699302416138053,
      "rewrite6_current_pass_fraction": 0.14767969465054234,
      "rewrite6_full_passes": 0.0,
      "step": 84490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.03232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.25e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.09009552001953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08157333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.014233629348723,
      "rewrite6_current_pass_fraction": 0.14833693598956205,
      "rewrite6_full_passes": 0.0,
      "step": 84500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.25e-06,
      "eval_loss": 3.0825366973876953,
      "eval_perplexity": 21.813666943797855,
      "eval_runtime": 0.0495,
      "eval_samples_per_second": 647.113,
      "eval_steps_per_second": 20.222,
      "step": 84500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0347776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.255e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.142459869384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0625,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08196,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.380942759123343,
      "rewrite6_current_pass_fraction": 0.14890167669568266,
      "rewrite6_full_passes": 0.0,
      "step": 84510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0372352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.26e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.142459869384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0572,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08222666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.267923528021466,
      "rewrite6_current_pass_fraction": 0.1494274697668984,
      "rewrite6_full_passes": 0.0,
      "step": 84520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0396928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.265e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.142459869384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0346,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08242666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.792659172172286,
      "rewrite6_current_pass_fraction": 0.15009931646900743,
      "rewrite6_full_passes": 0.0,
      "step": 84530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0421504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.27e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.142459869384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0335,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08276,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.769799822030453,
      "rewrite6_current_pass_fraction": 0.15057155654222898,
      "rewrite6_full_passes": 0.0,
      "step": 84540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.044608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.275e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.142459869384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08305333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.049988209300682,
      "rewrite6_current_pass_fraction": 0.15117037642889136,
      "rewrite6_full_passes": 0.0,
      "step": 84550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0470656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.28e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.142459869384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0711,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08328,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.56561180557606,
      "rewrite6_current_pass_fraction": 0.15171077486319642,
      "rewrite6_full_passes": 0.0,
      "step": 84560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0495232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.285e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.142459869384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0298,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08368,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.69309355678868,
      "rewrite6_current_pass_fraction": 0.152363147747853,
      "rewrite6_full_passes": 0.0,
      "step": 84570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0519808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.29e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.142459869384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08394666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.678613459924104,
      "rewrite6_current_pass_fraction": 0.15298144145196782,
      "rewrite6_full_passes": 0.0,
      "step": 84580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0544384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.295e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.142494201660156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08422666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.15358999824735642,
      "rewrite6_full_passes": 0.0,
      "step": 84590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.056896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.142494201660156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0401,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08449333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.90733386394596,
      "rewrite6_current_pass_fraction": 0.15420829195147126,
      "rewrite6_full_passes": 0.0,
      "step": 84600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0593536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.305e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.526111602783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0368,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08490666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.838453377506596,
      "rewrite6_current_pass_fraction": 0.15471461120523455,
      "rewrite6_full_passes": 0.0,
      "step": 84610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0618112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.31e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.527225494384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0171,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08528,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.431953020956982,
      "rewrite6_current_pass_fraction": 0.15532316800062315,
      "rewrite6_full_passes": 0.0,
      "step": 84620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0642688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.315e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.528236389160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08553333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.1558927771611069,
      "rewrite6_full_passes": 0.0,
      "step": 84630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0667264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.32e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.529083251953125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08577333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.986932874965774,
      "rewrite6_current_pass_fraction": 0.1565062024108586,
      "rewrite6_full_passes": 0.0,
      "step": 84640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.069184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.325e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.52962875366211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0206,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08621333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.503590148373757,
      "rewrite6_current_pass_fraction": 0.15713910147806273,
      "rewrite6_full_passes": 0.0,
      "step": 84650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0716416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.33e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.529830932617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0747,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08646666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.643387921085868,
      "rewrite6_current_pass_fraction": 0.15766489454927848,
      "rewrite6_full_passes": 0.0,
      "step": 84660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0740992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.335e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.529830932617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0346,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08684,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.792659172172286,
      "rewrite6_current_pass_fraction": 0.15824424061848844,
      "rewrite6_full_passes": 0.0,
      "step": 84670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0765568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.34e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.529949188232422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0773,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08722666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.699733947773797,
      "rewrite6_current_pass_fraction": 0.15886253432260325,
      "rewrite6_full_passes": 0.0,
      "step": 84680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0790144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.345e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.530094146728516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.02,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08750666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.49129168419294,
      "rewrite6_current_pass_fraction": 0.15947595957235497,
      "rewrite6_full_passes": 0.0,
      "step": 84690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.081472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.35e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.53079605102539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08774666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.028948742578034,
      "rewrite6_current_pass_fraction": 0.16008451636774357,
      "rewrite6_full_passes": 0.0,
      "step": 84700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0839296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.355e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.531097412109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.042,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08797333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.947095559937008,
      "rewrite6_current_pass_fraction": 0.1606687308913166,
      "rewrite6_full_passes": 0.0,
      "step": 84710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0863872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.36e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 15.531097412109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0323,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08838666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.744891010519986,
      "rewrite6_current_pass_fraction": 0.1613308406846994,
      "rewrite6_full_passes": 0.0,
      "step": 84720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0888448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.365e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.531097412109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08873333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.97853978054658,
      "rewrite6_current_pass_fraction": 0.161939397480088,
      "rewrite6_full_passes": 0.0,
      "step": 84730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0913024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.37e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.531097412109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08901333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.974344492133298,
      "rewrite6_current_pass_fraction": 0.16256742809292907,
      "rewrite6_full_passes": 0.0,
      "step": 84740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.09376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.375e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.53182601928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0892,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.197855056943403,
      "rewrite6_current_pass_fraction": 0.1632100640688594,
      "rewrite6_full_passes": 0.0,
      "step": 84750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0962176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.38e-06,
      "grad_norm": 0.55078125,
      "host_rss_gib": 15.53182601928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0627,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08948,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.38521937532253,
      "rewrite6_current_pass_fraction": 0.163862436953516,
      "rewrite6_full_passes": 0.0,
      "step": 84760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.0986752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.385e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.53182601928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0373,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08985333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.848875209436216,
      "rewrite6_current_pass_fraction": 0.16442717765963663,
      "rewrite6_full_passes": 0.0,
      "step": 84770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1011328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.39e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.53182601928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0902,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.16505520827247766,
      "rewrite6_full_passes": 0.0,
      "step": 84780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1035904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.395e-06,
      "grad_norm": 0.5546875,
      "host_rss_gib": 15.53182601928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0269,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09050666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.63317051587937,
      "rewrite6_current_pass_fraction": 0.16561021206987206,
      "rewrite6_full_passes": 0.0,
      "step": 84790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.106048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.53182601928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0908,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.974344492133298,
      "rewrite6_current_pass_fraction": 0.166189558139082,
      "rewrite6_full_passes": 0.0,
      "step": 84800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1085056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.405e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.53182601928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09108,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.210577586354717,
      "rewrite6_current_pass_fraction": 0.16672995657338707,
      "rewrite6_full_passes": 0.0,
      "step": 84810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1109632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.41e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.53182601928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0556,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09128,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.233922058805657,
      "rewrite6_current_pass_fraction": 0.16730443418823393,
      "rewrite6_full_passes": 0.0,
      "step": 84820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1134208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.415e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.53182601928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09162666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.16792272789234874,
      "rewrite6_full_passes": 0.0,
      "step": 84830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1158784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.42e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.53182601928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0696,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09188,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.53328763705487,
      "rewrite6_current_pass_fraction": 0.16848746859846936,
      "rewrite6_full_passes": 0.0,
      "step": 84840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.118336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.425e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.531829833984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09217333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.16902299857841133,
      "rewrite6_full_passes": 0.0,
      "step": 84850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1207936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.43e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.531829833984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0608,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09246666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34462603439513,
      "rewrite6_current_pass_fraction": 0.16964616073688926,
      "rewrite6_full_passes": 0.0,
      "step": 84860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1232512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.435e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.531829833984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09269333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.039465846006077,
      "rewrite6_current_pass_fraction": 0.17024011216918852,
      "rewrite6_full_passes": 0.0,
      "step": 84870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1257088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.44e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.53188705444336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0223,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09306666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.538475896109915,
      "rewrite6_current_pass_fraction": 0.17069774687932077,
      "rewrite6_full_passes": 0.0,
      "step": 84880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1281664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.445e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.53188705444336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0588,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09344,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.301979443133135,
      "rewrite6_current_pass_fraction": 0.17135011976397732,
      "rewrite6_full_passes": 0.0,
      "step": 84890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.130624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.45e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.53188705444336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09366666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.898972602764104,
      "rewrite6_current_pass_fraction": 0.17190025510700863,
      "rewrite6_full_passes": 0.0,
      "step": 84900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1330816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.455e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.53188705444336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.094,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.265796842004743,
      "rewrite6_current_pass_fraction": 0.17252828571984966,
      "rewrite6_full_passes": 0.0,
      "step": 84910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1355392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.46e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.53188705444336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0489,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09426666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.092130313776398,
      "rewrite6_current_pass_fraction": 0.17311736869778582,
      "rewrite6_full_passes": 0.0,
      "step": 84920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1379968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.465e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.531967163085938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0487,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09456,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08791230952813,
      "rewrite6_current_pass_fraction": 0.17376487312807928,
      "rewrite6_full_passes": 0.0,
      "step": 84930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1404544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.47e-06,
      "grad_norm": 0.57421875,
      "host_rss_gib": 15.531967163085938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0334,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09498666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.767722945893784,
      "rewrite6_current_pass_fraction": 0.17444645673891454,
      "rewrite6_full_passes": 0.0,
      "step": 84940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.142912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.475e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.531967163085938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0631,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09533333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.393775174118343,
      "rewrite6_current_pass_fraction": 0.17506475044302935,
      "rewrite6_full_passes": 0.0,
      "step": 84950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1453696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.48e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.822429656982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0331,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09557333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.761493563464107,
      "rewrite6_current_pass_fraction": 0.17577067632568014,
      "rewrite6_full_passes": 0.0,
      "step": 84960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1478272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.485e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.453720092773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0502,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09577333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11956791376016,
      "rewrite6_current_pass_fraction": 0.17635975930361628,
      "rewrite6_full_passes": 0.0,
      "step": 84970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1502848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.49e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.453720092773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09614666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.014233629348723,
      "rewrite6_current_pass_fraction": 0.17693423691846313,
      "rewrite6_full_passes": 0.0,
      "step": 84980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1527424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.495e-06,
      "grad_norm": 0.53515625,
      "host_rss_gib": 16.453720092773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0711,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09630666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.56561180557606,
      "rewrite6_current_pass_fraction": 0.17748924071585753,
      "rewrite6_full_passes": 0.0,
      "step": 84990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.453720092773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09669333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.155501714289947,
      "rewrite6_current_pass_fraction": 0.178092929056883,
      "rewrite6_full_passes": 0.0,
      "step": 85000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.5e-06,
      "eval_loss": 3.0828537940979004,
      "eval_perplexity": 21.820585082624806,
      "eval_runtime": 0.051,
      "eval_samples_per_second": 628.013,
      "eval_steps_per_second": 19.625,
      "step": 85000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1576576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.505e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 16.47747039794922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0279,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.097,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.653814006420227,
      "rewrite6_current_pass_fraction": 0.1787501703959027,
      "rewrite6_full_passes": 0.0,
      "step": 85010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1601152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.51e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.47747039794922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0688,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09718666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.51606789576013,
      "rewrite6_current_pass_fraction": 0.17931491110202333,
      "rewrite6_full_passes": 0.0,
      "step": 85020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1625728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.515e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.47747039794922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0761,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09746666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.67370988459726,
      "rewrite6_current_pass_fraction": 0.1799672839866799,
      "rewrite6_full_passes": 0.0,
      "step": 85030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1650304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.52e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.477771759033203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09773333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.913607005029263,
      "rewrite6_current_pass_fraction": 0.1806245253256996,
      "rewrite6_full_passes": 0.0,
      "step": 85040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.167488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.525e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0406,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09794666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.917790144730294,
      "rewrite6_current_pass_fraction": 0.18119900294054644,
      "rewrite6_full_passes": 0.0,
      "step": 85050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1699456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.53e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0304,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09841333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.705513138424664,
      "rewrite6_current_pass_fraction": 0.1817880859184826,
      "rewrite6_full_passes": 0.0,
      "step": 85060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1724032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.535e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09876,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.18237230044205566,
      "rewrite6_full_passes": 0.0,
      "step": 85070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1748608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.54e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0556,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09909333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.233922058805657,
      "rewrite6_current_pass_fraction": 0.18293704114817627,
      "rewrite6_full_passes": 0.0,
      "step": 85080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1773184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.545e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09954666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.18355533485229109,
      "rewrite6_full_passes": 0.0,
      "step": 85090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.179776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.55e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0366,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09994666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.83428610357238,
      "rewrite6_current_pass_fraction": 0.18414441783022725,
      "rewrite6_full_passes": 0.0,
      "step": 85100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1822336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.555e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10024,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.240293191045346,
      "rewrite6_current_pass_fraction": 0.18470429008198477,
      "rewrite6_full_passes": 0.0,
      "step": 85110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1846912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.56e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10049333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.064728359497874,
      "rewrite6_current_pass_fraction": 0.18523008315320053,
      "rewrite6_full_passes": 0.0,
      "step": 85120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1871488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.565e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1008,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.18583863994858912,
      "rewrite6_full_passes": 0.0,
      "step": 85130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1896064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.57e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0516,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10118666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14915601567805,
      "rewrite6_current_pass_fraction": 0.18643259138088839,
      "rewrite6_full_passes": 0.0,
      "step": 85140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.192064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.575e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10153333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.125904734609893,
      "rewrite6_current_pass_fraction": 0.18711904344608674,
      "rewrite6_full_passes": 0.0,
      "step": 85150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1945216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.58e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10189333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.18786391696364238,
      "rewrite6_full_passes": 0.0,
      "step": 85160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1969792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.585e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10226666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.18846760530466788,
      "rewrite6_full_passes": 0.0,
      "step": 85170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.1994368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.59e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1026,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.188234818531463,
      "rewrite6_current_pass_fraction": 0.18901287219333607,
      "rewrite6_full_passes": 0.0,
      "step": 85180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2018944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.595e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10290666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.18958248135381978,
      "rewrite6_full_passes": 0.0,
      "step": 85190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.204352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10317333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.945000955113006,
      "rewrite6_current_pass_fraction": 0.19020077505793462,
      "rewrite6_full_passes": 0.0,
      "step": 85200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2068096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.605e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0239,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10345333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.57136376081938,
      "rewrite6_current_pass_fraction": 0.19081906876204943,
      "rewrite6_full_passes": 0.0,
      "step": 85210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2092672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.61e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0643,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10368,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.419463114008664,
      "rewrite6_current_pass_fraction": 0.19133999337890206,
      "rewrite6_full_passes": 0.0,
      "step": 85220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2117248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.615e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0174,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10412,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.4380835263931,
      "rewrite6_current_pass_fraction": 0.19190473408502268,
      "rewrite6_full_passes": 0.0,
      "step": 85230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2141824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.62e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10441333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.19244513251932777,
      "rewrite6_full_passes": 0.0,
      "step": 85240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.21664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.625e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0579,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10469333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28281628634838,
      "rewrite6_current_pass_fraction": 0.19305368931471636,
      "rewrite6_full_passes": 0.0,
      "step": 85250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2190976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.63e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10494666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.871821590362348,
      "rewrite6_current_pass_fraction": 0.19365737765574184,
      "rewrite6_full_passes": 0.0,
      "step": 85260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2215552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.635e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0466,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10528,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.043674159992648,
      "rewrite6_current_pass_fraction": 0.1943000136316722,
      "rewrite6_full_passes": 0.0,
      "step": 85270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2240128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.64e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10557333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.19487449124651907,
      "rewrite6_full_passes": 0.0,
      "step": 85280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2264704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.645e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0743,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10582666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.63473229715763,
      "rewrite6_current_pass_fraction": 0.19541975813518725,
      "rewrite6_full_passes": 0.0,
      "step": 85290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.228928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.65e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0362,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10609333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.825954055651632,
      "rewrite6_current_pass_fraction": 0.19593094584331366,
      "rewrite6_full_passes": 0.0,
      "step": 85300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2313856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.655e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0577,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10632,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27856014871906,
      "rewrite6_current_pass_fraction": 0.19644213355144008,
      "rewrite6_full_passes": 0.0,
      "step": 85310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2338432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.66e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0338,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.77603169671152,
      "rewrite6_current_pass_fraction": 0.19700687425756072,
      "rewrite6_full_passes": 0.0,
      "step": 85320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2363008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.665e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10709333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.19763490487040175,
      "rewrite6_full_passes": 0.0,
      "step": 85330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2387584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.67e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0374,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10750666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.850960201205005,
      "rewrite6_current_pass_fraction": 0.19824346166579035,
      "rewrite6_full_passes": 0.0,
      "step": 85340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.241216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.675e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0428,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10789333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.96385994124339,
      "rewrite6_current_pass_fraction": 0.19883254464372652,
      "rewrite6_full_passes": 0.0,
      "step": 85350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2436736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.68e-06,
      "grad_norm": 0.60546875,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0155,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10826666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.399288035080676,
      "rewrite6_current_pass_fraction": 0.199392416895484,
      "rewrite6_full_passes": 0.0,
      "step": 85360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2461312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.685e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.477977752685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10864,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10267901588904,
      "rewrite6_current_pass_fraction": 0.19995228914724153,
      "rewrite6_full_passes": 0.0,
      "step": 85370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2485888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.69e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.47803497314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10902666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.20050242449027283,
      "rewrite6_full_passes": 0.0,
      "step": 85380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2510464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.695e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.478042602539062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10930666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.039465846006077,
      "rewrite6_current_pass_fraction": 0.2011499289205663,
      "rewrite6_full_passes": 0.0,
      "step": 85390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.253504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.47820281982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0711,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10954666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.56561180557606,
      "rewrite6_current_pass_fraction": 0.20167085353741895,
      "rewrite6_full_passes": 0.0,
      "step": 85400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2559616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.705e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.47820281982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10986666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.817625339883573,
      "rewrite6_current_pass_fraction": 0.20219177815427158,
      "rewrite6_full_passes": 0.0,
      "step": 85410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2584192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.71e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.47820281982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11010666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.20283928258456504,
      "rewrite6_full_passes": 0.0,
      "step": 85420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2608768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.715e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.47820281982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.027,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1104,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.635233936100253,
      "rewrite6_current_pass_fraction": 0.20338941792759635,
      "rewrite6_full_passes": 0.0,
      "step": 85430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2633344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.72e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.47820281982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11062666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.383080960307538,
      "rewrite6_current_pass_fraction": 0.2039054740900859,
      "rewrite6_full_passes": 0.0,
      "step": 85440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.265792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.725e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.47820281982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.084,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.111,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.84561030532294,
      "rewrite6_current_pass_fraction": 0.20439718798075987,
      "rewrite6_full_passes": 0.0,
      "step": 85450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2682496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.73e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.479347229003906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11137333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.934531072324265,
      "rewrite6_current_pass_fraction": 0.2050252185936009,
      "rewrite6_full_passes": 0.0,
      "step": 85460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2707072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.735e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.911033630371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0666,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11170666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.468784577110917,
      "rewrite6_current_pass_fraction": 0.20563864384335262,
      "rewrite6_full_passes": 0.0,
      "step": 85470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2731648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.74e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.911914825439453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11202666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.2061741738232946,
      "rewrite6_full_passes": 0.0,
      "step": 85480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2756224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.745e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.912696838378906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0562,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11232,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.246666234911444,
      "rewrite6_current_pass_fraction": 0.20670483534887343,
      "rewrite6_full_passes": 0.0,
      "step": 85490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.27808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.75e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.914005279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0328,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11257333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.755266049568853,
      "rewrite6_current_pass_fraction": 0.20733773441607758,
      "rewrite6_full_passes": 0.0,
      "step": 85500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.75e-06,
      "eval_loss": 3.0820508003234863,
      "eval_perplexity": 21.803070321712497,
      "eval_runtime": 0.0488,
      "eval_samples_per_second": 655.136,
      "eval_steps_per_second": 20.473,
      "step": 85500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2805376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.755e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.963069915771484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0765,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11285333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.682381102679095,
      "rewrite6_current_pass_fraction": 0.2078878697591089,
      "rewrite6_full_passes": 0.0,
      "step": 85510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2829952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.76e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.963069915771484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.03,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11312,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.6972325893895,
      "rewrite6_current_pass_fraction": 0.20859866409612277,
      "rewrite6_full_passes": 0.0,
      "step": 85520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2854528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.765e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.964885711669922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11342666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.20923643161769,
      "rewrite6_full_passes": 0.0,
      "step": 85530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2879104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.77e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.964885711669922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0351,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11370666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.803058101273997,
      "rewrite6_current_pass_fraction": 0.20984985686744173,
      "rewrite6_full_passes": 0.0,
      "step": 85540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.290368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.775e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.964885711669922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11392,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.039465846006077,
      "rewrite6_current_pass_fraction": 0.21038051839302058,
      "rewrite6_full_passes": 0.0,
      "step": 85550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2928256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.78e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.964885711669922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0364,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11432,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.83011966300961,
      "rewrite6_current_pass_fraction": 0.21101341746022473,
      "rewrite6_full_passes": 0.0,
      "step": 85560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2952832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.785e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.964897155761719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0682,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11454666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.503162127140435,
      "rewrite6_current_pass_fraction": 0.21151973671398805,
      "rewrite6_full_passes": 0.0,
      "step": 85570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.2977408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.79e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.965007781982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11493333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86556098302329,
      "rewrite6_current_pass_fraction": 0.21218671496173394,
      "rewrite6_full_passes": 0.0,
      "step": 85580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3001984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.795e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.965007781982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0337,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11536,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.77395419741855,
      "rewrite6_current_pass_fraction": 0.2127757979396701,
      "rewrite6_full_passes": 0.0,
      "step": 85590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.302656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8e-06,
      "grad_norm": 0.5546875,
      "host_rss_gib": 15.965057373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0428,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11564,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.96385994124339,
      "rewrite6_current_pass_fraction": 0.2133210648283383,
      "rewrite6_full_passes": 0.0,
      "step": 85600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3051136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.805e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.965339660644531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11592,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.191496654234413,
      "rewrite6_current_pass_fraction": 0.2138712001713696,
      "rewrite6_full_passes": 0.0,
      "step": 85610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3075712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.81e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.965339660644531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11624,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.193615909860856,
      "rewrite6_current_pass_fraction": 0.2143677825164067,
      "rewrite6_full_passes": 0.0,
      "step": 85620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3100288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.815e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.965435028076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0909,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.996866249329003,
      "rewrite6_current_pass_fraction": 0.21495686549434287,
      "rewrite6_full_passes": 0.0,
      "step": 85630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3124864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.82e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.965435028076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0383,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11694666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.869734512558935,
      "rewrite6_current_pass_fraction": 0.21559463301591011,
      "rewrite6_full_passes": 0.0,
      "step": 85640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.314944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.825e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.965435028076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11724,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.054198627970376,
      "rewrite6_current_pass_fraction": 0.21627134817238222,
      "rewrite6_full_passes": 0.0,
      "step": 85650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3174016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.83e-06,
      "grad_norm": 0.431640625,
      "host_rss_gib": 15.965435028076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0741,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11756,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.630405783364,
      "rewrite6_current_pass_fraction": 0.21678740433487176,
      "rewrite6_full_passes": 0.0,
      "step": 85660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3198592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.835e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.965877532958984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0616,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11790666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.361708567324747,
      "rewrite6_current_pass_fraction": 0.21728885513427196,
      "rewrite6_full_passes": 0.0,
      "step": 85670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3223168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.84e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.965877532958984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11824,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86347453124931,
      "rewrite6_current_pass_fraction": 0.2178682012034819,
      "rewrite6_full_passes": 0.0,
      "step": 85680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3247744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.845e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.965877532958984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11854666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.678613459924104,
      "rewrite6_current_pass_fraction": 0.2184767579988705,
      "rewrite6_full_passes": 0.0,
      "step": 85690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.327232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.85e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.965877532958984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11894666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86556098302329,
      "rewrite6_current_pass_fraction": 0.2190853147942591,
      "rewrite6_full_passes": 0.0,
      "step": 85700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3296896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.855e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.965877532958984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1192,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.74904040364757,
      "rewrite6_current_pass_fraction": 0.21955755486748066,
      "rewrite6_full_passes": 0.0,
      "step": 85710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3321472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.86e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.965877532958984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0564,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11953333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25091599312008,
      "rewrite6_current_pass_fraction": 0.22008821639305953,
      "rewrite6_full_passes": 0.0,
      "step": 85720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3346048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.865e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.968170166015625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11989333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.024743373380456,
      "rewrite6_current_pass_fraction": 0.22065295709918015,
      "rewrite6_full_passes": 0.0,
      "step": 85730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3370624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.87e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.968170166015625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0468,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12025333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.04788331572619,
      "rewrite6_current_pass_fraction": 0.22121282935093767,
      "rewrite6_full_passes": 0.0,
      "step": 85740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.33952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.875e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.968204498291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12058666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.22185546532686803,
      "rewrite6_full_passes": 0.0,
      "step": 85750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3419776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.88e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.968204498291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0351,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12109333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.803058101273997,
      "rewrite6_current_pass_fraction": 0.22248836439407216,
      "rewrite6_full_passes": 0.0,
      "step": 85760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3444352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.885e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.968204498291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0532,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12158666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18302175066643,
      "rewrite6_current_pass_fraction": 0.22306284200891902,
      "rewrite6_full_passes": 0.0,
      "step": 85770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3468928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.89e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.968204498291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12185333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.132243456791066,
      "rewrite6_current_pass_fraction": 0.2237249518023018,
      "rewrite6_full_passes": 0.0,
      "step": 85780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3493504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.895e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.968204498291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12221333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.22436758777823218,
      "rewrite6_full_passes": 0.0,
      "step": 85790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.351808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.9e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.968204498291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0591,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12261333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.30837099565101,
      "rewrite6_current_pass_fraction": 0.2248933808494479,
      "rewrite6_full_passes": 0.0,
      "step": 85800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3542656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.905e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.00491714477539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12294666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.32969002438444,
      "rewrite6_current_pass_fraction": 0.2254922007361103,
      "rewrite6_full_passes": 0.0,
      "step": 85810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3567232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.91e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.00491714477539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0608,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12330666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34462603439513,
      "rewrite6_current_pass_fraction": 0.2260520729878678,
      "rewrite6_full_passes": 0.0,
      "step": 85820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3591808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.915e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.004920959472656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.03,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12368,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.6972325893895,
      "rewrite6_current_pass_fraction": 0.2266022083308991,
      "rewrite6_full_passes": 0.0,
      "step": 85830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3616384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.92e-06,
      "grad_norm": 0.55078125,
      "host_rss_gib": 16.004920959472656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0691,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12397333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.52252368444875,
      "rewrite6_current_pass_fraction": 0.22717668594574594,
      "rewrite6_full_passes": 0.0,
      "step": 85840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.364096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.925e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.004920959472656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0683,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12421333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50531255087254,
      "rewrite6_current_pass_fraction": 0.22766839983641993,
      "rewrite6_full_passes": 0.0,
      "step": 85850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3665536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.93e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.004920959472656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0622,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12449333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.37452943834182,
      "rewrite6_current_pass_fraction": 0.22814063990964148,
      "rewrite6_full_passes": 0.0,
      "step": 85860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3690112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.935e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 16.005050659179688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0398,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12477333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.901062604522732,
      "rewrite6_current_pass_fraction": 0.2287540651593932,
      "rewrite6_full_passes": 0.0,
      "step": 85870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3714688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.94e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.00506591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0332,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12504,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.763569816631378,
      "rewrite6_current_pass_fraction": 0.2294210434071391,
      "rewrite6_full_passes": 0.0,
      "step": 85880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3739264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.945e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.00506591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0367,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1254,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.836369636357645,
      "rewrite6_current_pass_fraction": 0.23001986329380147,
      "rewrite6_full_passes": 0.0,
      "step": 85890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.376384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.95e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.00506591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12569333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.89479322619514,
      "rewrite6_current_pass_fraction": 0.23055539327374344,
      "rewrite6_full_passes": 0.0,
      "step": 85900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3788416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.955e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.00506591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12605333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.989031673191437,
      "rewrite6_current_pass_fraction": 0.2311834238865845,
      "rewrite6_full_passes": 0.0,
      "step": 85910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3812992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.96e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 16.00506591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0382,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12634666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86764764345287,
      "rewrite6_current_pass_fraction": 0.2317919806819731,
      "rewrite6_full_passes": 0.0,
      "step": 85920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3837568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.965e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.00506591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12654666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.828036755190443,
      "rewrite6_current_pass_fraction": 0.23242487974917722,
      "rewrite6_full_passes": 0.0,
      "step": 85930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3862144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.97e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.00506591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0608,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12682666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34462603439513,
      "rewrite6_current_pass_fraction": 0.2330236996358396,
      "rewrite6_full_passes": 0.0,
      "step": 85940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.388672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.975e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.00506591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0308,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12710666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.713797000341962,
      "rewrite6_current_pass_fraction": 0.23371502015540108,
      "rewrite6_full_passes": 0.0,
      "step": 85950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3911296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.98e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.005172729492188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0312,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12746666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.72208417646683,
      "rewrite6_current_pass_fraction": 0.23435765613133142,
      "rewrite6_full_passes": 0.0,
      "step": 85960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3935872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.985e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.005176544189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0558,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12768,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23816926792417,
      "rewrite6_current_pass_fraction": 0.23489318611127338,
      "rewrite6_full_passes": 0.0,
      "step": 85970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3960448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.99e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.005176544189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12806666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.016335157786333,
      "rewrite6_current_pass_fraction": 0.23548226908920955,
      "rewrite6_full_passes": 0.0,
      "step": 85980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.3985024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.995e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.005176544189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12838666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.23606648361278262,
      "rewrite6_full_passes": 0.0,
      "step": 85990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.40096,
      "curriculum_late_phase": 1.0,
      "epoch": 2e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.005176544189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0483,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12866666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.079478831412377,
      "rewrite6_current_pass_fraction": 0.23667504040817122,
      "rewrite6_full_passes": 0.0,
      "step": 86000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 2e-06,
      "eval_loss": 3.082476854324341,
      "eval_perplexity": 21.812361586203632,
      "eval_runtime": 0.0518,
      "eval_samples_per_second": 618.082,
      "eval_steps_per_second": 19.315,
      "step": 86000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4034176,
      "curriculum_late_phase": 1.0,
      "epoch": 2.005e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.101173400878906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12901333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.997428965207153,
      "rewrite6_current_pass_fraction": 0.23718135966193452,
      "rewrite6_full_passes": 0.0,
      "step": 86010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4058752,
      "curriculum_late_phase": 1.0,
      "epoch": 2.01e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0767,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12929333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.686718012576172,
      "rewrite6_current_pass_fraction": 0.23770228427878717,
      "rewrite6_full_passes": 0.0,
      "step": 86020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4083328,
      "curriculum_late_phase": 1.0,
      "epoch": 2.015e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0486,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12962666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.085803623733216,
      "rewrite6_current_pass_fraction": 0.23831570952853887,
      "rewrite6_full_passes": 0.0,
      "step": 86030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4107904,
      "curriculum_late_phase": 1.0,
      "epoch": 2.02e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0228,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12998666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.5487477017954,
      "rewrite6_current_pass_fraction": 0.23890479250647503,
      "rewrite6_full_passes": 0.0,
      "step": 86040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.413248,
      "curriculum_late_phase": 1.0,
      "epoch": 2.025e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13014666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.210577586354717,
      "rewrite6_current_pass_fraction": 0.23950361239313742,
      "rewrite6_full_passes": 0.0,
      "step": 86050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4157056,
      "curriculum_late_phase": 1.0,
      "epoch": 2.03e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13052,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.24013651146034157,
      "rewrite6_full_passes": 0.0,
      "step": 86060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4181632,
      "curriculum_late_phase": 1.0,
      "epoch": 2.035e-06,
      "grad_norm": 0.431640625,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0301,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13092,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.699302416138053,
      "rewrite6_current_pass_fraction": 0.24066717298592044,
      "rewrite6_full_passes": 0.0,
      "step": 86070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4206208,
      "curriculum_late_phase": 1.0,
      "epoch": 2.04e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13113333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.24116862378532064,
      "rewrite6_full_passes": 0.0,
      "step": 86080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4230784,
      "curriculum_late_phase": 1.0,
      "epoch": 2.045e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0466,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13149333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.043674159992648,
      "rewrite6_current_pass_fraction": 0.24177231212634612,
      "rewrite6_full_passes": 0.0,
      "step": 86090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.425536,
      "curriculum_late_phase": 1.0,
      "epoch": 2.05e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13178666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.24244415882845513,
      "rewrite6_full_passes": 0.0,
      "step": 86100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4279936,
      "curriculum_late_phase": 1.0,
      "epoch": 2.055e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13222666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.024743373380456,
      "rewrite6_current_pass_fraction": 0.24298455726276022,
      "rewrite6_full_passes": 0.0,
      "step": 86110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4304512,
      "curriculum_late_phase": 1.0,
      "epoch": 2.06e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13268,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.24350061342524976,
      "rewrite6_full_passes": 0.0,
      "step": 86120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4329088,
      "curriculum_late_phase": 1.0,
      "epoch": 2.065e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0658,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13296,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45161641762865,
      "rewrite6_current_pass_fraction": 0.2440799594944597,
      "rewrite6_full_passes": 0.0,
      "step": 86130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4353664,
      "curriculum_late_phase": 1.0,
      "epoch": 2.07e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.09741973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.961763660065063,
      "rewrite6_current_pass_fraction": 0.2446398317462172,
      "rewrite6_full_passes": 0.0,
      "step": 86140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.437824,
      "curriculum_late_phase": 1.0,
      "epoch": 2.075e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.068,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13362666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.498861924749583,
      "rewrite6_current_pass_fraction": 0.24522891472415337,
      "rewrite6_full_passes": 0.0,
      "step": 86150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4402816,
      "curriculum_late_phase": 1.0,
      "epoch": 2.08e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0364,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13393333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.83011966300961,
      "rewrite6_current_pass_fraction": 0.2459591828786197,
      "rewrite6_full_passes": 0.0,
      "step": 86160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4427392,
      "curriculum_late_phase": 1.0,
      "epoch": 2.085e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13434666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.896882809995212,
      "rewrite6_current_pass_fraction": 0.24664563494381803,
      "rewrite6_full_passes": 0.0,
      "step": 86170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4451968,
      "curriculum_late_phase": 1.0,
      "epoch": 2.09e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13477333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01003120287951,
      "rewrite6_current_pass_fraction": 0.24721524410430176,
      "rewrite6_full_passes": 0.0,
      "step": 86180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4476544,
      "curriculum_late_phase": 1.0,
      "epoch": 2.095e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0355,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13510666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.811380988981075,
      "rewrite6_current_pass_fraction": 0.24779459017351171,
      "rewrite6_full_passes": 0.0,
      "step": 86190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.450112,
      "curriculum_late_phase": 1.0,
      "epoch": 2.1e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0522,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13529333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.161849316897023,
      "rewrite6_current_pass_fraction": 0.24847617378434694,
      "rewrite6_full_passes": 0.0,
      "step": 86200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4525696,
      "curriculum_late_phase": 1.0,
      "epoch": 2.105e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13566666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.024743373380456,
      "rewrite6_current_pass_fraction": 0.24909446748846176,
      "rewrite6_full_passes": 0.0,
      "step": 86210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4550272,
      "curriculum_late_phase": 1.0,
      "epoch": 2.11e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13598666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.031051742600535,
      "rewrite6_current_pass_fraction": 0.2496738135576717,
      "rewrite6_full_passes": 0.0,
      "step": 86220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4574848,
      "curriculum_late_phase": 1.0,
      "epoch": 2.115e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13632,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.25030671262487586,
      "rewrite6_full_passes": 0.0,
      "step": 86230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4599424,
      "curriculum_late_phase": 1.0,
      "epoch": 2.12e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13662666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.926158934434543,
      "rewrite6_current_pass_fraction": 0.25101750696188974,
      "rewrite6_full_passes": 0.0,
      "step": 86240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4624,
      "curriculum_late_phase": 1.0,
      "epoch": 2.125e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.097503662109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13702666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.2515919845767366,
      "rewrite6_full_passes": 0.0,
      "step": 86250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4648576,
      "curriculum_late_phase": 1.0,
      "epoch": 2.13e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.09839630126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13728,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.25220054137212516,
      "rewrite6_full_passes": 0.0,
      "step": 86260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4673152,
      "curriculum_late_phase": 1.0,
      "epoch": 2.135e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 16.09839630126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0409,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1376,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.924066423168412,
      "rewrite6_current_pass_fraction": 0.25278962435006136,
      "rewrite6_full_passes": 0.0,
      "step": 86270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4697728,
      "curriculum_late_phase": 1.0,
      "epoch": 2.14e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.09839630126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0149,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13794666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.387052133397212,
      "rewrite6_current_pass_fraction": 0.25333975969309264,
      "rewrite6_full_passes": 0.0,
      "step": 86280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4722304,
      "curriculum_late_phase": 1.0,
      "epoch": 2.145e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.09839630126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1382,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.041569897791508,
      "rewrite6_current_pass_fraction": 0.2538898950361239,
      "rewrite6_full_passes": 0.0,
      "step": 86290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.474688,
      "curriculum_late_phase": 1.0,
      "epoch": 2.15e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.09839630126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13852,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13647032815547,
      "rewrite6_current_pass_fraction": 0.25449358337714945,
      "rewrite6_full_passes": 0.0,
      "step": 86300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4771456,
      "curriculum_late_phase": 1.0,
      "epoch": 2.155e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.09839630126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0611,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13882666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.351030382809682,
      "rewrite6_current_pass_fraction": 0.2550193764483652,
      "rewrite6_full_passes": 0.0,
      "step": 86310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4796032,
      "curriculum_late_phase": 1.0,
      "epoch": 2.16e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.09839630126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0236,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13906666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.565193277309945,
      "rewrite6_current_pass_fraction": 0.2555354326108547,
      "rewrite6_full_passes": 0.0,
      "step": 86320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4820608,
      "curriculum_late_phase": 1.0,
      "epoch": 2.165e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.09839630126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.068,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1394,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.498861924749583,
      "rewrite6_current_pass_fraction": 0.2561001733169753,
      "rewrite6_full_passes": 0.0,
      "step": 86330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4845184,
      "curriculum_late_phase": 1.0,
      "epoch": 2.17e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.528614044189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0324,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13984,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.746965603348944,
      "rewrite6_current_pass_fraction": 0.2567038616580008,
      "rewrite6_full_passes": 0.0,
      "step": 86340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.486976,
      "curriculum_late_phase": 1.0,
      "epoch": 2.175e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.531509399414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0991,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14010666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.17798211276216,
      "rewrite6_current_pass_fraction": 0.2573659714513836,
      "rewrite6_full_passes": 0.0,
      "step": 86350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4894336,
      "curriculum_late_phase": 1.0,
      "epoch": 2.18e-06,
      "grad_norm": 0.5546875,
      "host_rss_gib": 15.531509399414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0712,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14048,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.567768474588277,
      "rewrite6_current_pass_fraction": 0.25800860742731396,
      "rewrite6_full_passes": 0.0,
      "step": 86360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4918912,
      "curriculum_late_phase": 1.0,
      "epoch": 2.185e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.531509399414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0379,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14078666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.861388288110085,
      "rewrite6_current_pass_fraction": 0.2585344004985297,
      "rewrite6_full_passes": 0.0,
      "step": 86370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4943488,
      "curriculum_late_phase": 1.0,
      "epoch": 2.19e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.533218383789062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0685,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14105333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50961404351764,
      "rewrite6_current_pass_fraction": 0.25906506202410856,
      "rewrite6_full_passes": 0.0,
      "step": 86380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.4968064,
      "curriculum_late_phase": 1.0,
      "epoch": 2.195e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.533462524414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0495,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14134666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.104789389307548,
      "rewrite6_current_pass_fraction": 0.25961032891277674,
      "rewrite6_full_passes": 0.0,
      "step": 86390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.499264,
      "curriculum_late_phase": 1.0,
      "epoch": 2.2e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.533462524414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14158666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.738668476643575,
      "rewrite6_current_pass_fraction": 0.2601215166209032,
      "rewrite6_full_passes": 0.0,
      "step": 86400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5017216,
      "curriculum_late_phase": 1.0,
      "epoch": 2.205e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.534317016601562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14204,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.248791007771803,
      "rewrite6_current_pass_fraction": 0.26063270432902963,
      "rewrite6_full_passes": 0.0,
      "step": 86410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5041792,
      "curriculum_late_phase": 1.0,
      "epoch": 2.21e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.534317016601562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14242666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.892703851343008,
      "rewrite6_current_pass_fraction": 0.2612899456680493,
      "rewrite6_full_passes": 0.0,
      "step": 86420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5066368,
      "curriculum_late_phase": 1.0,
      "epoch": 2.215e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.534603118896484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14282666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14704120581873,
      "rewrite6_current_pass_fraction": 0.2619374500983428,
      "rewrite6_full_passes": 0.0,
      "step": 86430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5090944,
      "curriculum_late_phase": 1.0,
      "epoch": 2.22e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.853599548339844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0796,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14321333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.749700775678726,
      "rewrite6_current_pass_fraction": 0.262619033709178,
      "rewrite6_full_passes": 0.0,
      "step": 86440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.511552,
      "curriculum_late_phase": 1.0,
      "epoch": 2.225e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.525646209716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14354666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.974344492133298,
      "rewrite6_current_pass_fraction": 0.26335417031800745,
      "rewrite6_full_passes": 0.0,
      "step": 86450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5140096,
      "curriculum_late_phase": 1.0,
      "epoch": 2.23e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.856128692626953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0728,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14386666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.602304525620752,
      "rewrite6_current_pass_fraction": 0.26387022648049696,
      "rewrite6_full_passes": 0.0,
      "step": 86460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5164672,
      "curriculum_late_phase": 1.0,
      "epoch": 2.235e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.857559204101562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0348,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14418666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.796818119887625,
      "rewrite6_current_pass_fraction": 0.26441062491480205,
      "rewrite6_full_passes": 0.0,
      "step": 86470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5189248,
      "curriculum_late_phase": 1.0,
      "epoch": 2.24e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.857559204101562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14453333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.26497536562092266,
      "rewrite6_full_passes": 0.0,
      "step": 86480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5213824,
      "curriculum_late_phase": 1.0,
      "epoch": 2.245e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.857864379882812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0289,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14492,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.67447815077682,
      "rewrite6_current_pass_fraction": 0.26552063250959085,
      "rewrite6_full_passes": 0.0,
      "step": 86490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.52384,
      "curriculum_late_phase": 1.0,
      "epoch": 2.25e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.857952117919922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14521333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.259418059817307,
      "rewrite6_current_pass_fraction": 0.26605616248953284,
      "rewrite6_full_passes": 0.0,
      "step": 86500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 2.25e-06,
      "eval_loss": 3.082371711730957,
      "eval_perplexity": 21.810068298501854,
      "eval_runtime": 0.0507,
      "eval_samples_per_second": 631.399,
      "eval_steps_per_second": 19.731,
      "step": 86500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5262976,
      "curriculum_late_phase": 1.0,
      "epoch": 2.255e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.86175537109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14548,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.07737098892312,
      "rewrite6_current_pass_fraction": 0.2668205098245409,
      "rewrite6_full_passes": 0.0,
      "step": 86510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5287552,
      "curriculum_late_phase": 1.0,
      "epoch": 2.26e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.86175537109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14570666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.007930304805882,
      "rewrite6_current_pass_fraction": 0.26741446125684015,
      "rewrite6_full_passes": 0.0,
      "step": 86520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5312128,
      "curriculum_late_phase": 1.0,
      "epoch": 2.265e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.019,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14593333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.470810634740232,
      "rewrite6_current_pass_fraction": 0.2679938073260501,
      "rewrite6_full_passes": 0.0,
      "step": 86530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5336704,
      "curriculum_late_phase": 1.0,
      "epoch": 2.27e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14625333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86556098302329,
      "rewrite6_current_pass_fraction": 0.26860236412143873,
      "rewrite6_full_passes": 0.0,
      "step": 86540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.536128,
      "curriculum_late_phase": 1.0,
      "epoch": 2.275e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14657333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.997428965207153,
      "rewrite6_current_pass_fraction": 0.2693131584584526,
      "rewrite6_full_passes": 0.0,
      "step": 86550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5385856,
      "curriculum_late_phase": 1.0,
      "epoch": 2.28e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.069,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14678666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.520371539689332,
      "rewrite6_current_pass_fraction": 0.2698000038947635,
      "rewrite6_full_passes": 0.0,
      "step": 86560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5410432,
      "curriculum_late_phase": 1.0,
      "epoch": 2.285e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14725333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.210577586354717,
      "rewrite6_current_pass_fraction": 0.2702917177854375,
      "rewrite6_full_passes": 0.0,
      "step": 86570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5435008,
      "curriculum_late_phase": 1.0,
      "epoch": 2.29e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0491,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14770666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09634916170988,
      "rewrite6_current_pass_fraction": 0.27086619540028434,
      "rewrite6_full_passes": 0.0,
      "step": 86580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5459584,
      "curriculum_late_phase": 1.0,
      "epoch": 2.295e-06,
      "grad_norm": 0.54296875,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0907,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14814666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.992467315987135,
      "rewrite6_current_pass_fraction": 0.2714017253802263,
      "rewrite6_full_passes": 0.0,
      "step": 86590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.548416,
      "curriculum_late_phase": 1.0,
      "epoch": 2.3e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0553,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14848,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.22755283761896,
      "rewrite6_current_pass_fraction": 0.27205409826488286,
      "rewrite6_full_passes": 0.0,
      "step": 86600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5508736,
      "curriculum_late_phase": 1.0,
      "epoch": 2.305e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14870666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.144926607429838,
      "rewrite6_current_pass_fraction": 0.272686997332087,
      "rewrite6_full_passes": 0.0,
      "step": 86610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5533312,
      "curriculum_late_phase": 1.0,
      "epoch": 2.31e-06,
      "grad_norm": 0.609375,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14912,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.27321765885766586,
      "rewrite6_full_passes": 0.0,
      "step": 86620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5557888,
      "curriculum_late_phase": 1.0,
      "epoch": 2.315e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14952,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.913607005029263,
      "rewrite6_current_pass_fraction": 0.2738262156530545,
      "rewrite6_full_passes": 0.0,
      "step": 86630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5582464,
      "curriculum_late_phase": 1.0,
      "epoch": 2.32e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0525,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14986666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.16819882407055,
      "rewrite6_current_pass_fraction": 0.2744542462658955,
      "rewrite6_full_passes": 0.0,
      "step": 86640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.560704,
      "curriculum_late_phase": 1.0,
      "epoch": 2.325e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0321,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15034666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.740742447188033,
      "rewrite6_current_pass_fraction": 0.275014118517653,
      "rewrite6_full_passes": 0.0,
      "step": 86650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5631616,
      "curriculum_late_phase": 1.0,
      "epoch": 2.33e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0599,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15070666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.325424512944927,
      "rewrite6_current_pass_fraction": 0.2756129384043154,
      "rewrite6_full_passes": 0.0,
      "step": 86660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5656192,
      "curriculum_late_phase": 1.0,
      "epoch": 2.335e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0277,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15098666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.649683656667687,
      "rewrite6_current_pass_fraction": 0.2763334696500555,
      "rewrite6_full_passes": 0.0,
      "step": 86670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5680768,
      "curriculum_late_phase": 1.0,
      "epoch": 2.34e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.861804962158203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15125333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.054198627970376,
      "rewrite6_current_pass_fraction": 0.2768836049930868,
      "rewrite6_full_passes": 0.0,
      "step": 86680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5705344,
      "curriculum_late_phase": 1.0,
      "epoch": 2.345e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.861808776855469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.069,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15154666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.520371539689332,
      "rewrite6_current_pass_fraction": 0.277472687971023,
      "rewrite6_full_passes": 0.0,
      "step": 86690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.572992,
      "curriculum_late_phase": 1.0,
      "epoch": 2.35e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.861808776855469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15182666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.210577586354717,
      "rewrite6_current_pass_fraction": 0.27807637631204846,
      "rewrite6_full_passes": 0.0,
      "step": 86700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5754496,
      "curriculum_late_phase": 1.0,
      "epoch": 2.355e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.861808776855469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15213333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.17878556994849,
      "rewrite6_current_pass_fraction": 0.2786313801094428,
      "rewrite6_full_passes": 0.0,
      "step": 86710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5779072,
      "curriculum_late_phase": 1.0,
      "epoch": 2.36e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.861808776855469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15250666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.035258373598133,
      "rewrite6_current_pass_fraction": 0.27921559463301593,
      "rewrite6_full_passes": 0.0,
      "step": 86720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5803648,
      "curriculum_late_phase": 1.0,
      "epoch": 2.365e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.861808776855469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15290666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.2798290198827676,
      "rewrite6_full_passes": 0.0,
      "step": 86730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5828224,
      "curriculum_late_phase": 1.0,
      "epoch": 2.37e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.861808776855469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0551,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15317333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.22330775157419,
      "rewrite6_current_pass_fraction": 0.28042783976943,
      "rewrite6_full_passes": 0.0,
      "step": 86740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.58528,
      "curriculum_late_phase": 1.0,
      "epoch": 2.375e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.861808776855469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0255,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15338666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.60430428823131,
      "rewrite6_current_pass_fraction": 0.28101692274736617,
      "rewrite6_full_passes": 0.0,
      "step": 86750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5877376,
      "curriculum_late_phase": 1.0,
      "epoch": 2.38e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.861808776855469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1538,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.2815816634534868,
      "rewrite6_full_passes": 0.0,
      "step": 86760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5901952,
      "curriculum_late_phase": 1.0,
      "epoch": 2.385e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0593,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15410666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.312633096045975,
      "rewrite6_current_pass_fraction": 0.2821999571576016,
      "rewrite6_full_passes": 0.0,
      "step": 86770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5926528,
      "curriculum_late_phase": 1.0,
      "epoch": 2.39e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.027,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15449333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.635233936100253,
      "rewrite6_current_pass_fraction": 0.282798777044264,
      "rewrite6_full_passes": 0.0,
      "step": 86780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.5951104,
      "curriculum_late_phase": 1.0,
      "epoch": 2.395e-06,
      "grad_norm": 0.4296875,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0211,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15486666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.513844506823926,
      "rewrite6_current_pass_fraction": 0.2833050962980273,
      "rewrite6_full_passes": 0.0,
      "step": 86790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.597568,
      "curriculum_late_phase": 1.0,
      "epoch": 2.4e-06,
      "grad_norm": 0.55078125,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0287,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15518666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.670343668608663,
      "rewrite6_current_pass_fraction": 0.2839233900021421,
      "rewrite6_full_passes": 0.0,
      "step": 86800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6000256,
      "curriculum_late_phase": 1.0,
      "epoch": 2.405e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0567,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15550666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.257292224304877,
      "rewrite6_current_pass_fraction": 0.28454655216062,
      "rewrite6_full_passes": 0.0,
      "step": 86810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6024832,
      "curriculum_late_phase": 1.0,
      "epoch": 2.41e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0319,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15581333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.736594713485793,
      "rewrite6_current_pass_fraction": 0.2850626083231096,
      "rewrite6_full_passes": 0.0,
      "step": 86820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6049408,
      "curriculum_late_phase": 1.0,
      "epoch": 2.415e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0361,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15618666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.82387156437236,
      "rewrite6_current_pass_fraction": 0.28566629666413507,
      "rewrite6_full_passes": 0.0,
      "step": 86830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6073984,
      "curriculum_late_phase": 1.0,
      "epoch": 2.42e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.049,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15652,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.094239632271936,
      "rewrite6_current_pass_fraction": 0.28615801055480905,
      "rewrite6_full_passes": 0.0,
      "step": 86840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.609856,
      "curriculum_late_phase": 1.0,
      "epoch": 2.425e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15684,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91988202833721,
      "rewrite6_current_pass_fraction": 0.28682985725691806,
      "rewrite6_full_passes": 0.0,
      "step": 86850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6123136,
      "curriculum_late_phase": 1.0,
      "epoch": 2.43e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15709333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10689997377394,
      "rewrite6_current_pass_fraction": 0.2874676247784853,
      "rewrite6_full_passes": 0.0,
      "step": 86860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6147712,
      "curriculum_late_phase": 1.0,
      "epoch": 2.435e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0335,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15734666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.769799822030453,
      "rewrite6_current_pass_fraction": 0.2879496017604331,
      "rewrite6_full_passes": 0.0,
      "step": 86870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6172288,
      "curriculum_late_phase": 1.0,
      "epoch": 2.44e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15768,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.859302253584733,
      "rewrite6_current_pass_fraction": 0.2885143424665537,
      "rewrite6_full_passes": 0.0,
      "step": 86880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6196864,
      "curriculum_late_phase": 1.0,
      "epoch": 2.445e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0767,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15794666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.686718012576172,
      "rewrite6_current_pass_fraction": 0.28924461062102,
      "rewrite6_full_passes": 0.0,
      "step": 86890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.622144,
      "curriculum_late_phase": 1.0,
      "epoch": 2.45e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.86184310913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0849,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1582,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.865280204724744,
      "rewrite6_current_pass_fraction": 0.2897314560573309,
      "rewrite6_full_passes": 0.0,
      "step": 86900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6246016,
      "curriculum_late_phase": 1.0,
      "epoch": 2.455e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.861923217773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15856,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10267901588904,
      "rewrite6_current_pass_fraction": 0.2903254074896302,
      "rewrite6_full_passes": 0.0,
      "step": 86910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6270592,
      "curriculum_late_phase": 1.0,
      "epoch": 2.46e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.861923217773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0228,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1588,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.5487477017954,
      "rewrite6_current_pass_fraction": 0.290987517283013,
      "rewrite6_full_passes": 0.0,
      "step": 86920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6295168,
      "curriculum_late_phase": 1.0,
      "epoch": 2.465e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.861923217773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15910666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.291571731806586,
      "rewrite6_full_passes": 0.0,
      "step": 86930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6319744,
      "curriculum_late_phase": 1.0,
      "epoch": 2.47e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.861923217773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15933333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.29217542014761155,
      "rewrite6_full_passes": 0.0,
      "step": 86940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.634432,
      "curriculum_late_phase": 1.0,
      "epoch": 2.475e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.861923217773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0761,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15966666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.67370988459726,
      "rewrite6_current_pass_fraction": 0.29274016085373217,
      "rewrite6_full_passes": 0.0,
      "step": 86950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6368896,
      "curriculum_late_phase": 1.0,
      "epoch": 2.48e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.861923217773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0607,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15994666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34249167851127,
      "rewrite6_current_pass_fraction": 0.2933049015598528,
      "rewrite6_full_passes": 0.0,
      "step": 86960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6393472,
      "curriculum_late_phase": 1.0,
      "epoch": 2.485e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.861923217773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16030666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.29385016844852097,
      "rewrite6_full_passes": 0.0,
      "step": 86970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6418048,
      "curriculum_late_phase": 1.0,
      "epoch": 2.49e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.861923217773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16062666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.974344492133298,
      "rewrite6_current_pass_fraction": 0.29445385678954644,
      "rewrite6_full_passes": 0.0,
      "step": 86980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6442624,
      "curriculum_late_phase": 1.0,
      "epoch": 2.495e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.861923217773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16090666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09845890211131,
      "rewrite6_current_pass_fraction": 0.2951305719460186,
      "rewrite6_full_passes": 0.0,
      "step": 86990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.64672,
      "curriculum_late_phase": 1.0,
      "epoch": 2.5e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.861923217773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0345,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16121333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.790580010214892,
      "rewrite6_current_pass_fraction": 0.2957391287414072,
      "rewrite6_full_passes": 0.0,
      "step": 87000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 2.5e-06,
      "eval_loss": 3.0827138423919678,
      "eval_perplexity": 21.817531468202304,
      "eval_runtime": 0.05,
      "eval_samples_per_second": 639.882,
      "eval_steps_per_second": 19.996,
      "step": 87000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6491776,
      "curriculum_late_phase": 1.0,
      "epoch": 2.505e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.861957550048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16150666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.29626492181262293,
      "rewrite6_full_passes": 0.0,
      "step": 87010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6516352,
      "curriculum_late_phase": 1.0,
      "epoch": 2.51e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.861957550048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0584,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16181333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.29346035528703,
      "rewrite6_current_pass_fraction": 0.2968393994274698,
      "rewrite6_full_passes": 0.0,
      "step": 87020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6540928,
      "curriculum_late_phase": 1.0,
      "epoch": 2.515e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.861957550048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0342,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1622,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.784343771694374,
      "rewrite6_current_pass_fraction": 0.2974625615859477,
      "rewrite6_full_passes": 0.0,
      "step": 87030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6565504,
      "curriculum_late_phase": 1.0,
      "epoch": 2.52e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.862194061279297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0512,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16248,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14069804497869,
      "rewrite6_current_pass_fraction": 0.2980175653833421,
      "rewrite6_full_passes": 0.0,
      "step": 87040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.659008,
      "curriculum_late_phase": 1.0,
      "epoch": 2.525e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.862194061279297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16284,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.265796842004743,
      "rewrite6_current_pass_fraction": 0.2985384900001947,
      "rewrite6_full_passes": 0.0,
      "step": 87050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6614656,
      "curriculum_late_phase": 1.0,
      "epoch": 2.53e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.862194061279297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0102,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16297333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.29145781100103,
      "rewrite6_current_pass_fraction": 0.29914704679558335,
      "rewrite6_full_passes": 0.0,
      "step": 87060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6639232,
      "curriculum_late_phase": 1.0,
      "epoch": 2.535e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.862194061279297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0489,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16337333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.092130313776398,
      "rewrite6_current_pass_fraction": 0.29971665595606706,
      "rewrite6_full_passes": 0.0,
      "step": 87070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6663808,
      "curriculum_late_phase": 1.0,
      "epoch": 2.54e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.862194061279297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16382666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.30028139666218767,
      "rewrite6_full_passes": 0.0,
      "step": 87080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6688384,
      "curriculum_late_phase": 1.0,
      "epoch": 2.545e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.8631591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16418666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.300787715915951,
      "rewrite6_full_passes": 0.0,
      "step": 87090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.671296,
      "curriculum_late_phase": 1.0,
      "epoch": 2.55e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.8631591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0743,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16449333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.63473229715763,
      "rewrite6_current_pass_fraction": 0.3013427197133454,
      "rewrite6_full_passes": 0.0,
      "step": 87100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6737536,
      "curriculum_late_phase": 1.0,
      "epoch": 2.555e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.8631591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0353,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16482666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.80721912898315,
      "rewrite6_current_pass_fraction": 0.3020632509590855,
      "rewrite6_full_passes": 0.0,
      "step": 87110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6762112,
      "curriculum_late_phase": 1.0,
      "epoch": 2.56e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.8631591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16518666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.24241732656946,
      "rewrite6_current_pass_fraction": 0.3026182547564799,
      "rewrite6_full_passes": 0.0,
      "step": 87120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6786688,
      "curriculum_late_phase": 1.0,
      "epoch": 2.565e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.8631591796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16553333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.054198627970376,
      "rewrite6_current_pass_fraction": 0.30320246928005296,
      "rewrite6_full_passes": 0.0,
      "step": 87130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6811264,
      "curriculum_late_phase": 1.0,
      "epoch": 2.57e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.863277435302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16588,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.189377610522946,
      "rewrite6_current_pass_fraction": 0.3036747093532745,
      "rewrite6_full_passes": 0.0,
      "step": 87140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.683584,
      "curriculum_late_phase": 1.0,
      "epoch": 2.575e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.863277435302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16613333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.3042637923312107,
      "rewrite6_full_passes": 0.0,
      "step": 87150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6860416,
      "curriculum_late_phase": 1.0,
      "epoch": 2.58e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.863277435302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0516,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16661333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14915601567805,
      "rewrite6_current_pass_fraction": 0.3048772175809624,
      "rewrite6_full_passes": 0.0,
      "step": 87160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6884992,
      "curriculum_late_phase": 1.0,
      "epoch": 2.585e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.863277435302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0429,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16693333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.9659564320603,
      "rewrite6_current_pass_fraction": 0.305485774376351,
      "rewrite6_full_passes": 0.0,
      "step": 87170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6909568,
      "curriculum_late_phase": 1.0,
      "epoch": 2.59e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.863277435302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16728,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.007930304805882,
      "rewrite6_current_pass_fraction": 0.3060553835368347,
      "rewrite6_full_passes": 0.0,
      "step": 87180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6934144,
      "curriculum_late_phase": 1.0,
      "epoch": 2.595e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.863277435302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16757333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.997428965207153,
      "rewrite6_current_pass_fraction": 0.3065665712449612,
      "rewrite6_full_passes": 0.0,
      "step": 87190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.695872,
      "curriculum_late_phase": 1.0,
      "epoch": 2.6e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 14.54690170288086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0533,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16784,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18514015876014,
      "rewrite6_current_pass_fraction": 0.30719947031216527,
      "rewrite6_full_passes": 0.0,
      "step": 87200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.6983296,
      "curriculum_late_phase": 1.0,
      "epoch": 2.605e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 14.55007553100586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0682,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16814666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.503162127140435,
      "rewrite6_current_pass_fraction": 0.30773013183774417,
      "rewrite6_full_passes": 0.0,
      "step": 87210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7007872,
      "curriculum_late_phase": 1.0,
      "epoch": 2.61e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.55227279663086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16849333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.3083094779069541,
      "rewrite6_full_passes": 0.0,
      "step": 87220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7032448,
      "curriculum_late_phase": 1.0,
      "epoch": 2.615e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 14.55301284790039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0627,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16898666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.38521937532253,
      "rewrite6_current_pass_fraction": 0.3090251406983311,
      "rewrite6_full_passes": 0.0,
      "step": 87230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7057024,
      "curriculum_late_phase": 1.0,
      "epoch": 2.62e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.556922912597656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0331,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16933333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.761493563464107,
      "rewrite6_current_pass_fraction": 0.30957527604136237,
      "rewrite6_full_passes": 0.0,
      "step": 87240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.70816,
      "curriculum_late_phase": 1.0,
      "epoch": 2.625e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.557605743408203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0217,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16965333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.52615650675864,
      "rewrite6_current_pass_fraction": 0.310183832836751,
      "rewrite6_full_passes": 0.0,
      "step": 87250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7106176,
      "curriculum_late_phase": 1.0,
      "epoch": 2.63e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.557640075683594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0629,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17010666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3894968469305,
      "rewrite6_current_pass_fraction": 0.31075831045159785,
      "rewrite6_full_passes": 0.0,
      "step": 87260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7130752,
      "curriculum_late_phase": 1.0,
      "epoch": 2.635e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.559196472167969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17046666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.873908876883966,
      "rewrite6_current_pass_fraction": 0.31136199879262333,
      "rewrite6_full_passes": 0.0,
      "step": 87270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7155328,
      "curriculum_late_phase": 1.0,
      "epoch": 2.64e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 14.559196472167969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17077333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.968053132536788,
      "rewrite6_current_pass_fraction": 0.3119462133161964,
      "rewrite6_full_passes": 0.0,
      "step": 87280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7179904,
      "curriculum_late_phase": 1.0,
      "epoch": 2.645e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 13.842369079589844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0716,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17113333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.57639730762967,
      "rewrite6_current_pass_fraction": 0.3124038480263286,
      "rewrite6_full_passes": 0.0,
      "step": 87290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.720448,
      "curriculum_late_phase": 1.0,
      "epoch": 2.65e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.78341293334961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0721,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17156,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.587188203782706,
      "rewrite6_current_pass_fraction": 0.3130124048217172,
      "rewrite6_full_passes": 0.0,
      "step": 87300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7229056,
      "curriculum_late_phase": 1.0,
      "epoch": 2.655e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.30929946899414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0541,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17184,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20209505194016,
      "rewrite6_current_pass_fraction": 0.31357227707347474,
      "rewrite6_full_passes": 0.0,
      "step": 87310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7253632,
      "curriculum_late_phase": 1.0,
      "epoch": 2.66e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.30929946899414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0511,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17217333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.138584080874157,
      "rewrite6_current_pass_fraction": 0.3141078070534167,
      "rewrite6_full_passes": 0.0,
      "step": 87320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7278208,
      "curriculum_late_phase": 1.0,
      "epoch": 2.665e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.30929946899414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17248,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.3146384685789955,
      "rewrite6_full_passes": 0.0,
      "step": 87330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7302784,
      "curriculum_late_phase": 1.0,
      "epoch": 2.67e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.30929946899414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17274666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.049988209300682,
      "rewrite6_current_pass_fraction": 0.3151642616502113,
      "rewrite6_full_passes": 0.0,
      "step": 87340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.732736,
      "curriculum_late_phase": 1.0,
      "epoch": 2.675e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.30929946899414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0521,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17302666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.159733237771047,
      "rewrite6_current_pass_fraction": 0.3157436077194212,
      "rewrite6_full_passes": 0.0,
      "step": 87350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7351936,
      "curriculum_late_phase": 1.0,
      "epoch": 2.68e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.309310913085938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17334666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.3164251913302565,
      "rewrite6_full_passes": 0.0,
      "step": 87360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7376512,
      "curriculum_late_phase": 1.0,
      "epoch": 2.685e-06,
      "grad_norm": 0.57421875,
      "host_rss_gib": 15.309310913085938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0657,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17365333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.449471363241397,
      "rewrite6_current_pass_fraction": 0.31695585285583533,
      "rewrite6_full_passes": 0.0,
      "step": 87370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7401088,
      "curriculum_late_phase": 1.0,
      "epoch": 2.69e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0594,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17398666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3147644659223,
      "rewrite6_current_pass_fraction": 0.31745730365523556,
      "rewrite6_full_passes": 0.0,
      "step": 87380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7425664,
      "curriculum_late_phase": 1.0,
      "epoch": 2.695e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0629,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1744,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3894968469305,
      "rewrite6_current_pass_fraction": 0.3181583610835232,
      "rewrite6_full_passes": 0.0,
      "step": 87390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.745024,
      "curriculum_late_phase": 1.0,
      "epoch": 2.7e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17468,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.054198627970376,
      "rewrite6_current_pass_fraction": 0.31874257560709623,
      "rewrite6_full_passes": 0.0,
      "step": 87400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7474816,
      "curriculum_late_phase": 1.0,
      "epoch": 2.705e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0682,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17490666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.503162127140435,
      "rewrite6_current_pass_fraction": 0.31930244785885376,
      "rewrite6_full_passes": 0.0,
      "step": 87410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7499392,
      "curriculum_late_phase": 1.0,
      "epoch": 2.71e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0688,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17518666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.51606789576013,
      "rewrite6_current_pass_fraction": 0.31995482074351034,
      "rewrite6_full_passes": 0.0,
      "step": 87420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7523968,
      "curriculum_late_phase": 1.0,
      "epoch": 2.715e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.035,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17550666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.800977899475697,
      "rewrite6_current_pass_fraction": 0.3205536406301727,
      "rewrite6_full_passes": 0.0,
      "step": 87430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7548544,
      "curriculum_late_phase": 1.0,
      "epoch": 2.72e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17584,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.817625339883573,
      "rewrite6_current_pass_fraction": 0.32111838133629333,
      "rewrite6_full_passes": 0.0,
      "step": 87440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.757312,
      "curriculum_late_phase": 1.0,
      "epoch": 2.725e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0606,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17617333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.340357536052313,
      "rewrite6_current_pass_fraction": 0.32166851667932467,
      "rewrite6_full_passes": 0.0,
      "step": 87450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7597696,
      "curriculum_late_phase": 1.0,
      "epoch": 2.73e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0579,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17648,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28281628634838,
      "rewrite6_current_pass_fraction": 0.3223452318357968,
      "rewrite6_full_passes": 0.0,
      "step": 87460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7622272,
      "curriculum_late_phase": 1.0,
      "epoch": 2.735e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17670666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.028948742578034,
      "rewrite6_current_pass_fraction": 0.3229245779050067,
      "rewrite6_full_passes": 0.0,
      "step": 87470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7646848,
      "curriculum_late_phase": 1.0,
      "epoch": 2.74e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.309455871582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0328,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17710666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.755266049568853,
      "rewrite6_current_pass_fraction": 0.3234698447936749,
      "rewrite6_full_passes": 0.0,
      "step": 87480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7671424,
      "curriculum_late_phase": 1.0,
      "epoch": 2.745e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.31097412109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0596,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17736,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.319027845139193,
      "rewrite6_current_pass_fraction": 0.32410274386087906,
      "rewrite6_full_passes": 0.0,
      "step": 87490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7696,
      "curriculum_late_phase": 1.0,
      "epoch": 2.75e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.31097412109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17773333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11534442254061,
      "rewrite6_current_pass_fraction": 0.32459932620591614,
      "rewrite6_full_passes": 0.0,
      "step": 87500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 2.75e-06,
      "eval_loss": 3.0833146572113037,
      "eval_perplexity": 21.830643703047794,
      "eval_runtime": 0.048,
      "eval_samples_per_second": 666.771,
      "eval_steps_per_second": 20.837,
      "step": 87500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7720576,
      "curriculum_late_phase": 1.0,
      "epoch": 2.755e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.370018005371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.178,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20845663464543,
      "rewrite6_current_pass_fraction": 0.32521761991003095,
      "rewrite6_full_passes": 0.0,
      "step": 87510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7745152,
      "curriculum_late_phase": 1.0,
      "epoch": 2.76e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.370018005371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17826666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.961763660065063,
      "rewrite6_current_pass_fraction": 0.3257336760725205,
      "rewrite6_full_passes": 0.0,
      "step": 87520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7769728,
      "curriculum_late_phase": 1.0,
      "epoch": 2.765e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.370018005371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0751,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17862666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.65204700795622,
      "rewrite6_current_pass_fraction": 0.3263471013222722,
      "rewrite6_full_passes": 0.0,
      "step": 87530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7794304,
      "curriculum_late_phase": 1.0,
      "epoch": 2.77e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.370018005371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0681,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17898666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.501011918439943,
      "rewrite6_current_pass_fraction": 0.3270140795700181,
      "rewrite6_full_passes": 0.0,
      "step": 87540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.781888,
      "curriculum_late_phase": 1.0,
      "epoch": 2.775e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.370018005371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0399,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17941333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.903152815291975,
      "rewrite6_current_pass_fraction": 0.3275642149130494,
      "rewrite6_full_passes": 0.0,
      "step": 87550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7843456,
      "curriculum_late_phase": 1.0,
      "epoch": 2.78e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.370018005371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0194,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17965333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.479000596877352,
      "rewrite6_current_pass_fraction": 0.3282311931607953,
      "rewrite6_full_passes": 0.0,
      "step": 87560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7868032,
      "curriculum_late_phase": 1.0,
      "epoch": 2.785e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.370018005371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0489,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17994666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.092130313776398,
      "rewrite6_current_pass_fraction": 0.3288933029541781,
      "rewrite6_full_passes": 0.0,
      "step": 87570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7892608,
      "curriculum_late_phase": 1.0,
      "epoch": 2.79e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.370018005371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0636,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18018666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.40447473637306,
      "rewrite6_current_pass_fraction": 0.32940449066230454,
      "rewrite6_full_passes": 0.0,
      "step": 87580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7917184,
      "curriculum_late_phase": 1.0,
      "epoch": 2.795e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.370018005371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18050666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.035258373598133,
      "rewrite6_current_pass_fraction": 0.3299497575509727,
      "rewrite6_full_passes": 0.0,
      "step": 87590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.794176,
      "curriculum_late_phase": 1.0,
      "epoch": 2.8e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.370018005371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18078666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.33056318280072444,
      "rewrite6_full_passes": 0.0,
      "step": 87600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7966336,
      "curriculum_late_phase": 1.0,
      "epoch": 2.805e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18117333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.125904734609893,
      "rewrite6_current_pass_fraction": 0.3312252925941072,
      "rewrite6_full_passes": 0.0,
      "step": 87610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.7990912,
      "curriculum_late_phase": 1.0,
      "epoch": 2.81e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0249,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18153333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.591945413691505,
      "rewrite6_current_pass_fraction": 0.33190687620494247,
      "rewrite6_full_passes": 0.0,
      "step": 87620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8015488,
      "curriculum_late_phase": 1.0,
      "epoch": 2.815e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18185333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.989031673191437,
      "rewrite6_current_pass_fraction": 0.3325251699090573,
      "rewrite6_full_passes": 0.0,
      "step": 87630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8040064,
      "curriculum_late_phase": 1.0,
      "epoch": 2.82e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18216,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.819707206509154,
      "rewrite6_current_pass_fraction": 0.3330655683433624,
      "rewrite6_full_passes": 0.0,
      "step": 87640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.806464,
      "curriculum_late_phase": 1.0,
      "epoch": 2.825e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0679,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18242666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.49671214604783,
      "rewrite6_current_pass_fraction": 0.33357188759712564,
      "rewrite6_full_passes": 0.0,
      "step": 87650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8089216,
      "curriculum_late_phase": 1.0,
      "epoch": 2.83e-06,
      "grad_norm": 0.53515625,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0582,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18269333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28920208905679,
      "rewrite6_current_pass_fraction": 0.3341950497556036,
      "rewrite6_full_passes": 0.0,
      "step": 87660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8113792,
      "curriculum_late_phase": 1.0,
      "epoch": 2.835e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0409,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18293333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.924066423168412,
      "rewrite6_current_pass_fraction": 0.33473057973554554,
      "rewrite6_full_passes": 0.0,
      "step": 87670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8138368,
      "curriculum_late_phase": 1.0,
      "epoch": 2.84e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0262,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18333333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.618732350465706,
      "rewrite6_current_pass_fraction": 0.33529532044166616,
      "rewrite6_full_passes": 0.0,
      "step": 87680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8162944,
      "curriculum_late_phase": 1.0,
      "epoch": 2.845e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18369333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.028948742578034,
      "rewrite6_current_pass_fraction": 0.3358113766041557,
      "rewrite6_full_passes": 0.0,
      "step": 87690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.818752,
      "curriculum_late_phase": 1.0,
      "epoch": 2.85e-06,
      "grad_norm": 0.4296875,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0236,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18398666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.565193277309945,
      "rewrite6_current_pass_fraction": 0.3364101964908181,
      "rewrite6_full_passes": 0.0,
      "step": 87700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8212096,
      "curriculum_late_phase": 1.0,
      "epoch": 2.855e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0867,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18421333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.904673152109797,
      "rewrite6_current_pass_fraction": 0.33694085801639695,
      "rewrite6_full_passes": 0.0,
      "step": 87710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8236672,
      "curriculum_late_phase": 1.0,
      "epoch": 2.86e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18438666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.87808407616535,
      "rewrite6_current_pass_fraction": 0.3376370469903215,
      "rewrite6_full_passes": 0.0,
      "step": 87720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8261248,
      "curriculum_late_phase": 1.0,
      "epoch": 2.865e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18470666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01213231105346,
      "rewrite6_current_pass_fraction": 0.33826994605752564,
      "rewrite6_full_passes": 0.0,
      "step": 87730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8285824,
      "curriculum_late_phase": 1.0,
      "epoch": 2.87e-06,
      "grad_norm": 0.5546875,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18505333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86556098302329,
      "rewrite6_current_pass_fraction": 0.33894179275963465,
      "rewrite6_full_passes": 0.0,
      "step": 87740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.83104,
      "curriculum_late_phase": 1.0,
      "epoch": 2.875e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18533333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.945000955113006,
      "rewrite6_current_pass_fraction": 0.3395698233724757,
      "rewrite6_full_passes": 0.0,
      "step": 87750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8334976,
      "curriculum_late_phase": 1.0,
      "epoch": 2.88e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0702,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18561333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.54621148640419,
      "rewrite6_current_pass_fraction": 0.34020759089404295,
      "rewrite6_full_passes": 0.0,
      "step": 87760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8359552,
      "curriculum_late_phase": 1.0,
      "epoch": 2.885e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0628,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18581333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.38735800418973,
      "rewrite6_current_pass_fraction": 0.3407382524196218,
      "rewrite6_full_passes": 0.0,
      "step": 87770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8384128,
      "curriculum_late_phase": 1.0,
      "epoch": 2.89e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18601333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.986932874965774,
      "rewrite6_current_pass_fraction": 0.3413516776693735,
      "rewrite6_full_passes": 0.0,
      "step": 87780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8408704,
      "curriculum_late_phase": 1.0,
      "epoch": 2.895e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18634666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10267901588904,
      "rewrite6_current_pass_fraction": 0.34196997137348834,
      "rewrite6_full_passes": 0.0,
      "step": 87790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.843328,
      "curriculum_late_phase": 1.0,
      "epoch": 2.9e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.37014389038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0565,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18678666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25304119097752,
      "rewrite6_current_pass_fraction": 0.34253471207960895,
      "rewrite6_full_passes": 0.0,
      "step": 87800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8457856,
      "curriculum_late_phase": 1.0,
      "epoch": 2.905e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.370189666748047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0345,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18710666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.790580010214892,
      "rewrite6_current_pass_fraction": 0.34323576950789664,
      "rewrite6_full_passes": 0.0,
      "step": 87810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8482432,
      "curriculum_late_phase": 1.0,
      "epoch": 2.91e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.370189666748047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18746666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.132243456791066,
      "rewrite6_current_pass_fraction": 0.3438443263032852,
      "rewrite6_full_passes": 0.0,
      "step": 87820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8507008,
      "curriculum_late_phase": 1.0,
      "epoch": 2.915e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.370189666748047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18778666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.911515748893304,
      "rewrite6_current_pass_fraction": 0.34436525092013787,
      "rewrite6_full_passes": 0.0,
      "step": 87830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8531584,
      "curriculum_late_phase": 1.0,
      "epoch": 2.92e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.370189666748047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0333,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18810666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.765646277434353,
      "rewrite6_current_pass_fraction": 0.3449397285349847,
      "rewrite6_full_passes": 0.0,
      "step": 87840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.855616,
      "curriculum_late_phase": 1.0,
      "epoch": 2.925e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1884,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.961763660065063,
      "rewrite6_current_pass_fraction": 0.3454655216062005,
      "rewrite6_full_passes": 0.0,
      "step": 87850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8580736,
      "curriculum_late_phase": 1.0,
      "epoch": 2.93e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18868,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.3460399992210473,
      "rewrite6_full_passes": 0.0,
      "step": 87860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8605312,
      "curriculum_late_phase": 1.0,
      "epoch": 2.935e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0244,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18905333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.581652014548887,
      "rewrite6_current_pass_fraction": 0.3465901345640786,
      "rewrite6_full_passes": 0.0,
      "step": 87870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8629888,
      "curriculum_late_phase": 1.0,
      "epoch": 2.94e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.048,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1894,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.073155936244657,
      "rewrite6_current_pass_fraction": 0.347101322272205,
      "rewrite6_full_passes": 0.0,
      "step": 87880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8654464,
      "curriculum_late_phase": 1.0,
      "epoch": 2.945e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18969333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.3477050106132305,
      "rewrite6_full_passes": 0.0,
      "step": 87890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.867904,
      "curriculum_late_phase": 1.0,
      "epoch": 2.95e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18998666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.016335157786333,
      "rewrite6_current_pass_fraction": 0.348308698954256,
      "rewrite6_full_passes": 0.0,
      "step": 87900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8703616,
      "curriculum_late_phase": 1.0,
      "epoch": 2.955e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.064,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19022666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.413038238853925,
      "rewrite6_current_pass_fraction": 0.3489123872952815,
      "rewrite6_full_passes": 0.0,
      "step": 87910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8728192,
      "curriculum_late_phase": 1.0,
      "epoch": 2.96e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19048,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.041569897791508,
      "rewrite6_current_pass_fraction": 0.3495209440906701,
      "rewrite6_full_passes": 0.0,
      "step": 87920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8752768,
      "curriculum_late_phase": 1.0,
      "epoch": 2.965e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1908,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.898972602764104,
      "rewrite6_current_pass_fraction": 0.35011489552296937,
      "rewrite6_full_passes": 0.0,
      "step": 87930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8777344,
      "curriculum_late_phase": 1.0,
      "epoch": 2.97e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19106666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21482012611182,
      "rewrite6_current_pass_fraction": 0.35076239995326286,
      "rewrite6_full_passes": 0.0,
      "step": 87940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.880192,
      "curriculum_late_phase": 1.0,
      "epoch": 2.975e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0303,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19142666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.70344269063493,
      "rewrite6_current_pass_fraction": 0.3513027983875679,
      "rewrite6_full_passes": 0.0,
      "step": 87950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8826496,
      "curriculum_late_phase": 1.0,
      "epoch": 2.98e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19185333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.974344492133298,
      "rewrite6_current_pass_fraction": 0.3519600397265876,
      "rewrite6_full_passes": 0.0,
      "step": 87960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8851072,
      "curriculum_late_phase": 1.0,
      "epoch": 2.985e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0266,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19212,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.626981493124443,
      "rewrite6_current_pass_fraction": 0.3525101750696189,
      "rewrite6_full_passes": 0.0,
      "step": 87970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8875648,
      "curriculum_late_phase": 1.0,
      "epoch": 2.99e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19237333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.3531187318650075,
      "rewrite6_full_passes": 0.0,
      "step": 87980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8900224,
      "curriculum_late_phase": 1.0,
      "epoch": 2.995e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0347,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19265333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.79473854205626,
      "rewrite6_current_pass_fraction": 0.3536493933905864,
      "rewrite6_full_passes": 0.0,
      "step": 87990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.89248,
      "curriculum_late_phase": 1.0,
      "epoch": 3e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.370227813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0942,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.193,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.06957581274751,
      "rewrite6_current_pass_fraction": 0.3542433448228856,
      "rewrite6_full_passes": 0.0,
      "step": 88000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 3e-06,
      "eval_loss": 3.0838468074798584,
      "eval_perplexity": 21.84226397754897,
      "eval_runtime": 0.0498,
      "eval_samples_per_second": 642.049,
      "eval_steps_per_second": 20.064,
      "step": 88000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8949376,
      "curriculum_late_phase": 1.0,
      "epoch": 3.005e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.451774597167969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19337333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.3547740063484645,
      "rewrite6_full_passes": 0.0,
      "step": 88010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8973952,
      "curriculum_late_phase": 1.0,
      "epoch": 3.01e-06,
      "grad_norm": 0.5625,
      "host_rss_gib": 15.451801300048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19374666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.144926607429838,
      "rewrite6_current_pass_fraction": 0.3553436155089482,
      "rewrite6_full_passes": 0.0,
      "step": 88020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.8998528,
      "curriculum_late_phase": 1.0,
      "epoch": 3.015e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.451801300048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0623,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19394666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.376666998161866,
      "rewrite6_current_pass_fraction": 0.35589861930634265,
      "rewrite6_full_passes": 0.0,
      "step": 88030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9023104,
      "curriculum_late_phase": 1.0,
      "epoch": 3.02e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.451801300048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19425333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.248791007771803,
      "rewrite6_current_pass_fraction": 0.3564584915581001,
      "rewrite6_full_passes": 0.0,
      "step": 88040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.904768,
      "curriculum_late_phase": 1.0,
      "epoch": 3.025e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.451801300048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0368,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19456,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.838453377506596,
      "rewrite6_current_pass_fraction": 0.35708652217094117,
      "rewrite6_full_passes": 0.0,
      "step": 88050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9072256,
      "curriculum_late_phase": 1.0,
      "epoch": 3.03e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.451801300048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19488,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20845663464543,
      "rewrite6_current_pass_fraction": 0.35774376350996084,
      "rewrite6_full_passes": 0.0,
      "step": 88060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9096832,
      "curriculum_late_phase": 1.0,
      "epoch": 3.035e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.803199768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0486,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1952,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.085803623733216,
      "rewrite6_current_pass_fraction": 0.35824521430936107,
      "rewrite6_full_passes": 0.0,
      "step": 88070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9121408,
      "curriculum_late_phase": 1.0,
      "epoch": 3.04e-06,
      "grad_norm": 0.57421875,
      "host_rss_gib": 14.807064056396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19549333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10267901588904,
      "rewrite6_current_pass_fraction": 0.35878561274366616,
      "rewrite6_full_passes": 0.0,
      "step": 88080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9145984,
      "curriculum_late_phase": 1.0,
      "epoch": 3.045e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.808082580566406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19586666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.84679042615617,
      "rewrite6_current_pass_fraction": 0.3594818017175907,
      "rewrite6_full_passes": 0.0,
      "step": 88090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.917056,
      "curriculum_late_phase": 1.0,
      "epoch": 3.05e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.816009521484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19610666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.066834937660975,
      "rewrite6_current_pass_fraction": 0.36005627933243756,
      "rewrite6_full_passes": 0.0,
      "step": 88100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9195136,
      "curriculum_late_phase": 1.0,
      "epoch": 3.055e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.816032409667969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0647,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1964,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.42803261303982,
      "rewrite6_current_pass_fraction": 0.36062102003855817,
      "rewrite6_full_passes": 0.0,
      "step": 88110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9219712,
      "curriculum_late_phase": 1.0,
      "epoch": 3.06e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.816032409667969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0589,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19670666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.304109747590893,
      "rewrite6_current_pass_fraction": 0.36122470837958365,
      "rewrite6_full_passes": 0.0,
      "step": 88120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9244288,
      "curriculum_late_phase": 1.0,
      "epoch": 3.065e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.816043853759766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0438,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.197,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.98483428660945,
      "rewrite6_current_pass_fraction": 0.361867344355514,
      "rewrite6_full_passes": 0.0,
      "step": 88130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9268864,
      "curriculum_late_phase": 1.0,
      "epoch": 3.07e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.555953979492188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19736,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.942906559739004,
      "rewrite6_current_pass_fraction": 0.36249537496835504,
      "rewrite6_full_passes": 0.0,
      "step": 88140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.929344,
      "curriculum_late_phase": 1.0,
      "epoch": 3.075e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.522777557373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0679,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19758666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.49671214604783,
      "rewrite6_current_pass_fraction": 0.36311853712683295,
      "rewrite6_full_passes": 0.0,
      "step": 88150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9318016,
      "curriculum_late_phase": 1.0,
      "epoch": 3.08e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.929924011230469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19797333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.229675699044027,
      "rewrite6_current_pass_fraction": 0.36365893556113804,
      "rewrite6_full_passes": 0.0,
      "step": 88160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9342592,
      "curriculum_late_phase": 1.0,
      "epoch": 3.085e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.929924011230469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0561,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1982,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.244541674517738,
      "rewrite6_current_pass_fraction": 0.3642967030827053,
      "rewrite6_full_passes": 0.0,
      "step": 88170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9367168,
      "curriculum_late_phase": 1.0,
      "epoch": 3.09e-06,
      "grad_norm": 0.59765625,
      "host_rss_gib": 15.930339813232422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0793,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19869333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.743176844084687,
      "rewrite6_current_pass_fraction": 0.36492473369554634,
      "rewrite6_full_passes": 0.0,
      "step": 88180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9391744,
      "curriculum_late_phase": 1.0,
      "epoch": 3.095e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.931327819824219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0486,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19904,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.085803623733216,
      "rewrite6_current_pass_fraction": 0.36560631730638155,
      "rewrite6_full_passes": 0.0,
      "step": 88190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.941632,
      "curriculum_late_phase": 1.0,
      "epoch": 3.1e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.931877136230469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19937333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.974344492133298,
      "rewrite6_current_pass_fraction": 0.36621974255613327,
      "rewrite6_full_passes": 0.0,
      "step": 88200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9440896,
      "curriculum_late_phase": 1.0,
      "epoch": 3.105e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.931877136230469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0318,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19970666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.734521157693965,
      "rewrite6_current_pass_fraction": 0.3667942201709801,
      "rewrite6_full_passes": 0.0,
      "step": 88210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9465472,
      "curriculum_late_phase": 1.0,
      "epoch": 3.11e-06,
      "grad_norm": 0.423828125,
      "host_rss_gib": 15.931877136230469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20013333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01003120287951,
      "rewrite6_current_pass_fraction": 0.36740764542073184,
      "rewrite6_full_passes": 0.0,
      "step": 88220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9490048,
      "curriculum_late_phase": 1.0,
      "epoch": 3.115e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.931877136230469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20041333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13435678680148,
      "rewrite6_current_pass_fraction": 0.36796751767248936,
      "rewrite6_full_passes": 0.0,
      "step": 88230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9514624,
      "curriculum_late_phase": 1.0,
      "epoch": 3.12e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.931877136230469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0625,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20076,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.380942759123343,
      "rewrite6_current_pass_fraction": 0.3686150221027828,
      "rewrite6_full_passes": 0.0,
      "step": 88240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.95392,
      "curriculum_late_phase": 1.0,
      "epoch": 3.125e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0369,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20094666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.840537327040092,
      "rewrite6_current_pass_fraction": 0.36918463126326656,
      "rewrite6_full_passes": 0.0,
      "step": 88250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9563776,
      "curriculum_late_phase": 1.0,
      "epoch": 3.13e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0715,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20121333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.574239775777293,
      "rewrite6_current_pass_fraction": 0.369744503515024,
      "rewrite6_full_passes": 0.0,
      "step": 88260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9588352,
      "curriculum_late_phase": 1.0,
      "epoch": 3.135e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20168,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.7801873185992,
      "rewrite6_current_pass_fraction": 0.3703920079453175,
      "rewrite6_full_passes": 0.0,
      "step": 88270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9612928,
      "curriculum_late_phase": 1.0,
      "epoch": 3.14e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0481,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20204,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.075263357207568,
      "rewrite6_current_pass_fraction": 0.3710638546474265,
      "rewrite6_full_passes": 0.0,
      "step": 88280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9637504,
      "curriculum_late_phase": 1.0,
      "epoch": 3.145e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20229333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.24241732656946,
      "rewrite6_current_pass_fraction": 0.3715896477186423,
      "rewrite6_full_passes": 0.0,
      "step": 88290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.966208,
      "curriculum_late_phase": 1.0,
      "epoch": 3.15e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0604,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20262666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3360898913238,
      "rewrite6_current_pass_fraction": 0.3722566259663882,
      "rewrite6_full_passes": 0.0,
      "step": 88300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9686656,
      "curriculum_late_phase": 1.0,
      "epoch": 3.155e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20290666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.26367036864598,
      "rewrite6_current_pass_fraction": 0.3728262351268719,
      "rewrite6_full_passes": 0.0,
      "step": 88310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9711232,
      "curriculum_late_phase": 1.0,
      "epoch": 3.16e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20324,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35530101593532,
      "rewrite6_current_pass_fraction": 0.37341044965044495,
      "rewrite6_full_passes": 0.0,
      "step": 88320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9735808,
      "curriculum_late_phase": 1.0,
      "epoch": 3.165e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20370666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.37395084808475004,
      "rewrite6_full_passes": 0.0,
      "step": 88330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9760384,
      "curriculum_late_phase": 1.0,
      "epoch": 3.17e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0405,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20408,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91569847030129,
      "rewrite6_current_pass_fraction": 0.3746080894237697,
      "rewrite6_full_passes": 0.0,
      "step": 88340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.978496,
      "curriculum_late_phase": 1.0,
      "epoch": 3.175e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2044,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01213231105346,
      "rewrite6_current_pass_fraction": 0.3751825670386166,
      "rewrite6_full_passes": 0.0,
      "step": 88350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9809536,
      "curriculum_late_phase": 1.0,
      "epoch": 3.18e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0669,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20468,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.475226178675968,
      "rewrite6_current_pass_fraction": 0.3758203345601838,
      "rewrite6_full_passes": 0.0,
      "step": 88360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9834112,
      "curriculum_late_phase": 1.0,
      "epoch": 3.185e-06,
      "grad_norm": 0.53515625,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0366,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20498666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.83428610357238,
      "rewrite6_current_pass_fraction": 0.37648731280792974,
      "rewrite6_full_passes": 0.0,
      "step": 88370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9858688,
      "curriculum_late_phase": 1.0,
      "epoch": 3.19e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0529,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20529333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.176667797281898,
      "rewrite6_current_pass_fraction": 0.3770910011489552,
      "rewrite6_full_passes": 0.0,
      "step": 88380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9883264,
      "curriculum_late_phase": 1.0,
      "epoch": 3.195e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20554666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20845663464543,
      "rewrite6_current_pass_fraction": 0.3776946894899807,
      "rewrite6_full_passes": 0.0,
      "step": 88390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.990784,
      "curriculum_late_phase": 1.0,
      "epoch": 3.2e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20592,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.892703851343008,
      "rewrite6_current_pass_fraction": 0.37832272010282175,
      "rewrite6_full_passes": 0.0,
      "step": 88400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9932416,
      "curriculum_late_phase": 1.0,
      "epoch": 3.205e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20630666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.295589807793416,
      "rewrite6_current_pass_fraction": 0.3788777239002162,
      "rewrite6_full_passes": 0.0,
      "step": 88410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9956992,
      "curriculum_late_phase": 1.0,
      "epoch": 3.21e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.075,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.649881911512058,
      "rewrite6_current_pass_fraction": 0.37945706996942613,
      "rewrite6_full_passes": 0.0,
      "step": 88420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 15.9981568,
      "curriculum_late_phase": 1.0,
      "epoch": 3.215e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20693333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.71586848361443,
      "rewrite6_current_pass_fraction": 0.3801483904889876,
      "rewrite6_full_passes": 0.0,
      "step": 88430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0006144,
      "curriculum_late_phase": 1.0,
      "epoch": 3.22e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0311,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20724,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.720012071656146,
      "rewrite6_current_pass_fraction": 0.3807423419212868,
      "rewrite6_full_passes": 0.0,
      "step": 88440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.003072,
      "curriculum_late_phase": 1.0,
      "epoch": 3.225e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20754666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.959667588504388,
      "rewrite6_current_pass_fraction": 0.3813070826274074,
      "rewrite6_full_passes": 0.0,
      "step": 88450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0055296,
      "curriculum_late_phase": 1.0,
      "epoch": 3.23e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.931880950927734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20793333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.3818864286966174,
      "rewrite6_full_passes": 0.0,
      "step": 88460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0079872,
      "curriculum_late_phase": 1.0,
      "epoch": 3.235e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.932201385498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20825333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.3825193277638215,
      "rewrite6_full_passes": 0.0,
      "step": 88470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0104448,
      "curriculum_late_phase": 1.0,
      "epoch": 3.24e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.932201385498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20848,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.38302564701758485,
      "rewrite6_full_passes": 0.0,
      "step": 88480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0129024,
      "curriculum_late_phase": 1.0,
      "epoch": 3.245e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.932201385498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0583,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2088,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.29133111571526,
      "rewrite6_current_pass_fraction": 0.3835757823606161,
      "rewrite6_full_passes": 0.0,
      "step": 88490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.01536,
      "curriculum_late_phase": 1.0,
      "epoch": 3.25e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.932201385498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0107,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20912,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.301606076761544,
      "rewrite6_current_pass_fraction": 0.3842038129734572,
      "rewrite6_full_passes": 0.0,
      "step": 88500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 3.25e-06,
      "eval_loss": 3.083131790161133,
      "eval_perplexity": 21.82665196262051,
      "eval_runtime": 0.0484,
      "eval_samples_per_second": 660.982,
      "eval_steps_per_second": 20.656,
      "step": 88500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0178176,
      "curriculum_late_phase": 1.0,
      "epoch": 3.255e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 15.992794036865234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0155,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2094,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.399288035080676,
      "rewrite6_current_pass_fraction": 0.38472473759030984,
      "rewrite6_full_passes": 0.0,
      "step": 88510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0202752,
      "curriculum_late_phase": 1.0,
      "epoch": 3.26e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.992794036865234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20981333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21482012611182,
      "rewrite6_current_pass_fraction": 0.38531868902260913,
      "rewrite6_full_passes": 0.0,
      "step": 88520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0227328,
      "curriculum_late_phase": 1.0,
      "epoch": 3.265e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.06,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21004,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.327557162026903,
      "rewrite6_current_pass_fraction": 0.3859175089092715,
      "rewrite6_full_passes": 0.0,
      "step": 88530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0251904,
      "curriculum_late_phase": 1.0,
      "epoch": 3.27e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21032,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01003120287951,
      "rewrite6_current_pass_fraction": 0.38655527643083876,
      "rewrite6_full_passes": 0.0,
      "step": 88540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.027648,
      "curriculum_late_phase": 1.0,
      "epoch": 3.275e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21068,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.74904040364757,
      "rewrite6_current_pass_fraction": 0.3871151486825962,
      "rewrite6_full_passes": 0.0,
      "step": 88550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0301056,
      "curriculum_late_phase": 1.0,
      "epoch": 3.28e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0364,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21096,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.83011966300961,
      "rewrite6_current_pass_fraction": 0.3877042316605324,
      "rewrite6_full_passes": 0.0,
      "step": 88560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0325632,
      "curriculum_late_phase": 1.0,
      "epoch": 3.285e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21130666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.701372449879624,
      "rewrite6_current_pass_fraction": 0.38834199918209966,
      "rewrite6_full_passes": 0.0,
      "step": 88570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0350208,
      "curriculum_late_phase": 1.0,
      "epoch": 3.29e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0311,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21164,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.720012071656146,
      "rewrite6_current_pass_fraction": 0.38894081906876204,
      "rewrite6_full_passes": 0.0,
      "step": 88580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0374784,
      "curriculum_late_phase": 1.0,
      "epoch": 3.295e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0365,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21202666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.832202779129986,
      "rewrite6_current_pass_fraction": 0.3894666121399778,
      "rewrite6_full_passes": 0.0,
      "step": 88590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.039936,
      "curriculum_late_phase": 1.0,
      "epoch": 3.3e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21230666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.678613459924104,
      "rewrite6_current_pass_fraction": 0.3901433272964499,
      "rewrite6_full_passes": 0.0,
      "step": 88600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0423936,
      "curriculum_late_phase": 1.0,
      "epoch": 3.305e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.068,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21258666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.498861924749583,
      "rewrite6_current_pass_fraction": 0.39077135790929096,
      "rewrite6_full_passes": 0.0,
      "step": 88610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0448512,
      "curriculum_late_phase": 1.0,
      "epoch": 3.31e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0398,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21314666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.901062604522732,
      "rewrite6_current_pass_fraction": 0.39132149325232224,
      "rewrite6_full_passes": 0.0,
      "step": 88620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0473088,
      "curriculum_late_phase": 1.0,
      "epoch": 3.315e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21342666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.957571726540376,
      "rewrite6_current_pass_fraction": 0.39193978695643705,
      "rewrite6_full_passes": 0.0,
      "step": 88630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0497664,
      "curriculum_late_phase": 1.0,
      "epoch": 3.32e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0491,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21369333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09634916170988,
      "rewrite6_current_pass_fraction": 0.3924607115732897,
      "rewrite6_full_passes": 0.0,
      "step": 88640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.052224,
      "curriculum_late_phase": 1.0,
      "epoch": 3.325e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2142,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.896882809995212,
      "rewrite6_current_pass_fraction": 0.39301571537068414,
      "rewrite6_full_passes": 0.0,
      "step": 88650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0546816,
      "curriculum_late_phase": 1.0,
      "epoch": 3.33e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.99288558959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21449333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.89479322619514,
      "rewrite6_current_pass_fraction": 0.3936145352573465,
      "rewrite6_full_passes": 0.0,
      "step": 88660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0571392,
      "curriculum_late_phase": 1.0,
      "epoch": 3.335e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21486666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.07737098892312,
      "rewrite6_current_pass_fraction": 0.3942133551440089,
      "rewrite6_full_passes": 0.0,
      "step": 88670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0595968,
      "curriculum_late_phase": 1.0,
      "epoch": 3.34e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21521333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.39473914821522466,
      "rewrite6_full_passes": 0.0,
      "step": 88680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0620544,
      "curriculum_late_phase": 1.0,
      "epoch": 3.345e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.058,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21554666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.284944674394637,
      "rewrite6_current_pass_fraction": 0.39528928355825593,
      "rewrite6_full_passes": 0.0,
      "step": 88690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.064512,
      "curriculum_late_phase": 1.0,
      "epoch": 3.35e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21582666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.959667588504388,
      "rewrite6_current_pass_fraction": 0.3958637611731028,
      "rewrite6_full_passes": 0.0,
      "step": 88700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0669696,
      "curriculum_late_phase": 1.0,
      "epoch": 3.355e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0718,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21618666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.580713018647913,
      "rewrite6_current_pass_fraction": 0.3963749488812292,
      "rewrite6_full_passes": 0.0,
      "step": 88710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0694272,
      "curriculum_late_phase": 1.0,
      "epoch": 3.36e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21652,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.240293191045346,
      "rewrite6_current_pass_fraction": 0.3969396895873498,
      "rewrite6_full_passes": 0.0,
      "step": 88720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0718848,
      "curriculum_late_phase": 1.0,
      "epoch": 3.365e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0428,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21688,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.96385994124339,
      "rewrite6_current_pass_fraction": 0.39761153628945883,
      "rewrite6_full_passes": 0.0,
      "step": 88730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0743424,
      "curriculum_late_phase": 1.0,
      "epoch": 3.37e-06,
      "grad_norm": 0.546875,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21734666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.898972602764104,
      "rewrite6_current_pass_fraction": 0.39812759245194834,
      "rewrite6_full_passes": 0.0,
      "step": 88740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0768,
      "curriculum_late_phase": 1.0,
      "epoch": 3.375e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21768,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.058409888808015,
      "rewrite6_current_pass_fraction": 0.3987020700667952,
      "rewrite6_full_passes": 0.0,
      "step": 88750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0792576,
      "curriculum_late_phase": 1.0,
      "epoch": 3.38e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0655,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.218,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.44518189792958,
      "rewrite6_current_pass_fraction": 0.39927654768164206,
      "rewrite6_full_passes": 0.0,
      "step": 88760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0817152,
      "curriculum_late_phase": 1.0,
      "epoch": 3.385e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.992889404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0238,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21836,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.56930672729669,
      "rewrite6_current_pass_fraction": 0.3998607622052151,
      "rewrite6_full_passes": 0.0,
      "step": 88770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0841728,
      "curriculum_late_phase": 1.0,
      "epoch": 3.39e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.992912292480469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21866666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.4004741874549668,
      "rewrite6_full_passes": 0.0,
      "step": 88780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0866304,
      "curriculum_late_phase": 1.0,
      "epoch": 3.395e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.992912292480469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.042,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21890666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.947095559937008,
      "rewrite6_current_pass_fraction": 0.40109734961344473,
      "rewrite6_full_passes": 0.0,
      "step": 88790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.089088,
      "curriculum_late_phase": 1.0,
      "epoch": 3.4e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.317131042480469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0525,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21922666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.16819882407055,
      "rewrite6_current_pass_fraction": 0.40168643259138087,
      "rewrite6_full_passes": 0.0,
      "step": 88800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0915456,
      "curriculum_late_phase": 1.0,
      "epoch": 3.405e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.317131042480469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21948,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.844705851344035,
      "rewrite6_current_pass_fraction": 0.4021099881209714,
      "rewrite6_full_passes": 0.0,
      "step": 88810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0940032,
      "curriculum_late_phase": 1.0,
      "epoch": 3.41e-06,
      "grad_norm": 0.5625,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0641,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21985333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.415179649746566,
      "rewrite6_current_pass_fraction": 0.40276236100562796,
      "rewrite6_full_passes": 0.0,
      "step": 88820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0964608,
      "curriculum_late_phase": 1.0,
      "epoch": 3.415e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22010666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.976442031457733,
      "rewrite6_current_pass_fraction": 0.4034244707990107,
      "rewrite6_full_passes": 0.0,
      "step": 88830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.0989184,
      "curriculum_late_phase": 1.0,
      "epoch": 3.42e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0456,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22045333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.022641004163326,
      "rewrite6_current_pass_fraction": 0.4039600007789527,
      "rewrite6_full_passes": 0.0,
      "step": 88840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.101376,
      "curriculum_late_phase": 1.0,
      "epoch": 3.425e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0611,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2208,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.351030382809682,
      "rewrite6_current_pass_fraction": 0.40455395221125195,
      "rewrite6_full_passes": 0.0,
      "step": 88850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1038336,
      "curriculum_late_phase": 1.0,
      "epoch": 3.43e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0673,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22104,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.483817987394623,
      "rewrite6_current_pass_fraction": 0.4051332982804619,
      "rewrite6_full_passes": 0.0,
      "step": 88860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1062912,
      "curriculum_late_phase": 1.0,
      "epoch": 3.435e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0565,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22144,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25304119097752,
      "rewrite6_current_pass_fraction": 0.4056785651691301,
      "rewrite6_full_passes": 0.0,
      "step": 88870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1087488,
      "curriculum_late_phase": 1.0,
      "epoch": 3.44e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0649,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2218,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.432318648151654,
      "rewrite6_current_pass_fraction": 0.40633093805378667,
      "rewrite6_full_passes": 0.0,
      "step": 88880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1112064,
      "curriculum_late_phase": 1.0,
      "epoch": 3.445e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22218666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.819707206509154,
      "rewrite6_current_pass_fraction": 0.40675936203774027,
      "rewrite6_full_passes": 0.0,
      "step": 88890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.113664,
      "curriculum_late_phase": 1.0,
      "epoch": 3.45e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0696,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22238666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.53328763705487,
      "rewrite6_current_pass_fraction": 0.40735331347003956,
      "rewrite6_full_passes": 0.0,
      "step": 88900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1161216,
      "curriculum_late_phase": 1.0,
      "epoch": 3.455e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0319,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22282666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.736594713485793,
      "rewrite6_current_pass_fraction": 0.4079667387197912,
      "rewrite6_full_passes": 0.0,
      "step": 88910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1185792,
      "curriculum_late_phase": 1.0,
      "epoch": 3.46e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.31717300415039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22316,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06894172649243,
      "rewrite6_current_pass_fraction": 0.4085120056084594,
      "rewrite6_full_passes": 0.0,
      "step": 88920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1210368,
      "curriculum_late_phase": 1.0,
      "epoch": 3.465e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.60122299194336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0827,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22344,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.817229463470188,
      "rewrite6_current_pass_fraction": 0.4091059570407587,
      "rewrite6_full_passes": 0.0,
      "step": 88930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1234944,
      "curriculum_late_phase": 1.0,
      "epoch": 3.47e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.601951599121094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22374666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.031051742600535,
      "rewrite6_current_pass_fraction": 0.40973885610796285,
      "rewrite6_full_passes": 0.0,
      "step": 88940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.125952,
      "curriculum_late_phase": 1.0,
      "epoch": 3.475e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.602333068847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22409333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.7801873185992,
      "rewrite6_current_pass_fraction": 0.41033767599462523,
      "rewrite6_full_passes": 0.0,
      "step": 88950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1284096,
      "curriculum_late_phase": 1.0,
      "epoch": 3.48e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.602596282958984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0349,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22446666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.798897905687166,
      "rewrite6_current_pass_fraction": 0.4109754435161925,
      "rewrite6_full_passes": 0.0,
      "step": 88960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1308672,
      "curriculum_late_phase": 1.0,
      "epoch": 3.485e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.603446960449219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0606,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22478666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.340357536052313,
      "rewrite6_current_pass_fraction": 0.41163268485521215,
      "rewrite6_full_passes": 0.0,
      "step": 88970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1333248,
      "curriculum_late_phase": 1.0,
      "epoch": 3.49e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.604682922363281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22513333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.999528813094322,
      "rewrite6_current_pass_fraction": 0.41219742556133276,
      "rewrite6_full_passes": 0.0,
      "step": 88980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1357824,
      "curriculum_late_phase": 1.0,
      "epoch": 3.495e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 14.424457550048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0808,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22542666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.77581608265989,
      "rewrite6_current_pass_fraction": 0.41270861326945923,
      "rewrite6_full_passes": 0.0,
      "step": 88990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.13824,
      "curriculum_late_phase": 1.0,
      "epoch": 3.5e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.4739990234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22576,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.240293191045346,
      "rewrite6_current_pass_fraction": 0.4133463807910265,
      "rewrite6_full_passes": 0.0,
      "step": 89000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 3.5e-06,
      "eval_loss": 3.0828261375427246,
      "eval_perplexity": 21.819981608754542,
      "eval_runtime": 0.0633,
      "eval_samples_per_second": 505.681,
      "eval_steps_per_second": 15.803,
      "step": 89000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1406976,
      "curriculum_late_phase": 1.0,
      "epoch": 3.505e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.979045867919922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22618666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.132243456791066,
      "rewrite6_current_pass_fraction": 0.413906253042784,
      "rewrite6_full_passes": 0.0,
      "step": 89010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1431552,
      "curriculum_late_phase": 1.0,
      "epoch": 3.51e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0696,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22654666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.53328763705487,
      "rewrite6_current_pass_fraction": 0.41460244201670854,
      "rewrite6_full_passes": 0.0,
      "step": 89020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1456128,
      "curriculum_late_phase": 1.0,
      "epoch": 3.515e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22685333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.229675699044027,
      "rewrite6_current_pass_fraction": 0.4152596833557282,
      "rewrite6_full_passes": 0.0,
      "step": 89030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1480704,
      "curriculum_late_phase": 1.0,
      "epoch": 3.52e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0525,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22726666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.16819882407055,
      "rewrite6_current_pass_fraction": 0.41575626570076535,
      "rewrite6_full_passes": 0.0,
      "step": 89040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.150528,
      "curriculum_late_phase": 1.0,
      "epoch": 3.525e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22765333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.41628205877198105,
      "rewrite6_full_passes": 0.0,
      "step": 89050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1529856,
      "curriculum_late_phase": 1.0,
      "epoch": 3.53e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0269,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22804,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.63317051587937,
      "rewrite6_current_pass_fraction": 0.416861404841191,
      "rewrite6_full_passes": 0.0,
      "step": 89060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1554432,
      "curriculum_late_phase": 1.0,
      "epoch": 3.535e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0495,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22844,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.104789389307548,
      "rewrite6_current_pass_fraction": 0.41746509318221653,
      "rewrite6_full_passes": 0.0,
      "step": 89070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1579008,
      "curriculum_late_phase": 1.0,
      "epoch": 3.54e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2288,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.4180152285252478,
      "rewrite6_full_passes": 0.0,
      "step": 89080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1603584,
      "curriculum_late_phase": 1.0,
      "epoch": 3.545e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22924,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.225430188469428,
      "rewrite6_current_pass_fraction": 0.4185556269595529,
      "rewrite6_full_passes": 0.0,
      "step": 89090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.162816,
      "curriculum_late_phase": 1.0,
      "epoch": 3.55e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0226,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22964,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.544638363202598,
      "rewrite6_current_pass_fraction": 0.4190765515764055,
      "rewrite6_full_passes": 0.0,
      "step": 89100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1652736,
      "curriculum_late_phase": 1.0,
      "epoch": 3.555e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.078,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.22990666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.71492907921277,
      "rewrite6_current_pass_fraction": 0.4196315553737999,
      "rewrite6_full_passes": 0.0,
      "step": 89110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1677312,
      "curriculum_late_phase": 1.0,
      "epoch": 3.56e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0219,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23021333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.53026214861049,
      "rewrite6_current_pass_fraction": 0.42032774434772446,
      "rewrite6_full_passes": 0.0,
      "step": 89120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1701888,
      "curriculum_late_phase": 1.0,
      "epoch": 3.565e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0551,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23050666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.22330775157419,
      "rewrite6_current_pass_fraction": 0.420887616599482,
      "rewrite6_full_passes": 0.0,
      "step": 89130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1726464,
      "curriculum_late_phase": 1.0,
      "epoch": 3.57e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23090666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86556098302329,
      "rewrite6_current_pass_fraction": 0.4214377519425133,
      "rewrite6_full_passes": 0.0,
      "step": 89140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.175104,
      "curriculum_late_phase": 1.0,
      "epoch": 3.575e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0643,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23132,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.419463114008664,
      "rewrite6_current_pass_fraction": 0.4220365718291757,
      "rewrite6_full_passes": 0.0,
      "step": 89150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1775616,
      "curriculum_late_phase": 1.0,
      "epoch": 3.58e-06,
      "grad_norm": 0.55078125,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23154666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.738668476643575,
      "rewrite6_current_pass_fraction": 0.422586707172207,
      "rewrite6_full_passes": 0.0,
      "step": 89160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1800192,
      "curriculum_late_phase": 1.0,
      "epoch": 3.585e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0639,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23177333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.41089704209166,
      "rewrite6_current_pass_fraction": 0.4232293431481373,
      "rewrite6_full_passes": 0.0,
      "step": 89170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1824768,
      "curriculum_late_phase": 1.0,
      "epoch": 3.59e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0635,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23210666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.402334395918224,
      "rewrite6_current_pass_fraction": 0.4237210570388113,
      "rewrite6_full_passes": 0.0,
      "step": 89180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1849344,
      "curriculum_late_phase": 1.0,
      "epoch": 3.595e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0472,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23246666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.05630415310768,
      "rewrite6_current_pass_fraction": 0.42434421919728926,
      "rewrite6_full_passes": 0.0,
      "step": 89190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.187392,
      "curriculum_late_phase": 1.0,
      "epoch": 3.6e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.052,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23285333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.157617370242413,
      "rewrite6_current_pass_fraction": 0.42488948608595745,
      "rewrite6_full_passes": 0.0,
      "step": 89200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1898496,
      "curriculum_late_phase": 1.0,
      "epoch": 3.605e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.979785919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0811,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23309333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.782349807494416,
      "rewrite6_current_pass_fraction": 0.42543475297462563,
      "rewrite6_full_passes": 0.0,
      "step": 89210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1923072,
      "curriculum_late_phase": 1.0,
      "epoch": 3.61e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.980968475341797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23345333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.42600923058947243,
      "rewrite6_full_passes": 0.0,
      "step": 89220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1947648,
      "curriculum_late_phase": 1.0,
      "epoch": 3.615e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.980968475341797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0373,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23378666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.848875209436216,
      "rewrite6_current_pass_fraction": 0.42656423438686686,
      "rewrite6_full_passes": 0.0,
      "step": 89230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.1972224,
      "curriculum_late_phase": 1.0,
      "epoch": 3.62e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.980968475341797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0479,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23410666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.071048726013295,
      "rewrite6_current_pass_fraction": 0.42705107982317775,
      "rewrite6_full_passes": 0.0,
      "step": 89240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.19968,
      "curriculum_late_phase": 1.0,
      "epoch": 3.625e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.980968475341797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23438666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.4276547681642032,
      "rewrite6_full_passes": 0.0,
      "step": 89250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2021376,
      "curriculum_late_phase": 1.0,
      "epoch": 3.63e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.980968475341797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0848,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23476,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.863093786027026,
      "rewrite6_current_pass_fraction": 0.4283558255924909,
      "rewrite6_full_passes": 0.0,
      "step": 89260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2045952,
      "curriculum_late_phase": 1.0,
      "epoch": 3.635e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.981800079345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23504,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.892703851343008,
      "rewrite6_current_pass_fraction": 0.4289303032073377,
      "rewrite6_full_passes": 0.0,
      "step": 89270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2070528,
      "curriculum_late_phase": 1.0,
      "epoch": 3.64e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.981800079345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0267,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23534666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.629044294412097,
      "rewrite6_current_pass_fraction": 0.4296605713618041,
      "rewrite6_full_passes": 0.0,
      "step": 89280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2095104,
      "curriculum_late_phase": 1.0,
      "epoch": 3.645e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.981807708740234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23566666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.064728359497874,
      "rewrite6_current_pass_fraction": 0.43020583825047226,
      "rewrite6_full_passes": 0.0,
      "step": 89290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.211968,
      "curriculum_late_phase": 1.0,
      "epoch": 3.65e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.982524871826172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.066,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23596,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.455907169973106,
      "rewrite6_current_pass_fraction": 0.4308533426807657,
      "rewrite6_full_passes": 0.0,
      "step": 89300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2144256,
      "curriculum_late_phase": 1.0,
      "epoch": 3.655e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.982524871826172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23625333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.225430188469428,
      "rewrite6_current_pass_fraction": 0.4315154524741485,
      "rewrite6_full_passes": 0.0,
      "step": 89310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2168832,
      "curriculum_late_phase": 1.0,
      "epoch": 3.66e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.982524871826172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0315,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.728301734306807,
      "rewrite6_current_pass_fraction": 0.43218729917625753,
      "rewrite6_full_passes": 0.0,
      "step": 89320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2193408,
      "curriculum_late_phase": 1.0,
      "epoch": 3.665e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.982528686523438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23697333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.69516296961328,
      "rewrite6_current_pass_fraction": 0.43280072442600925,
      "rewrite6_full_passes": 0.0,
      "step": 89330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2217984,
      "curriculum_late_phase": 1.0,
      "epoch": 3.67e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23730666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86556098302329,
      "rewrite6_current_pass_fraction": 0.43342875503885026,
      "rewrite6_full_passes": 0.0,
      "step": 89340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.224256,
      "curriculum_late_phase": 1.0,
      "epoch": 3.675e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23754666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.678613459924104,
      "rewrite6_current_pass_fraction": 0.4339788903818816,
      "rewrite6_full_passes": 0.0,
      "step": 89350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2267136,
      "curriculum_late_phase": 1.0,
      "epoch": 3.68e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0367,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23784,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.836369636357645,
      "rewrite6_current_pass_fraction": 0.43453389417927596,
      "rewrite6_full_passes": 0.0,
      "step": 89360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2291712,
      "curriculum_late_phase": 1.0,
      "epoch": 3.685e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0779,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23821333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.71275769487588,
      "rewrite6_current_pass_fraction": 0.4351473194290277,
      "rewrite6_full_passes": 0.0,
      "step": 89370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2316288,
      "curriculum_late_phase": 1.0,
      "epoch": 3.69e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0293,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23853333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.68274959621593,
      "rewrite6_current_pass_fraction": 0.4357120601351483,
      "rewrite6_full_passes": 0.0,
      "step": 89380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2340864,
      "curriculum_late_phase": 1.0,
      "epoch": 3.695e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23886666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21482012611182,
      "rewrite6_current_pass_fraction": 0.4363060115674476,
      "rewrite6_full_passes": 0.0,
      "step": 89390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.236544,
      "curriculum_late_phase": 1.0,
      "epoch": 3.7e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0377,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23918666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.857216427652418,
      "rewrite6_current_pass_fraction": 0.43689022609102063,
      "rewrite6_full_passes": 0.0,
      "step": 89400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2390016,
      "curriculum_late_phase": 1.0,
      "epoch": 3.705e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23941333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.039465846006077,
      "rewrite6_current_pass_fraction": 0.43756694124749274,
      "rewrite6_full_passes": 0.0,
      "step": 89410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2414592,
      "curriculum_late_phase": 1.0,
      "epoch": 3.71e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.23972,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35530101593532,
      "rewrite6_current_pass_fraction": 0.4381414188623396,
      "rewrite6_full_passes": 0.0,
      "step": 89420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2439168,
      "curriculum_late_phase": 1.0,
      "epoch": 3.715e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24008,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.828036755190443,
      "rewrite6_current_pass_fraction": 0.43877431792954374,
      "rewrite6_full_passes": 0.0,
      "step": 89430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2463744,
      "curriculum_late_phase": 1.0,
      "epoch": 3.72e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.058,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24041333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.284944674394637,
      "rewrite6_current_pass_fraction": 0.4393293217269381,
      "rewrite6_full_passes": 0.0,
      "step": 89440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.248832,
      "curriculum_late_phase": 1.0,
      "epoch": 3.725e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0412,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24069333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.93034458477251,
      "rewrite6_current_pass_fraction": 0.43988919397869564,
      "rewrite6_full_passes": 0.0,
      "step": 89450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2512896,
      "curriculum_late_phase": 1.0,
      "epoch": 3.73e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0516,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.241,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14915601567805,
      "rewrite6_current_pass_fraction": 0.44045393468481625,
      "rewrite6_full_passes": 0.0,
      "step": 89460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2537472,
      "curriculum_late_phase": 1.0,
      "epoch": 3.735e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0549,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24138666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.219063514461734,
      "rewrite6_current_pass_fraction": 0.4410235438453,
      "rewrite6_full_passes": 0.0,
      "step": 89470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2562048,
      "curriculum_late_phase": 1.0,
      "epoch": 3.74e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2418,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.974344492133298,
      "rewrite6_current_pass_fraction": 0.4416369690950517,
      "rewrite6_full_passes": 0.0,
      "step": 89480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2586624,
      "curriculum_late_phase": 1.0,
      "epoch": 3.745e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24209333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.4421286829857257,
      "rewrite6_full_passes": 0.0,
      "step": 89490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.26112,
      "curriculum_late_phase": 1.0,
      "epoch": 3.75e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.98257064819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0695,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24250666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.531134415954018,
      "rewrite6_current_pass_fraction": 0.44278105587038225,
      "rewrite6_full_passes": 0.0,
      "step": 89500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 3.75e-06,
      "eval_loss": 3.083512306213379,
      "eval_perplexity": 21.8349589344269,
      "eval_runtime": 0.0496,
      "eval_samples_per_second": 645.743,
      "eval_steps_per_second": 20.179,
      "step": 89500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2635776,
      "curriculum_late_phase": 1.0,
      "epoch": 3.755e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2428,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.44334092812213977,
      "rewrite6_full_passes": 0.0,
      "step": 89510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2660352,
      "curriculum_late_phase": 1.0,
      "epoch": 3.76e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0228,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24324,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.5487477017954,
      "rewrite6_current_pass_fraction": 0.4439154057369866,
      "rewrite6_full_passes": 0.0,
      "step": 89520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2684928,
      "curriculum_late_phase": 1.0,
      "epoch": 3.765e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24350666666666668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.4444411988082024,
      "rewrite6_full_passes": 0.0,
      "step": 89530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2709504,
      "curriculum_late_phase": 1.0,
      "epoch": 3.77e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0538,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24389333333333332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.19573537742345,
      "rewrite6_current_pass_fraction": 0.44496212342505503,
      "rewrite6_full_passes": 0.0,
      "step": 89540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.273408,
      "curriculum_late_phase": 1.0,
      "epoch": 3.775e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0546,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24424,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21269875016978,
      "rewrite6_current_pass_fraction": 0.44563883858152714,
      "rewrite6_full_passes": 0.0,
      "step": 89550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2758656,
      "curriculum_late_phase": 1.0,
      "epoch": 3.78e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24466666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.4462376584681895,
      "rewrite6_full_passes": 0.0,
      "step": 89560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2783232,
      "curriculum_late_phase": 1.0,
      "epoch": 3.785e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0842,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24510666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.849979864325338,
      "rewrite6_current_pass_fraction": 0.4468364783548519,
      "rewrite6_full_passes": 0.0,
      "step": 89570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2807808,
      "curriculum_late_phase": 1.0,
      "epoch": 3.79e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24538666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.69516296961328,
      "rewrite6_current_pass_fraction": 0.4474450351502405,
      "rewrite6_full_passes": 0.0,
      "step": 89580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2832384,
      "curriculum_late_phase": 1.0,
      "epoch": 3.795e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0436,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24566666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.980637739420835,
      "rewrite6_current_pass_fraction": 0.4480633288543553,
      "rewrite6_full_passes": 0.0,
      "step": 89590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.285696,
      "curriculum_late_phase": 1.0,
      "epoch": 3.8e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24601333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.873908876883966,
      "rewrite6_current_pass_fraction": 0.4487010963759226,
      "rewrite6_full_passes": 0.0,
      "step": 89600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2881536,
      "curriculum_late_phase": 1.0,
      "epoch": 3.805e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0526,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24638666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.170315749797474,
      "rewrite6_current_pass_fraction": 0.44920254717532276,
      "rewrite6_full_passes": 0.0,
      "step": 89610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2906112,
      "curriculum_late_phase": 1.0,
      "epoch": 3.81e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24674666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.191496654234413,
      "rewrite6_current_pass_fraction": 0.4498208408794376,
      "rewrite6_full_passes": 0.0,
      "step": 89620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2930688,
      "curriculum_late_phase": 1.0,
      "epoch": 3.815e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24709333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86556098302329,
      "rewrite6_current_pass_fraction": 0.4504050554030107,
      "rewrite6_full_passes": 0.0,
      "step": 89630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.2955264,
      "curriculum_late_phase": 1.0,
      "epoch": 3.82e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24756,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08158688469643,
      "rewrite6_current_pass_fraction": 0.4509357169285895,
      "rewrite6_full_passes": 0.0,
      "step": 89640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.297984,
      "curriculum_late_phase": 1.0,
      "epoch": 3.825e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2478,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.821789281331817,
      "rewrite6_current_pass_fraction": 0.4515150629977995,
      "rewrite6_full_passes": 0.0,
      "step": 89650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3004416,
      "curriculum_late_phase": 1.0,
      "epoch": 3.83e-06,
      "grad_norm": 0.55078125,
      "host_rss_gib": 15.990673065185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0368,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24812,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.838453377506596,
      "rewrite6_current_pass_fraction": 0.4520213822515628,
      "rewrite6_full_passes": 0.0,
      "step": 89660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3028992,
      "curriculum_late_phase": 1.0,
      "epoch": 3.835e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.990699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0619,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24845333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.36811804126797,
      "rewrite6_current_pass_fraction": 0.4525861229576834,
      "rewrite6_full_passes": 0.0,
      "step": 89670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3053568,
      "curriculum_late_phase": 1.0,
      "epoch": 3.84e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.990699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24870666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.45312652139198845,
      "rewrite6_full_passes": 0.0,
      "step": 89680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3078144,
      "curriculum_late_phase": 1.0,
      "epoch": 3.845e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.990699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24898666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86347453124931,
      "rewrite6_current_pass_fraction": 0.45373507818737707,
      "rewrite6_full_passes": 0.0,
      "step": 89690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.310272,
      "curriculum_late_phase": 1.0,
      "epoch": 3.85e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.990699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24934666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.454270608167319,
      "rewrite6_full_passes": 0.0,
      "step": 89700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3127296,
      "curriculum_late_phase": 1.0,
      "epoch": 3.855e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.990726470947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0598,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24965333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.323292077117205,
      "rewrite6_current_pass_fraction": 0.45483048041907653,
      "rewrite6_full_passes": 0.0,
      "step": 89710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3151872,
      "curriculum_late_phase": 1.0,
      "epoch": 3.86e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.990726470947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.24998666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.049988209300682,
      "rewrite6_current_pass_fraction": 0.4554633794862807,
      "rewrite6_full_passes": 0.0,
      "step": 89720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3176448,
      "curriculum_late_phase": 1.0,
      "epoch": 3.865e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.990726470947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0428,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2503066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.96385994124339,
      "rewrite6_current_pass_fraction": 0.45611575237093727,
      "rewrite6_full_passes": 0.0,
      "step": 89730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3201024,
      "curriculum_late_phase": 1.0,
      "epoch": 3.87e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.990726470947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0644,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2505866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.421605167420953,
      "rewrite6_current_pass_fraction": 0.4566999668945103,
      "rewrite6_full_passes": 0.0,
      "step": 89740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.32256,
      "curriculum_late_phase": 1.0,
      "epoch": 3.875e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.990726470947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0646,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2509733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.425889916915104,
      "rewrite6_current_pass_fraction": 0.45730365523553584,
      "rewrite6_full_passes": 0.0,
      "step": 89750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3250176,
      "curriculum_late_phase": 1.0,
      "epoch": 3.88e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.990726470947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0653,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25117333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.440893290425038,
      "rewrite6_current_pass_fraction": 0.45792681739401375,
      "rewrite6_full_passes": 0.0,
      "step": 89760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3274752,
      "curriculum_late_phase": 1.0,
      "epoch": 3.885e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.990726470947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0648,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25156,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.430175523444856,
      "rewrite6_current_pass_fraction": 0.45852076882631304,
      "rewrite6_full_passes": 0.0,
      "step": 89770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3299328,
      "curriculum_late_phase": 1.0,
      "epoch": 3.89e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.990726470947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2518,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.4591828786196958,
      "rewrite6_full_passes": 0.0,
      "step": 89780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3323904,
      "curriculum_late_phase": 1.0,
      "epoch": 3.895e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.990726470947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0622,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25212,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.37452943834182,
      "rewrite6_current_pass_fraction": 0.45975735623454267,
      "rewrite6_full_passes": 0.0,
      "step": 89790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.334848,
      "curriculum_late_phase": 1.0,
      "epoch": 3.9e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.437938690185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2524266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.69516296961328,
      "rewrite6_current_pass_fraction": 0.4604340713910148,
      "rewrite6_full_passes": 0.0,
      "step": 89800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3373056,
      "curriculum_late_phase": 1.0,
      "epoch": 3.905e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.4407958984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25264,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.045778632630523,
      "rewrite6_current_pass_fraction": 0.46100854900586163,
      "rewrite6_full_passes": 0.0,
      "step": 89810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3397632,
      "curriculum_late_phase": 1.0,
      "epoch": 3.91e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.4407958984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25308,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.07737098892312,
      "rewrite6_current_pass_fraction": 0.4615343420770774,
      "rewrite6_full_passes": 0.0,
      "step": 89820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3422208,
      "curriculum_late_phase": 1.0,
      "epoch": 3.915e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.4427490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0683,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2534133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50531255087254,
      "rewrite6_current_pass_fraction": 0.462099082783198,
      "rewrite6_full_passes": 0.0,
      "step": 89830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3446784,
      "curriculum_late_phase": 1.0,
      "epoch": 3.92e-06,
      "grad_norm": 0.5625,
      "host_rss_gib": 15.131576538085938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0278,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25377333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.651748728285217,
      "rewrite6_current_pass_fraction": 0.462770929485307,
      "rewrite6_full_passes": 0.0,
      "step": 89840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.347136,
      "curriculum_late_phase": 1.0,
      "epoch": 3.925e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.585494995117188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2541466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.26367036864598,
      "rewrite6_current_pass_fraction": 0.4633746178263325,
      "rewrite6_full_passes": 0.0,
      "step": 89850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3495936,
      "curriculum_late_phase": 1.0,
      "epoch": 3.93e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.585494995117188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0249,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2545733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.591945413691505,
      "rewrite6_current_pass_fraction": 0.4640367276197153,
      "rewrite6_full_passes": 0.0,
      "step": 89860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3520512,
      "curriculum_late_phase": 1.0,
      "epoch": 3.935e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.585494995117188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0374,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25489333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.850960201205005,
      "rewrite6_current_pass_fraction": 0.4647183112305505,
      "rewrite6_full_passes": 0.0,
      "step": 89870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3545088,
      "curriculum_late_phase": 1.0,
      "epoch": 3.94e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.585796356201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0331,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2551333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.761493563464107,
      "rewrite6_current_pass_fraction": 0.46524410430176627,
      "rewrite6_full_passes": 0.0,
      "step": 89880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3569664,
      "curriculum_late_phase": 1.0,
      "epoch": 3.945e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.585796356201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25569333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.17878556994849,
      "rewrite6_current_pass_fraction": 0.465857529551518,
      "rewrite6_full_passes": 0.0,
      "step": 89890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.359424,
      "curriculum_late_phase": 1.0,
      "epoch": 3.95e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.58602523803711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2558933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.46646608634690656,
      "rewrite6_full_passes": 0.0,
      "step": 89900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3618816,
      "curriculum_late_phase": 1.0,
      "epoch": 3.955e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.58602523803711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0621,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25625333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.372392092267074,
      "rewrite6_current_pass_fraction": 0.4669675371463068,
      "rewrite6_full_passes": 0.0,
      "step": 89910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3643392,
      "curriculum_late_phase": 1.0,
      "epoch": 3.96e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.586639404296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0297,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25652,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.691024350895024,
      "rewrite6_current_pass_fraction": 0.46764912075714204,
      "rewrite6_full_passes": 0.0,
      "step": 89920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3667968,
      "curriculum_late_phase": 1.0,
      "epoch": 3.965e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.587615966796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25676,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.46827228291561995,
      "rewrite6_full_passes": 0.0,
      "step": 89930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3692544,
      "curriculum_late_phase": 1.0,
      "epoch": 3.97e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.587635040283203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2570266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08369514879635,
      "rewrite6_current_pass_fraction": 0.4688272867130144,
      "rewrite6_full_passes": 0.0,
      "step": 89940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.371712,
      "curriculum_late_phase": 1.0,
      "epoch": 3.975e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.587635040283203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0737,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2574,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.621755351252418,
      "rewrite6_current_pass_fraction": 0.46944071196276604,
      "rewrite6_full_passes": 0.0,
      "step": 89950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3741696,
      "curriculum_late_phase": 1.0,
      "epoch": 3.98e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.587635040283203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0521,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25764,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.159733237771047,
      "rewrite6_current_pass_fraction": 0.4699032151272614,
      "rewrite6_full_passes": 0.0,
      "step": 89960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3766272,
      "curriculum_late_phase": 1.0,
      "epoch": 3.985e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.587635040283203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25788,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.968053132536788,
      "rewrite6_current_pass_fraction": 0.47051664037701313,
      "rewrite6_full_passes": 0.0,
      "step": 89970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3790848,
      "curriculum_late_phase": 1.0,
      "epoch": 3.99e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.587635040283203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0802,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25821333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.76275451187338,
      "rewrite6_current_pass_fraction": 0.47113980253549104,
      "rewrite6_full_passes": 0.0,
      "step": 89980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3815424,
      "curriculum_late_phase": 1.0,
      "epoch": 3.995e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.587635040283203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2584533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.17878556994849,
      "rewrite6_current_pass_fraction": 0.4716899378785223,
      "rewrite6_full_passes": 0.0,
      "step": 89990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.384,
      "curriculum_late_phase": 1.0,
      "epoch": 4e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.589588165283203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.054,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25876,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.199974948441902,
      "rewrite6_current_pass_fraction": 0.4724007322155362,
      "rewrite6_full_passes": 0.0,
      "step": 90000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 4e-06,
      "eval_loss": 3.082961082458496,
      "eval_perplexity": 21.822926303016164,
      "eval_runtime": 0.0527,
      "eval_samples_per_second": 606.818,
      "eval_steps_per_second": 18.963,
      "step": 90000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3864576,
      "curriculum_late_phase": 1.0,
      "epoch": 4.005e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.805618286132812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0782,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2591466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.719272499356148,
      "rewrite6_current_pass_fraction": 0.4730141574652879,
      "rewrite6_full_passes": 0.0,
      "step": 90010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3889152,
      "curriculum_late_phase": 1.0,
      "epoch": 4.01e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.798625946044922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0834,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2594533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.832506870563275,
      "rewrite6_current_pass_fraction": 0.4736567934412183,
      "rewrite6_full_passes": 0.0,
      "step": 90020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3913728,
      "curriculum_late_phase": 1.0,
      "epoch": 4.015e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.798625946044922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0671,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.25978666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.47952165344486,
      "rewrite6_current_pass_fraction": 0.47427021869096997,
      "rewrite6_full_passes": 0.0,
      "step": 90030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3938304,
      "curriculum_late_phase": 1.0,
      "epoch": 4.02e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.798625946044922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0655,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26012,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.44518189792958,
      "rewrite6_current_pass_fraction": 0.47485930166890616,
      "rewrite6_full_passes": 0.0,
      "step": 90040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.396288,
      "curriculum_late_phase": 1.0,
      "epoch": 4.025e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.798831939697266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0374,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26044,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.850960201205005,
      "rewrite6_current_pass_fraction": 0.4754922007361103,
      "rewrite6_full_passes": 0.0,
      "step": 90050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.3987456,
      "curriculum_late_phase": 1.0,
      "epoch": 4.03e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.798858642578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0521,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26070666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.159733237771047,
      "rewrite6_current_pass_fraction": 0.47605207298786784,
      "rewrite6_full_passes": 0.0,
      "step": 90060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4012032,
      "curriculum_late_phase": 1.0,
      "epoch": 4.035e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.79913330078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26112,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.47669957741816127,
      "rewrite6_full_passes": 0.0,
      "step": 90070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4036608,
      "curriculum_late_phase": 1.0,
      "epoch": 4.04e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.79913330078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0639,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2612933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.41089704209166,
      "rewrite6_current_pass_fraction": 0.47729839730482365,
      "rewrite6_full_passes": 0.0,
      "step": 90080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4061184,
      "curriculum_late_phase": 1.0,
      "epoch": 4.045e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.79913330078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26168,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.942906559739004,
      "rewrite6_current_pass_fraction": 0.47789721719148603,
      "rewrite6_full_passes": 0.0,
      "step": 90090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.408576,
      "curriculum_late_phase": 1.0,
      "epoch": 4.05e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.79913330078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0731,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.262,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.608786189179362,
      "rewrite6_current_pass_fraction": 0.4785836692566844,
      "rewrite6_full_passes": 0.0,
      "step": 90100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4110336,
      "curriculum_late_phase": 1.0,
      "epoch": 4.055e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.79913330078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2623333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.47913380459971566,
      "rewrite6_full_passes": 0.0,
      "step": 90110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4134912,
      "curriculum_late_phase": 1.0,
      "epoch": 4.06e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.79913330078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0339,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2626266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.77810940376482,
      "rewrite6_current_pass_fraction": 0.4798494673910927,
      "rewrite6_full_passes": 0.0,
      "step": 90120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4159488,
      "curriculum_late_phase": 1.0,
      "epoch": 4.065e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.79913330078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0448,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2629066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.005829616811546,
      "rewrite6_current_pass_fraction": 0.480443418823392,
      "rewrite6_full_passes": 0.0,
      "step": 90130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4184064,
      "curriculum_late_phase": 1.0,
      "epoch": 4.07e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.79913330078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0232,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2633066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.556968844995144,
      "rewrite6_current_pass_fraction": 0.48089618507916104,
      "rewrite6_full_passes": 0.0,
      "step": 90140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.420864,
      "curriculum_late_phase": 1.0,
      "epoch": 4.075e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.79913330078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26366666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11534442254061,
      "rewrite6_current_pass_fraction": 0.4815242156920021,
      "rewrite6_full_passes": 0.0,
      "step": 90150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4233216,
      "curriculum_late_phase": 1.0,
      "epoch": 4.08e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.79913330078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0278,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26398666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.651748728285217,
      "rewrite6_current_pass_fraction": 0.4821230355786645,
      "rewrite6_full_passes": 0.0,
      "step": 90160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4257792,
      "curriculum_late_phase": 1.0,
      "epoch": 4.085e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.799530029296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2643066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.928251654962274,
      "rewrite6_current_pass_fraction": 0.482819224552589,
      "rewrite6_full_passes": 0.0,
      "step": 90170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4282368,
      "curriculum_late_phase": 1.0,
      "epoch": 4.09e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.799530029296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0261,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2647333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.61667058032089,
      "rewrite6_current_pass_fraction": 0.48332554380635234,
      "rewrite6_full_passes": 0.0,
      "step": 90180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4306944,
      "curriculum_late_phase": 1.0,
      "epoch": 4.095e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.799530029296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2650666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.898972602764104,
      "rewrite6_current_pass_fraction": 0.48386594224065743,
      "rewrite6_full_passes": 0.0,
      "step": 90190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.433152,
      "curriculum_late_phase": 1.0,
      "epoch": 4.1e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.799530029296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0586,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2654533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.297719473255697,
      "rewrite6_current_pass_fraction": 0.4843966037662363,
      "rewrite6_full_passes": 0.0,
      "step": 90200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4356096,
      "curriculum_late_phase": 1.0,
      "epoch": 4.105e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.799530029296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0737,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26568,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.621755351252418,
      "rewrite6_current_pass_fraction": 0.48502463437907734,
      "rewrite6_full_passes": 0.0,
      "step": 90210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4380672,
      "curriculum_late_phase": 1.0,
      "epoch": 4.11e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.799530029296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0633,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26608,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.398054357057195,
      "rewrite6_current_pass_fraction": 0.4855747697221086,
      "rewrite6_full_passes": 0.0,
      "step": 90220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4405248,
      "curriculum_late_phase": 1.0,
      "epoch": 4.115e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.799530029296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26648,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10689997377394,
      "rewrite6_current_pass_fraction": 0.48609569433896127,
      "rewrite6_full_passes": 0.0,
      "step": 90230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4429824,
      "curriculum_late_phase": 1.0,
      "epoch": 4.12e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.79995346069336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0772,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.69756408287407,
      "rewrite6_current_pass_fraction": 0.48669451422562365,
      "rewrite6_full_passes": 0.0,
      "step": 90240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.44544,
      "curriculum_late_phase": 1.0,
      "epoch": 4.125e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.79995346069336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2672,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.045778632630523,
      "rewrite6_current_pass_fraction": 0.4872738602948336,
      "rewrite6_full_passes": 0.0,
      "step": 90250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4478976,
      "curriculum_late_phase": 1.0,
      "epoch": 4.13e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.79995346069336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26757333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.48780939027477555,
      "rewrite6_full_passes": 0.0,
      "step": 90260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4503552,
      "curriculum_late_phase": 1.0,
      "epoch": 4.135e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.799995422363281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0588,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2678933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.301979443133135,
      "rewrite6_current_pass_fraction": 0.4884471577963428,
      "rewrite6_full_passes": 0.0,
      "step": 90270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4528128,
      "curriculum_late_phase": 1.0,
      "epoch": 4.14e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.799995422363281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26824,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.88017198896687,
      "rewrite6_current_pass_fraction": 0.488948608595743,
      "rewrite6_full_passes": 0.0,
      "step": 90280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4552704,
      "curriculum_late_phase": 1.0,
      "epoch": 4.145e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.799995422363281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0569,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26869333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.26154410792393,
      "rewrite6_current_pass_fraction": 0.4895912445716734,
      "rewrite6_full_passes": 0.0,
      "step": 90290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.457728,
      "curriculum_late_phase": 1.0,
      "epoch": 4.15e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.799995422363281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2690666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.49018519600397265,
      "rewrite6_full_passes": 0.0,
      "step": 90300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4601856,
      "curriculum_late_phase": 1.0,
      "epoch": 4.155e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26941333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.090021206202152,
      "rewrite6_current_pass_fraction": 0.4908619111604448,
      "rewrite6_full_passes": 0.0,
      "step": 90310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4626432,
      "curriculum_late_phase": 1.0,
      "epoch": 4.16e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26965333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.49154349477128,
      "rewrite6_full_passes": 0.0,
      "step": 90320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4651008,
      "curriculum_late_phase": 1.0,
      "epoch": 4.165e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0672,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.26998666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.481669713011396,
      "rewrite6_current_pass_fraction": 0.49204981402504333,
      "rewrite6_full_passes": 0.0,
      "step": 90330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4675584,
      "curriculum_late_phase": 1.0,
      "epoch": 4.17e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0487,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27036,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08791230952813,
      "rewrite6_current_pass_fraction": 0.49266810772915814,
      "rewrite6_full_passes": 0.0,
      "step": 90340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.470016,
      "curriculum_late_phase": 1.0,
      "epoch": 4.175e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0334,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27065333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.767722945893784,
      "rewrite6_current_pass_fraction": 0.4931452162567428,
      "rewrite6_full_passes": 0.0,
      "step": 90350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4724736,
      "curriculum_late_phase": 1.0,
      "epoch": 4.18e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2710266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.132243456791066,
      "rewrite6_current_pass_fraction": 0.49367100932795854,
      "rewrite6_full_passes": 0.0,
      "step": 90360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4749312,
      "curriculum_late_phase": 1.0,
      "epoch": 4.185e-06,
      "grad_norm": 0.734375,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0689,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27130666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.518219610133638,
      "rewrite6_current_pass_fraction": 0.4942406184884423,
      "rewrite6_full_passes": 0.0,
      "step": 90370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4773888,
      "curriculum_late_phase": 1.0,
      "epoch": 4.19e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0533,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2716133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18514015876014,
      "rewrite6_current_pass_fraction": 0.49482483301201535,
      "rewrite6_full_passes": 0.0,
      "step": 90380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4798464,
      "curriculum_late_phase": 1.0,
      "epoch": 4.195e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27196,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.4953554945375942,
      "rewrite6_full_passes": 0.0,
      "step": 90390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.482304,
      "curriculum_late_phase": 1.0,
      "epoch": 4.2e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27230666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.844705851344035,
      "rewrite6_current_pass_fraction": 0.4958033923390002,
      "rewrite6_full_passes": 0.0,
      "step": 90400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4847616,
      "curriculum_late_phase": 1.0,
      "epoch": 4.205e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2727866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.49638760686257327,
      "rewrite6_full_passes": 0.0,
      "step": 90410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4872192,
      "curriculum_late_phase": 1.0,
      "epoch": 4.21e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27309333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.007930304805882,
      "rewrite6_current_pass_fraction": 0.49686958384452107,
      "rewrite6_full_passes": 0.0,
      "step": 90420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4896768,
      "curriculum_late_phase": 1.0,
      "epoch": 4.215e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0267,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2733466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.629044294412097,
      "rewrite6_current_pass_fraction": 0.4974343245506417,
      "rewrite6_full_passes": 0.0,
      "step": 90430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4921344,
      "curriculum_late_phase": 1.0,
      "epoch": 4.22e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2736266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.54258400208604,
      "rewrite6_current_pass_fraction": 0.4980574867091196,
      "rewrite6_full_passes": 0.0,
      "step": 90440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.494592,
      "curriculum_late_phase": 1.0,
      "epoch": 4.225e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.800106048583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2740533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.4986173589608771,
      "rewrite6_full_passes": 0.0,
      "step": 90450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4970496,
      "curriculum_late_phase": 1.0,
      "epoch": 4.23e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.80010986328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0455,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27432,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.020538845172617,
      "rewrite6_current_pass_fraction": 0.4991334151233666,
      "rewrite6_full_passes": 0.0,
      "step": 90460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.4995072,
      "curriculum_late_phase": 1.0,
      "epoch": 4.235e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.80010986328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27454666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.265796842004743,
      "rewrite6_current_pass_fraction": 0.49974684037311834,
      "rewrite6_full_passes": 0.0,
      "step": 90470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5019648,
      "curriculum_late_phase": 1.0,
      "epoch": 4.24e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.80010986328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.054,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27482666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.199974948441902,
      "rewrite6_current_pass_fraction": 0.5003067126248758,
      "rewrite6_full_passes": 0.0,
      "step": 90480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5044224,
      "curriculum_late_phase": 1.0,
      "epoch": 4.245e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.80010986328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27528,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.5009493486008062,
      "rewrite6_full_passes": 0.0,
      "step": 90490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.50688,
      "curriculum_late_phase": 1.0,
      "epoch": 4.25e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.346477508544922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0608,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2756266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34462603439513,
      "rewrite6_current_pass_fraction": 0.5014702732176589,
      "rewrite6_full_passes": 0.0,
      "step": 90500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 4.25e-06,
      "eval_loss": 3.08317232131958,
      "eval_perplexity": 21.82753664003796,
      "eval_runtime": 0.06,
      "eval_samples_per_second": 533.533,
      "eval_steps_per_second": 16.673,
      "step": 90500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5093376,
      "curriculum_late_phase": 1.0,
      "epoch": 4.255e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.927764892578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27604,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34676060372526,
      "rewrite6_current_pass_fraction": 0.5020836984674105,
      "rewrite6_full_passes": 0.0,
      "step": 90510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5117952,
      "curriculum_late_phase": 1.0,
      "epoch": 4.26e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.120407104492188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0551,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27625333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.22330775157419,
      "rewrite6_current_pass_fraction": 0.5027701505326089,
      "rewrite6_full_passes": 0.0,
      "step": 90520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5142528,
      "curriculum_late_phase": 1.0,
      "epoch": 4.265e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.120407104492188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2766266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.7801873185992,
      "rewrite6_current_pass_fraction": 0.5034371287803548,
      "rewrite6_full_passes": 0.0,
      "step": 90530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5167104,
      "curriculum_late_phase": 1.0,
      "epoch": 4.27e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.120407104492188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0541,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27696,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20209505194016,
      "rewrite6_current_pass_fraction": 0.5040992385737376,
      "rewrite6_full_passes": 0.0,
      "step": 90540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.519168,
      "curriculum_late_phase": 1.0,
      "epoch": 4.275e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.120407104492188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0087,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2774,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.2610434407649,
      "rewrite6_current_pass_fraction": 0.5047613483671204,
      "rewrite6_full_passes": 0.0,
      "step": 90550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5216256,
      "curriculum_late_phase": 1.0,
      "epoch": 4.28e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.120407104492188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27774666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.016335157786333,
      "rewrite6_current_pass_fraction": 0.5054429319779556,
      "rewrite6_full_passes": 0.0,
      "step": 90560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5240832,
      "curriculum_late_phase": 1.0,
      "epoch": 4.285e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.120410919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27801333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.5060174095928025,
      "rewrite6_full_passes": 0.0,
      "step": 90570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5265408,
      "curriculum_late_phase": 1.0,
      "epoch": 4.29e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 16.120410919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27836,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.066834937660975,
      "rewrite6_current_pass_fraction": 0.5067720200190844,
      "rewrite6_full_passes": 0.0,
      "step": 90580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5289984,
      "curriculum_late_phase": 1.0,
      "epoch": 4.295e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.120410919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0246,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27864,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.58576875661228,
      "rewrite6_current_pass_fraction": 0.5073805768144729,
      "rewrite6_full_passes": 0.0,
      "step": 90590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.531456,
      "curriculum_late_phase": 1.0,
      "epoch": 4.3e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.120410919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.279,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.058409888808015,
      "rewrite6_current_pass_fraction": 0.5079550544293198,
      "rewrite6_full_passes": 0.0,
      "step": 90600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5339136,
      "curriculum_late_phase": 1.0,
      "epoch": 4.305e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.120410919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0596,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27941333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.319027845139193,
      "rewrite6_current_pass_fraction": 0.508587953496524,
      "rewrite6_full_passes": 0.0,
      "step": 90610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5363712,
      "curriculum_late_phase": 1.0,
      "epoch": 4.31e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.120410919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0663,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27974666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.462344907736483,
      "rewrite6_current_pass_fraction": 0.5092403263811806,
      "rewrite6_full_passes": 0.0,
      "step": 90620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5388288,
      "curriculum_late_phase": 1.0,
      "epoch": 4.315e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.120410919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0189,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.27998666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.468763656027395,
      "rewrite6_current_pass_fraction": 0.5098780939027477,
      "rewrite6_full_passes": 0.0,
      "step": 90630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5412864,
      "curriculum_late_phase": 1.0,
      "epoch": 4.32e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.120410919189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0348,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2802266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.796818119887625,
      "rewrite6_current_pass_fraction": 0.5103503339759693,
      "rewrite6_full_passes": 0.0,
      "step": 90640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.543744,
      "curriculum_late_phase": 1.0,
      "epoch": 4.325e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.12041473388672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0511,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28054666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.138584080874157,
      "rewrite6_current_pass_fraction": 0.5109540223169948,
      "rewrite6_full_passes": 0.0,
      "step": 90650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5462016,
      "curriculum_late_phase": 1.0,
      "epoch": 4.33e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.164154052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28088,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.942906559739004,
      "rewrite6_current_pass_fraction": 0.5115090261143892,
      "rewrite6_full_passes": 0.0,
      "step": 90660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5486592,
      "curriculum_late_phase": 1.0,
      "epoch": 4.335e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.165313720703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0579,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28116,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28281628634838,
      "rewrite6_current_pass_fraction": 0.5121419251815933,
      "rewrite6_full_passes": 0.0,
      "step": 90670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5511168,
      "curriculum_late_phase": 1.0,
      "epoch": 4.34e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.166645050048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28141333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.049988209300682,
      "rewrite6_current_pass_fraction": 0.512755350431345,
      "rewrite6_full_passes": 0.0,
      "step": 90680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5535744,
      "curriculum_late_phase": 1.0,
      "epoch": 4.345e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.166645050048828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0279,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28174666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.653814006420227,
      "rewrite6_current_pass_fraction": 0.5132616696851083,
      "rewrite6_full_passes": 0.0,
      "step": 90690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.556032,
      "curriculum_late_phase": 1.0,
      "epoch": 4.35e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.167251586914062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0634,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28197333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.400194269486743,
      "rewrite6_current_pass_fraction": 0.5138604895717708,
      "rewrite6_full_passes": 0.0,
      "step": 90700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5584896,
      "curriculum_late_phase": 1.0,
      "epoch": 4.355e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.167362213134766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2823333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.5144836517302487,
      "rewrite6_full_passes": 0.0,
      "step": 90710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5609472,
      "curriculum_late_phase": 1.0,
      "epoch": 4.36e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 14.168048858642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0611,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28264,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.351030382809682,
      "rewrite6_current_pass_fraction": 0.5150873400712742,
      "rewrite6_full_passes": 0.0,
      "step": 90720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5634048,
      "curriculum_late_phase": 1.0,
      "epoch": 4.365e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.168182373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0662,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28298666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.460198780553853,
      "rewrite6_current_pass_fraction": 0.5155547116901327,
      "rewrite6_full_passes": 0.0,
      "step": 90730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5658624,
      "curriculum_late_phase": 1.0,
      "epoch": 4.37e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.169261932373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0623,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2833333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.376666998161866,
      "rewrite6_current_pass_fraction": 0.516197347666063,
      "rewrite6_full_passes": 0.0,
      "step": 90740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.56832,
      "curriculum_late_phase": 1.0,
      "epoch": 4.375e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.170124053955078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28376,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.959667588504388,
      "rewrite6_current_pass_fraction": 0.5168010360070885,
      "rewrite6_full_passes": 0.0,
      "step": 90750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5707776,
      "curriculum_late_phase": 1.0,
      "epoch": 4.38e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.170387268066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2839466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.97853978054658,
      "rewrite6_current_pass_fraction": 0.5174388035286557,
      "rewrite6_full_passes": 0.0,
      "step": 90760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5732352,
      "curriculum_late_phase": 1.0,
      "epoch": 4.385e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.170387268066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.066,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2842,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.455907169973106,
      "rewrite6_current_pass_fraction": 0.5179548596911453,
      "rewrite6_full_passes": 0.0,
      "step": 90770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5756928,
      "curriculum_late_phase": 1.0,
      "epoch": 4.39e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.170387268066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28462666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.265796842004743,
      "rewrite6_current_pass_fraction": 0.5185439426690814,
      "rewrite6_full_passes": 0.0,
      "step": 90780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5781504,
      "curriculum_late_phase": 1.0,
      "epoch": 4.395e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.170387268066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0631,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28492,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.393775174118343,
      "rewrite6_current_pass_fraction": 0.5191038149208389,
      "rewrite6_full_passes": 0.0,
      "step": 90790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.580608,
      "curriculum_late_phase": 1.0,
      "epoch": 4.4e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.170387268066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2852266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.299849351695165,
      "rewrite6_current_pass_fraction": 0.5197221086249537,
      "rewrite6_full_passes": 0.0,
      "step": 90800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5830656,
      "curriculum_late_phase": 1.0,
      "epoch": 4.405e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.170387268066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0533,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2855333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18514015876014,
      "rewrite6_current_pass_fraction": 0.5203257969659792,
      "rewrite6_full_passes": 0.0,
      "step": 90810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5855232,
      "curriculum_late_phase": 1.0,
      "epoch": 4.41e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.170387268066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28584,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.32969002438444,
      "rewrite6_current_pass_fraction": 0.520987906759362,
      "rewrite6_full_passes": 0.0,
      "step": 90820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5879808,
      "curriculum_late_phase": 1.0,
      "epoch": 4.415e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.170387268066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0569,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28614666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.26154410792393,
      "rewrite6_current_pass_fraction": 0.5216354111896555,
      "rewrite6_full_passes": 0.0,
      "step": 90830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5904384,
      "curriculum_late_phase": 1.0,
      "epoch": 4.42e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.170387268066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0297,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2864,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.691024350895024,
      "rewrite6_current_pass_fraction": 0.5222293626219547,
      "rewrite6_full_passes": 0.0,
      "step": 90840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.592896,
      "curriculum_late_phase": 1.0,
      "epoch": 4.425e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.170387268066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0564,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2867866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25091599312008,
      "rewrite6_current_pass_fraction": 0.5228135771455278,
      "rewrite6_full_passes": 0.0,
      "step": 90850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5953536,
      "curriculum_late_phase": 1.0,
      "epoch": 4.43e-06,
      "grad_norm": 0.54296875,
      "host_rss_gib": 14.171871185302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0306,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28698666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.709654655190217,
      "rewrite6_current_pass_fraction": 0.5234805553932738,
      "rewrite6_full_passes": 0.0,
      "step": 90860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.5978112,
      "curriculum_late_phase": 1.0,
      "epoch": 4.435e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0179,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28730666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.44830512334259,
      "rewrite6_current_pass_fraction": 0.5241670074584721,
      "rewrite6_full_passes": 0.0,
      "step": 90870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6002688,
      "curriculum_late_phase": 1.0,
      "epoch": 4.44e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28770666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.524741485073319,
      "rewrite6_full_passes": 0.0,
      "step": 90880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6027264,
      "curriculum_late_phase": 1.0,
      "epoch": 4.445e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28796,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.248791007771803,
      "rewrite6_current_pass_fraction": 0.5252478043270823,
      "rewrite6_full_passes": 0.0,
      "step": 90890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.605184,
      "curriculum_late_phase": 1.0,
      "epoch": 4.45e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.28828,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.5259001772117388,
      "rewrite6_full_passes": 0.0,
      "step": 90900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6076416,
      "curriculum_late_phase": 1.0,
      "epoch": 4.455e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2884933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.5264454441004071,
      "rewrite6_full_passes": 0.0,
      "step": 90910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6100992,
      "curriculum_late_phase": 1.0,
      "epoch": 4.46e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0795,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2887866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.747525914346035,
      "rewrite6_current_pass_fraction": 0.5270734747132481,
      "rewrite6_full_passes": 0.0,
      "step": 90920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6125568,
      "curriculum_late_phase": 1.0,
      "epoch": 4.465e-06,
      "grad_norm": 0.58203125,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0685,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2890933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50961404351764,
      "rewrite6_current_pass_fraction": 0.527696636871726,
      "rewrite6_full_passes": 0.0,
      "step": 90930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6150144,
      "curriculum_late_phase": 1.0,
      "epoch": 4.47e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0603,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2893733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.333956389011558,
      "rewrite6_current_pass_fraction": 0.5282857198496621,
      "rewrite6_full_passes": 0.0,
      "step": 90940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.617472,
      "curriculum_late_phase": 1.0,
      "epoch": 4.475e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2896666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.5288650659188721,
      "rewrite6_full_passes": 0.0,
      "step": 90950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6199296,
      "curriculum_late_phase": 1.0,
      "epoch": 4.48e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0699,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29013333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.53974859244084,
      "rewrite6_current_pass_fraction": 0.5294541488968082,
      "rewrite6_full_passes": 0.0,
      "step": 90960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6223872,
      "curriculum_late_phase": 1.0,
      "epoch": 4.485e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2905066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.155501714289947,
      "rewrite6_current_pass_fraction": 0.5300967848727386,
      "rewrite6_full_passes": 0.0,
      "step": 90970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6248448,
      "curriculum_late_phase": 1.0,
      "epoch": 4.49e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29092,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.997428965207153,
      "rewrite6_current_pass_fraction": 0.5307442893030321,
      "rewrite6_full_passes": 0.0,
      "step": 90980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6273024,
      "curriculum_late_phase": 1.0,
      "epoch": 4.495e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29124,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.265796842004743,
      "rewrite6_current_pass_fraction": 0.5312213978306167,
      "rewrite6_full_passes": 0.0,
      "step": 90990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.62976,
      "curriculum_late_phase": 1.0,
      "epoch": 4.5e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.049,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29142666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.094239632271936,
      "rewrite6_current_pass_fraction": 0.5317910069911005,
      "rewrite6_full_passes": 0.0,
      "step": 91000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 4.5e-06,
      "eval_loss": 3.0826895236968994,
      "eval_perplexity": 21.817000900758764,
      "eval_runtime": 0.0508,
      "eval_samples_per_second": 629.911,
      "eval_steps_per_second": 19.685,
      "step": 91000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6322176,
      "curriculum_late_phase": 1.0,
      "epoch": 4.505e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.240482330322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29176,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.159991065026112,
      "rewrite6_current_pass_fraction": 0.5324239060583046,
      "rewrite6_full_passes": 0.0,
      "step": 91010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6346752,
      "curriculum_late_phase": 1.0,
      "epoch": 4.51e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.240482330322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29213333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.225430188469428,
      "rewrite6_current_pass_fraction": 0.5330275943993301,
      "rewrite6_full_passes": 0.0,
      "step": 91020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6371328,
      "curriculum_late_phase": 1.0,
      "epoch": 4.515e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.240482330322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0631,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29245333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.393775174118343,
      "rewrite6_current_pass_fraction": 0.5335533874705458,
      "rewrite6_full_passes": 0.0,
      "step": 91030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6395904,
      "curriculum_late_phase": 1.0,
      "epoch": 4.52e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.240482330322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29276,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.5340743120873985,
      "rewrite6_full_passes": 0.0,
      "step": 91040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.642048,
      "curriculum_late_phase": 1.0,
      "epoch": 4.525e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 14.240482330322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0222,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2931066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.536422151209266,
      "rewrite6_current_pass_fraction": 0.5346731319740609,
      "rewrite6_full_passes": 0.0,
      "step": 91050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6445056,
      "curriculum_late_phase": 1.0,
      "epoch": 4.53e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.240482330322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29341333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.5352865572238126,
      "rewrite6_full_passes": 0.0,
      "step": 91060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6469632,
      "curriculum_late_phase": 1.0,
      "epoch": 4.535e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.240482330322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0795,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2936533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.747525914346035,
      "rewrite6_current_pass_fraction": 0.5358220872037546,
      "rewrite6_full_passes": 0.0,
      "step": 91070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6494208,
      "curriculum_late_phase": 1.0,
      "epoch": 4.54e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.240482330322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2939866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.536381959455512,
      "rewrite6_full_passes": 0.0,
      "step": 91080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6518784,
      "curriculum_late_phase": 1.0,
      "epoch": 4.545e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.240531921386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2942266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08158688469643,
      "rewrite6_current_pass_fraction": 0.5369953847052638,
      "rewrite6_full_passes": 0.0,
      "step": 91090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.654336,
      "curriculum_late_phase": 1.0,
      "epoch": 4.55e-06,
      "grad_norm": 0.5234375,
      "host_rss_gib": 14.240531921386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0491,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29462666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09634916170988,
      "rewrite6_current_pass_fraction": 0.5374919670503009,
      "rewrite6_full_passes": 0.0,
      "step": 91100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6567936,
      "curriculum_late_phase": 1.0,
      "epoch": 4.555e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.240531921386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2949333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.117456062563107,
      "rewrite6_current_pass_fraction": 0.5381638137524098,
      "rewrite6_full_passes": 0.0,
      "step": 91110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6592512,
      "curriculum_late_phase": 1.0,
      "epoch": 4.56e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.240531921386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0734,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29528,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.615269797528736,
      "rewrite6_current_pass_fraction": 0.5387869759108879,
      "rewrite6_full_passes": 0.0,
      "step": 91120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6617088,
      "curriculum_late_phase": 1.0,
      "epoch": 4.565e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.240531921386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2956,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.024743373380456,
      "rewrite6_current_pass_fraction": 0.5394004011606395,
      "rewrite6_full_passes": 0.0,
      "step": 91130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6641664,
      "curriculum_late_phase": 1.0,
      "epoch": 4.57e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.240531921386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.296,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.5400089579560281,
      "rewrite6_full_passes": 0.0,
      "step": 91140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.666624,
      "curriculum_late_phase": 1.0,
      "epoch": 4.575e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2964133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.871821590362348,
      "rewrite6_current_pass_fraction": 0.5406369885688691,
      "rewrite6_full_passes": 0.0,
      "step": 91150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6690816,
      "curriculum_late_phase": 1.0,
      "epoch": 4.58e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0612,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35316559260667,
      "rewrite6_current_pass_fraction": 0.5413137037253413,
      "rewrite6_full_passes": 0.0,
      "step": 91160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6715392,
      "curriculum_late_phase": 1.0,
      "epoch": 4.585e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29701333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.5418005491616522,
      "rewrite6_full_passes": 0.0,
      "step": 91170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6739968,
      "curriculum_late_phase": 1.0,
      "epoch": 4.59e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29732,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.031051742600535,
      "rewrite6_current_pass_fraction": 0.542462658955035,
      "rewrite6_full_passes": 0.0,
      "step": 91180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6764544,
      "curriculum_late_phase": 1.0,
      "epoch": 4.595e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0234,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29752,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.56108064993093,
      "rewrite6_current_pass_fraction": 0.5430420050242449,
      "rewrite6_full_passes": 0.0,
      "step": 91190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.678912,
      "curriculum_late_phase": 1.0,
      "epoch": 4.6e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29784,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.5436310880021811,
      "rewrite6_full_passes": 0.0,
      "step": 91200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6813696,
      "curriculum_late_phase": 1.0,
      "epoch": 4.605e-06,
      "grad_norm": 0.5390625,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0448,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29824,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.005829616811546,
      "rewrite6_current_pass_fraction": 0.544341882339195,
      "rewrite6_full_passes": 0.0,
      "step": 91210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6838272,
      "curriculum_late_phase": 1.0,
      "epoch": 4.61e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2985333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13435678680148,
      "rewrite6_current_pass_fraction": 0.5448676754104107,
      "rewrite6_full_passes": 0.0,
      "step": 91220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6862848,
      "curriculum_late_phase": 1.0,
      "epoch": 4.615e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29890666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.016335157786333,
      "rewrite6_current_pass_fraction": 0.5454859691145255,
      "rewrite6_full_passes": 0.0,
      "step": 91230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6887424,
      "curriculum_late_phase": 1.0,
      "epoch": 4.62e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0654,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2991066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.443037486962115,
      "rewrite6_current_pass_fraction": 0.5461286050904559,
      "rewrite6_full_passes": 0.0,
      "step": 91240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6912,
      "curriculum_late_phase": 1.0,
      "epoch": 4.625e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0564,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2992533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25091599312008,
      "rewrite6_current_pass_fraction": 0.5467322934314813,
      "rewrite6_full_passes": 0.0,
      "step": 91250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6936576,
      "curriculum_late_phase": 1.0,
      "epoch": 4.63e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29952,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.5473797978617748,
      "rewrite6_full_passes": 0.0,
      "step": 91260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6961152,
      "curriculum_late_phase": 1.0,
      "epoch": 4.635e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.29996,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.5480175653833421,
      "rewrite6_full_passes": 0.0,
      "step": 91270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.6985728,
      "curriculum_late_phase": 1.0,
      "epoch": 4.64e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30036,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.5486212537243675,
      "rewrite6_full_passes": 0.0,
      "step": 91280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7010304,
      "curriculum_late_phase": 1.0,
      "epoch": 4.645e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3006933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.5490934937975891,
      "rewrite6_full_passes": 0.0,
      "step": 91290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.703488,
      "curriculum_late_phase": 1.0,
      "epoch": 4.65e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0656,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3009866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.44732652334885,
      "rewrite6_current_pass_fraction": 0.549667971412436,
      "rewrite6_full_passes": 0.0,
      "step": 91300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7059456,
      "curriculum_late_phase": 1.0,
      "epoch": 4.655e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0709,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30118666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.561299114498432,
      "rewrite6_current_pass_fraction": 0.5502424490272828,
      "rewrite6_full_passes": 0.0,
      "step": 91310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7084032,
      "curriculum_late_phase": 1.0,
      "epoch": 4.66e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30146666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.035258373598133,
      "rewrite6_current_pass_fraction": 0.5507633736441354,
      "rewrite6_full_passes": 0.0,
      "step": 91320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7108608,
      "curriculum_late_phase": 1.0,
      "epoch": 4.665e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30178666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.921974121142934,
      "rewrite6_current_pass_fraction": 0.5513962727113396,
      "rewrite6_full_passes": 0.0,
      "step": 91330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7133184,
      "curriculum_late_phase": 1.0,
      "epoch": 4.67e-06,
      "grad_norm": 0.5390625,
      "host_rss_gib": 14.240699768066406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0541,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30204,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20209505194016,
      "rewrite6_current_pass_fraction": 0.5520535140503593,
      "rewrite6_full_passes": 0.0,
      "step": 91340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.715776,
      "curriculum_late_phase": 1.0,
      "epoch": 4.675e-06,
      "grad_norm": 0.671875,
      "host_rss_gib": 13.699199676513672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30236,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.210577586354717,
      "rewrite6_current_pass_fraction": 0.5526328601195692,
      "rewrite6_full_passes": 0.0,
      "step": 91350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7182336,
      "curriculum_late_phase": 1.0,
      "epoch": 4.68e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.726829528808594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30276,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.817625339883573,
      "rewrite6_current_pass_fraction": 0.5533923390002142,
      "rewrite6_full_passes": 0.0,
      "step": 91360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7206912,
      "curriculum_late_phase": 1.0,
      "epoch": 4.685e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.041770935058594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30324,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.738668476643575,
      "rewrite6_current_pass_fraction": 0.5539911588868766,
      "rewrite6_full_passes": 0.0,
      "step": 91370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7231488,
      "curriculum_late_phase": 1.0,
      "epoch": 4.69e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.041770935058594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30349333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.5545753734104496,
      "rewrite6_full_passes": 0.0,
      "step": 91380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7256064,
      "curriculum_late_phase": 1.0,
      "epoch": 4.695e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0675,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30384,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.488115180697108,
      "rewrite6_current_pass_fraction": 0.5552910362018266,
      "rewrite6_full_passes": 0.0,
      "step": 91390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.728064,
      "curriculum_late_phase": 1.0,
      "epoch": 4.7e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0602,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30417333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.33182310003889,
      "rewrite6_current_pass_fraction": 0.5559093299059414,
      "rewrite6_full_passes": 0.0,
      "step": 91400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7305216,
      "curriculum_late_phase": 1.0,
      "epoch": 4.705e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0331,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3044266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.761493563464107,
      "rewrite6_current_pass_fraction": 0.556561702790598,
      "rewrite6_full_passes": 0.0,
      "step": 91410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7329792,
      "curriculum_late_phase": 1.0,
      "epoch": 4.71e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.067,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3047066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.477373808673548,
      "rewrite6_current_pass_fraction": 0.5572092072208915,
      "rewrite6_full_passes": 0.0,
      "step": 91420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7354368,
      "curriculum_late_phase": 1.0,
      "epoch": 4.715e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3051066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.031051742600535,
      "rewrite6_current_pass_fraction": 0.5577885532901015,
      "rewrite6_full_passes": 0.0,
      "step": 91430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7378944,
      "curriculum_late_phase": 1.0,
      "epoch": 4.72e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0832,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3054,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.828140805810193,
      "rewrite6_current_pass_fraction": 0.5583825047224007,
      "rewrite6_full_passes": 0.0,
      "step": 91440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.740352,
      "curriculum_late_phase": 1.0,
      "epoch": 4.725e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0383,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30577333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.869734512558935,
      "rewrite6_current_pass_fraction": 0.5590446145157836,
      "rewrite6_full_passes": 0.0,
      "step": 91450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7428096,
      "curriculum_late_phase": 1.0,
      "epoch": 4.73e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.1022,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30605333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.246840532718544,
      "rewrite6_current_pass_fraction": 0.5596093552219041,
      "rewrite6_full_passes": 0.0,
      "step": 91460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7452672,
      "curriculum_late_phase": 1.0,
      "epoch": 4.735e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.068,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30638666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.498861924749583,
      "rewrite6_current_pass_fraction": 0.5601887012911141,
      "rewrite6_full_passes": 0.0,
      "step": 91470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7477248,
      "curriculum_late_phase": 1.0,
      "epoch": 4.74e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0374,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3066933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.850960201205005,
      "rewrite6_current_pass_fraction": 0.560719362816693,
      "rewrite6_full_passes": 0.0,
      "step": 91480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7501824,
      "curriculum_late_phase": 1.0,
      "epoch": 4.745e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0374,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3071333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.850960201205005,
      "rewrite6_current_pass_fraction": 0.5613619987926233,
      "rewrite6_full_passes": 0.0,
      "step": 91490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.75264,
      "curriculum_late_phase": 1.0,
      "epoch": 4.75e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.04195785522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0332,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30744,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.763569816631378,
      "rewrite6_current_pass_fraction": 0.5620241085860062,
      "rewrite6_full_passes": 0.0,
      "step": 91500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 4.75e-06,
      "eval_loss": 3.082914352416992,
      "eval_perplexity": 21.821906540591243,
      "eval_runtime": 0.051,
      "eval_samples_per_second": 627.115,
      "eval_steps_per_second": 19.597,
      "step": 91500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7550976,
      "curriculum_late_phase": 1.0,
      "epoch": 4.755e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.054988861083984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3077066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11534442254061,
      "rewrite6_current_pass_fraction": 0.5625742439290374,
      "rewrite6_full_passes": 0.0,
      "step": 91510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7575552,
      "curriculum_late_phase": 1.0,
      "epoch": 4.76e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.054988861083984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.308,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.99532932729427,
      "rewrite6_current_pass_fraction": 0.5631535899982474,
      "rewrite6_full_passes": 0.0,
      "step": 91520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7600128,
      "curriculum_late_phase": 1.0,
      "epoch": 4.765e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.41733169555664,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0678,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30834666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.494562582313208,
      "rewrite6_current_pass_fraction": 0.5637572783392728,
      "rewrite6_full_passes": 0.0,
      "step": 91530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7624704,
      "curriculum_late_phase": 1.0,
      "epoch": 4.77e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.417560577392578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30864,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.24241732656946,
      "rewrite6_current_pass_fraction": 0.5642538606843099,
      "rewrite6_full_passes": 0.0,
      "step": 91540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.764928,
      "curriculum_late_phase": 1.0,
      "epoch": 4.775e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.420425415039062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0572,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30905333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.267923528021466,
      "rewrite6_current_pass_fraction": 0.5648867597515141,
      "rewrite6_full_passes": 0.0,
      "step": 91550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7673856,
      "curriculum_late_phase": 1.0,
      "epoch": 4.78e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.420425415039062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30945333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.859302253584733,
      "rewrite6_current_pass_fraction": 0.5654368950945454,
      "rewrite6_full_passes": 0.0,
      "step": 91560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7698432,
      "curriculum_late_phase": 1.0,
      "epoch": 4.785e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.420425415039062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0657,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.30974666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.449471363241397,
      "rewrite6_current_pass_fraction": 0.5660308465268447,
      "rewrite6_full_passes": 0.0,
      "step": 91570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7723008,
      "curriculum_late_phase": 1.0,
      "epoch": 4.79e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 14.420429229736328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0647,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3099866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.42803261303982,
      "rewrite6_current_pass_fraction": 0.5666491402309595,
      "rewrite6_full_passes": 0.0,
      "step": 91580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7747584,
      "curriculum_late_phase": 1.0,
      "epoch": 4.795e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.4208984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31038666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.391635903566247,
      "rewrite6_current_pass_fraction": 0.5672284863001694,
      "rewrite6_full_passes": 0.0,
      "step": 91590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.777216,
      "curriculum_late_phase": 1.0,
      "epoch": 4.8e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 14.42116928100586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31076,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.216941714202072,
      "rewrite6_current_pass_fraction": 0.5678516484586473,
      "rewrite6_full_passes": 0.0,
      "step": 91600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7796736,
      "curriculum_late_phase": 1.0,
      "epoch": 4.805e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 14.421546936035156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0621,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31109333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.372392092267074,
      "rewrite6_current_pass_fraction": 0.5684796790714883,
      "rewrite6_full_passes": 0.0,
      "step": 91610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7821312,
      "curriculum_late_phase": 1.0,
      "epoch": 4.81e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.421558380126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0567,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31148,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.257292224304877,
      "rewrite6_current_pass_fraction": 0.569000603688341,
      "rewrite6_full_passes": 0.0,
      "step": 91620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7845888,
      "curriculum_late_phase": 1.0,
      "epoch": 4.815e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.421558380126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31176,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.035258373598133,
      "rewrite6_current_pass_fraction": 0.5696042920293665,
      "rewrite6_full_passes": 0.0,
      "step": 91630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7870464,
      "curriculum_late_phase": 1.0,
      "epoch": 4.82e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.421558380126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3120266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.997428965207153,
      "rewrite6_current_pass_fraction": 0.5701885065529396,
      "rewrite6_full_passes": 0.0,
      "step": 91640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.789504,
      "curriculum_late_phase": 1.0,
      "epoch": 4.825e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.421897888183594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0422,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31244,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.951285398018836,
      "rewrite6_current_pass_fraction": 0.5708165371657806,
      "rewrite6_full_passes": 0.0,
      "step": 91650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7919616,
      "curriculum_late_phase": 1.0,
      "epoch": 4.83e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.421897888183594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0586,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3127466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.297719473255697,
      "rewrite6_current_pass_fraction": 0.5714835154135265,
      "rewrite6_full_passes": 0.0,
      "step": 91660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7944192,
      "curriculum_late_phase": 1.0,
      "epoch": 4.835e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.42190170288086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31308,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.819707206509154,
      "rewrite6_current_pass_fraction": 0.5721699674787248,
      "rewrite6_full_passes": 0.0,
      "step": 91670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7968768,
      "curriculum_late_phase": 1.0,
      "epoch": 4.84e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.42190170288086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0306,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31344,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.709654655190217,
      "rewrite6_current_pass_fraction": 0.5727687873653873,
      "rewrite6_full_passes": 0.0,
      "step": 91680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.7993344,
      "curriculum_late_phase": 1.0,
      "epoch": 4.845e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.423213958740234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0422,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31389333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.951285398018836,
      "rewrite6_current_pass_fraction": 0.5734065548869545,
      "rewrite6_full_passes": 0.0,
      "step": 91690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.801792,
      "curriculum_late_phase": 1.0,
      "epoch": 4.85e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.423721313476562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0755,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3141333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.660709559154142,
      "rewrite6_current_pass_fraction": 0.5739615586843488,
      "rewrite6_full_passes": 0.0,
      "step": 91700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8042496,
      "curriculum_late_phase": 1.0,
      "epoch": 4.855e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.423721313476562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31444,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01213231105346,
      "rewrite6_current_pass_fraction": 0.5745019571186539,
      "rewrite6_full_passes": 0.0,
      "step": 91710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8067072,
      "curriculum_late_phase": 1.0,
      "epoch": 4.86e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.423721313476562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31482666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.5751202508227687,
      "rewrite6_full_passes": 0.0,
      "step": 91720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8091648,
      "curriculum_late_phase": 1.0,
      "epoch": 4.865e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.423721313476562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31504,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.5757044653463419,
      "rewrite6_full_passes": 0.0,
      "step": 91730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8116224,
      "curriculum_late_phase": 1.0,
      "epoch": 4.87e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.423721313476562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0487,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31537333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08791230952813,
      "rewrite6_current_pass_fraction": 0.5762497322350101,
      "rewrite6_full_passes": 0.0,
      "step": 91740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.81408,
      "curriculum_late_phase": 1.0,
      "epoch": 4.875e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.423721313476562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0674,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31562666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.48596647661604,
      "rewrite6_current_pass_fraction": 0.5768680259391249,
      "rewrite6_full_passes": 0.0,
      "step": 91750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8165376,
      "curriculum_late_phase": 1.0,
      "epoch": 4.88e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.423721313476562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0526,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31596,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.170315749797474,
      "rewrite6_current_pass_fraction": 0.5774911880976028,
      "rewrite6_full_passes": 0.0,
      "step": 91760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8189952,
      "curriculum_late_phase": 1.0,
      "epoch": 4.885e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.423721313476562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0401,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3164133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.90733386394596,
      "rewrite6_current_pass_fraction": 0.578085139529902,
      "rewrite6_full_passes": 0.0,
      "step": 91770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8214528,
      "curriculum_late_phase": 1.0,
      "epoch": 4.89e-06,
      "grad_norm": 0.5234375,
      "host_rss_gib": 14.423721313476562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3166,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.5786352748729333,
      "rewrite6_full_passes": 0.0,
      "step": 91780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8239104,
      "curriculum_late_phase": 1.0,
      "epoch": 4.895e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.423721313476562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3170133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.07737098892312,
      "rewrite6_current_pass_fraction": 0.5791951471246909,
      "rewrite6_full_passes": 0.0,
      "step": 91790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.826368,
      "curriculum_late_phase": 1.0,
      "epoch": 4.9e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31737333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.84679042615617,
      "rewrite6_current_pass_fraction": 0.5797890985569901,
      "rewrite6_full_passes": 0.0,
      "step": 91800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8288256,
      "curriculum_late_phase": 1.0,
      "epoch": 4.905e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3177066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.062621991982052,
      "rewrite6_current_pass_fraction": 0.5804219976241942,
      "rewrite6_full_passes": 0.0,
      "step": 91810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8312832,
      "curriculum_late_phase": 1.0,
      "epoch": 4.91e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0412,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3180266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.93034458477251,
      "rewrite6_current_pass_fraction": 0.5810938443263033,
      "rewrite6_full_passes": 0.0,
      "step": 91820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8337408,
      "curriculum_late_phase": 1.0,
      "epoch": 4.915e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0538,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3184266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.19573537742345,
      "rewrite6_current_pass_fraction": 0.581707269576055,
      "rewrite6_full_passes": 0.0,
      "step": 91830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8361984,
      "curriculum_late_phase": 1.0,
      "epoch": 4.92e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0491,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.31868,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09634916170988,
      "rewrite6_current_pass_fraction": 0.5823499055519854,
      "rewrite6_full_passes": 0.0,
      "step": 91840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.838656,
      "curriculum_late_phase": 1.0,
      "epoch": 4.925e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0565,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.319,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25304119097752,
      "rewrite6_current_pass_fraction": 0.5829000408950167,
      "rewrite6_full_passes": 0.0,
      "step": 91850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8411136,
      "curriculum_late_phase": 1.0,
      "epoch": 4.93e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0499,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3194266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.113232993671563,
      "rewrite6_current_pass_fraction": 0.5835280715078577,
      "rewrite6_full_passes": 0.0,
      "step": 91860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8435712,
      "curriculum_late_phase": 1.0,
      "epoch": 4.935e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3198533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.738668476643575,
      "rewrite6_current_pass_fraction": 0.5840489961247103,
      "rewrite6_full_passes": 0.0,
      "step": 91870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8460288,
      "curriculum_late_phase": 1.0,
      "epoch": 4.94e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0314,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3201866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.726229007771437,
      "rewrite6_current_pass_fraction": 0.5846575529200989,
      "rewrite6_full_passes": 0.0,
      "step": 91880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8484864,
      "curriculum_late_phase": 1.0,
      "epoch": 4.945e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3204,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.144926607429838,
      "rewrite6_current_pass_fraction": 0.5852807150785768,
      "rewrite6_full_passes": 0.0,
      "step": 91890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.850944,
      "curriculum_late_phase": 1.0,
      "epoch": 4.95e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0405,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32082666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91569847030129,
      "rewrite6_current_pass_fraction": 0.5858357188759713,
      "rewrite6_full_passes": 0.0,
      "step": 91900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8534016,
      "curriculum_late_phase": 1.0,
      "epoch": 4.955e-06,
      "grad_norm": 0.5390625,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0486,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3211866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.085803623733216,
      "rewrite6_current_pass_fraction": 0.586449144125723,
      "rewrite6_full_passes": 0.0,
      "step": 91910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8558592,
      "curriculum_late_phase": 1.0,
      "epoch": 4.96e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0717,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32149333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.57855505524601,
      "rewrite6_current_pass_fraction": 0.5870479640123853,
      "rewrite6_full_passes": 0.0,
      "step": 91920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8583168,
      "curriculum_late_phase": 1.0,
      "epoch": 4.965e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.42422103881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0599,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3217333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.325424512944927,
      "rewrite6_current_pass_fraction": 0.5876711261708633,
      "rewrite6_full_passes": 0.0,
      "step": 91930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8607744,
      "curriculum_late_phase": 1.0,
      "epoch": 4.97e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.424259185791016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.058,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3220266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.284944674394637,
      "rewrite6_current_pass_fraction": 0.5882942883293412,
      "rewrite6_full_passes": 0.0,
      "step": 91940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.863232,
      "curriculum_late_phase": 1.0,
      "epoch": 4.975e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.424259185791016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32228,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.942906559739004,
      "rewrite6_current_pass_fraction": 0.5889466612139977,
      "rewrite6_full_passes": 0.0,
      "step": 91950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8656896,
      "curriculum_late_phase": 1.0,
      "epoch": 4.98e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.424259185791016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0502,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3225866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11956791376016,
      "rewrite6_current_pass_fraction": 0.5895114019201184,
      "rewrite6_full_passes": 0.0,
      "step": 91960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8681472,
      "curriculum_late_phase": 1.0,
      "epoch": 4.985e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 14.424259185791016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0228,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32292,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.5487477017954,
      "rewrite6_current_pass_fraction": 0.5901832486222274,
      "rewrite6_full_passes": 0.0,
      "step": 91970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8706048,
      "curriculum_late_phase": 1.0,
      "epoch": 4.99e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.424263000488281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0163,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3231866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.415613995022007,
      "rewrite6_current_pass_fraction": 0.5906749625129014,
      "rewrite6_full_passes": 0.0,
      "step": 91980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8730624,
      "curriculum_late_phase": 1.0,
      "epoch": 4.995e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.424270629882812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0634,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3236266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.400194269486743,
      "rewrite6_current_pass_fraction": 0.5912250978559327,
      "rewrite6_full_passes": 0.0,
      "step": 91990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.87552,
      "curriculum_late_phase": 1.0,
      "epoch": 5e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.424270629882812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0502,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32406666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11956791376016,
      "rewrite6_current_pass_fraction": 0.5918287861969582,
      "rewrite6_full_passes": 0.0,
      "step": 92000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 5e-06,
      "eval_loss": 3.0827534198760986,
      "eval_perplexity": 21.818394968295227,
      "eval_runtime": 0.0504,
      "eval_samples_per_second": 635.498,
      "eval_steps_per_second": 19.859,
      "step": 92000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8779776,
      "curriculum_late_phase": 1.0,
      "epoch": 5.005e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.662155151367188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0667,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3244,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.470931562916125,
      "rewrite6_current_pass_fraction": 0.5923156316332691,
      "rewrite6_full_passes": 0.0,
      "step": 92010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8804352,
      "curriculum_late_phase": 1.0,
      "epoch": 5.01e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.65999984741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0817,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3247466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.79542313898616,
      "rewrite6_current_pass_fraction": 0.5928706354306634,
      "rewrite6_full_passes": 0.0,
      "step": 92020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8828928,
      "curriculum_late_phase": 1.0,
      "epoch": 5.015e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 14.65999984741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0675,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3250133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.488115180697108,
      "rewrite6_current_pass_fraction": 0.5934451130455103,
      "rewrite6_full_passes": 0.0,
      "step": 92030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8853504,
      "curriculum_late_phase": 1.0,
      "epoch": 5.02e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.65999984741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32532,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.5939368269361843,
      "rewrite6_full_passes": 0.0,
      "step": 92040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.887808,
      "curriculum_late_phase": 1.0,
      "epoch": 5.025e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.65999984741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0777,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3256133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.70841557756311,
      "rewrite6_current_pass_fraction": 0.5944966991879418,
      "rewrite6_full_passes": 0.0,
      "step": 92050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8902656,
      "curriculum_late_phase": 1.0,
      "epoch": 5.03e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.65999984741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32594666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.594998149987342,
      "rewrite6_full_passes": 0.0,
      "step": 92060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8927232,
      "curriculum_late_phase": 1.0,
      "epoch": 5.035e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.65999984741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0624,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3262,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.378804771748577,
      "rewrite6_current_pass_fraction": 0.5955190746041946,
      "rewrite6_full_passes": 0.0,
      "step": 92070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8951808,
      "curriculum_late_phase": 1.0,
      "epoch": 5.04e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 14.65999984741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0783,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32649333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.72144453520607,
      "rewrite6_current_pass_fraction": 0.5960886837646784,
      "rewrite6_full_passes": 0.0,
      "step": 92080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.8976384,
      "curriculum_late_phase": 1.0,
      "epoch": 5.045e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.65999984741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32682666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.151271037028916,
      "rewrite6_current_pass_fraction": 0.5967264512862457,
      "rewrite6_full_passes": 0.0,
      "step": 92090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.900096,
      "curriculum_late_phase": 1.0,
      "epoch": 5.05e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.65999984741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32726666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.5972717181749139,
      "rewrite6_full_passes": 0.0,
      "step": 92100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9025536,
      "curriculum_late_phase": 1.0,
      "epoch": 5.055e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 14.65999984741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32752,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.782265441235463,
      "rewrite6_current_pass_fraction": 0.5977196159763198,
      "rewrite6_full_passes": 0.0,
      "step": 92110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9050112,
      "curriculum_late_phase": 1.0,
      "epoch": 5.06e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.66025161743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0828,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3278,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.819411295506324,
      "rewrite6_current_pass_fraction": 0.5982989620455298,
      "rewrite6_full_passes": 0.0,
      "step": 92120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9074688,
      "curriculum_late_phase": 1.0,
      "epoch": 5.065e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.66025161743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3281866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.007930304805882,
      "rewrite6_current_pass_fraction": 0.5988101497536562,
      "rewrite6_full_passes": 0.0,
      "step": 92130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9099264,
      "curriculum_late_phase": 1.0,
      "epoch": 5.07e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.66025161743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32850666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.87808407616535,
      "rewrite6_current_pass_fraction": 0.5994625226383128,
      "rewrite6_full_passes": 0.0,
      "step": 92140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.912384,
      "curriculum_late_phase": 1.0,
      "epoch": 5.075e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.66025161743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32886666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.882260110570105,
      "rewrite6_current_pass_fraction": 0.5999542365289868,
      "rewrite6_full_passes": 0.0,
      "step": 92150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9148416,
      "curriculum_late_phase": 1.0,
      "epoch": 5.08e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.66025161743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0379,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3291866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.861388288110085,
      "rewrite6_current_pass_fraction": 0.6004848980545656,
      "rewrite6_full_passes": 0.0,
      "step": 92160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9172992,
      "curriculum_late_phase": 1.0,
      "epoch": 5.085e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.66025161743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0631,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32952,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.393775174118343,
      "rewrite6_current_pass_fraction": 0.6010642441237756,
      "rewrite6_full_passes": 0.0,
      "step": 92170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9197568,
      "curriculum_late_phase": 1.0,
      "epoch": 5.09e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.66025161743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0375,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.32988,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.853045401483406,
      "rewrite6_current_pass_fraction": 0.6016046425580807,
      "rewrite6_full_passes": 0.0,
      "step": 92180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9222144,
      "curriculum_late_phase": 1.0,
      "epoch": 5.095e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.66025161743164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0513,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33013333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.1428122204902,
      "rewrite6_current_pass_fraction": 0.6021742517185644,
      "rewrite6_full_passes": 0.0,
      "step": 92190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.924672,
      "curriculum_late_phase": 1.0,
      "epoch": 5.1e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.764678955078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0619,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33041333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.36811804126797,
      "rewrite6_current_pass_fraction": 0.6027487293334113,
      "rewrite6_full_passes": 0.0,
      "step": 92200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9271296,
      "curriculum_late_phase": 1.0,
      "epoch": 5.105e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33082666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.945000955113006,
      "rewrite6_current_pass_fraction": 0.6033280754026211,
      "rewrite6_full_passes": 0.0,
      "step": 92210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9295872,
      "curriculum_late_phase": 1.0,
      "epoch": 5.11e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33112,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.945000955113006,
      "rewrite6_current_pass_fraction": 0.6038684738369262,
      "rewrite6_full_passes": 0.0,
      "step": 92220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9320448,
      "curriculum_late_phase": 1.0,
      "epoch": 5.115e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33157333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.6043991353625051,
      "rewrite6_full_passes": 0.0,
      "step": 92230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9345024,
      "curriculum_late_phase": 1.0,
      "epoch": 5.12e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33192,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.604929796888084,
      "rewrite6_full_passes": 0.0,
      "step": 92240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.93696,
      "curriculum_late_phase": 1.0,
      "epoch": 5.125e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0644,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33210666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.421605167420953,
      "rewrite6_current_pass_fraction": 0.6054507215049366,
      "rewrite6_full_passes": 0.0,
      "step": 92250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9394176,
      "curriculum_late_phase": 1.0,
      "epoch": 5.13e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33230666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.6059813830305155,
      "rewrite6_full_passes": 0.0,
      "step": 92260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9418752,
      "curriculum_late_phase": 1.0,
      "epoch": 5.135e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3327466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.6066580981869876,
      "rewrite6_full_passes": 0.0,
      "step": 92270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9443328,
      "curriculum_late_phase": 1.0,
      "epoch": 5.14e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0633,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33304,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.398054357057195,
      "rewrite6_current_pass_fraction": 0.6071498120776616,
      "rewrite6_full_passes": 0.0,
      "step": 92280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9467904,
      "curriculum_late_phase": 1.0,
      "epoch": 5.145e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33341333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.6077778426905026,
      "rewrite6_full_passes": 0.0,
      "step": 92290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.949248,
      "curriculum_late_phase": 1.0,
      "epoch": 5.15e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0513,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33369333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.1428122204902,
      "rewrite6_current_pass_fraction": 0.6083815310315281,
      "rewrite6_full_passes": 0.0,
      "step": 92300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9517056,
      "curriculum_late_phase": 1.0,
      "epoch": 5.155e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.019,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33393333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.470810634740232,
      "rewrite6_current_pass_fraction": 0.6088732449222021,
      "rewrite6_full_passes": 0.0,
      "step": 92310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9541632,
      "curriculum_late_phase": 1.0,
      "epoch": 5.16e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0365,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3342133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.832202779129986,
      "rewrite6_current_pass_fraction": 0.6093357480866974,
      "rewrite6_full_passes": 0.0,
      "step": 92320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9566208,
      "curriculum_late_phase": 1.0,
      "epoch": 5.165e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0621,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33454666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.372392092267074,
      "rewrite6_current_pass_fraction": 0.6099004887928181,
      "rewrite6_full_passes": 0.0,
      "step": 92330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9590784,
      "curriculum_late_phase": 1.0,
      "epoch": 5.17e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33482666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.60662541860077,
      "rewrite6_current_pass_fraction": 0.6105431247687484,
      "rewrite6_full_passes": 0.0,
      "step": 92340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.961536,
      "curriculum_late_phase": 1.0,
      "epoch": 5.175e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3352,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.6110299702050593,
      "rewrite6_full_passes": 0.0,
      "step": 92350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9639936,
      "curriculum_late_phase": 1.0,
      "epoch": 5.18e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33554666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08158688469643,
      "rewrite6_current_pass_fraction": 0.6115947109111799,
      "rewrite6_full_passes": 0.0,
      "step": 92360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9664512,
      "curriculum_late_phase": 1.0,
      "epoch": 5.185e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0367,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3358,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.836369636357645,
      "rewrite6_current_pass_fraction": 0.6121058986193063,
      "rewrite6_full_passes": 0.0,
      "step": 92370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9689088,
      "curriculum_late_phase": 1.0,
      "epoch": 5.19e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.602481842041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0437,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33604,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.982735908101457,
      "rewrite6_current_pass_fraction": 0.6127047185059687,
      "rewrite6_full_passes": 0.0,
      "step": 92380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9713664,
      "curriculum_late_phase": 1.0,
      "epoch": 5.195e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.9124755859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0313,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3363466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.724156488498345,
      "rewrite6_current_pass_fraction": 0.6133960390255302,
      "rewrite6_full_passes": 0.0,
      "step": 92390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.973824,
      "curriculum_late_phase": 1.0,
      "epoch": 5.2e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 14.920425415039062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0804,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33654666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.76710749805986,
      "rewrite6_current_pass_fraction": 0.613965648186014,
      "rewrite6_full_passes": 0.0,
      "step": 92400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9762816,
      "curriculum_late_phase": 1.0,
      "epoch": 5.205e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.920963287353516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33670666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.6146034157075811,
      "rewrite6_full_passes": 0.0,
      "step": 92410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9787392,
      "curriculum_late_phase": 1.0,
      "epoch": 5.21e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.921073913574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.337,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.6152557885922377,
      "rewrite6_full_passes": 0.0,
      "step": 92420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9811968,
      "curriculum_late_phase": 1.0,
      "epoch": 5.215e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.921478271484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0751,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33737333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.65204700795622,
      "rewrite6_current_pass_fraction": 0.6158838192050787,
      "rewrite6_full_passes": 0.0,
      "step": 92430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9836544,
      "curriculum_late_phase": 1.0,
      "epoch": 5.22e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.921493530273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33764,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.6165313236353722,
      "rewrite6_full_passes": 0.0,
      "step": 92440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.986112,
      "curriculum_late_phase": 1.0,
      "epoch": 5.225e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.921493530273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0699,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33797333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.53974859244084,
      "rewrite6_current_pass_fraction": 0.6170327744347724,
      "rewrite6_full_passes": 0.0,
      "step": 92450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9885696,
      "curriculum_late_phase": 1.0,
      "epoch": 5.23e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.922348022460938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0357,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33816,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.81554368143424,
      "rewrite6_current_pass_fraction": 0.6176121205039824,
      "rewrite6_full_passes": 0.0,
      "step": 92460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9910272,
      "curriculum_late_phase": 1.0,
      "epoch": 5.235e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.922348022460938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33842666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.6182498880255497,
      "rewrite6_full_passes": 0.0,
      "step": 92470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9934848,
      "curriculum_late_phase": 1.0,
      "epoch": 5.24e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.922348022460938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0685,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33888,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50961404351764,
      "rewrite6_current_pass_fraction": 0.6188097602773072,
      "rewrite6_full_passes": 0.0,
      "step": 92480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9959424,
      "curriculum_late_phase": 1.0,
      "epoch": 5.245e-06,
      "grad_norm": 0.57421875,
      "host_rss_gib": 14.922592163085938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33909333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.6194475277988745,
      "rewrite6_full_passes": 0.0,
      "step": 92490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 16.9984,
      "curriculum_late_phase": 1.0,
      "epoch": 5.25e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 14.922592163085938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33941333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.942906559739004,
      "rewrite6_current_pass_fraction": 0.6200804268660786,
      "rewrite6_full_passes": 0.0,
      "step": 92500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 5.25e-06,
      "eval_loss": 3.082709312438965,
      "eval_perplexity": 21.817432636033963,
      "eval_runtime": 0.051,
      "eval_samples_per_second": 627.461,
      "eval_steps_per_second": 19.608,
      "step": 92500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0008576,
      "curriculum_late_phase": 1.0,
      "epoch": 5.255e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.996902465820312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0614,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.33972,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.357436652816972,
      "rewrite6_current_pass_fraction": 0.6206938521158303,
      "rewrite6_full_passes": 0.0,
      "step": 92510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0033152,
      "curriculum_late_phase": 1.0,
      "epoch": 5.26e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 14.996902465820312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.299849351695165,
      "rewrite6_current_pass_fraction": 0.6213316196373975,
      "rewrite6_full_passes": 0.0,
      "step": 92520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0057728,
      "curriculum_late_phase": 1.0,
      "epoch": 5.265e-06,
      "grad_norm": 6.59375,
      "host_rss_gib": 14.996902465820312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0342,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34026666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.784343771694374,
      "rewrite6_current_pass_fraction": 0.6218768865260657,
      "rewrite6_full_passes": 0.0,
      "step": 92530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0082304,
      "curriculum_late_phase": 1.0,
      "epoch": 5.27e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.996902465820312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.075,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34052,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.649881911512058,
      "rewrite6_current_pass_fraction": 0.622470837958365,
      "rewrite6_full_passes": 0.0,
      "step": 92540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.010688,
      "curriculum_late_phase": 1.0,
      "epoch": 5.275e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.997039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34085333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.6230161048470332,
      "rewrite6_full_passes": 0.0,
      "step": 92550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0131456,
      "curriculum_late_phase": 1.0,
      "epoch": 5.28e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 14.997039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0361,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34114666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.82387156437236,
      "rewrite6_current_pass_fraction": 0.6237415045471364,
      "rewrite6_full_passes": 0.0,
      "step": 92560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0156032,
      "curriculum_late_phase": 1.0,
      "epoch": 5.285e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.997039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34153333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.6244133512492454,
      "rewrite6_full_passes": 0.0,
      "step": 92570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0180608,
      "curriculum_late_phase": 1.0,
      "epoch": 5.29e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.997039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0337,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34185333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.77395419741855,
      "rewrite6_current_pass_fraction": 0.6250754610426282,
      "rewrite6_full_passes": 0.0,
      "step": 92580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0205184,
      "curriculum_late_phase": 1.0,
      "epoch": 5.295e-06,
      "grad_norm": 0.5390625,
      "host_rss_gib": 14.997039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34217333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.6257083601098323,
      "rewrite6_full_passes": 0.0,
      "step": 92590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.022976,
      "curriculum_late_phase": 1.0,
      "epoch": 5.3e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 14.997039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0699,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34236,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.53974859244084,
      "rewrite6_current_pass_fraction": 0.6262584954528636,
      "rewrite6_full_passes": 0.0,
      "step": 92600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0254336,
      "curriculum_late_phase": 1.0,
      "epoch": 5.305e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.997039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0546,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34256,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21269875016978,
      "rewrite6_current_pass_fraction": 0.6268767891569784,
      "rewrite6_full_passes": 0.0,
      "step": 92610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0278912,
      "curriculum_late_phase": 1.0,
      "epoch": 5.31e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.997039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34286666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.6274999513154563,
      "rewrite6_full_passes": 0.0,
      "step": 92620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0303488,
      "curriculum_late_phase": 1.0,
      "epoch": 5.315e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.997039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0502,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34310666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11956791376016,
      "rewrite6_current_pass_fraction": 0.6281425872913867,
      "rewrite6_full_passes": 0.0,
      "step": 92630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0328064,
      "curriculum_late_phase": 1.0,
      "epoch": 5.32e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 14.99746322631836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0271,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3434,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.637297562673467,
      "rewrite6_current_pass_fraction": 0.6287121964518705,
      "rewrite6_full_passes": 0.0,
      "step": 92640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.035264,
      "curriculum_late_phase": 1.0,
      "epoch": 5.325e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0594,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34376,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3147644659223,
      "rewrite6_current_pass_fraction": 0.6293743062452533,
      "rewrite6_full_passes": 0.0,
      "step": 92650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0377216,
      "curriculum_late_phase": 1.0,
      "epoch": 5.33e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34397333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.090021206202152,
      "rewrite6_current_pass_fraction": 0.6299341784970107,
      "rewrite6_full_passes": 0.0,
      "step": 92660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0401792,
      "curriculum_late_phase": 1.0,
      "epoch": 5.335e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34428,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.630571946018578,
      "rewrite6_full_passes": 0.0,
      "step": 92670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0426368,
      "curriculum_late_phase": 1.0,
      "epoch": 5.34e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0336,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34469333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.771876905865113,
      "rewrite6_current_pass_fraction": 0.6311464236334249,
      "rewrite6_full_passes": 0.0,
      "step": 92680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0450944,
      "curriculum_late_phase": 1.0,
      "epoch": 5.345e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3449466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.155501714289947,
      "rewrite6_current_pass_fraction": 0.6316819536133669,
      "rewrite6_full_passes": 0.0,
      "step": 92690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.047552,
      "curriculum_late_phase": 1.0,
      "epoch": 5.35e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3452133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.97853978054658,
      "rewrite6_current_pass_fraction": 0.6322856419543923,
      "rewrite6_full_passes": 0.0,
      "step": 92700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0500096,
      "curriculum_late_phase": 1.0,
      "epoch": 5.355e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0313,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34542666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.724156488498345,
      "rewrite6_current_pass_fraction": 0.6328552511148761,
      "rewrite6_full_passes": 0.0,
      "step": 92710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0524672,
      "curriculum_late_phase": 1.0,
      "epoch": 5.36e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34568,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08369514879635,
      "rewrite6_current_pass_fraction": 0.6335222293626219,
      "rewrite6_full_passes": 0.0,
      "step": 92720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0549248,
      "curriculum_late_phase": 1.0,
      "epoch": 5.365e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.061,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3459466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34889538652299,
      "rewrite6_current_pass_fraction": 0.6340821016143795,
      "rewrite6_full_passes": 0.0,
      "step": 92730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0573824,
      "curriculum_late_phase": 1.0,
      "epoch": 5.37e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0522,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34630666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.161849316897023,
      "rewrite6_current_pass_fraction": 0.6347247375903098,
      "rewrite6_full_passes": 0.0,
      "step": 92740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.05984,
      "curriculum_late_phase": 1.0,
      "epoch": 5.375e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.998245239257812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.025,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3466,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.59400471119603,
      "rewrite6_current_pass_fraction": 0.6353478997487878,
      "rewrite6_full_passes": 0.0,
      "step": 92750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0622976,
      "curriculum_late_phase": 1.0,
      "epoch": 5.38e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.998470306396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3469466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.193615909860856,
      "rewrite6_current_pass_fraction": 0.6360392202683492,
      "rewrite6_full_passes": 0.0,
      "step": 92760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0647552,
      "curriculum_late_phase": 1.0,
      "epoch": 5.385e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.998470306396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0351,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3472266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.803058101273997,
      "rewrite6_current_pass_fraction": 0.6367548830597262,
      "rewrite6_full_passes": 0.0,
      "step": 92770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0672128,
      "curriculum_late_phase": 1.0,
      "epoch": 5.39e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0618,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34754666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.365981336300866,
      "rewrite6_current_pass_fraction": 0.6373585714007517,
      "rewrite6_full_passes": 0.0,
      "step": 92780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0696704,
      "curriculum_late_phase": 1.0,
      "epoch": 5.395e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3478266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.153386269892504,
      "rewrite6_current_pass_fraction": 0.6379379174699616,
      "rewrite6_full_passes": 0.0,
      "step": 92790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.072128,
      "curriculum_late_phase": 1.0,
      "epoch": 5.4e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34813333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.968053132536788,
      "rewrite6_current_pass_fraction": 0.6385562111740765,
      "rewrite6_full_passes": 0.0,
      "step": 92800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0745856,
      "curriculum_late_phase": 1.0,
      "epoch": 5.405e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34852,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.74904040364757,
      "rewrite6_current_pass_fraction": 0.6391696364238282,
      "rewrite6_full_passes": 0.0,
      "step": 92810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0770432,
      "curriculum_late_phase": 1.0,
      "epoch": 5.41e-06,
      "grad_norm": 0.5390625,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34884,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.6397246402212226,
      "rewrite6_full_passes": 0.0,
      "step": 92820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0795008,
      "curriculum_late_phase": 1.0,
      "epoch": 5.415e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34925333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.299849351695165,
      "rewrite6_current_pass_fraction": 0.6404549083756889,
      "rewrite6_full_passes": 0.0,
      "step": 92830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0819584,
      "curriculum_late_phase": 1.0,
      "epoch": 5.42e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0549,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34958666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.219063514461734,
      "rewrite6_current_pass_fraction": 0.6410293859905357,
      "rewrite6_full_passes": 0.0,
      "step": 92840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.084416,
      "curriculum_late_phase": 1.0,
      "epoch": 5.425e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0498,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.34986666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.111121775934837,
      "rewrite6_current_pass_fraction": 0.6416379427859243,
      "rewrite6_full_passes": 0.0,
      "step": 92850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0868736,
      "curriculum_late_phase": 1.0,
      "epoch": 5.43e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0532,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3502,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18302175066643,
      "rewrite6_current_pass_fraction": 0.64216373585714,
      "rewrite6_full_passes": 0.0,
      "step": 92860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0893312,
      "curriculum_late_phase": 1.0,
      "epoch": 5.435e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0777,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35062666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.70841557756311,
      "rewrite6_current_pass_fraction": 0.6427625557438025,
      "rewrite6_full_passes": 0.0,
      "step": 92870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0917888,
      "curriculum_late_phase": 1.0,
      "epoch": 5.44e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35102666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.6433321649042862,
      "rewrite6_full_passes": 0.0,
      "step": 92880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0942464,
      "curriculum_late_phase": 1.0,
      "epoch": 5.445e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.059,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35132,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.306240265089755,
      "rewrite6_current_pass_fraction": 0.6439115109734961,
      "rewrite6_full_passes": 0.0,
      "step": 92890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.096704,
      "curriculum_late_phase": 1.0,
      "epoch": 5.45e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0523,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35170666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.16396560764148,
      "rewrite6_current_pass_fraction": 0.6444470409534381,
      "rewrite6_full_passes": 0.0,
      "step": 92900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.0991616,
      "curriculum_late_phase": 1.0,
      "epoch": 5.455e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0634,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35202666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.400194269486743,
      "rewrite6_current_pass_fraction": 0.6450507292944636,
      "rewrite6_full_passes": 0.0,
      "step": 92910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1016192,
      "curriculum_late_phase": 1.0,
      "epoch": 5.46e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0645,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3522133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.423747435049286,
      "rewrite6_current_pass_fraction": 0.6456884968160308,
      "rewrite6_full_passes": 0.0,
      "step": 92920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1040768,
      "curriculum_late_phase": 1.0,
      "epoch": 5.465e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0606,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3525333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.340357536052313,
      "rewrite6_current_pass_fraction": 0.6463895542443185,
      "rewrite6_full_passes": 0.0,
      "step": 92930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1065344,
      "curriculum_late_phase": 1.0,
      "epoch": 5.47e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.99898910522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0347,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3528133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.79473854205626,
      "rewrite6_current_pass_fraction": 0.6469689003135285,
      "rewrite6_full_passes": 0.0,
      "step": 92940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.108992,
      "curriculum_late_phase": 1.0,
      "epoch": 5.475e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.000133514404297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.075,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35306666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.649881911512058,
      "rewrite6_current_pass_fraction": 0.6475725886545539,
      "rewrite6_full_passes": 0.0,
      "step": 92950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1114496,
      "curriculum_late_phase": 1.0,
      "epoch": 5.48e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.000133514404297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35346666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.6481762769955794,
      "rewrite6_full_passes": 0.0,
      "step": 92960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1139072,
      "curriculum_late_phase": 1.0,
      "epoch": 5.485e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.000133514404297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3537866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.87808407616535,
      "rewrite6_current_pass_fraction": 0.6487945706996943,
      "rewrite6_full_passes": 0.0,
      "step": 92970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1163648,
      "curriculum_late_phase": 1.0,
      "epoch": 5.49e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.000133514404297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35412,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.649456680493077,
      "rewrite6_full_passes": 0.0,
      "step": 92980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1188224,
      "curriculum_late_phase": 1.0,
      "epoch": 5.495e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.000133514404297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35433333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.066834937660975,
      "rewrite6_current_pass_fraction": 0.6499483943837511,
      "rewrite6_full_passes": 0.0,
      "step": 92990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.12128,
      "curriculum_late_phase": 1.0,
      "epoch": 5.5e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.000133514404297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0638,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35468,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.40875605943837,
      "rewrite6_current_pass_fraction": 0.6505472142704134,
      "rewrite6_full_passes": 0.0,
      "step": 93000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 5.5e-06,
      "eval_loss": 3.0827836990356445,
      "eval_perplexity": 21.81905562095946,
      "eval_runtime": 0.0496,
      "eval_samples_per_second": 645.107,
      "eval_steps_per_second": 20.16,
      "step": 93000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1237376,
      "curriculum_late_phase": 1.0,
      "epoch": 5.505e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.002304077148438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0422,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35508,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.951285398018836,
      "rewrite6_current_pass_fraction": 0.6511022180678079,
      "rewrite6_full_passes": 0.0,
      "step": 93010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1261952,
      "curriculum_late_phase": 1.0,
      "epoch": 5.51e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.002304077148438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0538,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35544,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.19573537742345,
      "rewrite6_current_pass_fraction": 0.651735117135012,
      "rewrite6_full_passes": 0.0,
      "step": 93020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1286528,
      "curriculum_late_phase": 1.0,
      "epoch": 5.515e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.002304077148438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0682,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35570666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.503162127140435,
      "rewrite6_current_pass_fraction": 0.6523242001129481,
      "rewrite6_full_passes": 0.0,
      "step": 93030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1311104,
      "curriculum_late_phase": 1.0,
      "epoch": 5.52e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.002304077148438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35601333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.6528792039103426,
      "rewrite6_full_passes": 0.0,
      "step": 93040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.133568,
      "curriculum_late_phase": 1.0,
      "epoch": 5.525e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.547706604003906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3563866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.26367036864598,
      "rewrite6_current_pass_fraction": 0.6534098654359214,
      "rewrite6_full_passes": 0.0,
      "step": 93050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1360256,
      "curriculum_late_phase": 1.0,
      "epoch": 5.53e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.365806579589844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35662666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01213231105346,
      "rewrite6_current_pass_fraction": 0.6539210531440478,
      "rewrite6_full_passes": 0.0,
      "step": 93060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1384832,
      "curriculum_late_phase": 1.0,
      "epoch": 5.535e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0588,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35702666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.301979443133135,
      "rewrite6_current_pass_fraction": 0.6545636891199782,
      "rewrite6_full_passes": 0.0,
      "step": 93070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1409408,
      "curriculum_late_phase": 1.0,
      "epoch": 5.54e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0551,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35742666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.22330775157419,
      "rewrite6_current_pass_fraction": 0.6551479036435512,
      "rewrite6_full_passes": 0.0,
      "step": 93080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1433984,
      "curriculum_late_phase": 1.0,
      "epoch": 5.545e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35764,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.24241732656946,
      "rewrite6_current_pass_fraction": 0.6557564604389399,
      "rewrite6_full_passes": 0.0,
      "step": 93090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.145856,
      "curriculum_late_phase": 1.0,
      "epoch": 5.55e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35801333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.210577586354717,
      "rewrite6_current_pass_fraction": 0.656389359506144,
      "rewrite6_full_passes": 0.0,
      "step": 93100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1483136,
      "curriculum_late_phase": 1.0,
      "epoch": 5.555e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0616,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35830666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.361708567324747,
      "rewrite6_current_pass_fraction": 0.6569833109384433,
      "rewrite6_full_passes": 0.0,
      "step": 93110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1507712,
      "curriculum_late_phase": 1.0,
      "epoch": 5.56e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3585333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.02684595284501,
      "rewrite6_current_pass_fraction": 0.6576454207318261,
      "rewrite6_full_passes": 0.0,
      "step": 93120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1532288,
      "curriculum_late_phase": 1.0,
      "epoch": 5.565e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0883,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35878666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.939748682094375,
      "rewrite6_current_pass_fraction": 0.6583805573406555,
      "rewrite6_full_passes": 0.0,
      "step": 93130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1556864,
      "curriculum_late_phase": 1.0,
      "epoch": 5.57e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.35934666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11534442254061,
      "rewrite6_current_pass_fraction": 0.6590475355884013,
      "rewrite6_full_passes": 0.0,
      "step": 93140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.158144,
      "curriculum_late_phase": 1.0,
      "epoch": 5.575e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.049,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3596666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.094239632271936,
      "rewrite6_current_pass_fraction": 0.6595781971139802,
      "rewrite6_full_passes": 0.0,
      "step": 93150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1606016,
      "curriculum_late_phase": 1.0,
      "epoch": 5.58e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10689997377394,
      "rewrite6_current_pass_fraction": 0.6602208330899106,
      "rewrite6_full_passes": 0.0,
      "step": 93160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1630592,
      "curriculum_late_phase": 1.0,
      "epoch": 5.585e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0833,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36030666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.83032372903511,
      "rewrite6_current_pass_fraction": 0.6607125469805846,
      "rewrite6_full_passes": 0.0,
      "step": 93170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1655168,
      "curriculum_late_phase": 1.0,
      "epoch": 5.59e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.869804382324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0352,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3606666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.805138511102886,
      "rewrite6_current_pass_fraction": 0.661311366867247,
      "rewrite6_full_passes": 0.0,
      "step": 93180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1679744,
      "curriculum_late_phase": 1.0,
      "epoch": 5.595e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.869915008544922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.361,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35530101593532,
      "rewrite6_current_pass_fraction": 0.6619247921169987,
      "rewrite6_full_passes": 0.0,
      "step": 93190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.170432,
      "curriculum_late_phase": 1.0,
      "epoch": 5.6e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.869915008544922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36133333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20845663464543,
      "rewrite6_current_pass_fraction": 0.6625333489123872,
      "rewrite6_full_passes": 0.0,
      "step": 93200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1728896,
      "curriculum_late_phase": 1.0,
      "epoch": 5.605e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.869915008544922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0522,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3616,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.161849316897023,
      "rewrite6_current_pass_fraction": 0.6630445366205138,
      "rewrite6_full_passes": 0.0,
      "step": 93210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1753472,
      "curriculum_late_phase": 1.0,
      "epoch": 5.61e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.869915008544922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.077,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36193333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.693225003979848,
      "rewrite6_current_pass_fraction": 0.6636190142353605,
      "rewrite6_full_passes": 0.0,
      "step": 93220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1778048,
      "curriculum_late_phase": 1.0,
      "epoch": 5.615e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.869915008544922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3622,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.6642470448482016,
      "rewrite6_full_passes": 0.0,
      "step": 93230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1802624,
      "curriculum_late_phase": 1.0,
      "epoch": 5.62e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.869915008544922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0744,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36246666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.63689587856461,
      "rewrite6_current_pass_fraction": 0.6648069170999591,
      "rewrite6_full_passes": 0.0,
      "step": 93240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.18272,
      "curriculum_late_phase": 1.0,
      "epoch": 5.625e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.869915008544922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0808,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36274666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.77581608265989,
      "rewrite6_current_pass_fraction": 0.6654057369866215,
      "rewrite6_full_passes": 0.0,
      "step": 93250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1851776,
      "curriculum_late_phase": 1.0,
      "epoch": 5.63e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.319343566894531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0697,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3630933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.535441073488606,
      "rewrite6_current_pass_fraction": 0.6659558723296528,
      "rewrite6_full_passes": 0.0,
      "step": 93260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1876352,
      "curriculum_late_phase": 1.0,
      "epoch": 5.635e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.31960678100586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0366,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3634133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.83428610357238,
      "rewrite6_current_pass_fraction": 0.6665206130357734,
      "rewrite6_full_passes": 0.0,
      "step": 93270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1900928,
      "curriculum_late_phase": 1.0,
      "epoch": 5.64e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.319629669189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.072,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36373333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.585029592894678,
      "rewrite6_current_pass_fraction": 0.6671145644680727,
      "rewrite6_full_passes": 0.0,
      "step": 93280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1925504,
      "curriculum_late_phase": 1.0,
      "epoch": 5.645e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.319629669189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.054,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36406666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.199974948441902,
      "rewrite6_current_pass_fraction": 0.667620883721836,
      "rewrite6_full_passes": 0.0,
      "step": 93290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.195008,
      "curriculum_late_phase": 1.0,
      "epoch": 5.65e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.322998046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0511,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36436,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.138584080874157,
      "rewrite6_current_pass_fraction": 0.6682586512434032,
      "rewrite6_full_passes": 0.0,
      "step": 93300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1974656,
      "curriculum_late_phase": 1.0,
      "epoch": 5.655e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.325393676757812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36477333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.32969002438444,
      "rewrite6_current_pass_fraction": 0.668828260403887,
      "rewrite6_full_passes": 0.0,
      "step": 93310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.1999232,
      "curriculum_late_phase": 1.0,
      "epoch": 5.66e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.325557708740234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36512,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.896882809995212,
      "rewrite6_current_pass_fraction": 0.66941247492746,
      "rewrite6_full_passes": 0.0,
      "step": 93320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2023808,
      "curriculum_late_phase": 1.0,
      "epoch": 5.665e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.325992584228516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0558,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3654,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23816926792417,
      "rewrite6_current_pass_fraction": 0.6700356370859379,
      "rewrite6_full_passes": 0.0,
      "step": 93330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2048384,
      "curriculum_late_phase": 1.0,
      "epoch": 5.67e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.32674789428711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0685,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36572,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50961404351764,
      "rewrite6_current_pass_fraction": 0.6706734046075052,
      "rewrite6_full_passes": 0.0,
      "step": 93340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.207296,
      "curriculum_late_phase": 1.0,
      "epoch": 5.675e-06,
      "grad_norm": 0.5859375,
      "host_rss_gib": 15.327693939208984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0736,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3659733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.61959328382246,
      "rewrite6_current_pass_fraction": 0.6712089345874471,
      "rewrite6_full_passes": 0.0,
      "step": 93350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2097536,
      "curriculum_late_phase": 1.0,
      "epoch": 5.68e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.32769775390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0186,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36628,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.46262394793285,
      "rewrite6_current_pass_fraction": 0.6717152538412104,
      "rewrite6_full_passes": 0.0,
      "step": 93360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2122112,
      "curriculum_late_phase": 1.0,
      "epoch": 5.685e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.32769775390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36674666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.6722556522755155,
      "rewrite6_full_passes": 0.0,
      "step": 93370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2146688,
      "curriculum_late_phase": 1.0,
      "epoch": 5.69e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.331169128417969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0599,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3670933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.325424512944927,
      "rewrite6_current_pass_fraction": 0.6729274989776246,
      "rewrite6_full_passes": 0.0,
      "step": 93380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2171264,
      "curriculum_late_phase": 1.0,
      "epoch": 5.695e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.331169128417969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0761,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3674266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.67370988459726,
      "rewrite6_current_pass_fraction": 0.6734630289575666,
      "rewrite6_full_passes": 0.0,
      "step": 93390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.219584,
      "curriculum_late_phase": 1.0,
      "epoch": 5.7e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.331169128417969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0616,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36784,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.361708567324747,
      "rewrite6_current_pass_fraction": 0.6740229012093241,
      "rewrite6_full_passes": 0.0,
      "step": 93400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2220416,
      "curriculum_late_phase": 1.0,
      "epoch": 5.705e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.331169128417969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36802666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.016335157786333,
      "rewrite6_current_pass_fraction": 0.6746655371852545,
      "rewrite6_full_passes": 0.0,
      "step": 93410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2244992,
      "curriculum_late_phase": 1.0,
      "epoch": 5.71e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.331169128417969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.36837333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.6752448832544644,
      "rewrite6_full_passes": 0.0,
      "step": 93420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2269568,
      "curriculum_late_phase": 1.0,
      "epoch": 5.715e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.331192016601562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3686933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.041569897791508,
      "rewrite6_current_pass_fraction": 0.6758047555062219,
      "rewrite6_full_passes": 0.0,
      "step": 93430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2294144,
      "curriculum_late_phase": 1.0,
      "epoch": 5.72e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.331192016601562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0711,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.369,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.56561180557606,
      "rewrite6_current_pass_fraction": 0.6764912075714202,
      "rewrite6_full_passes": 0.0,
      "step": 93440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.231872,
      "curriculum_late_phase": 1.0,
      "epoch": 5.725e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.331192016601562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0602,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3692533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.33182310003889,
      "rewrite6_current_pass_fraction": 0.6770656851862671,
      "rewrite6_full_passes": 0.0,
      "step": 93450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2343296,
      "curriculum_late_phase": 1.0,
      "epoch": 5.73e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.331192016601562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3695466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.295589807793416,
      "rewrite6_current_pass_fraction": 0.677645031255477,
      "rewrite6_full_passes": 0.0,
      "step": 93460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2367872,
      "curriculum_late_phase": 1.0,
      "epoch": 5.735e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.331192016601562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0614,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3698666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.357436652816972,
      "rewrite6_current_pass_fraction": 0.678273061868318,
      "rewrite6_full_passes": 0.0,
      "step": 93470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2392448,
      "curriculum_late_phase": 1.0,
      "epoch": 5.74e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.331192016601562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0661,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3702,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45805286797322,
      "rewrite6_current_pass_fraction": 0.678852407937528,
      "rewrite6_full_passes": 0.0,
      "step": 93480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2417024,
      "curriculum_late_phase": 1.0,
      "epoch": 5.745e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.331192016601562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0521,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37050666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.159733237771047,
      "rewrite6_current_pass_fraction": 0.6794463593698272,
      "rewrite6_full_passes": 0.0,
      "step": 93490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.24416,
      "curriculum_late_phase": 1.0,
      "epoch": 5.75e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.331192016601562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37084,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.07737098892312,
      "rewrite6_current_pass_fraction": 0.6800987322544838,
      "rewrite6_full_passes": 0.0,
      "step": 93500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 5.75e-06,
      "eval_loss": 3.082920551300049,
      "eval_perplexity": 21.822041812457226,
      "eval_runtime": 0.0489,
      "eval_samples_per_second": 654.239,
      "eval_steps_per_second": 20.445,
      "step": 93500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2466176,
      "curriculum_late_phase": 1.0,
      "epoch": 5.755e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.377342224121094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0702,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3712933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.54621148640419,
      "rewrite6_current_pass_fraction": 0.6807559735935035,
      "rewrite6_full_passes": 0.0,
      "step": 93510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2490752,
      "curriculum_late_phase": 1.0,
      "epoch": 5.76e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.377674102783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0526,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3715333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.170315749797474,
      "rewrite6_current_pass_fraction": 0.6813353196627134,
      "rewrite6_full_passes": 0.0,
      "step": 93520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2515328,
      "curriculum_late_phase": 1.0,
      "epoch": 5.765e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.377674102783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37190666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.97853978054658,
      "rewrite6_current_pass_fraction": 0.6819779556386438,
      "rewrite6_full_passes": 0.0,
      "step": 93530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2539904,
      "curriculum_late_phase": 1.0,
      "epoch": 5.77e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.377674102783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0669,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3721466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.475226178675968,
      "rewrite6_current_pass_fraction": 0.6826157231602111,
      "rewrite6_full_passes": 0.0,
      "step": 93540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.256448,
      "curriculum_late_phase": 1.0,
      "epoch": 5.775e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.377674102783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0566,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37236,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25516660136536,
      "rewrite6_current_pass_fraction": 0.6832340168643259,
      "rewrite6_full_passes": 0.0,
      "step": 93550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2589056,
      "curriculum_late_phase": 1.0,
      "epoch": 5.78e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.377674102783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37273333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.229675699044027,
      "rewrite6_current_pass_fraction": 0.6840178380167864,
      "rewrite6_full_passes": 0.0,
      "step": 93560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2613632,
      "curriculum_late_phase": 1.0,
      "epoch": 5.785e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.377674102783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37308,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.921974121142934,
      "rewrite6_current_pass_fraction": 0.684533894179276,
      "rewrite6_full_passes": 0.0,
      "step": 93570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2638208,
      "curriculum_late_phase": 1.0,
      "epoch": 5.79e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.949190374231975,
      "rewrite6_current_pass_fraction": 0.6851667932464801,
      "rewrite6_full_passes": 0.0,
      "step": 93580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2662784,
      "curriculum_late_phase": 1.0,
      "epoch": 5.795e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3736933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.066834937660975,
      "rewrite6_current_pass_fraction": 0.6858532453116785,
      "rewrite6_full_passes": 0.0,
      "step": 93590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.268736,
      "curriculum_late_phase": 1.0,
      "epoch": 5.8e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37390666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.391635903566247,
      "rewrite6_current_pass_fraction": 0.6864764074701564,
      "rewrite6_full_passes": 0.0,
      "step": 93600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2711936,
      "curriculum_late_phase": 1.0,
      "epoch": 5.805e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37422666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.229675699044027,
      "rewrite6_current_pass_fraction": 0.6872018071702596,
      "rewrite6_full_passes": 0.0,
      "step": 93610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2736512,
      "curriculum_late_phase": 1.0,
      "epoch": 5.81e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37456,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.968053132536788,
      "rewrite6_current_pass_fraction": 0.687761679422017,
      "rewrite6_full_passes": 0.0,
      "step": 93620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2761088,
      "curriculum_late_phase": 1.0,
      "epoch": 5.815e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3749866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.662077184548252,
      "rewrite6_current_pass_fraction": 0.6883897100348582,
      "rewrite6_full_passes": 0.0,
      "step": 93630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2785664,
      "curriculum_late_phase": 1.0,
      "epoch": 5.82e-06,
      "grad_norm": 0.55078125,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0606,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3752666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.340357536052313,
      "rewrite6_current_pass_fraction": 0.6890323460107884,
      "rewrite6_full_passes": 0.0,
      "step": 93640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.281024,
      "curriculum_late_phase": 1.0,
      "epoch": 5.825e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0497,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37554666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10901076930934,
      "rewrite6_current_pass_fraction": 0.6896749819867188,
      "rewrite6_full_passes": 0.0,
      "step": 93650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2834816,
      "curriculum_late_phase": 1.0,
      "epoch": 5.83e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37594666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.928251654962274,
      "rewrite6_current_pass_fraction": 0.6902884072364706,
      "rewrite6_full_passes": 0.0,
      "step": 93660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2859392,
      "curriculum_late_phase": 1.0,
      "epoch": 5.835e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3763066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.018436896387286,
      "rewrite6_current_pass_fraction": 0.6908774902144067,
      "rewrite6_full_passes": 0.0,
      "step": 93670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2883968,
      "curriculum_late_phase": 1.0,
      "epoch": 5.84e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37656,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.02684595284501,
      "rewrite6_current_pass_fraction": 0.6914957839185215,
      "rewrite6_full_passes": 0.0,
      "step": 93680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2908544,
      "curriculum_late_phase": 1.0,
      "epoch": 5.845e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3768666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.7801873185992,
      "rewrite6_current_pass_fraction": 0.6921140776226363,
      "rewrite6_full_passes": 0.0,
      "step": 93690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.293312,
      "curriculum_late_phase": 1.0,
      "epoch": 5.85e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.377738952636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37722666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.039465846006077,
      "rewrite6_current_pass_fraction": 0.6927128975092988,
      "rewrite6_full_passes": 0.0,
      "step": 93700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2957696,
      "curriculum_late_phase": 1.0,
      "epoch": 5.855e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0337,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37756,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.77395419741855,
      "rewrite6_current_pass_fraction": 0.6933165858503242,
      "rewrite6_full_passes": 0.0,
      "step": 93710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.2982272,
      "curriculum_late_phase": 1.0,
      "epoch": 5.86e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3779866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35530101593532,
      "rewrite6_current_pass_fraction": 0.693886195010808,
      "rewrite6_full_passes": 0.0,
      "step": 93720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3006848,
      "curriculum_late_phase": 1.0,
      "epoch": 5.865e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3782,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.87808407616535,
      "rewrite6_current_pass_fraction": 0.6945434363498276,
      "rewrite6_full_passes": 0.0,
      "step": 93730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3031424,
      "curriculum_late_phase": 1.0,
      "epoch": 5.87e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0712,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3784,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.567768474588277,
      "rewrite6_current_pass_fraction": 0.6950448871492279,
      "rewrite6_full_passes": 0.0,
      "step": 93740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3056,
      "curriculum_late_phase": 1.0,
      "epoch": 5.875e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0409,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37877333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.924066423168412,
      "rewrite6_current_pass_fraction": 0.6956923915795213,
      "rewrite6_full_passes": 0.0,
      "step": 93750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3080576,
      "curriculum_late_phase": 1.0,
      "epoch": 5.88e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0618,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37905333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.365981336300866,
      "rewrite6_current_pass_fraction": 0.6962230531051002,
      "rewrite6_full_passes": 0.0,
      "step": 93760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3105152,
      "curriculum_late_phase": 1.0,
      "epoch": 5.885e-06,
      "grad_norm": 0.625,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3792933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.882260110570105,
      "rewrite6_current_pass_fraction": 0.6968364783548519,
      "rewrite6_full_passes": 0.0,
      "step": 93770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3129728,
      "curriculum_late_phase": 1.0,
      "epoch": 5.89e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0627,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.37961333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.38521937532253,
      "rewrite6_current_pass_fraction": 0.6973768767891569,
      "rewrite6_full_passes": 0.0,
      "step": 93780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3154304,
      "curriculum_late_phase": 1.0,
      "epoch": 5.895e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0533,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18514015876014,
      "rewrite6_current_pass_fraction": 0.6980389865825398,
      "rewrite6_full_passes": 0.0,
      "step": 93790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.317888,
      "curriculum_late_phase": 1.0,
      "epoch": 5.9e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3804133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.913607005029263,
      "rewrite6_current_pass_fraction": 0.6986718856497439,
      "rewrite6_full_passes": 0.0,
      "step": 93800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3203456,
      "curriculum_late_phase": 1.0,
      "epoch": 5.905e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0778,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3807333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.710586527666557,
      "rewrite6_current_pass_fraction": 0.6992609686276801,
      "rewrite6_full_passes": 0.0,
      "step": 93810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3228032,
      "curriculum_late_phase": 1.0,
      "epoch": 5.91e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0493,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38109333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.100568853497336,
      "rewrite6_current_pass_fraction": 0.6998208408794376,
      "rewrite6_full_passes": 0.0,
      "step": 93820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3252608,
      "curriculum_late_phase": 1.0,
      "epoch": 5.915e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38136,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.259418059817307,
      "rewrite6_current_pass_fraction": 0.7004001869486476,
      "rewrite6_full_passes": 0.0,
      "step": 93830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3277184,
      "curriculum_late_phase": 1.0,
      "epoch": 5.92e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0706,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38166666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.55483169492553,
      "rewrite6_current_pass_fraction": 0.7009697961091312,
      "rewrite6_full_passes": 0.0,
      "step": 93840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.330176,
      "curriculum_late_phase": 1.0,
      "epoch": 5.925e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38188,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.155501714289947,
      "rewrite6_current_pass_fraction": 0.7015345368152519,
      "rewrite6_full_passes": 0.0,
      "step": 93850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3326336,
      "curriculum_late_phase": 1.0,
      "epoch": 5.93e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38212,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35530101593532,
      "rewrite6_current_pass_fraction": 0.702118751338825,
      "rewrite6_full_passes": 0.0,
      "step": 93860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3350912,
      "curriculum_late_phase": 1.0,
      "epoch": 5.935e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0209,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3824266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.5097421481721,
      "rewrite6_current_pass_fraction": 0.7027127027711242,
      "rewrite6_full_passes": 0.0,
      "step": 93870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3375488,
      "curriculum_late_phase": 1.0,
      "epoch": 5.94e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0503,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38265333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.1216799761529,
      "rewrite6_current_pass_fraction": 0.703282311931608,
      "rewrite6_full_passes": 0.0,
      "step": 93880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3400064,
      "curriculum_late_phase": 1.0,
      "epoch": 5.945e-06,
      "grad_norm": 0.53515625,
      "host_rss_gib": 15.377769470214844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38294666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.7037983680940975,
      "rewrite6_full_passes": 0.0,
      "step": 93890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.342464,
      "curriculum_late_phase": 1.0,
      "epoch": 5.95e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.79098892211914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0739,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38328,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.626080134786605,
      "rewrite6_current_pass_fraction": 0.7043923195263968,
      "rewrite6_full_passes": 0.0,
      "step": 93900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3449216,
      "curriculum_late_phase": 1.0,
      "epoch": 5.955e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.121307373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0528,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38344,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.174550236381975,
      "rewrite6_current_pass_fraction": 0.7049327179607018,
      "rewrite6_full_passes": 0.0,
      "step": 93910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3473792,
      "curriculum_late_phase": 1.0,
      "epoch": 5.96e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.121307373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0549,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38382666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.219063514461734,
      "rewrite6_current_pass_fraction": 0.7054633794862807,
      "rewrite6_full_passes": 0.0,
      "step": 93920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3498368,
      "curriculum_late_phase": 1.0,
      "epoch": 5.965e-06,
      "grad_norm": 0.58203125,
      "host_rss_gib": 15.121307373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0511,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3842933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.138584080874157,
      "rewrite6_current_pass_fraction": 0.706193647640747,
      "rewrite6_full_passes": 0.0,
      "step": 93930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3522944,
      "curriculum_late_phase": 1.0,
      "epoch": 5.97e-06,
      "grad_norm": 0.5234375,
      "host_rss_gib": 15.121307373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0772,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38452,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.69756408287407,
      "rewrite6_current_pass_fraction": 0.7068362836166774,
      "rewrite6_full_passes": 0.0,
      "step": 93940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.354752,
      "curriculum_late_phase": 1.0,
      "epoch": 5.975e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.121307373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38481333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.295589807793416,
      "rewrite6_current_pass_fraction": 0.7075762886798699,
      "rewrite6_full_passes": 0.0,
      "step": 93950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3572096,
      "curriculum_late_phase": 1.0,
      "epoch": 5.98e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.121307373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38510666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.7081751085665323,
      "rewrite6_full_passes": 0.0,
      "step": 93960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3596672,
      "curriculum_late_phase": 1.0,
      "epoch": 5.985e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.121307373046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0653,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38522666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.440893290425038,
      "rewrite6_current_pass_fraction": 0.7088956398122724,
      "rewrite6_full_passes": 0.0,
      "step": 93970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3621248,
      "curriculum_late_phase": 1.0,
      "epoch": 5.99e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.249591827392578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0692,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38550666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.524676044433395,
      "rewrite6_current_pass_fraction": 0.709504196607661,
      "rewrite6_full_passes": 0.0,
      "step": 93980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3645824,
      "curriculum_late_phase": 1.0,
      "epoch": 5.995e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.100711822509766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0752,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38576,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.654212320920866,
      "rewrite6_current_pass_fraction": 0.7100981480399603,
      "rewrite6_full_passes": 0.0,
      "step": 93990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.36704,
      "curriculum_late_phase": 1.0,
      "epoch": 6e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.805908203125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38616,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.898972602764104,
      "rewrite6_current_pass_fraction": 0.7107992054682479,
      "rewrite6_full_passes": 0.0,
      "step": 94000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 6e-06,
      "eval_loss": 3.0833892822265625,
      "eval_perplexity": 21.832272875955017,
      "eval_runtime": 0.0483,
      "eval_samples_per_second": 662.683,
      "eval_steps_per_second": 20.709,
      "step": 94000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3694976,
      "curriculum_late_phase": 1.0,
      "epoch": 6.005e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3864,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.821789281331817,
      "rewrite6_current_pass_fraction": 0.7114905259878094,
      "rewrite6_full_passes": 0.0,
      "step": 94010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3719552,
      "curriculum_late_phase": 1.0,
      "epoch": 6.01e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3867333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.882260110570105,
      "rewrite6_current_pass_fraction": 0.7120650036026562,
      "rewrite6_full_passes": 0.0,
      "step": 94020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3744128,
      "curriculum_late_phase": 1.0,
      "epoch": 6.015e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0538,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38710666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.19573537742345,
      "rewrite6_current_pass_fraction": 0.712678428852408,
      "rewrite6_full_passes": 0.0,
      "step": 94030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3768704,
      "curriculum_late_phase": 1.0,
      "epoch": 6.02e-06,
      "grad_norm": 0.431640625,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0481,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3873066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.075263357207568,
      "rewrite6_current_pass_fraction": 0.7133161963739751,
      "rewrite6_full_passes": 0.0,
      "step": 94040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.379328,
      "curriculum_late_phase": 1.0,
      "epoch": 6.025e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0522,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3876,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.161849316897023,
      "rewrite6_current_pass_fraction": 0.713841989445191,
      "rewrite6_full_passes": 0.0,
      "step": 94050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3817856,
      "curriculum_late_phase": 1.0,
      "epoch": 6.03e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0591,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38792,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.30837099565101,
      "rewrite6_current_pass_fraction": 0.7144408093318533,
      "rewrite6_full_passes": 0.0,
      "step": 94060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3842432,
      "curriculum_late_phase": 1.0,
      "epoch": 6.035e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.028,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38821333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.655879491093387,
      "rewrite6_current_pass_fraction": 0.7149422601312535,
      "rewrite6_full_passes": 0.0,
      "step": 94070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3867008,
      "curriculum_late_phase": 1.0,
      "epoch": 6.04e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38865333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.7156043699246363,
      "rewrite6_full_passes": 0.0,
      "step": 94080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3891584,
      "curriculum_late_phase": 1.0,
      "epoch": 6.045e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38898666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.7161983213569356,
      "rewrite6_full_passes": 0.0,
      "step": 94090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.391616,
      "curriculum_late_phase": 1.0,
      "epoch": 6.05e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0185,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38941333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.460577787847765,
      "rewrite6_current_pass_fraction": 0.7167727989717825,
      "rewrite6_full_passes": 0.0,
      "step": 94100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3940736,
      "curriculum_late_phase": 1.0,
      "epoch": 6.055e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.95122528076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0742,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.38981333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.632568932097964,
      "rewrite6_current_pass_fraction": 0.7174154349477128,
      "rewrite6_full_passes": 0.0,
      "step": 94110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3965312,
      "curriculum_late_phase": 1.0,
      "epoch": 6.06e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.187026977539062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0657,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39004,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.449471363241397,
      "rewrite6_current_pass_fraction": 0.7179753071994703,
      "rewrite6_full_passes": 0.0,
      "step": 94120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.3989888,
      "curriculum_late_phase": 1.0,
      "epoch": 6.065e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.18956756591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3903333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.153386269892504,
      "rewrite6_current_pass_fraction": 0.7185303109968647,
      "rewrite6_full_passes": 0.0,
      "step": 94130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4014464,
      "curriculum_late_phase": 1.0,
      "epoch": 6.07e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.19347381591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39066666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.7190609725224436,
      "rewrite6_full_passes": 0.0,
      "step": 94140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.403904,
      "curriculum_late_phase": 1.0,
      "epoch": 6.075e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.194839477539062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39092,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.058409888808015,
      "rewrite6_current_pass_fraction": 0.7197133454071002,
      "rewrite6_full_passes": 0.0,
      "step": 94150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4063616,
      "curriculum_late_phase": 1.0,
      "epoch": 6.08e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.194839477539062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0844,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3913066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.854350297326935,
      "rewrite6_current_pass_fraction": 0.7202683492044946,
      "rewrite6_full_passes": 0.0,
      "step": 94160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4088192,
      "curriculum_late_phase": 1.0,
      "epoch": 6.085e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.195175170898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39166666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.058409888808015,
      "rewrite6_current_pass_fraction": 0.7207698000038948,
      "rewrite6_full_passes": 0.0,
      "step": 94170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4112768,
      "curriculum_late_phase": 1.0,
      "epoch": 6.09e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 16.195175170898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0697,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3920666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.535441073488606,
      "rewrite6_current_pass_fraction": 0.7213734883449202,
      "rewrite6_full_passes": 0.0,
      "step": 94180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4137344,
      "curriculum_late_phase": 1.0,
      "epoch": 6.095e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.195175170898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39236,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01213231105346,
      "rewrite6_current_pass_fraction": 0.7219528344141302,
      "rewrite6_full_passes": 0.0,
      "step": 94190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.416192,
      "curriculum_late_phase": 1.0,
      "epoch": 6.1e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.195194244384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3926266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.016335157786333,
      "rewrite6_current_pass_fraction": 0.7224737590309829,
      "rewrite6_full_passes": 0.0,
      "step": 94200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4186496,
      "curriculum_late_phase": 1.0,
      "epoch": 6.105e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.195194244384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.393,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.189377610522946,
      "rewrite6_current_pass_fraction": 0.7231115265525501,
      "rewrite6_full_passes": 0.0,
      "step": 94210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4211072,
      "curriculum_late_phase": 1.0,
      "epoch": 6.11e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.195331573486328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0583,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.29133111571526,
      "rewrite6_current_pass_fraction": 0.7235253451734144,
      "rewrite6_full_passes": 0.0,
      "step": 94220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4235648,
      "curriculum_late_phase": 1.0,
      "epoch": 6.115e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.195331573486328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0292,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3937866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.680681424666606,
      "rewrite6_current_pass_fraction": 0.7241825865124341,
      "rewrite6_full_passes": 0.0,
      "step": 94230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4260224,
      "curriculum_late_phase": 1.0,
      "epoch": 6.12e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.195331573486328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39404,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.7246889057661974,
      "rewrite6_full_passes": 0.0,
      "step": 94240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.42848,
      "curriculum_late_phase": 1.0,
      "epoch": 6.125e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.19533920288086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0191,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39441333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.472857818161167,
      "rewrite6_current_pass_fraction": 0.7253607524683063,
      "rewrite6_full_passes": 0.0,
      "step": 94250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4309376,
      "curriculum_late_phase": 1.0,
      "epoch": 6.13e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.19533920288086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0258,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39470666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.610486506804197,
      "rewrite6_current_pass_fraction": 0.7259839146267842,
      "rewrite6_full_passes": 0.0,
      "step": 94260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4333952,
      "curriculum_late_phase": 1.0,
      "epoch": 6.135e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.19533920288086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.395,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.016335157786333,
      "rewrite6_current_pass_fraction": 0.726553523787268,
      "rewrite6_full_passes": 0.0,
      "step": 94270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4358528,
      "curriculum_late_phase": 1.0,
      "epoch": 6.14e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.19533920288086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39524,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09845890211131,
      "rewrite6_current_pass_fraction": 0.7270939222215731,
      "rewrite6_full_passes": 0.0,
      "step": 94280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4383104,
      "curriculum_late_phase": 1.0,
      "epoch": 6.145e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.195667266845703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3956133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.028948742578034,
      "rewrite6_current_pass_fraction": 0.7276537944733306,
      "rewrite6_full_passes": 0.0,
      "step": 94290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.440768,
      "curriculum_late_phase": 1.0,
      "epoch": 6.15e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.196613311767578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.396,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.7281795875445464,
      "rewrite6_full_passes": 0.0,
      "step": 94300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4432256,
      "curriculum_late_phase": 1.0,
      "epoch": 6.155e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.196613311767578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0357,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39630666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.81554368143424,
      "rewrite6_current_pass_fraction": 0.7287394597963038,
      "rewrite6_full_passes": 0.0,
      "step": 94310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4456832,
      "curriculum_late_phase": 1.0,
      "epoch": 6.16e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.196895599365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0666,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39658666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.468784577110917,
      "rewrite6_current_pass_fraction": 0.7293772273178711,
      "rewrite6_full_passes": 0.0,
      "step": 94320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4481408,
      "curriculum_late_phase": 1.0,
      "epoch": 6.165e-06,
      "grad_norm": 0.58984375,
      "host_rss_gib": 16.197025299072266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39694666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.265796842004743,
      "rewrite6_current_pass_fraction": 0.7300588109287064,
      "rewrite6_full_passes": 0.0,
      "step": 94330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4505984,
      "curriculum_late_phase": 1.0,
      "epoch": 6.17e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.197673797607422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39728,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.191496654234413,
      "rewrite6_current_pass_fraction": 0.7306527623610056,
      "rewrite6_full_passes": 0.0,
      "step": 94340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.453056,
      "curriculum_late_phase": 1.0,
      "epoch": 6.175e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.197677612304688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0724,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39754666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.593665331764466,
      "rewrite6_current_pass_fraction": 0.7311980292496738,
      "rewrite6_full_passes": 0.0,
      "step": 94350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4555136,
      "curriculum_late_phase": 1.0,
      "epoch": 6.18e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.198165893554688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0324,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39796,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.746965603348944,
      "rewrite6_current_pass_fraction": 0.7317968491363362,
      "rewrite6_full_passes": 0.0,
      "step": 94360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4579712,
      "curriculum_late_phase": 1.0,
      "epoch": 6.185e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.19910430908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0528,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.3982266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.174550236381975,
      "rewrite6_current_pass_fraction": 0.7324784327471714,
      "rewrite6_full_passes": 0.0,
      "step": 94370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4604288,
      "curriculum_late_phase": 1.0,
      "epoch": 6.19e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.19910430908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39848,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.391635903566247,
      "rewrite6_current_pass_fraction": 0.7330431734532921,
      "rewrite6_full_passes": 0.0,
      "step": 94380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4628864,
      "curriculum_late_phase": 1.0,
      "epoch": 6.195e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.19910430908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0623,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39885333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.376666998161866,
      "rewrite6_current_pass_fraction": 0.7336858094292225,
      "rewrite6_full_passes": 0.0,
      "step": 94390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.465344,
      "curriculum_late_phase": 1.0,
      "epoch": 6.2e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.296024322509766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39912,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.018436896387286,
      "rewrite6_current_pass_fraction": 0.7342262078635274,
      "rewrite6_full_passes": 0.0,
      "step": 94400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4678016,
      "curriculum_late_phase": 1.0,
      "epoch": 6.205e-06,
      "grad_norm": 0.5546875,
      "host_rss_gib": 15.802608489990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39953333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01003120287951,
      "rewrite6_current_pass_fraction": 0.7348980545656365,
      "rewrite6_full_passes": 0.0,
      "step": 94410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4702592,
      "curriculum_late_phase": 1.0,
      "epoch": 6.21e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.802608489990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.39974666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.859302253584733,
      "rewrite6_current_pass_fraction": 0.7355163482697513,
      "rewrite6_full_passes": 0.0,
      "step": 94420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4727168,
      "curriculum_late_phase": 1.0,
      "epoch": 6.215e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.802608489990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4000666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.89479322619514,
      "rewrite6_current_pass_fraction": 0.736081088975872,
      "rewrite6_full_passes": 0.0,
      "step": 94430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4751744,
      "curriculum_late_phase": 1.0,
      "epoch": 6.22e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.802608489990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40038666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.7365825397752721,
      "rewrite6_full_passes": 0.0,
      "step": 94440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.477632,
      "curriculum_late_phase": 1.0,
      "epoch": 6.225e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.802608489990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0278,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4006266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.651748728285217,
      "rewrite6_current_pass_fraction": 0.7371813596619345,
      "rewrite6_full_passes": 0.0,
      "step": 94450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4800896,
      "curriculum_late_phase": 1.0,
      "epoch": 6.23e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.802608489990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0639,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40101333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.41089704209166,
      "rewrite6_current_pass_fraction": 0.7377850480029601,
      "rewrite6_full_passes": 0.0,
      "step": 94460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4825472,
      "curriculum_late_phase": 1.0,
      "epoch": 6.235e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.802608489990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4013466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.999528813094322,
      "rewrite6_current_pass_fraction": 0.7383643940721699,
      "rewrite6_full_passes": 0.0,
      "step": 94470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4850048,
      "curriculum_late_phase": 1.0,
      "epoch": 6.24e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.802608489990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0523,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4016266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.16396560764148,
      "rewrite6_current_pass_fraction": 0.7389145294152013,
      "rewrite6_full_passes": 0.0,
      "step": 94480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4874624,
      "curriculum_late_phase": 1.0,
      "epoch": 6.245e-06,
      "grad_norm": 0.5390625,
      "host_rss_gib": 15.802616119384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0406,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40197333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.917790144730294,
      "rewrite6_current_pass_fraction": 0.7395571653911316,
      "rewrite6_full_passes": 0.0,
      "step": 94490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.48992,
      "curriculum_late_phase": 1.0,
      "epoch": 6.25e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.802616119384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0314,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40232,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.726229007771437,
      "rewrite6_current_pass_fraction": 0.7400926953710736,
      "rewrite6_full_passes": 0.0,
      "step": 94500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 6.25e-06,
      "eval_loss": 3.083383798599243,
      "eval_perplexity": 21.83215315623528,
      "eval_runtime": 0.0496,
      "eval_samples_per_second": 645.706,
      "eval_steps_per_second": 20.178,
      "step": 94500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4923776,
      "curriculum_late_phase": 1.0,
      "epoch": 6.255e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.860466003417969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4026533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.871821590362348,
      "rewrite6_current_pass_fraction": 0.7406379622597418,
      "rewrite6_full_passes": 0.0,
      "step": 94510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4948352,
      "curriculum_late_phase": 1.0,
      "epoch": 6.26e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.860466003417969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40284,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.7412513875094935,
      "rewrite6_full_passes": 0.0,
      "step": 94520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4972928,
      "curriculum_late_phase": 1.0,
      "epoch": 6.265e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.862071990966797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0582,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40310666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28920208905679,
      "rewrite6_current_pass_fraction": 0.7419524449377811,
      "rewrite6_full_passes": 0.0,
      "step": 94530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.4997504,
      "curriculum_late_phase": 1.0,
      "epoch": 6.27e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.803779602050781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40338666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09845890211131,
      "rewrite6_current_pass_fraction": 0.7424587641915444,
      "rewrite6_full_passes": 0.0,
      "step": 94540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.502208,
      "curriculum_late_phase": 1.0,
      "epoch": 6.275e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.31005859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0676,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40386666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.49026409965934,
      "rewrite6_current_pass_fraction": 0.7430575840782069,
      "rewrite6_full_passes": 0.0,
      "step": 94550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5046656,
      "curriculum_late_phase": 1.0,
      "epoch": 6.28e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0682,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40416,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.503162127140435,
      "rewrite6_current_pass_fraction": 0.7438024575957625,
      "rewrite6_full_passes": 0.0,
      "step": 94560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5071232,
      "curriculum_late_phase": 1.0,
      "epoch": 6.285e-06,
      "grad_norm": 0.427734375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0342,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4045066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.784343771694374,
      "rewrite6_current_pass_fraction": 0.744357461393157,
      "rewrite6_full_passes": 0.0,
      "step": 94570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5095808,
      "curriculum_late_phase": 1.0,
      "epoch": 6.29e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0654,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4047866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.443037486962115,
      "rewrite6_current_pass_fraction": 0.7449270705536406,
      "rewrite6_full_passes": 0.0,
      "step": 94580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5120384,
      "curriculum_late_phase": 1.0,
      "epoch": 6.295e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0595,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40510666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.316896048946262,
      "rewrite6_current_pass_fraction": 0.7455502327121185,
      "rewrite6_full_passes": 0.0,
      "step": 94590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.514496,
      "curriculum_late_phase": 1.0,
      "epoch": 6.3e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.061,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4054,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34889538652299,
      "rewrite6_current_pass_fraction": 0.7462172109598645,
      "rewrite6_full_passes": 0.0,
      "step": 94600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5169536,
      "curriculum_late_phase": 1.0,
      "epoch": 6.305e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0588,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40570666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.301979443133135,
      "rewrite6_current_pass_fraction": 0.7467916885747113,
      "rewrite6_full_passes": 0.0,
      "step": 94610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5194112,
      "curriculum_late_phase": 1.0,
      "epoch": 6.31e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0623,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40589333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.376666998161866,
      "rewrite6_current_pass_fraction": 0.7474781406399097,
      "rewrite6_full_passes": 0.0,
      "step": 94620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5218688,
      "curriculum_late_phase": 1.0,
      "epoch": 6.315e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0806,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40614666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.771461354930647,
      "rewrite6_current_pass_fraction": 0.7480234075285779,
      "rewrite6_full_passes": 0.0,
      "step": 94630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5243264,
      "curriculum_late_phase": 1.0,
      "epoch": 6.32e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0723,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40646666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.591506073196012,
      "rewrite6_current_pass_fraction": 0.7485930166890615,
      "rewrite6_full_passes": 0.0,
      "step": 94640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.526784,
      "curriculum_late_phase": 1.0,
      "epoch": 6.325e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0226,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4068,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.544638363202598,
      "rewrite6_current_pass_fraction": 0.749104204397188,
      "rewrite6_full_passes": 0.0,
      "step": 94650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5292416,
      "curriculum_late_phase": 1.0,
      "epoch": 6.33e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40708,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.299849351695165,
      "rewrite6_current_pass_fraction": 0.7497614457362076,
      "rewrite6_full_passes": 0.0,
      "step": 94660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5316992,
      "curriculum_late_phase": 1.0,
      "epoch": 6.335e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0313,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40753333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.724156488498345,
      "rewrite6_current_pass_fraction": 0.7502872388074234,
      "rewrite6_full_passes": 0.0,
      "step": 94670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5341568,
      "curriculum_late_phase": 1.0,
      "epoch": 6.34e-06,
      "grad_norm": 0.55859375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0715,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40792,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.574239775777293,
      "rewrite6_current_pass_fraction": 0.7509590855095324,
      "rewrite6_full_passes": 0.0,
      "step": 94680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5366144,
      "curriculum_late_phase": 1.0,
      "epoch": 6.345e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4083466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.871821590362348,
      "rewrite6_current_pass_fraction": 0.7515871161223735,
      "rewrite6_full_passes": 0.0,
      "step": 94690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.539072,
      "curriculum_late_phase": 1.0,
      "epoch": 6.35e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4086533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.7521567252828572,
      "rewrite6_full_passes": 0.0,
      "step": 94700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5415296,
      "curriculum_late_phase": 1.0,
      "epoch": 6.355e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4090533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.892703851343008,
      "rewrite6_current_pass_fraction": 0.7527944928044245,
      "rewrite6_full_passes": 0.0,
      "step": 94710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5439872,
      "curriculum_late_phase": 1.0,
      "epoch": 6.36e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0481,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.40948,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.075263357207568,
      "rewrite6_current_pass_fraction": 0.7534225234172655,
      "rewrite6_full_passes": 0.0,
      "step": 94720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5464448,
      "curriculum_late_phase": 1.0,
      "epoch": 6.365e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0278,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4098266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.651748728285217,
      "rewrite6_current_pass_fraction": 0.7540602909388328,
      "rewrite6_full_passes": 0.0,
      "step": 94730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5489024,
      "curriculum_late_phase": 1.0,
      "epoch": 6.37e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4102266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.7545860840100485,
      "rewrite6_full_passes": 0.0,
      "step": 94740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.55136,
      "curriculum_late_phase": 1.0,
      "epoch": 6.375e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0226,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41054666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.544638363202598,
      "rewrite6_current_pass_fraction": 0.7551702985336215,
      "rewrite6_full_passes": 0.0,
      "step": 94750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5538176,
      "curriculum_late_phase": 1.0,
      "epoch": 6.38e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0885,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41085333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.94413707065502,
      "rewrite6_current_pass_fraction": 0.7557934606920995,
      "rewrite6_full_passes": 0.0,
      "step": 94760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5562752,
      "curriculum_late_phase": 1.0,
      "epoch": 6.385e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0405,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41125333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91569847030129,
      "rewrite6_current_pass_fraction": 0.7564701758485716,
      "rewrite6_full_passes": 0.0,
      "step": 94770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5587328,
      "curriculum_late_phase": 1.0,
      "epoch": 6.39e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41145333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.911515748893304,
      "rewrite6_current_pass_fraction": 0.7570836010983233,
      "rewrite6_full_passes": 0.0,
      "step": 94780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5611904,
      "curriculum_late_phase": 1.0,
      "epoch": 6.395e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4117866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.7576337364413546,
      "rewrite6_full_passes": 0.0,
      "step": 94790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.563648,
      "curriculum_late_phase": 1.0,
      "epoch": 6.4e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0593,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41208,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.312633096045975,
      "rewrite6_current_pass_fraction": 0.7581546610582073,
      "rewrite6_full_passes": 0.0,
      "step": 94800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5661056,
      "curriculum_late_phase": 1.0,
      "epoch": 6.405e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0187,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4123866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.46467031264417,
      "rewrite6_current_pass_fraction": 0.7587388755817803,
      "rewrite6_full_passes": 0.0,
      "step": 94810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5685632,
      "curriculum_late_phase": 1.0,
      "epoch": 6.41e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0377,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41261333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.857216427652418,
      "rewrite6_current_pass_fraction": 0.7592598001986329,
      "rewrite6_full_passes": 0.0,
      "step": 94820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5710208,
      "curriculum_late_phase": 1.0,
      "epoch": 6.415e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0591,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4129333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.30837099565101,
      "rewrite6_current_pass_fraction": 0.7597758563611224,
      "rewrite6_full_passes": 0.0,
      "step": 94830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5734784,
      "curriculum_late_phase": 1.0,
      "epoch": 6.42e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0456,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41318666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.022641004163326,
      "rewrite6_current_pass_fraction": 0.7605353352417674,
      "rewrite6_full_passes": 0.0,
      "step": 94840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.575936,
      "curriculum_late_phase": 1.0,
      "epoch": 6.425e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4135066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.7611049444022512,
      "rewrite6_full_passes": 0.0,
      "step": 94850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5783936,
      "curriculum_late_phase": 1.0,
      "epoch": 6.43e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0286,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4138266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.668276737590073,
      "rewrite6_current_pass_fraction": 0.7617816595587233,
      "rewrite6_full_passes": 0.0,
      "step": 94860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5808512,
      "curriculum_late_phase": 1.0,
      "epoch": 6.435e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4141066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.17878556994849,
      "rewrite6_current_pass_fraction": 0.7624535062608323,
      "rewrite6_full_passes": 0.0,
      "step": 94870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5833088,
      "curriculum_late_phase": 1.0,
      "epoch": 6.44e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4144,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.7631545636891199,
      "rewrite6_full_passes": 0.0,
      "step": 94880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5857664,
      "curriculum_late_phase": 1.0,
      "epoch": 6.445e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0604,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41473333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3360898913238,
      "rewrite6_current_pass_fraction": 0.7637874627563241,
      "rewrite6_full_passes": 0.0,
      "step": 94890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.588224,
      "curriculum_late_phase": 1.0,
      "epoch": 6.45e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4150933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.7644690463671594,
      "rewrite6_full_passes": 0.0,
      "step": 94900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5906816,
      "curriculum_late_phase": 1.0,
      "epoch": 6.455e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.049,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4153733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.094239632271936,
      "rewrite6_current_pass_fraction": 0.7649997078927382,
      "rewrite6_full_passes": 0.0,
      "step": 94910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5931392,
      "curriculum_late_phase": 1.0,
      "epoch": 6.46e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0724,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41570666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.593665331764466,
      "rewrite6_current_pass_fraction": 0.7655644485988589,
      "rewrite6_full_passes": 0.0,
      "step": 94920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5955968,
      "curriculum_late_phase": 1.0,
      "epoch": 6.465e-06,
      "grad_norm": 0.5859375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.416,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.7661291893049794,
      "rewrite6_full_passes": 0.0,
      "step": 94930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.5980544,
      "curriculum_late_phase": 1.0,
      "epoch": 6.47e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0622,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41636,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.37452943834182,
      "rewrite6_current_pass_fraction": 0.7667426145547311,
      "rewrite6_full_passes": 0.0,
      "step": 94940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.600512,
      "curriculum_late_phase": 1.0,
      "epoch": 6.475e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0456,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4166666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.022641004163326,
      "rewrite6_current_pass_fraction": 0.7673901189850246,
      "rewrite6_full_passes": 0.0,
      "step": 94950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6029696,
      "curriculum_late_phase": 1.0,
      "epoch": 6.48e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4169466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.155501714289947,
      "rewrite6_current_pass_fraction": 0.7680084126891394,
      "rewrite6_full_passes": 0.0,
      "step": 94960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6054272,
      "curriculum_late_phase": 1.0,
      "epoch": 6.485e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4171066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.299849351695165,
      "rewrite6_current_pass_fraction": 0.7685974956670756,
      "rewrite6_full_passes": 0.0,
      "step": 94970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6078848,
      "curriculum_late_phase": 1.0,
      "epoch": 6.49e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.660717010498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41748,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.225430188469428,
      "rewrite6_current_pass_fraction": 0.7691914470993749,
      "rewrite6_full_passes": 0.0,
      "step": 94980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6103424,
      "curriculum_late_phase": 1.0,
      "epoch": 6.495e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.003704071044922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0558,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41785333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23816926792417,
      "rewrite6_current_pass_fraction": 0.7698340830753052,
      "rewrite6_full_passes": 0.0,
      "step": 94990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6128,
      "curriculum_late_phase": 1.0,
      "epoch": 6.5e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.003902435302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4182133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.125904734609893,
      "rewrite6_current_pass_fraction": 0.7705156666861405,
      "rewrite6_full_passes": 0.0,
      "step": 95000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 6.5e-06,
      "eval_loss": 3.083516836166382,
      "eval_perplexity": 21.83505784598873,
      "eval_runtime": 0.0494,
      "eval_samples_per_second": 647.3,
      "eval_steps_per_second": 20.228,
      "step": 95000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6152576,
      "curriculum_late_phase": 1.0,
      "epoch": 6.505e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.085556030273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4185466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.989031673191437,
      "rewrite6_current_pass_fraction": 0.7711339603902553,
      "rewrite6_full_passes": 0.0,
      "step": 95010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6177152,
      "curriculum_late_phase": 1.0,
      "epoch": 6.51e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.085769653320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4189466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.144926607429838,
      "rewrite6_current_pass_fraction": 0.7716500165527448,
      "rewrite6_full_passes": 0.0,
      "step": 95020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6201728,
      "curriculum_late_phase": 1.0,
      "epoch": 6.515e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.085769653320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41912,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.151271037028916,
      "rewrite6_current_pass_fraction": 0.772234231076318,
      "rewrite6_full_passes": 0.0,
      "step": 95030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6226304,
      "curriculum_late_phase": 1.0,
      "epoch": 6.52e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.085769653320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0538,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41944,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.19573537742345,
      "rewrite6_current_pass_fraction": 0.7727405503300812,
      "rewrite6_full_passes": 0.0,
      "step": 95040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.625088,
      "curriculum_late_phase": 1.0,
      "epoch": 6.525e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.087318420410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0643,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.41986666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.419463114008664,
      "rewrite6_current_pass_fraction": 0.7733004225818387,
      "rewrite6_full_passes": 0.0,
      "step": 95050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6275456,
      "curriculum_late_phase": 1.0,
      "epoch": 6.53e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.088619232177734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0677,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42014666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.492413233524204,
      "rewrite6_current_pass_fraction": 0.7739089793772274,
      "rewrite6_full_passes": 0.0,
      "step": 95060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6300032,
      "curriculum_late_phase": 1.0,
      "epoch": 6.535e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.088726043701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0731,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42041333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.608786189179362,
      "rewrite6_current_pass_fraction": 0.7746051683511519,
      "rewrite6_full_passes": 0.0,
      "step": 95070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6324608,
      "curriculum_late_phase": 1.0,
      "epoch": 6.54e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.088752746582031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0591,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4208,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.30837099565101,
      "rewrite6_current_pass_fraction": 0.7752137251465405,
      "rewrite6_full_passes": 0.0,
      "step": 95080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6349184,
      "curriculum_late_phase": 1.0,
      "epoch": 6.545e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.089286804199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42104,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.7757541235808455,
      "rewrite6_full_passes": 0.0,
      "step": 95090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.637376,
      "curriculum_late_phase": 1.0,
      "epoch": 6.55e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.089286804199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0634,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4214133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.400194269486743,
      "rewrite6_current_pass_fraction": 0.7763870226480497,
      "rewrite6_full_passes": 0.0,
      "step": 95100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6398336,
      "curriculum_late_phase": 1.0,
      "epoch": 6.555e-06,
      "grad_norm": 0.55078125,
      "host_rss_gib": 15.089286804199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0684,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42174666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50746318965778,
      "rewrite6_current_pass_fraction": 0.7770053163521645,
      "rewrite6_full_passes": 0.0,
      "step": 95110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6422912,
      "curriculum_late_phase": 1.0,
      "epoch": 6.56e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.089286804199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4219733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.07737098892312,
      "rewrite6_current_pass_fraction": 0.7775457147864696,
      "rewrite6_full_passes": 0.0,
      "step": 95120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6447488,
      "curriculum_late_phase": 1.0,
      "epoch": 6.565e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 15.089286804199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42221333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.7780666394033222,
      "rewrite6_full_passes": 0.0,
      "step": 95130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6472064,
      "curriculum_late_phase": 1.0,
      "epoch": 6.57e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.08956527709961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0481,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4225466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.075263357207568,
      "rewrite6_current_pass_fraction": 0.7786898015618001,
      "rewrite6_full_passes": 0.0,
      "step": 95140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.649664,
      "curriculum_late_phase": 1.0,
      "epoch": 6.575e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.090641021728516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42285333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.7793080952659149,
      "rewrite6_full_passes": 0.0,
      "step": 95150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6521216,
      "curriculum_late_phase": 1.0,
      "epoch": 6.58e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.091136932373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42314666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.295589807793416,
      "rewrite6_current_pass_fraction": 0.779931257424393,
      "rewrite6_full_passes": 0.0,
      "step": 95160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6545792,
      "curriculum_late_phase": 1.0,
      "epoch": 6.585e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.091136932373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0373,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4235333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.848875209436216,
      "rewrite6_current_pass_fraction": 0.7805738934003232,
      "rewrite6_full_passes": 0.0,
      "step": 95170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6570368,
      "curriculum_late_phase": 1.0,
      "epoch": 6.59e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.091136932373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0297,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42386666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.691024350895024,
      "rewrite6_current_pass_fraction": 0.7812213978306167,
      "rewrite6_full_passes": 0.0,
      "step": 95180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6594944,
      "curriculum_late_phase": 1.0,
      "epoch": 6.595e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.091136932373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0784,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4242533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.72361678827043,
      "rewrite6_current_pass_fraction": 0.7817958754454636,
      "rewrite6_full_passes": 0.0,
      "step": 95190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.661952,
      "curriculum_late_phase": 1.0,
      "epoch": 6.6e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.091136932373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0625,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42462666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.380942759123343,
      "rewrite6_current_pass_fraction": 0.7823703530603104,
      "rewrite6_full_passes": 0.0,
      "step": 95200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6644096,
      "curriculum_late_phase": 1.0,
      "epoch": 6.605e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.091136932373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0632,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42494666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3959146586082,
      "rewrite6_current_pass_fraction": 0.7829983836731514,
      "rewrite6_full_passes": 0.0,
      "step": 95210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6668672,
      "curriculum_late_phase": 1.0,
      "epoch": 6.61e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.091136932373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0634,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42522666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.400194269486743,
      "rewrite6_current_pass_fraction": 0.7835923351054507,
      "rewrite6_full_passes": 0.0,
      "step": 95220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6693248,
      "curriculum_late_phase": 1.0,
      "epoch": 6.615e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.091136932373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0319,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42552,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.736594713485793,
      "rewrite6_current_pass_fraction": 0.7841619442659344,
      "rewrite6_full_passes": 0.0,
      "step": 95230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6717824,
      "curriculum_late_phase": 1.0,
      "epoch": 6.62e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.091136932373047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4258133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.921974121142934,
      "rewrite6_current_pass_fraction": 0.7846974742458764,
      "rewrite6_full_passes": 0.0,
      "step": 95240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.67424,
      "curriculum_late_phase": 1.0,
      "epoch": 6.625e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42608,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.871821590362348,
      "rewrite6_current_pass_fraction": 0.785213530408366,
      "rewrite6_full_passes": 0.0,
      "step": 95250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6766976,
      "curriculum_late_phase": 1.0,
      "epoch": 6.63e-06,
      "grad_norm": 0.5546875,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0621,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42632,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.372392092267074,
      "rewrite6_current_pass_fraction": 0.78570524429904,
      "rewrite6_full_passes": 0.0,
      "step": 95260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6791552,
      "curriculum_late_phase": 1.0,
      "epoch": 6.635e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42658666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.7864209070904169,
      "rewrite6_full_passes": 0.0,
      "step": 95270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6816128,
      "curriculum_late_phase": 1.0,
      "epoch": 6.64e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0541,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42696,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20209505194016,
      "rewrite6_current_pass_fraction": 0.7869953847052638,
      "rewrite6_full_passes": 0.0,
      "step": 95280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6840704,
      "curriculum_late_phase": 1.0,
      "epoch": 6.645e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0671,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4272,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.47952165344486,
      "rewrite6_current_pass_fraction": 0.7875601254113844,
      "rewrite6_full_passes": 0.0,
      "step": 95290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.686528,
      "curriculum_late_phase": 1.0,
      "epoch": 6.65e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4275466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.016335157786333,
      "rewrite6_current_pass_fraction": 0.7882319721134934,
      "rewrite6_full_passes": 0.0,
      "step": 95300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6889856,
      "curriculum_late_phase": 1.0,
      "epoch": 6.655e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0623,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42789333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.376666998161866,
      "rewrite6_current_pass_fraction": 0.7889427664505073,
      "rewrite6_full_passes": 0.0,
      "step": 95310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6914432,
      "curriculum_late_phase": 1.0,
      "epoch": 6.66e-06,
      "grad_norm": 0.5859375,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0334,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4281333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.767722945893784,
      "rewrite6_current_pass_fraction": 0.7895026387022648,
      "rewrite6_full_passes": 0.0,
      "step": 95320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6939008,
      "curriculum_late_phase": 1.0,
      "epoch": 6.665e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0411,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4285466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.928251654962274,
      "rewrite6_current_pass_fraction": 0.7900868532258378,
      "rewrite6_full_passes": 0.0,
      "step": 95330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.6963584,
      "curriculum_late_phase": 1.0,
      "epoch": 6.67e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0578,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4288,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.280688111130274,
      "rewrite6_current_pass_fraction": 0.7907586999279469,
      "rewrite6_full_passes": 0.0,
      "step": 95340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.698816,
      "curriculum_late_phase": 1.0,
      "epoch": 6.675e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0566,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42929333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25516660136536,
      "rewrite6_current_pass_fraction": 0.7913721251776986,
      "rewrite6_full_passes": 0.0,
      "step": 95350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7012736,
      "curriculum_late_phase": 1.0,
      "epoch": 6.68e-06,
      "grad_norm": 0.431640625,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.052,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4295733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.157617370242413,
      "rewrite6_current_pass_fraction": 0.7918589706140094,
      "rewrite6_full_passes": 0.0,
      "step": 95360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7037312,
      "curriculum_late_phase": 1.0,
      "epoch": 6.685e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.42992,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.240293191045346,
      "rewrite6_current_pass_fraction": 0.7924577905006719,
      "rewrite6_full_passes": 0.0,
      "step": 95370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7061888,
      "curriculum_late_phase": 1.0,
      "epoch": 6.69e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.049,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4302666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.094239632271936,
      "rewrite6_current_pass_fraction": 0.7930663472960604,
      "rewrite6_full_passes": 0.0,
      "step": 95380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7086464,
      "curriculum_late_phase": 1.0,
      "epoch": 6.695e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.091724395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43050666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.873908876883966,
      "rewrite6_current_pass_fraction": 0.7937138517263539,
      "rewrite6_full_passes": 0.0,
      "step": 95390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.711104,
      "curriculum_late_phase": 1.0,
      "epoch": 6.7e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.321910858154297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4308133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.84679042615617,
      "rewrite6_current_pass_fraction": 0.7943564877022843,
      "rewrite6_full_passes": 0.0,
      "step": 95400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7135616,
      "curriculum_late_phase": 1.0,
      "epoch": 6.705e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.104358673095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0589,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43106666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.304109747590893,
      "rewrite6_current_pass_fraction": 0.7949845183151253,
      "rewrite6_full_passes": 0.0,
      "step": 95410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7160192,
      "curriculum_late_phase": 1.0,
      "epoch": 6.71e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.104358673095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43137333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.7955346536581566,
      "rewrite6_full_passes": 0.0,
      "step": 95420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7184768,
      "curriculum_late_phase": 1.0,
      "epoch": 6.715e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.104358673095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4316,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.7961432104535452,
      "rewrite6_full_passes": 0.0,
      "step": 95430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7209344,
      "curriculum_late_phase": 1.0,
      "epoch": 6.72e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.104358673095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43189333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.295589807793416,
      "rewrite6_current_pass_fraction": 0.7968101887012912,
      "rewrite6_full_passes": 0.0,
      "step": 95440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.723392,
      "curriculum_late_phase": 1.0,
      "epoch": 6.725e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.104358673095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43224,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.7973213764094176,
      "rewrite6_full_passes": 0.0,
      "step": 95450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7258496,
      "curriculum_late_phase": 1.0,
      "epoch": 6.73e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.104358673095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0705,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43250666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.552676319526597,
      "rewrite6_current_pass_fraction": 0.7978617748437226,
      "rewrite6_full_passes": 0.0,
      "step": 95460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7283072,
      "curriculum_late_phase": 1.0,
      "epoch": 6.735e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.104358673095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4328266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.014233629348723,
      "rewrite6_current_pass_fraction": 0.7984070417323909,
      "rewrite6_full_passes": 0.0,
      "step": 95470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7307648,
      "curriculum_late_phase": 1.0,
      "epoch": 6.74e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.104358673095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.7990009931646901,
      "rewrite6_full_passes": 0.0,
      "step": 95480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7332224,
      "curriculum_late_phase": 1.0,
      "epoch": 6.745e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 16.104358673095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0794,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43365333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.74535127048861,
      "rewrite6_current_pass_fraction": 0.7996582345037098,
      "rewrite6_full_passes": 0.0,
      "step": 95490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.73568,
      "curriculum_late_phase": 1.0,
      "epoch": 6.75e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.104358673095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.049,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43402666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.094239632271936,
      "rewrite6_current_pass_fraction": 0.8003057389340033,
      "rewrite6_full_passes": 0.0,
      "step": 95500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 6.75e-06,
      "eval_loss": 3.0835976600646973,
      "eval_perplexity": 21.83682271180449,
      "eval_runtime": 0.0482,
      "eval_samples_per_second": 663.846,
      "eval_steps_per_second": 20.745,
      "step": 95500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7381376,
      "curriculum_late_phase": 1.0,
      "epoch": 6.755e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.104408264160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43434666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.155501714289947,
      "rewrite6_current_pass_fraction": 0.8008510058226714,
      "rewrite6_full_passes": 0.0,
      "step": 95510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7405952,
      "curriculum_late_phase": 1.0,
      "epoch": 6.76e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 16.104408264160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.089,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43464,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.955111882664713,
      "rewrite6_current_pass_fraction": 0.8014206149831552,
      "rewrite6_full_passes": 0.0,
      "step": 95520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7430528,
      "curriculum_late_phase": 1.0,
      "epoch": 6.765e-06,
      "grad_norm": 0.5546875,
      "host_rss_gib": 16.104408264160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43484,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23604555718468,
      "rewrite6_current_pass_fraction": 0.8020583825047224,
      "rewrite6_full_passes": 0.0,
      "step": 95530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7455104,
      "curriculum_late_phase": 1.0,
      "epoch": 6.77e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.104408264160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4352533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.686886559817744,
      "rewrite6_current_pass_fraction": 0.8027302292068315,
      "rewrite6_full_passes": 0.0,
      "step": 95540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.747968,
      "curriculum_late_phase": 1.0,
      "epoch": 6.775e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.104408264160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43566666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.189377610522946,
      "rewrite6_current_pass_fraction": 0.8033095752760414,
      "rewrite6_full_passes": 0.0,
      "step": 95550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7504256,
      "curriculum_late_phase": 1.0,
      "epoch": 6.78e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.104408264160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4360266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.8038548421647096,
      "rewrite6_full_passes": 0.0,
      "step": 95560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7528832,
      "curriculum_late_phase": 1.0,
      "epoch": 6.785e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.104408264160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43634666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.804502346595003,
      "rewrite6_full_passes": 0.0,
      "step": 95570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7553408,
      "curriculum_late_phase": 1.0,
      "epoch": 6.79e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.104408264160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0775,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.704074328586966,
      "rewrite6_current_pass_fraction": 0.8050573503923975,
      "rewrite6_full_passes": 0.0,
      "step": 95580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7577984,
      "curriculum_late_phase": 1.0,
      "epoch": 6.795e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43696,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.090021206202152,
      "rewrite6_current_pass_fraction": 0.8055539327374345,
      "rewrite6_full_passes": 0.0,
      "step": 95590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.760256,
      "curriculum_late_phase": 1.0,
      "epoch": 6.8e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0752,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4372666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.654212320920866,
      "rewrite6_current_pass_fraction": 0.8060748573542872,
      "rewrite6_full_passes": 0.0,
      "step": 95600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7627136,
      "curriculum_late_phase": 1.0,
      "epoch": 6.805e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0546,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4376,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21269875016978,
      "rewrite6_current_pass_fraction": 0.8066736772409495,
      "rewrite6_full_passes": 0.0,
      "step": 95610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7651712,
      "curriculum_late_phase": 1.0,
      "epoch": 6.81e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0516,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4379866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14915601567805,
      "rewrite6_current_pass_fraction": 0.8072043387665284,
      "rewrite6_full_passes": 0.0,
      "step": 95620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7676288,
      "curriculum_late_phase": 1.0,
      "epoch": 6.815e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0831,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4383066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.825958100866675,
      "rewrite6_current_pass_fraction": 0.8077398687464704,
      "rewrite6_full_passes": 0.0,
      "step": 95630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7700864,
      "curriculum_late_phase": 1.0,
      "epoch": 6.82e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0697,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43861333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.535441073488606,
      "rewrite6_current_pass_fraction": 0.8082997409982279,
      "rewrite6_full_passes": 0.0,
      "step": 95640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.772544,
      "curriculum_late_phase": 1.0,
      "epoch": 6.825e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43889333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09845890211131,
      "rewrite6_current_pass_fraction": 0.8090251406983311,
      "rewrite6_full_passes": 0.0,
      "step": 95650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7750016,
      "curriculum_late_phase": 1.0,
      "epoch": 6.83e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0831,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4392,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.825958100866675,
      "rewrite6_current_pass_fraction": 0.8096629082198984,
      "rewrite6_full_passes": 0.0,
      "step": 95660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7774592,
      "curriculum_late_phase": 1.0,
      "epoch": 6.835e-06,
      "grad_norm": 0.625,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43946666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.8103201495589181,
      "rewrite6_full_passes": 0.0,
      "step": 95670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7799168,
      "curriculum_late_phase": 1.0,
      "epoch": 6.84e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0717,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.43978666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.57855505524601,
      "rewrite6_current_pass_fraction": 0.8108069949952289,
      "rewrite6_full_passes": 0.0,
      "step": 95680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7823744,
      "curriculum_late_phase": 1.0,
      "epoch": 6.845e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.068,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.498861924749583,
      "rewrite6_current_pass_fraction": 0.8114350256080699,
      "rewrite6_full_passes": 0.0,
      "step": 95690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.784832,
      "curriculum_late_phase": 1.0,
      "epoch": 6.85e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4402933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.32969002438444,
      "rewrite6_current_pass_fraction": 0.811882923409476,
      "rewrite6_full_passes": 0.0,
      "step": 95700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7872896,
      "curriculum_late_phase": 1.0,
      "epoch": 6.855e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0607,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4406,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34249167851127,
      "rewrite6_current_pass_fraction": 0.8124525325699596,
      "rewrite6_full_passes": 0.0,
      "step": 95710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7897472,
      "curriculum_late_phase": 1.0,
      "epoch": 6.86e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44097333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91988202833721,
      "rewrite6_current_pass_fraction": 0.8131000370002531,
      "rewrite6_full_passes": 0.0,
      "step": 95720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7922048,
      "curriculum_late_phase": 1.0,
      "epoch": 6.865e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44121333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.817625339883573,
      "rewrite6_current_pass_fraction": 0.8136258300714689,
      "rewrite6_full_passes": 0.0,
      "step": 95730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7946624,
      "curriculum_late_phase": 1.0,
      "epoch": 6.87e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44148,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.090021206202152,
      "rewrite6_current_pass_fraction": 0.8142051761406789,
      "rewrite6_full_passes": 0.0,
      "step": 95740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.79712,
      "curriculum_late_phase": 1.0,
      "epoch": 6.875e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0924,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44189333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.029886307555408,
      "rewrite6_current_pass_fraction": 0.8148526805709724,
      "rewrite6_full_passes": 0.0,
      "step": 95750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.7995776,
      "curriculum_late_phase": 1.0,
      "epoch": 6.88e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0328,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44212,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.755266049568853,
      "rewrite6_current_pass_fraction": 0.8154368950945454,
      "rewrite6_full_passes": 0.0,
      "step": 95760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8020352,
      "curriculum_late_phase": 1.0,
      "epoch": 6.885e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0599,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44249333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.325424512944927,
      "rewrite6_current_pass_fraction": 0.8160454518899339,
      "rewrite6_full_passes": 0.0,
      "step": 95770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8044928,
      "curriculum_late_phase": 1.0,
      "epoch": 6.89e-06,
      "grad_norm": 0.55078125,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0836,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44282666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.836873808616637,
      "rewrite6_current_pass_fraction": 0.8165322973262449,
      "rewrite6_full_passes": 0.0,
      "step": 95780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8069504,
      "curriculum_late_phase": 1.0,
      "epoch": 6.895e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4430266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.855130810292252,
      "rewrite6_current_pass_fraction": 0.8171311172129072,
      "rewrite6_full_passes": 0.0,
      "step": 95790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.809408,
      "curriculum_late_phase": 1.0,
      "epoch": 6.9e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.054,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4434133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.199974948441902,
      "rewrite6_current_pass_fraction": 0.8177202001908435,
      "rewrite6_full_passes": 0.0,
      "step": 95800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8118656,
      "curriculum_late_phase": 1.0,
      "epoch": 6.905e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44366666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13647032815547,
      "rewrite6_current_pass_fraction": 0.8182265194446068,
      "rewrite6_full_passes": 0.0,
      "step": 95810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8143232,
      "curriculum_late_phase": 1.0,
      "epoch": 6.91e-06,
      "grad_norm": 0.578125,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0714,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44404,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.572082459667325,
      "rewrite6_current_pass_fraction": 0.8188253393312691,
      "rewrite6_full_passes": 0.0,
      "step": 95820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8167808,
      "curriculum_late_phase": 1.0,
      "epoch": 6.915e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44448,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.229675699044027,
      "rewrite6_current_pass_fraction": 0.8193949484917529,
      "rewrite6_full_passes": 0.0,
      "step": 95830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8192384,
      "curriculum_late_phase": 1.0,
      "epoch": 6.92e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.928627014160156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0218,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4447333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.528209225043515,
      "rewrite6_current_pass_fraction": 0.8200181106502308,
      "rewrite6_full_passes": 0.0,
      "step": 95840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.821696,
      "curriculum_late_phase": 1.0,
      "epoch": 6.925e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.298065185546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4451333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.8205828513563513,
      "rewrite6_full_passes": 0.0,
      "step": 95850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8241536,
      "curriculum_late_phase": 1.0,
      "epoch": 6.93e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.304088592529297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44544,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.99532932729427,
      "rewrite6_current_pass_fraction": 0.8211183813362933,
      "rewrite6_full_passes": 0.0,
      "step": 95860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8266112,
      "curriculum_late_phase": 1.0,
      "epoch": 6.935e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.304641723632812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0663,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44581333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.462344907736483,
      "rewrite6_current_pass_fraction": 0.8216782535880509,
      "rewrite6_full_passes": 0.0,
      "step": 95870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8290688,
      "curriculum_late_phase": 1.0,
      "epoch": 6.94e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.305721282958984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44622666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.898972602764104,
      "rewrite6_current_pass_fraction": 0.8222527312028977,
      "rewrite6_full_passes": 0.0,
      "step": 95880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8315264,
      "curriculum_late_phase": 1.0,
      "epoch": 6.945e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.305953979492188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44650666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.855130810292252,
      "rewrite6_current_pass_fraction": 0.8227493135479348,
      "rewrite6_full_passes": 0.0,
      "step": 95890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.833984,
      "curriculum_late_phase": 1.0,
      "epoch": 6.95e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.307365417480469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0602,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4469066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.33182310003889,
      "rewrite6_current_pass_fraction": 0.823250764347335,
      "rewrite6_full_passes": 0.0,
      "step": 95900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8364416,
      "curriculum_late_phase": 1.0,
      "epoch": 6.955e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.307365417480469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0491,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4472,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09634916170988,
      "rewrite6_current_pass_fraction": 0.823922611049444,
      "rewrite6_full_passes": 0.0,
      "step": 95910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8388992,
      "curriculum_late_phase": 1.0,
      "epoch": 6.96e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.307365417480469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0555,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4474,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.231798772765842,
      "rewrite6_current_pass_fraction": 0.8245701154797375,
      "rewrite6_full_passes": 0.0,
      "step": 95920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8413568,
      "curriculum_late_phase": 1.0,
      "epoch": 6.965e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.308074951171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0612,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44770666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35316559260667,
      "rewrite6_current_pass_fraction": 0.8252078830013048,
      "rewrite6_full_passes": 0.0,
      "step": 95930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8438144,
      "curriculum_late_phase": 1.0,
      "epoch": 6.97e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.308074951171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0598,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44806666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.323292077117205,
      "rewrite6_current_pass_fraction": 0.8257774921617885,
      "rewrite6_full_passes": 0.0,
      "step": 95940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.846272,
      "curriculum_late_phase": 1.0,
      "epoch": 6.975e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.308074951171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0641,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44853333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.415179649746566,
      "rewrite6_current_pass_fraction": 0.8264249965920819,
      "rewrite6_full_passes": 0.0,
      "step": 95950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8487296,
      "curriculum_late_phase": 1.0,
      "epoch": 6.98e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.308074951171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0354,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44892,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.809299954935618,
      "rewrite6_current_pass_fraction": 0.8269605265720239,
      "rewrite6_full_passes": 0.0,
      "step": 95960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8511872,
      "curriculum_late_phase": 1.0,
      "epoch": 6.985e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.308074951171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0323,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4491866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.744891010519986,
      "rewrite6_current_pass_fraction": 0.8276275048197698,
      "rewrite6_full_passes": 0.0,
      "step": 95970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8536448,
      "curriculum_late_phase": 1.0,
      "epoch": 6.99e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.308074951171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.44952,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34676060372526,
      "rewrite6_current_pass_fraction": 0.8282117193433429,
      "rewrite6_full_passes": 0.0,
      "step": 95980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8561024,
      "curriculum_late_phase": 1.0,
      "epoch": 6.995e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.308338165283203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4499066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13435678680148,
      "rewrite6_current_pass_fraction": 0.828844618410547,
      "rewrite6_full_passes": 0.0,
      "step": 95990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.85856,
      "curriculum_late_phase": 1.0,
      "epoch": 7e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.308338165283203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0334,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4503066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.767722945893784,
      "rewrite6_current_pass_fraction": 0.829423964479757,
      "rewrite6_full_passes": 0.0,
      "step": 96000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 7e-06,
      "eval_loss": 3.083017349243164,
      "eval_perplexity": 21.824154243457087,
      "eval_runtime": 0.0497,
      "eval_samples_per_second": 643.993,
      "eval_steps_per_second": 20.125,
      "step": 96000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8610176,
      "curriculum_late_phase": 1.0,
      "epoch": 7.005e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.489200592041016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45069333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.8299351521878834,
      "rewrite6_full_passes": 0.0,
      "step": 96010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8634752,
      "curriculum_late_phase": 1.0,
      "epoch": 7.01e-06,
      "grad_norm": 0.5390625,
      "host_rss_gib": 15.485401153564453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0769,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4510266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.691055789941966,
      "rewrite6_current_pass_fraction": 0.8305242351658195,
      "rewrite6_full_passes": 0.0,
      "step": 96020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8659328,
      "curriculum_late_phase": 1.0,
      "epoch": 7.015e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.485401153564453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0368,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4513466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.838453377506596,
      "rewrite6_current_pass_fraction": 0.8310646336001246,
      "rewrite6_full_passes": 0.0,
      "step": 96030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8683904,
      "curriculum_late_phase": 1.0,
      "epoch": 7.02e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.485431671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0812,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45165333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.78452815139054,
      "rewrite6_current_pass_fraction": 0.8317559541196861,
      "rewrite6_full_passes": 0.0,
      "step": 96040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.870848,
      "curriculum_late_phase": 1.0,
      "epoch": 7.025e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.485431671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0516,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45198666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14915601567805,
      "rewrite6_current_pass_fraction": 0.8324618800023369,
      "rewrite6_full_passes": 0.0,
      "step": 96050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8733056,
      "curriculum_late_phase": 1.0,
      "epoch": 7.03e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.485431671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45225333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14704120581873,
      "rewrite6_current_pass_fraction": 0.8330899106151779,
      "rewrite6_full_passes": 0.0,
      "step": 96060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8757632,
      "curriculum_late_phase": 1.0,
      "epoch": 7.035e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.485431671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0881,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45264,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.93536117112368,
      "rewrite6_current_pass_fraction": 0.8336643882300248,
      "rewrite6_full_passes": 0.0,
      "step": 96070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8782208,
      "curriculum_late_phase": 1.0,
      "epoch": 7.04e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.485431671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.453,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.913607005029263,
      "rewrite6_current_pass_fraction": 0.8342486027535978,
      "rewrite6_full_passes": 0.0,
      "step": 96080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8806784,
      "curriculum_late_phase": 1.0,
      "epoch": 7.045e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.485431671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0566,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45337333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25516660136536,
      "rewrite6_current_pass_fraction": 0.8348815018208019,
      "rewrite6_full_passes": 0.0,
      "step": 96090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.883136,
      "curriculum_late_phase": 1.0,
      "epoch": 7.05e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.485580444335938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45369333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.8354608478900118,
      "rewrite6_full_passes": 0.0,
      "step": 96100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8855936,
      "curriculum_late_phase": 1.0,
      "epoch": 7.055e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.485580444335938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45398666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21482012611182,
      "rewrite6_current_pass_fraction": 0.8361424315008471,
      "rewrite6_full_passes": 0.0,
      "step": 96110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8880512,
      "curriculum_late_phase": 1.0,
      "epoch": 7.06e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.485580444335938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0297,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4544,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.691024350895024,
      "rewrite6_current_pass_fraction": 0.8367315144787832,
      "rewrite6_full_passes": 0.0,
      "step": 96120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8905088,
      "curriculum_late_phase": 1.0,
      "epoch": 7.065e-06,
      "grad_norm": 0.54296875,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45476,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13435678680148,
      "rewrite6_current_pass_fraction": 0.837344939728535,
      "rewrite6_full_passes": 0.0,
      "step": 96130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8929664,
      "curriculum_late_phase": 1.0,
      "epoch": 7.07e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0841,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45512,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.84779497558516,
      "rewrite6_current_pass_fraction": 0.8379194173433818,
      "rewrite6_full_passes": 0.0,
      "step": 96140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.895424,
      "curriculum_late_phase": 1.0,
      "epoch": 7.075e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0807,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45542666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.77363860992708,
      "rewrite6_current_pass_fraction": 0.8384549473233238,
      "rewrite6_full_passes": 0.0,
      "step": 96150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.8978816,
      "curriculum_late_phase": 1.0,
      "epoch": 7.08e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4557866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.896882809995212,
      "rewrite6_current_pass_fraction": 0.8390781094818017,
      "rewrite6_full_passes": 0.0,
      "step": 96160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9003392,
      "curriculum_late_phase": 1.0,
      "epoch": 7.085e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45612,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.882260110570105,
      "rewrite6_current_pass_fraction": 0.8396331132791961,
      "rewrite6_full_passes": 0.0,
      "step": 96170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9027968,
      "curriculum_late_phase": 1.0,
      "epoch": 7.09e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0318,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45636,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.734521157693965,
      "rewrite6_current_pass_fraction": 0.8401248271698701,
      "rewrite6_full_passes": 0.0,
      "step": 96180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9052544,
      "curriculum_late_phase": 1.0,
      "epoch": 7.095e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4567466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.064728359497874,
      "rewrite6_current_pass_fraction": 0.8407139101478063,
      "rewrite6_full_passes": 0.0,
      "step": 96190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.907712,
      "curriculum_late_phase": 1.0,
      "epoch": 7.1e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0333,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4570666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.765646277434353,
      "rewrite6_current_pass_fraction": 0.8413175984888318,
      "rewrite6_full_passes": 0.0,
      "step": 96200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9101696,
      "curriculum_late_phase": 1.0,
      "epoch": 7.105e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.059,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45742666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.306240265089755,
      "rewrite6_current_pass_fraction": 0.8419310237385835,
      "rewrite6_full_passes": 0.0,
      "step": 96210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9126272,
      "curriculum_late_phase": 1.0,
      "epoch": 7.11e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0741,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45766666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.630405783364,
      "rewrite6_current_pass_fraction": 0.8425639228057876,
      "rewrite6_full_passes": 0.0,
      "step": 96220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9150848,
      "curriculum_late_phase": 1.0,
      "epoch": 7.115e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45812,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.882260110570105,
      "rewrite6_current_pass_fraction": 0.8430410313333723,
      "rewrite6_full_passes": 0.0,
      "step": 96230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9175424,
      "curriculum_late_phase": 1.0,
      "epoch": 7.12e-06,
      "grad_norm": 0.54296875,
      "host_rss_gib": 15.485809326171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4585066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.045778632630523,
      "rewrite6_current_pass_fraction": 0.8436057720394929,
      "rewrite6_full_passes": 0.0,
      "step": 96240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.92,
      "curriculum_late_phase": 1.0,
      "epoch": 7.125e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.988540649414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0368,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4587866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.838453377506596,
      "rewrite6_current_pass_fraction": 0.8442289341979708,
      "rewrite6_full_passes": 0.0,
      "step": 96250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9224576,
      "curriculum_late_phase": 1.0,
      "epoch": 7.13e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.683204650878906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45912,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.8448277540846332,
      "rewrite6_full_passes": 0.0,
      "step": 96260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9249152,
      "curriculum_late_phase": 1.0,
      "epoch": 7.135e-06,
      "grad_norm": 0.5859375,
      "host_rss_gib": 16.683204650878906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4593466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.248791007771803,
      "rewrite6_current_pass_fraction": 0.8454119686082062,
      "rewrite6_full_passes": 0.0,
      "step": 96270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9273728,
      "curriculum_late_phase": 1.0,
      "epoch": 7.14e-06,
      "grad_norm": 0.54296875,
      "host_rss_gib": 16.683204650878906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0428,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4596133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.96385994124339,
      "rewrite6_current_pass_fraction": 0.8459474985881482,
      "rewrite6_full_passes": 0.0,
      "step": 96280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9298304,
      "curriculum_late_phase": 1.0,
      "epoch": 7.145e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.45992,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.24241732656946,
      "rewrite6_current_pass_fraction": 0.8466436875620728,
      "rewrite6_full_passes": 0.0,
      "step": 96290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.932288,
      "curriculum_late_phase": 1.0,
      "epoch": 7.15e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0498,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46029333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.111121775934837,
      "rewrite6_current_pass_fraction": 0.8472571128118245,
      "rewrite6_full_passes": 0.0,
      "step": 96300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9347456,
      "curriculum_late_phase": 1.0,
      "epoch": 7.155e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46056,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.999528813094322,
      "rewrite6_current_pass_fraction": 0.8478997487877549,
      "rewrite6_full_passes": 0.0,
      "step": 96310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9372032,
      "curriculum_late_phase": 1.0,
      "epoch": 7.16e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0664,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4608933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.464491249542554,
      "rewrite6_current_pass_fraction": 0.8484693579482386,
      "rewrite6_full_passes": 0.0,
      "step": 96320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9396608,
      "curriculum_late_phase": 1.0,
      "epoch": 7.165e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.035,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4612,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.800977899475697,
      "rewrite6_current_pass_fraction": 0.8491071254698058,
      "rewrite6_full_passes": 0.0,
      "step": 96330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9421184,
      "curriculum_late_phase": 1.0,
      "epoch": 7.17e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0661,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46148,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45805286797322,
      "rewrite6_current_pass_fraction": 0.8497643668088255,
      "rewrite6_full_passes": 0.0,
      "step": 96340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.944576,
      "curriculum_late_phase": 1.0,
      "epoch": 7.175e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46182666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.125904734609893,
      "rewrite6_current_pass_fraction": 0.8505189772351074,
      "rewrite6_full_passes": 0.0,
      "step": 96350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9470336,
      "curriculum_late_phase": 1.0,
      "epoch": 7.18e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0455,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46214666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.020538845172617,
      "rewrite6_current_pass_fraction": 0.8511031917586804,
      "rewrite6_full_passes": 0.0,
      "step": 96360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9494912,
      "curriculum_late_phase": 1.0,
      "epoch": 7.185e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4625066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.8517409592802477,
      "rewrite6_full_passes": 0.0,
      "step": 96370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9519488,
      "curriculum_late_phase": 1.0,
      "epoch": 7.19e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4628933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.8523154368950946,
      "rewrite6_full_passes": 0.0,
      "step": 96380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9544064,
      "curriculum_late_phase": 1.0,
      "epoch": 7.195e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46312,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.701372449879624,
      "rewrite6_current_pass_fraction": 0.8529434675079356,
      "rewrite6_full_passes": 0.0,
      "step": 96390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.956864,
      "curriculum_late_phase": 1.0,
      "epoch": 7.2e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46353333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.87808407616535,
      "rewrite6_current_pass_fraction": 0.8535033397596931,
      "rewrite6_full_passes": 0.0,
      "step": 96400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9593216,
      "curriculum_late_phase": 1.0,
      "epoch": 7.205e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0336,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4638933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.771876905865113,
      "rewrite6_current_pass_fraction": 0.8540778173745399,
      "rewrite6_full_passes": 0.0,
      "step": 96410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9617792,
      "curriculum_late_phase": 1.0,
      "epoch": 7.21e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46424,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.8546133473544819,
      "rewrite6_full_passes": 0.0,
      "step": 96420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9642368,
      "curriculum_late_phase": 1.0,
      "epoch": 7.215e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4646266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.193615909860856,
      "rewrite6_current_pass_fraction": 0.8551586142431501,
      "rewrite6_full_passes": 0.0,
      "step": 96430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9666944,
      "curriculum_late_phase": 1.0,
      "epoch": 7.22e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0646,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4649333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.425889916915104,
      "rewrite6_current_pass_fraction": 0.8557671710385387,
      "rewrite6_full_passes": 0.0,
      "step": 96440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.969152,
      "curriculum_late_phase": 1.0,
      "epoch": 7.225e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0648,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46524,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.430175523444856,
      "rewrite6_current_pass_fraction": 0.8563075694728438,
      "rewrite6_full_passes": 0.0,
      "step": 96450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9716096,
      "curriculum_late_phase": 1.0,
      "epoch": 7.23e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46557333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.189377610522946,
      "rewrite6_current_pass_fraction": 0.8569258631769586,
      "rewrite6_full_passes": 0.0,
      "step": 96460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9740672,
      "curriculum_late_phase": 1.0,
      "epoch": 7.235e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.06,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4659466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.327557162026903,
      "rewrite6_current_pass_fraction": 0.8575344199723471,
      "rewrite6_full_passes": 0.0,
      "step": 96470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9765248,
      "curriculum_late_phase": 1.0,
      "epoch": 7.24e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0773,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4661866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.699733947773797,
      "rewrite6_current_pass_fraction": 0.8580115284999319,
      "rewrite6_full_passes": 0.0,
      "step": 96480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9789824,
      "curriculum_late_phase": 1.0,
      "epoch": 7.245e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.064,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46653333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.413038238853925,
      "rewrite6_current_pass_fraction": 0.8585908745691417,
      "rewrite6_full_passes": 0.0,
      "step": 96490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.98144,
      "curriculum_late_phase": 1.0,
      "epoch": 7.25e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0583,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46685333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.29133111571526,
      "rewrite6_current_pass_fraction": 0.8592529843625246,
      "rewrite6_full_passes": 0.0,
      "step": 96500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 7.25e-06,
      "eval_loss": 3.0828633308410645,
      "eval_perplexity": 21.820793180932718,
      "eval_runtime": 0.0504,
      "eval_samples_per_second": 635.233,
      "eval_steps_per_second": 19.851,
      "step": 96500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9838976,
      "curriculum_late_phase": 1.0,
      "epoch": 7.255e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.683334350585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0767,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46716,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.686718012576172,
      "rewrite6_current_pass_fraction": 0.859988120971354,
      "rewrite6_full_passes": 0.0,
      "step": 96510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9863552,
      "curriculum_late_phase": 1.0,
      "epoch": 7.26e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0583,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46736,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.29133111571526,
      "rewrite6_current_pass_fraction": 0.8605285194056591,
      "rewrite6_full_passes": 0.0,
      "step": 96520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9888128,
      "curriculum_late_phase": 1.0,
      "epoch": 7.265e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0716,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46764,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.57639730762967,
      "rewrite6_current_pass_fraction": 0.861102997020506,
      "rewrite6_full_passes": 0.0,
      "step": 96530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9912704,
      "curriculum_late_phase": 1.0,
      "epoch": 7.27e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0598,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46781333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.323292077117205,
      "rewrite6_current_pass_fraction": 0.8616920799984421,
      "rewrite6_full_passes": 0.0,
      "step": 96540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.993728,
      "curriculum_late_phase": 1.0,
      "epoch": 7.275e-06,
      "grad_norm": 0.57421875,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0853,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46814666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.8740280662623,
      "rewrite6_current_pass_fraction": 0.8621691885260268,
      "rewrite6_full_passes": 0.0,
      "step": 96550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9961856,
      "curriculum_late_phase": 1.0,
      "epoch": 7.28e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0836,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46848,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.836873808616637,
      "rewrite6_current_pass_fraction": 0.8627047185059687,
      "rewrite6_full_passes": 0.0,
      "step": 96560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 17.9986432,
      "curriculum_late_phase": 1.0,
      "epoch": 7.285e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46876,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.8633132753013574,
      "rewrite6_full_passes": 0.0,
      "step": 96570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0011008,
      "curriculum_late_phase": 1.0,
      "epoch": 7.29e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46908,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.007930304805882,
      "rewrite6_current_pass_fraction": 0.8638049891920313,
      "rewrite6_full_passes": 0.0,
      "step": 96580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0035584,
      "curriculum_late_phase": 1.0,
      "epoch": 7.295e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0481,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46945333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.075263357207568,
      "rewrite6_current_pass_fraction": 0.8643502560806995,
      "rewrite6_full_passes": 0.0,
      "step": 96590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.006016,
      "curriculum_late_phase": 1.0,
      "epoch": 7.3e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0611,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4696666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.351030382809682,
      "rewrite6_current_pass_fraction": 0.8650074974197192,
      "rewrite6_full_passes": 0.0,
      "step": 96600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0084736,
      "curriculum_late_phase": 1.0,
      "epoch": 7.305e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0641,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.46992,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.415179649746566,
      "rewrite6_current_pass_fraction": 0.8655722381258398,
      "rewrite6_full_passes": 0.0,
      "step": 96610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0109312,
      "curriculum_late_phase": 1.0,
      "epoch": 7.31e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0399,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47013333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.903152815291975,
      "rewrite6_current_pass_fraction": 0.8663122431890323,
      "rewrite6_full_passes": 0.0,
      "step": 96620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0133888,
      "curriculum_late_phase": 1.0,
      "epoch": 7.315e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47046666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.97853978054658,
      "rewrite6_current_pass_fraction": 0.8669840898911414,
      "rewrite6_full_passes": 0.0,
      "step": 96630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0158464,
      "curriculum_late_phase": 1.0,
      "epoch": 7.32e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.683799743652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0561,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4708,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.244541674517738,
      "rewrite6_current_pass_fraction": 0.8675731728690775,
      "rewrite6_full_passes": 0.0,
      "step": 96640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.018304,
      "curriculum_late_phase": 1.0,
      "epoch": 7.325e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.683853149414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47117333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.855130810292252,
      "rewrite6_current_pass_fraction": 0.8682109403906447,
      "rewrite6_full_passes": 0.0,
      "step": 96650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0207616,
      "curriculum_late_phase": 1.0,
      "epoch": 7.33e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.683853149414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0721,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47141333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.587188203782706,
      "rewrite6_current_pass_fraction": 0.8687075227356819,
      "rewrite6_full_passes": 0.0,
      "step": 96660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0232192,
      "curriculum_late_phase": 1.0,
      "epoch": 7.335e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.683853149414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0783,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47170666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.72144453520607,
      "rewrite6_current_pass_fraction": 0.8691797628089034,
      "rewrite6_full_passes": 0.0,
      "step": 96670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0256768,
      "curriculum_late_phase": 1.0,
      "epoch": 7.34e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.683853149414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4720666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.216941714202072,
      "rewrite6_current_pass_fraction": 0.8697737142412028,
      "rewrite6_full_passes": 0.0,
      "step": 96680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0281344,
      "curriculum_late_phase": 1.0,
      "epoch": 7.345e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.683853149414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47241333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.216941714202072,
      "rewrite6_current_pass_fraction": 0.8703725341278651,
      "rewrite6_full_passes": 0.0,
      "step": 96690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.030592,
      "curriculum_late_phase": 1.0,
      "epoch": 7.35e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 16.683853149414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0838,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47264,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.841241620144956,
      "rewrite6_current_pass_fraction": 0.8709616171058012,
      "rewrite6_full_passes": 0.0,
      "step": 96700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0330496,
      "curriculum_late_phase": 1.0,
      "epoch": 7.355e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.683853149414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0491,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4729333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09634916170988,
      "rewrite6_current_pass_fraction": 0.871531226266285,
      "rewrite6_full_passes": 0.0,
      "step": 96710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0355072,
      "curriculum_late_phase": 1.0,
      "epoch": 7.36e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.97951889038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47330666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.151271037028916,
      "rewrite6_current_pass_fraction": 0.8721251776985842,
      "rewrite6_full_passes": 0.0,
      "step": 96720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0379648,
      "curriculum_late_phase": 1.0,
      "epoch": 7.365e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.981880187988281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47357333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.8726607076785262,
      "rewrite6_full_passes": 0.0,
      "step": 96730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0404224,
      "curriculum_late_phase": 1.0,
      "epoch": 7.37e-06,
      "grad_norm": 0.54296875,
      "host_rss_gib": 15.983573913574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47390666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.32969002438444,
      "rewrite6_current_pass_fraction": 0.8732838698370041,
      "rewrite6_full_passes": 0.0,
      "step": 96740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.04288,
      "curriculum_late_phase": 1.0,
      "epoch": 7.375e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.984294891357422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0264,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4742133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.622856509337947,
      "rewrite6_current_pass_fraction": 0.8738924266323927,
      "rewrite6_full_passes": 0.0,
      "step": 96750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0453376,
      "curriculum_late_phase": 1.0,
      "epoch": 7.38e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.984306335449219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0656,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4743866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.44732652334885,
      "rewrite6_current_pass_fraction": 0.8744961149734183,
      "rewrite6_full_passes": 0.0,
      "step": 96760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0477952,
      "curriculum_late_phase": 1.0,
      "epoch": 7.385e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.985218048095703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4748,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.89479322619514,
      "rewrite6_current_pass_fraction": 0.8752361200366108,
      "rewrite6_full_passes": 0.0,
      "step": 96770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0502528,
      "curriculum_late_phase": 1.0,
      "epoch": 7.39e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.986137390136719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0555,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4753866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.231798772765842,
      "rewrite6_current_pass_fraction": 0.8757911238340051,
      "rewrite6_full_passes": 0.0,
      "step": 96780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0527104,
      "curriculum_late_phase": 1.0,
      "epoch": 7.395e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.988418579101562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0723,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47572,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.591506073196012,
      "rewrite6_current_pass_fraction": 0.8764727074448404,
      "rewrite6_full_passes": 0.0,
      "step": 96790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.055168,
      "curriculum_late_phase": 1.0,
      "epoch": 7.4e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.988426208496094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47604,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.997428965207153,
      "rewrite6_current_pass_fraction": 0.8771153434207708,
      "rewrite6_full_passes": 0.0,
      "step": 96800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0576256,
      "curriculum_late_phase": 1.0,
      "epoch": 7.405e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.98843002319336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47630666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.8777141633074331,
      "rewrite6_full_passes": 0.0,
      "step": 96810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0600832,
      "curriculum_late_phase": 1.0,
      "epoch": 7.41e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.99428939819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0168,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47657333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.425824354396642,
      "rewrite6_current_pass_fraction": 0.8783178516484587,
      "rewrite6_full_passes": 0.0,
      "step": 96820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0625408,
      "curriculum_late_phase": 1.0,
      "epoch": 7.415e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.99428939819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4769466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.8788582500827637,
      "rewrite6_full_passes": 0.0,
      "step": 96830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0649984,
      "curriculum_late_phase": 1.0,
      "epoch": 7.42e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.99428939819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0598,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4772666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.323292077117205,
      "rewrite6_current_pass_fraction": 0.8794619384237892,
      "rewrite6_full_passes": 0.0,
      "step": 96840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.067456,
      "curriculum_late_phase": 1.0,
      "epoch": 7.425e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.99428939819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4776,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.8800315475842729,
      "rewrite6_full_passes": 0.0,
      "step": 96850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0699136,
      "curriculum_late_phase": 1.0,
      "epoch": 7.43e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.99428939819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.097,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47786666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.13145721856217,
      "rewrite6_current_pass_fraction": 0.8805670775642149,
      "rewrite6_full_passes": 0.0,
      "step": 96860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0723712,
      "curriculum_late_phase": 1.0,
      "epoch": 7.435e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.99428939819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47817333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.07737098892312,
      "rewrite6_current_pass_fraction": 0.8812291873575977,
      "rewrite6_full_passes": 0.0,
      "step": 96870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0748288,
      "curriculum_late_phase": 1.0,
      "epoch": 7.44e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.994434356689453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0497,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4783733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10901076930934,
      "rewrite6_current_pass_fraction": 0.8817598488831766,
      "rewrite6_full_passes": 0.0,
      "step": 96880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0772864,
      "curriculum_late_phase": 1.0,
      "epoch": 7.445e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.994434356689453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.068,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47862666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.498861924749583,
      "rewrite6_current_pass_fraction": 0.8823197211349341,
      "rewrite6_full_passes": 0.0,
      "step": 96890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.079744,
      "curriculum_late_phase": 1.0,
      "epoch": 7.45e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.994434356689453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.47889333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.986932874965774,
      "rewrite6_current_pass_fraction": 0.8829039356585071,
      "rewrite6_full_passes": 0.0,
      "step": 96900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0822016,
      "curriculum_late_phase": 1.0,
      "epoch": 7.455e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.994697570800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0766,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4792133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.684549449204887,
      "rewrite6_current_pass_fraction": 0.8835806508149793,
      "rewrite6_full_passes": 0.0,
      "step": 96910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0846592,
      "curriculum_late_phase": 1.0,
      "epoch": 7.46e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.994697570800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0775,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4794933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.704074328586966,
      "rewrite6_current_pass_fraction": 0.8841551284298261,
      "rewrite6_full_passes": 0.0,
      "step": 96920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0871168,
      "curriculum_late_phase": 1.0,
      "epoch": 7.465e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.994697570800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4798266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.936624630107648,
      "rewrite6_current_pass_fraction": 0.884641973866137,
      "rewrite6_full_passes": 0.0,
      "step": 96930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0895744,
      "curriculum_late_phase": 1.0,
      "epoch": 7.47e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.994697570800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0562,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48028,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.246666234911444,
      "rewrite6_current_pass_fraction": 0.8853040836595197,
      "rewrite6_full_passes": 0.0,
      "step": 96940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.092032,
      "curriculum_late_phase": 1.0,
      "epoch": 7.475e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.994697570800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0448,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48057333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.005829616811546,
      "rewrite6_current_pass_fraction": 0.8859175089092715,
      "rewrite6_full_passes": 0.0,
      "step": 96950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0944896,
      "curriculum_late_phase": 1.0,
      "epoch": 7.48e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.994697570800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48078666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.007930304805882,
      "rewrite6_current_pass_fraction": 0.8865552764308388,
      "rewrite6_full_passes": 0.0,
      "step": 96960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0969472,
      "curriculum_late_phase": 1.0,
      "epoch": 7.485e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.994697570800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48106666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.132243456791066,
      "rewrite6_current_pass_fraction": 0.887193043952406,
      "rewrite6_full_passes": 0.0,
      "step": 96970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.0994048,
      "curriculum_late_phase": 1.0,
      "epoch": 7.49e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.994697570800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48156,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.035258373598133,
      "rewrite6_current_pass_fraction": 0.8878210745652471,
      "rewrite6_full_passes": 0.0,
      "step": 96980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1018624,
      "curriculum_late_phase": 1.0,
      "epoch": 7.495e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.994697570800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0729,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48186666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.604464864088442,
      "rewrite6_current_pass_fraction": 0.8884296313606356,
      "rewrite6_full_passes": 0.0,
      "step": 96990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.10432,
      "curriculum_late_phase": 1.0,
      "epoch": 7.5e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.994712829589844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48217333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.8889262137056727,
      "rewrite6_full_passes": 0.0,
      "step": 97000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 7.5e-06,
      "eval_loss": 3.082991123199463,
      "eval_perplexity": 21.823581889739476,
      "eval_runtime": 0.0494,
      "eval_samples_per_second": 648.417,
      "eval_steps_per_second": 20.263,
      "step": 97000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1067776,
      "curriculum_late_phase": 1.0,
      "epoch": 7.505e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.068893432617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48248,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.913607005029263,
      "rewrite6_current_pass_fraction": 0.8894812175030671,
      "rewrite6_full_passes": 0.0,
      "step": 97010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1092352,
      "curriculum_late_phase": 1.0,
      "epoch": 7.51e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.068893432617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0667,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48284,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.470931562916125,
      "rewrite6_current_pass_fraction": 0.8900800373897295,
      "rewrite6_full_passes": 0.0,
      "step": 97020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1116928,
      "curriculum_late_phase": 1.0,
      "epoch": 7.515e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.068893432617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48318666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.007930304805882,
      "rewrite6_current_pass_fraction": 0.890639909641487,
      "rewrite6_full_passes": 0.0,
      "step": 97030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1141504,
      "curriculum_late_phase": 1.0,
      "epoch": 7.52e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.068897247314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0705,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48350666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.552676319526597,
      "rewrite6_current_pass_fraction": 0.8913507039785009,
      "rewrite6_full_passes": 0.0,
      "step": 97040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.116608,
      "curriculum_late_phase": 1.0,
      "epoch": 7.525e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.068897247314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0549,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48374666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.219063514461734,
      "rewrite6_current_pass_fraction": 0.8919154446846215,
      "rewrite6_full_passes": 0.0,
      "step": 97050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1190656,
      "curriculum_late_phase": 1.0,
      "epoch": 7.53e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.068897247314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4841066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.8924947907538314,
      "rewrite6_full_passes": 0.0,
      "step": 97060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1215232,
      "curriculum_late_phase": 1.0,
      "epoch": 7.535e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.068897247314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0185,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48450666666666664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.460577787847765,
      "rewrite6_current_pass_fraction": 0.8930595314599521,
      "rewrite6_full_passes": 0.0,
      "step": 97070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1239808,
      "curriculum_late_phase": 1.0,
      "epoch": 7.54e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.06983184814453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48488,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23604555718468,
      "rewrite6_current_pass_fraction": 0.893726509707698,
      "rewrite6_full_passes": 0.0,
      "step": 97080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1264384,
      "curriculum_late_phase": 1.0,
      "epoch": 7.545e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.06983184814453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48524,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.8942863819594555,
      "rewrite6_full_passes": 0.0,
      "step": 97090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.128896,
      "curriculum_late_phase": 1.0,
      "epoch": 7.55e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.75516128540039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48552,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.8948705964830286,
      "rewrite6_full_passes": 0.0,
      "step": 97100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1313536,
      "curriculum_late_phase": 1.0,
      "epoch": 7.555e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.678421020507812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0586,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4858,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.297719473255697,
      "rewrite6_current_pass_fraction": 0.8954402056435123,
      "rewrite6_full_passes": 0.0,
      "step": 97110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1338112,
      "curriculum_late_phase": 1.0,
      "epoch": 7.56e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0362,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48616,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.825954055651632,
      "rewrite6_current_pass_fraction": 0.8960292886214485,
      "rewrite6_full_passes": 0.0,
      "step": 97120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1362688,
      "curriculum_late_phase": 1.0,
      "epoch": 7.565e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48646666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.911515748893304,
      "rewrite6_current_pass_fraction": 0.8966329769624739,
      "rewrite6_full_passes": 0.0,
      "step": 97130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1387264,
      "curriculum_late_phase": 1.0,
      "epoch": 7.57e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0618,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4868266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.365981336300866,
      "rewrite6_current_pass_fraction": 0.8973340343907616,
      "rewrite6_full_passes": 0.0,
      "step": 97140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.141184,
      "curriculum_late_phase": 1.0,
      "epoch": 7.575e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0448,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48718666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.005829616811546,
      "rewrite6_current_pass_fraction": 0.8979377227317871,
      "rewrite6_full_passes": 0.0,
      "step": 97150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1436416,
      "curriculum_late_phase": 1.0,
      "epoch": 7.58e-06,
      "grad_norm": 0.54296875,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48745333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.018436896387286,
      "rewrite6_current_pass_fraction": 0.8985268057097233,
      "rewrite6_full_passes": 0.0,
      "step": 97160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1460992,
      "curriculum_late_phase": 1.0,
      "epoch": 7.585e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0491,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4876933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09634916170988,
      "rewrite6_current_pass_fraction": 0.8991743101400167,
      "rewrite6_full_passes": 0.0,
      "step": 97170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1485568,
      "curriculum_late_phase": 1.0,
      "epoch": 7.59e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0594,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48808,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3147644659223,
      "rewrite6_current_pass_fraction": 0.8997001032112325,
      "rewrite6_full_passes": 0.0,
      "step": 97180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1510144,
      "curriculum_late_phase": 1.0,
      "epoch": 7.595e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48833333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86556098302329,
      "rewrite6_current_pass_fraction": 0.9003865552764309,
      "rewrite6_full_passes": 0.0,
      "step": 97190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.153472,
      "curriculum_late_phase": 1.0,
      "epoch": 7.6e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48857333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10689997377394,
      "rewrite6_current_pass_fraction": 0.901107086522171,
      "rewrite6_full_passes": 0.0,
      "step": 97200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1559296,
      "curriculum_late_phase": 1.0,
      "epoch": 7.605e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48893333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.9016620903195653,
      "rewrite6_full_passes": 0.0,
      "step": 97210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1583872,
      "curriculum_late_phase": 1.0,
      "epoch": 7.61e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48933333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20845663464543,
      "rewrite6_current_pass_fraction": 0.9022073572082335,
      "rewrite6_full_passes": 0.0,
      "step": 97220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1608448,
      "curriculum_late_phase": 1.0,
      "epoch": 7.615e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0468,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48962666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.04788331572619,
      "rewrite6_current_pass_fraction": 0.9028207824579852,
      "rewrite6_full_passes": 0.0,
      "step": 97230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1633024,
      "curriculum_late_phase": 1.0,
      "epoch": 7.62e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0836,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.48989333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.836873808616637,
      "rewrite6_current_pass_fraction": 0.9033806547097427,
      "rewrite6_full_passes": 0.0,
      "step": 97240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.16576,
      "curriculum_late_phase": 1.0,
      "epoch": 7.625e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4902533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.07737098892312,
      "rewrite6_current_pass_fraction": 0.9040525014118518,
      "rewrite6_full_passes": 0.0,
      "step": 97250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1682176,
      "curriculum_late_phase": 1.0,
      "epoch": 7.63e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49058666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.9046659266616035,
      "rewrite6_full_passes": 0.0,
      "step": 97260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1706752,
      "curriculum_late_phase": 1.0,
      "epoch": 7.635e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0879,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4908133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.930974537567433,
      "rewrite6_current_pass_fraction": 0.9051868512784561,
      "rewrite6_full_passes": 0.0,
      "step": 97270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1731328,
      "curriculum_late_phase": 1.0,
      "epoch": 7.64e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49116,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.299849351695165,
      "rewrite6_current_pass_fraction": 0.9058781717980176,
      "rewrite6_full_passes": 0.0,
      "step": 97280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1755904,
      "curriculum_late_phase": 1.0,
      "epoch": 7.645e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0466,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49158666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.043674159992648,
      "rewrite6_current_pass_fraction": 0.9065548869544897,
      "rewrite6_full_passes": 0.0,
      "step": 97290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.178048,
      "curriculum_late_phase": 1.0,
      "epoch": 7.65e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0272,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49189333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.639361395619666,
      "rewrite6_current_pass_fraction": 0.9071780491129676,
      "rewrite6_full_passes": 0.0,
      "step": 97300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1805056,
      "curriculum_late_phase": 1.0,
      "epoch": 7.655e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.064,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49222666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.413038238853925,
      "rewrite6_current_pass_fraction": 0.9078012112714455,
      "rewrite6_full_passes": 0.0,
      "step": 97310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1829632,
      "curriculum_late_phase": 1.0,
      "epoch": 7.66e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49248,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06894172649243,
      "rewrite6_current_pass_fraction": 0.9083610835232031,
      "rewrite6_full_passes": 0.0,
      "step": 97320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1854208,
      "curriculum_late_phase": 1.0,
      "epoch": 7.665e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 16.957454681396484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0712,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49278666666666665,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.567768474588277,
      "rewrite6_current_pass_fraction": 0.9089160873205975,
      "rewrite6_full_passes": 0.0,
      "step": 97330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1878784,
      "curriculum_late_phase": 1.0,
      "epoch": 7.67e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.95760726928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0227,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4932533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.54669292976553,
      "rewrite6_current_pass_fraction": 0.9095343810247123,
      "rewrite6_full_passes": 0.0,
      "step": 97340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.190336,
      "curriculum_late_phase": 1.0,
      "epoch": 7.675e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.95760726928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4935866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.17878556994849,
      "rewrite6_current_pass_fraction": 0.9101770170006427,
      "rewrite6_full_passes": 0.0,
      "step": 97350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1927936,
      "curriculum_late_phase": 1.0,
      "epoch": 7.68e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.95760726928711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49392,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.87808407616535,
      "rewrite6_current_pass_fraction": 0.9108537321571147,
      "rewrite6_full_passes": 0.0,
      "step": 97360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1952512,
      "curriculum_late_phase": 1.0,
      "epoch": 7.685e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.957927703857422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0429,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4942,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.9659564320603,
      "rewrite6_current_pass_fraction": 0.9114282097719616,
      "rewrite6_full_passes": 0.0,
      "step": 97370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.1977088,
      "curriculum_late_phase": 1.0,
      "epoch": 7.69e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.957927703857422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0217,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4946,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.52615650675864,
      "rewrite6_current_pass_fraction": 0.9120416350217133,
      "rewrite6_full_passes": 0.0,
      "step": 97380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2001664,
      "curriculum_late_phase": 1.0,
      "epoch": 7.695e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.957927703857422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49493333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.882260110570105,
      "rewrite6_current_pass_fraction": 0.9125479542754766,
      "rewrite6_full_passes": 0.0,
      "step": 97390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.202624,
      "curriculum_late_phase": 1.0,
      "epoch": 7.7e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.957927703857422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0382,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49533333333333335,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86764764345287,
      "rewrite6_current_pass_fraction": 0.9131857217970438,
      "rewrite6_full_passes": 0.0,
      "step": 97400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2050816,
      "curriculum_late_phase": 1.0,
      "epoch": 7.705e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.957927703857422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0619,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4957066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.36811804126797,
      "rewrite6_current_pass_fraction": 0.913818620864248,
      "rewrite6_full_passes": 0.0,
      "step": 97410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2075392,
      "curriculum_late_phase": 1.0,
      "epoch": 7.71e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.957927703857422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49602666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11534442254061,
      "rewrite6_current_pass_fraction": 0.9144271776596367,
      "rewrite6_full_passes": 0.0,
      "step": 97420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2099968,
      "curriculum_late_phase": 1.0,
      "epoch": 7.715e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.957927703857422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49637333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.9149870499113941,
      "rewrite6_full_passes": 0.0,
      "step": 97430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2124544,
      "curriculum_late_phase": 1.0,
      "epoch": 7.72e-06,
      "grad_norm": 0.54296875,
      "host_rss_gib": 16.957927703857422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13435678680148,
      "rewrite6_current_pass_fraction": 0.9155420537087885,
      "rewrite6_full_passes": 0.0,
      "step": 97440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.214912,
      "curriculum_late_phase": 1.0,
      "epoch": 7.725e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.958049774169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0472,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49692,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.05630415310768,
      "rewrite6_current_pass_fraction": 0.9160629783256412,
      "rewrite6_full_passes": 0.0,
      "step": 97450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2173696,
      "curriculum_late_phase": 1.0,
      "epoch": 7.73e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.958049774169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0239,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4974133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.57136376081938,
      "rewrite6_current_pass_fraction": 0.9166277190317618,
      "rewrite6_full_passes": 0.0,
      "step": 97460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2198272,
      "curriculum_late_phase": 1.0,
      "epoch": 7.735e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.958049774169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49781333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.828036755190443,
      "rewrite6_current_pass_fraction": 0.9172508811902397,
      "rewrite6_full_passes": 0.0,
      "step": 97470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2222848,
      "curriculum_late_phase": 1.0,
      "epoch": 7.74e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.958049774169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49821333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.02684595284501,
      "rewrite6_current_pass_fraction": 0.9178497010769021,
      "rewrite6_full_passes": 0.0,
      "step": 97480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2247424,
      "curriculum_late_phase": 1.0,
      "epoch": 7.745e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.958049774169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0586,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4985866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.297719473255697,
      "rewrite6_current_pass_fraction": 0.9183900995112072,
      "rewrite6_full_passes": 0.0,
      "step": 97490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2272,
      "curriculum_late_phase": 1.0,
      "epoch": 7.75e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.958057403564453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0281,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4988533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.65794518232533,
      "rewrite6_current_pass_fraction": 0.9189402348542385,
      "rewrite6_full_passes": 0.0,
      "step": 97500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 7.75e-06,
      "eval_loss": 3.0828518867492676,
      "eval_perplexity": 21.82054346320137,
      "eval_runtime": 0.0493,
      "eval_samples_per_second": 648.755,
      "eval_steps_per_second": 20.274,
      "step": 97500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2296576,
      "curriculum_late_phase": 1.0,
      "epoch": 7.755e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 17.022010803222656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4992,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.9194611594710911,
      "rewrite6_full_passes": 0.0,
      "step": 97510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2321152,
      "curriculum_late_phase": 1.0,
      "epoch": 7.76e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 17.022010803222656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.4995866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.859302253584733,
      "rewrite6_current_pass_fraction": 0.9200794531752059,
      "rewrite6_full_passes": 0.0,
      "step": 97520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2345728,
      "curriculum_late_phase": 1.0,
      "epoch": 7.765e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 17.022010803222656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0758,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.49989333333333336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.667208746851305,
      "rewrite6_current_pass_fraction": 0.9206149831551479,
      "rewrite6_full_passes": 0.0,
      "step": 97530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2370304,
      "curriculum_late_phase": 1.0,
      "epoch": 7.77e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 17.022464752197266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0636,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5000666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.40447473637306,
      "rewrite6_current_pass_fraction": 0.9211845923156317,
      "rewrite6_full_passes": 0.0,
      "step": 97540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.239488,
      "curriculum_late_phase": 1.0,
      "epoch": 7.775e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 17.022464752197266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50036,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.896882809995212,
      "rewrite6_current_pass_fraction": 0.9217347276586629,
      "rewrite6_full_passes": 0.0,
      "step": 97550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2419456,
      "curriculum_late_phase": 1.0,
      "epoch": 7.78e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 17.022464752197266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.078,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5006133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.71492907921277,
      "rewrite6_current_pass_fraction": 0.9222994683647836,
      "rewrite6_full_passes": 0.0,
      "step": 97560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2444032,
      "curriculum_late_phase": 1.0,
      "epoch": 7.785e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 17.022464752197266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0605,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5009866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.338223606996944,
      "rewrite6_current_pass_fraction": 0.9228349983447255,
      "rewrite6_full_passes": 0.0,
      "step": 97570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2468608,
      "curriculum_late_phase": 1.0,
      "epoch": 7.79e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.42340850830078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50132,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.039465846006077,
      "rewrite6_current_pass_fraction": 0.9233267122353995,
      "rewrite6_full_passes": 0.0,
      "step": 97580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2493184,
      "curriculum_late_phase": 1.0,
      "epoch": 7.795e-06,
      "grad_norm": 0.54296875,
      "host_rss_gib": 16.427623748779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5015733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.265796842004743,
      "rewrite6_current_pass_fraction": 0.9238719791240677,
      "rewrite6_full_passes": 0.0,
      "step": 97590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.251776,
      "curriculum_late_phase": 1.0,
      "epoch": 7.8e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.43047332763672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5017866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.9244172460127359,
      "rewrite6_full_passes": 0.0,
      "step": 97600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2542336,
      "curriculum_late_phase": 1.0,
      "epoch": 7.805e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.4324951171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.074,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50208,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.628242850934082,
      "rewrite6_current_pass_fraction": 0.9249965920819458,
      "rewrite6_full_passes": 0.0,
      "step": 97610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2566912,
      "curriculum_late_phase": 1.0,
      "epoch": 7.81e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.4324951171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5024133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08158688469643,
      "rewrite6_current_pass_fraction": 0.9256197542404238,
      "rewrite6_full_passes": 0.0,
      "step": 97620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2591488,
      "curriculum_late_phase": 1.0,
      "epoch": 7.815e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.432716369628906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.035,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50272,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.800977899475697,
      "rewrite6_current_pass_fraction": 0.9262185741270861,
      "rewrite6_full_passes": 0.0,
      "step": 97630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2616064,
      "curriculum_late_phase": 1.0,
      "epoch": 7.82e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.433048248291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5029866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.193615909860856,
      "rewrite6_current_pass_fraction": 0.9269050261922844,
      "rewrite6_full_passes": 0.0,
      "step": 97640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.264064,
      "curriculum_late_phase": 1.0,
      "epoch": 7.825e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.433048248291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0383,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5034666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.869734512558935,
      "rewrite6_current_pass_fraction": 0.9275281883507625,
      "rewrite6_full_passes": 0.0,
      "step": 97650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2665216,
      "curriculum_late_phase": 1.0,
      "epoch": 7.83e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.43305206298828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5039333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.828036755190443,
      "rewrite6_current_pass_fraction": 0.9280296391501626,
      "rewrite6_full_passes": 0.0,
      "step": 97660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2689792,
      "curriculum_late_phase": 1.0,
      "epoch": 7.835e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.433795928955078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0686,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50424,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.51176511247365,
      "rewrite6_current_pass_fraction": 0.9286235905824619,
      "rewrite6_full_passes": 0.0,
      "step": 97670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2714368,
      "curriculum_late_phase": 1.0,
      "epoch": 7.84e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.434066772460938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.059,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50452,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.306240265089755,
      "rewrite6_current_pass_fraction": 0.9292321473778504,
      "rewrite6_full_passes": 0.0,
      "step": 97680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2738944,
      "curriculum_late_phase": 1.0,
      "epoch": 7.845e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.434066772460938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5047733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.828036755190443,
      "rewrite6_current_pass_fraction": 0.9298260988101498,
      "rewrite6_full_passes": 0.0,
      "step": 97690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.276352,
      "curriculum_late_phase": 1.0,
      "epoch": 7.85e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.434066772460938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0633,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5050933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.398054357057195,
      "rewrite6_current_pass_fraction": 0.930463866331717,
      "rewrite6_full_passes": 0.0,
      "step": 97700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2788096,
      "curriculum_late_phase": 1.0,
      "epoch": 7.855e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.434066772460938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5054533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.031051742600535,
      "rewrite6_current_pass_fraction": 0.9309847909485697,
      "rewrite6_full_passes": 0.0,
      "step": 97710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2812672,
      "curriculum_late_phase": 1.0,
      "epoch": 7.86e-06,
      "grad_norm": 0.5234375,
      "host_rss_gib": 16.434066772460938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5058133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.9316420322875894,
      "rewrite6_full_passes": 0.0,
      "step": 97720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2837248,
      "curriculum_late_phase": 1.0,
      "epoch": 7.865e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.434131622314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0491,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50616,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09634916170988,
      "rewrite6_current_pass_fraction": 0.9323041420809721,
      "rewrite6_full_passes": 0.0,
      "step": 97730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2861824,
      "curriculum_late_phase": 1.0,
      "epoch": 7.87e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.434131622314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0615,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5064666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.359572503273004,
      "rewrite6_current_pass_fraction": 0.9329808572374443,
      "rewrite6_full_passes": 0.0,
      "step": 97740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.28864,
      "curriculum_late_phase": 1.0,
      "epoch": 7.875e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.434131622314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0683,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50676,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50531255087254,
      "rewrite6_current_pass_fraction": 0.9336040193959222,
      "rewrite6_full_passes": 0.0,
      "step": 97750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2910976,
      "curriculum_late_phase": 1.0,
      "epoch": 7.88e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.434131622314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0354,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50712,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.809299954935618,
      "rewrite6_current_pass_fraction": 0.9342174446456739,
      "rewrite6_full_passes": 0.0,
      "step": 97760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2935552,
      "curriculum_late_phase": 1.0,
      "epoch": 7.885e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.434131622314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0631,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50748,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.393775174118343,
      "rewrite6_current_pass_fraction": 0.934850343712878,
      "rewrite6_full_passes": 0.0,
      "step": 97770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2960128,
      "curriculum_late_phase": 1.0,
      "epoch": 7.89e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.434131622314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0406,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5077866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.917790144730294,
      "rewrite6_current_pass_fraction": 0.9354102159646356,
      "rewrite6_full_passes": 0.0,
      "step": 97780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.2984704,
      "curriculum_late_phase": 1.0,
      "epoch": 7.895e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.434131622314453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0282,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5080933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.660011080136737,
      "rewrite6_current_pass_fraction": 0.935877587583494,
      "rewrite6_full_passes": 0.0,
      "step": 97790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.300928,
      "curriculum_late_phase": 1.0,
      "epoch": 7.9e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.43413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50832,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35530101593532,
      "rewrite6_current_pass_fraction": 0.9364520651983408,
      "rewrite6_full_passes": 0.0,
      "step": 97800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3033856,
      "curriculum_late_phase": 1.0,
      "epoch": 7.905e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.43413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0315,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5087066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.728301734306807,
      "rewrite6_current_pass_fraction": 0.9370265428131876,
      "rewrite6_full_passes": 0.0,
      "step": 97810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3058432,
      "curriculum_late_phase": 1.0,
      "epoch": 7.91e-06,
      "grad_norm": 0.5390625,
      "host_rss_gib": 16.43413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0594,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50892,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3147644659223,
      "rewrite6_current_pass_fraction": 0.9377032579696598,
      "rewrite6_full_passes": 0.0,
      "step": 97820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3083008,
      "curriculum_late_phase": 1.0,
      "epoch": 7.915e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.43413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50928,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.892703851343008,
      "rewrite6_current_pass_fraction": 0.9383604993086795,
      "rewrite6_full_passes": 0.0,
      "step": 97830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3107584,
      "curriculum_late_phase": 1.0,
      "epoch": 7.92e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.43413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.50956,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.986932874965774,
      "rewrite6_current_pass_fraction": 0.9389349769235263,
      "rewrite6_full_passes": 0.0,
      "step": 97840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.313216,
      "curriculum_late_phase": 1.0,
      "epoch": 7.925e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.43413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0405,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5099066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91569847030129,
      "rewrite6_current_pass_fraction": 0.9397090611672606,
      "rewrite6_full_passes": 0.0,
      "step": 97850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3156736,
      "curriculum_late_phase": 1.0,
      "epoch": 7.93e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.43413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.51024,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.9402786703277444,
      "rewrite6_full_passes": 0.0,
      "step": 97860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3181312,
      "curriculum_late_phase": 1.0,
      "epoch": 7.935e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.43413543701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5105066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.9408823586687698,
      "rewrite6_full_passes": 0.0,
      "step": 97870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3205888,
      "curriculum_late_phase": 1.0,
      "epoch": 7.94e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.43505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0346,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5108933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.792659172172286,
      "rewrite6_current_pass_fraction": 0.941427625557438,
      "rewrite6_full_passes": 0.0,
      "step": 97880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3230464,
      "curriculum_late_phase": 1.0,
      "epoch": 7.945e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 16.43505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5112533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.759417517911764,
      "rewrite6_current_pass_fraction": 0.942050787715916,
      "rewrite6_full_passes": 0.0,
      "step": 97890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.325504,
      "curriculum_late_phase": 1.0,
      "epoch": 7.95e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 16.43505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0154,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5116266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.397248208270213,
      "rewrite6_current_pass_fraction": 0.9427177659636619,
      "rewrite6_full_passes": 0.0,
      "step": 97900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3279616,
      "curriculum_late_phase": 1.0,
      "epoch": 7.955e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 16.43505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0766,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.512,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.684549449204887,
      "rewrite6_current_pass_fraction": 0.9433360596677767,
      "rewrite6_full_passes": 0.0,
      "step": 97910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3304192,
      "curriculum_late_phase": 1.0,
      "epoch": 7.96e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.43505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5123466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.221185526912038,
      "rewrite6_current_pass_fraction": 0.9438764581020818,
      "rewrite6_full_passes": 0.0,
      "step": 97920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3328768,
      "curriculum_late_phase": 1.0,
      "epoch": 7.965e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.43505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5127066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.9443827773558451,
      "rewrite6_full_passes": 0.0,
      "step": 97930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3353344,
      "curriculum_late_phase": 1.0,
      "epoch": 7.97e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 16.43505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.51308,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.9449572549706919,
      "rewrite6_full_passes": 0.0,
      "step": 97940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.337792,
      "curriculum_late_phase": 1.0,
      "epoch": 7.975e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.43505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5133066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.945633970127164,
      "rewrite6_full_passes": 0.0,
      "step": 97950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3402496,
      "curriculum_late_phase": 1.0,
      "epoch": 7.98e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.456295013427734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5136266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.9461354209265642,
      "rewrite6_full_passes": 0.0,
      "step": 97960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3427072,
      "curriculum_late_phase": 1.0,
      "epoch": 7.985e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.456295013427734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5138666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01003120287951,
      "rewrite6_current_pass_fraction": 0.9467439777219528,
      "rewrite6_full_passes": 0.0,
      "step": 97970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3451648,
      "curriculum_late_phase": 1.0,
      "epoch": 7.99e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.456295013427734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5142,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.9473427976086152,
      "rewrite6_full_passes": 0.0,
      "step": 97980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3476224,
      "curriculum_late_phase": 1.0,
      "epoch": 7.995e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.456295013427734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5146266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.9478880644972834,
      "rewrite6_full_passes": 0.0,
      "step": 97990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.35008,
      "curriculum_late_phase": 1.0,
      "epoch": 8e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.456295013427734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.052,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5150266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.157617370242413,
      "rewrite6_current_pass_fraction": 0.9484381998403147,
      "rewrite6_full_passes": 0.0,
      "step": 98000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 8e-06,
      "eval_loss": 3.0826072692871094,
      "eval_perplexity": 21.81520643002886,
      "eval_runtime": 0.0481,
      "eval_samples_per_second": 665.541,
      "eval_steps_per_second": 20.798,
      "step": 98000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3525376,
      "curriculum_late_phase": 1.0,
      "epoch": 8.005e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.759090423583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0441,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5154133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.991130681307407,
      "rewrite6_current_pass_fraction": 0.9490516250900664,
      "rewrite6_full_passes": 0.0,
      "step": 98010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3549952,
      "curriculum_late_phase": 1.0,
      "epoch": 8.01e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5157333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.144926607429838,
      "rewrite6_current_pass_fraction": 0.9495871550700083,
      "rewrite6_full_passes": 0.0,
      "step": 98020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3574528,
      "curriculum_late_phase": 1.0,
      "epoch": 8.015e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0702,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5159866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.54621148640419,
      "rewrite6_current_pass_fraction": 0.9501665011392183,
      "rewrite6_full_passes": 0.0,
      "step": 98030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3599104,
      "curriculum_late_phase": 1.0,
      "epoch": 8.02e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5162266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.60662541860077,
      "rewrite6_current_pass_fraction": 0.9506728203929816,
      "rewrite6_full_passes": 0.0,
      "step": 98040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.362368,
      "curriculum_late_phase": 1.0,
      "epoch": 8.025e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0604,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5166133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3360898913238,
      "rewrite6_current_pass_fraction": 0.951315456368912,
      "rewrite6_full_passes": 0.0,
      "step": 98050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3648256,
      "curriculum_late_phase": 1.0,
      "epoch": 8.03e-06,
      "grad_norm": 0.5234375,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5170533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.9519288816186637,
      "rewrite6_full_passes": 0.0,
      "step": 98060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3672832,
      "curriculum_late_phase": 1.0,
      "epoch": 8.035e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0298,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.51732,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.69309355678868,
      "rewrite6_current_pass_fraction": 0.9524400693267902,
      "rewrite6_full_passes": 0.0,
      "step": 98070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3697408,
      "curriculum_late_phase": 1.0,
      "epoch": 8.04e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0512,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5175866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14069804497869,
      "rewrite6_current_pass_fraction": 0.9529415201261904,
      "rewrite6_full_passes": 0.0,
      "step": 98080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3721984,
      "curriculum_late_phase": 1.0,
      "epoch": 8.045e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5178666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.989031673191437,
      "rewrite6_current_pass_fraction": 0.9535598138303052,
      "rewrite6_full_passes": 0.0,
      "step": 98090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.374656,
      "curriculum_late_phase": 1.0,
      "epoch": 8.05e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0311,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5181866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.720012071656146,
      "rewrite6_current_pass_fraction": 0.9541440283538782,
      "rewrite6_full_passes": 0.0,
      "step": 98100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3771136,
      "curriculum_late_phase": 1.0,
      "epoch": 8.055e-06,
      "grad_norm": 0.435546875,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.054,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5185066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.199974948441902,
      "rewrite6_current_pass_fraction": 0.9547964012385348,
      "rewrite6_full_passes": 0.0,
      "step": 98110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3795712,
      "curriculum_late_phase": 1.0,
      "epoch": 8.06e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5190133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.045778632630523,
      "rewrite6_current_pass_fraction": 0.9554585110319176,
      "rewrite6_full_passes": 0.0,
      "step": 98120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3820288,
      "curriculum_late_phase": 1.0,
      "epoch": 8.065e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0546,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.51936,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21269875016978,
      "rewrite6_current_pass_fraction": 0.9559940410118596,
      "rewrite6_full_passes": 0.0,
      "step": 98130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3844864,
      "curriculum_late_phase": 1.0,
      "epoch": 8.07e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0638,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5196666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.40875605943837,
      "rewrite6_current_pass_fraction": 0.9564711495394442,
      "rewrite6_full_passes": 0.0,
      "step": 98140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.386944,
      "curriculum_late_phase": 1.0,
      "epoch": 8.075e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5200266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.9571186539697377,
      "rewrite6_full_passes": 0.0,
      "step": 98150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3894016,
      "curriculum_late_phase": 1.0,
      "epoch": 8.08e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0455,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.52028,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.020538845172617,
      "rewrite6_current_pass_fraction": 0.9577466845825787,
      "rewrite6_full_passes": 0.0,
      "step": 98160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3918592,
      "curriculum_late_phase": 1.0,
      "epoch": 8.085e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0703,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5205866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.548366215287484,
      "rewrite6_current_pass_fraction": 0.9584428735565033,
      "rewrite6_full_passes": 0.0,
      "step": 98170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3943168,
      "curriculum_late_phase": 1.0,
      "epoch": 8.09e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0263,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.52092,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.620794326797853,
      "rewrite6_current_pass_fraction": 0.9590416934431657,
      "rewrite6_full_passes": 0.0,
      "step": 98180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.3967744,
      "curriculum_late_phase": 1.0,
      "epoch": 8.095e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5212666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14704120581873,
      "rewrite6_current_pass_fraction": 0.9596015656949232,
      "rewrite6_full_passes": 0.0,
      "step": 98190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.399232,
      "curriculum_late_phase": 1.0,
      "epoch": 8.1e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0289,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.52156,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.67447815077682,
      "rewrite6_current_pass_fraction": 0.9601127534030496,
      "rewrite6_full_passes": 0.0,
      "step": 98200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4016896,
      "curriculum_late_phase": 1.0,
      "epoch": 8.105e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.750377655029297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0694,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5217866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.5289814101645,
      "rewrite6_current_pass_fraction": 0.9607067048353489,
      "rewrite6_full_passes": 0.0,
      "step": 98210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4041472,
      "curriculum_late_phase": 1.0,
      "epoch": 8.11e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.52216,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.71586848361443,
      "rewrite6_current_pass_fraction": 0.9612617086327433,
      "rewrite6_full_passes": 0.0,
      "step": 98220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4066048,
      "curriculum_late_phase": 1.0,
      "epoch": 8.115e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0422,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5225333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.951285398018836,
      "rewrite6_current_pass_fraction": 0.9619822398784834,
      "rewrite6_full_passes": 0.0,
      "step": 98230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4090624,
      "curriculum_late_phase": 1.0,
      "epoch": 8.12e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5228,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.9626005335825982,
      "rewrite6_full_passes": 0.0,
      "step": 98240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.41152,
      "curriculum_late_phase": 1.0,
      "epoch": 8.125e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0217,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5230266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.52615650675864,
      "rewrite6_current_pass_fraction": 0.96321395883235,
      "rewrite6_full_passes": 0.0,
      "step": 98250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4139776,
      "curriculum_late_phase": 1.0,
      "epoch": 8.13e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0605,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5232533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.338223606996944,
      "rewrite6_current_pass_fraction": 0.9637884364471967,
      "rewrite6_full_passes": 0.0,
      "step": 98260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4164352,
      "curriculum_late_phase": 1.0,
      "epoch": 8.135e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0472,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5235866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.05630415310768,
      "rewrite6_current_pass_fraction": 0.9644164670600378,
      "rewrite6_full_passes": 0.0,
      "step": 98270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4188928,
      "curriculum_late_phase": 1.0,
      "epoch": 8.14e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5239466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34676060372526,
      "rewrite6_current_pass_fraction": 0.9650347607641526,
      "rewrite6_full_passes": 0.0,
      "step": 98280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4213504,
      "curriculum_late_phase": 1.0,
      "epoch": 8.145e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0648,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5243066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.430175523444856,
      "rewrite6_current_pass_fraction": 0.9655313431091896,
      "rewrite6_full_passes": 0.0,
      "step": 98290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.423808,
      "curriculum_late_phase": 1.0,
      "epoch": 8.15e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5246533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.730374668125208,
      "rewrite6_current_pass_fraction": 0.966125294541489,
      "rewrite6_full_passes": 0.0,
      "step": 98300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4262656,
      "curriculum_late_phase": 1.0,
      "epoch": 8.155e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0379,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5248666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.861388288110085,
      "rewrite6_current_pass_fraction": 0.9666802983388834,
      "rewrite6_full_passes": 0.0,
      "step": 98310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4287232,
      "curriculum_late_phase": 1.0,
      "epoch": 8.16e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0312,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5252933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.72208417646683,
      "rewrite6_current_pass_fraction": 0.9672937235886351,
      "rewrite6_full_passes": 0.0,
      "step": 98320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4311808,
      "curriculum_late_phase": 1.0,
      "epoch": 8.165e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0658,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5256,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45161641762865,
      "rewrite6_current_pass_fraction": 0.9678779381122081,
      "rewrite6_full_passes": 0.0,
      "step": 98330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4336384,
      "curriculum_late_phase": 1.0,
      "epoch": 8.17e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0616,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5258933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.361708567324747,
      "rewrite6_current_pass_fraction": 0.9683939942746976,
      "rewrite6_full_passes": 0.0,
      "step": 98340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.436096,
      "curriculum_late_phase": 1.0,
      "epoch": 8.175e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0625,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.52616,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.380942759123343,
      "rewrite6_current_pass_fraction": 0.9690317617962649,
      "rewrite6_full_passes": 0.0,
      "step": 98350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4385536,
      "curriculum_late_phase": 1.0,
      "epoch": 8.18e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.52648,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.049988209300682,
      "rewrite6_current_pass_fraction": 0.9696257132285642,
      "rewrite6_full_passes": 0.0,
      "step": 98360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4410112,
      "curriculum_late_phase": 1.0,
      "epoch": 8.185e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0516,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5268933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14915601567805,
      "rewrite6_current_pass_fraction": 0.9701320324823275,
      "rewrite6_full_passes": 0.0,
      "step": 98370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4434688,
      "curriculum_late_phase": 1.0,
      "epoch": 8.19e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5272666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.32969002438444,
      "rewrite6_current_pass_fraction": 0.9707405892777161,
      "rewrite6_full_passes": 0.0,
      "step": 98380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4459264,
      "curriculum_late_phase": 1.0,
      "epoch": 8.195e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0577,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.52752,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27856014871906,
      "rewrite6_current_pass_fraction": 0.9713394091643784,
      "rewrite6_full_passes": 0.0,
      "step": 98390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.448384,
      "curriculum_late_phase": 1.0,
      "epoch": 8.2e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0565,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5278666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25304119097752,
      "rewrite6_current_pass_fraction": 0.9719090183248622,
      "rewrite6_full_passes": 0.0,
      "step": 98400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4508416,
      "curriculum_late_phase": 1.0,
      "epoch": 8.205e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0567,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5281066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.257292224304877,
      "rewrite6_current_pass_fraction": 0.9724688905766198,
      "rewrite6_full_passes": 0.0,
      "step": 98410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4532992,
      "curriculum_late_phase": 1.0,
      "epoch": 8.21e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5283333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.049988209300682,
      "rewrite6_current_pass_fraction": 0.9730336312827403,
      "rewrite6_full_passes": 0.0,
      "step": 98420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4557568,
      "curriculum_late_phase": 1.0,
      "epoch": 8.215e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0572,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5286133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.267923528021466,
      "rewrite6_current_pass_fraction": 0.97369087262176,
      "rewrite6_full_passes": 0.0,
      "step": 98430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4582144,
      "curriculum_late_phase": 1.0,
      "epoch": 8.22e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 16.75045394897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0634,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5289866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.400194269486743,
      "rewrite6_current_pass_fraction": 0.9743091663258748,
      "rewrite6_full_passes": 0.0,
      "step": 98440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.460672,
      "curriculum_late_phase": 1.0,
      "epoch": 8.225e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.13532257080078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5293066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.9749518023018052,
      "rewrite6_full_passes": 0.0,
      "step": 98450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4631296,
      "curriculum_late_phase": 1.0,
      "epoch": 8.23e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.135536193847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0351,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5296133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.803058101273997,
      "rewrite6_current_pass_fraction": 0.9755749644602831,
      "rewrite6_full_passes": 0.0,
      "step": 98460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4655872,
      "curriculum_late_phase": 1.0,
      "epoch": 8.235e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.13597869873047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0549,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5299733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.219063514461734,
      "rewrite6_current_pass_fraction": 0.9761640474382193,
      "rewrite6_full_passes": 0.0,
      "step": 98470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4680448,
      "curriculum_late_phase": 1.0,
      "epoch": 8.24e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.13988494873047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0572,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.53044,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.267923528021466,
      "rewrite6_current_pass_fraction": 0.9768407625946914,
      "rewrite6_full_passes": 0.0,
      "step": 98480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4705024,
      "curriculum_late_phase": 1.0,
      "epoch": 8.245e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.141075134277344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0595,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5309333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.316896048946262,
      "rewrite6_current_pass_fraction": 0.9775174777511636,
      "rewrite6_full_passes": 0.0,
      "step": 98490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.47296,
      "curriculum_late_phase": 1.0,
      "epoch": 8.25e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.141075134277344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0567,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5313733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.257292224304877,
      "rewrite6_current_pass_fraction": 0.9780627446398318,
      "rewrite6_full_passes": 0.0,
      "step": 98500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 8.25e-06,
      "eval_loss": 3.082427501678467,
      "eval_perplexity": 21.81128511501016,
      "eval_runtime": 0.0508,
      "eval_samples_per_second": 629.843,
      "eval_steps_per_second": 19.683,
      "step": 98500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4754176,
      "curriculum_late_phase": 1.0,
      "epoch": 8.255e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.19921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5316666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.9786420907090417,
      "rewrite6_full_passes": 0.0,
      "step": 98510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4778752,
      "curriculum_late_phase": 1.0,
      "epoch": 8.26e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.200233459472656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0569,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5321333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.26154410792393,
      "rewrite6_current_pass_fraction": 0.979236042141341,
      "rewrite6_full_passes": 0.0,
      "step": 98520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4803328,
      "curriculum_late_phase": 1.0,
      "epoch": 8.265e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.201358795166016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0307,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.53244,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.711725724207458,
      "rewrite6_current_pass_fraction": 0.9797569667581936,
      "rewrite6_full_passes": 0.0,
      "step": 98530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4827904,
      "curriculum_late_phase": 1.0,
      "epoch": 8.27e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.201427459716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.53264,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.9803655235535822,
      "rewrite6_full_passes": 0.0,
      "step": 98540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.485248,
      "curriculum_late_phase": 1.0,
      "epoch": 8.275e-06,
      "grad_norm": 0.5234375,
      "host_rss_gib": 16.201431274414062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0815,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5329866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.791064490237765,
      "rewrite6_current_pass_fraction": 0.9808961850791611,
      "rewrite6_full_passes": 0.0,
      "step": 98550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4877056,
      "curriculum_late_phase": 1.0,
      "epoch": 8.28e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.202007293701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5332,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13435678680148,
      "rewrite6_current_pass_fraction": 0.981470662694008,
      "rewrite6_full_passes": 0.0,
      "step": 98560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4901632,
      "curriculum_late_phase": 1.0,
      "epoch": 8.285e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.2025146484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5334933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.9821230355786645,
      "rewrite6_full_passes": 0.0,
      "step": 98570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4926208,
      "curriculum_late_phase": 1.0,
      "epoch": 8.29e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.436492919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0716,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5338,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.57639730762967,
      "rewrite6_current_pass_fraction": 0.9827023816478744,
      "rewrite6_full_passes": 0.0,
      "step": 98580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4950784,
      "curriculum_late_phase": 1.0,
      "epoch": 8.295e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.827842712402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0654,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.53412,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.443037486962115,
      "rewrite6_current_pass_fraction": 0.9833742283499834,
      "rewrite6_full_passes": 0.0,
      "step": 98590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.497536,
      "curriculum_late_phase": 1.0,
      "epoch": 8.3e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.280853271484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5344133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.9840752857782711,
      "rewrite6_full_passes": 0.0,
      "step": 98600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.4999936,
      "curriculum_late_phase": 1.0,
      "epoch": 8.305e-06,
      "grad_norm": 0.5,
      "host_rss_gib": 16.28226089477539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0813,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5346266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.786706713131952,
      "rewrite6_current_pass_fraction": 0.9846935794823859,
      "rewrite6_full_passes": 0.0,
      "step": 98610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5024512,
      "curriculum_late_phase": 1.0,
      "epoch": 8.31e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.282413482666016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5347733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.9851755564643337,
      "rewrite6_full_passes": 0.0,
      "step": 98620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5049088,
      "curriculum_late_phase": 1.0,
      "epoch": 8.315e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0489,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.53508,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.092130313776398,
      "rewrite6_current_pass_fraction": 0.9857743763509961,
      "rewrite6_full_passes": 0.0,
      "step": 98630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5073664,
      "curriculum_late_phase": 1.0,
      "epoch": 8.32e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5352933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.117456062563107,
      "rewrite6_current_pass_fraction": 0.9862417479698545,
      "rewrite6_full_passes": 0.0,
      "step": 98640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.509824,
      "curriculum_late_phase": 1.0,
      "epoch": 8.325e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0441,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.53564,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.991130681307407,
      "rewrite6_current_pass_fraction": 0.9868649101283324,
      "rewrite6_full_passes": 0.0,
      "step": 98650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5122816,
      "curriculum_late_phase": 1.0,
      "epoch": 8.33e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.07,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5359333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.54190267500241,
      "rewrite6_current_pass_fraction": 0.9875270199217152,
      "rewrite6_full_passes": 0.0,
      "step": 98660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5147392,
      "curriculum_late_phase": 1.0,
      "epoch": 8.335e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5362133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.02684595284501,
      "rewrite6_current_pass_fraction": 0.9881745243520087,
      "rewrite6_full_passes": 0.0,
      "step": 98670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5171968,
      "curriculum_late_phase": 1.0,
      "epoch": 8.34e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0572,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5364933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.267923528021466,
      "rewrite6_current_pass_fraction": 0.9887149227863138,
      "rewrite6_full_passes": 0.0,
      "step": 98680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5196544,
      "curriculum_late_phase": 1.0,
      "epoch": 8.345e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5368266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.240293191045346,
      "rewrite6_current_pass_fraction": 0.989347821853518,
      "rewrite6_full_passes": 0.0,
      "step": 98690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.522112,
      "curriculum_late_phase": 1.0,
      "epoch": 8.35e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0785,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5371466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.725789258570966,
      "rewrite6_current_pass_fraction": 0.9899076941052755,
      "rewrite6_full_passes": 0.0,
      "step": 98700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5245696,
      "curriculum_late_phase": 1.0,
      "epoch": 8.355e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0597,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5373333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.321159854522396,
      "rewrite6_current_pass_fraction": 0.9904724348113961,
      "rewrite6_full_passes": 0.0,
      "step": 98710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5270272,
      "curriculum_late_phase": 1.0,
      "epoch": 8.36e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0579,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.53756,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28281628634838,
      "rewrite6_current_pass_fraction": 0.9911150707873264,
      "rewrite6_full_passes": 0.0,
      "step": 98720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5294848,
      "curriculum_late_phase": 1.0,
      "epoch": 8.365e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.53792,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.248791007771803,
      "rewrite6_current_pass_fraction": 0.9917382329458043,
      "rewrite6_full_passes": 0.0,
      "step": 98730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5319424,
      "curriculum_late_phase": 1.0,
      "epoch": 8.37e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0512,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5383066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14069804497869,
      "rewrite6_current_pass_fraction": 0.9922688944713832,
      "rewrite6_full_passes": 0.0,
      "step": 98740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5344,
      "curriculum_late_phase": 1.0,
      "epoch": 8.375e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5386533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.9927995559969621,
      "rewrite6_full_passes": 0.0,
      "step": 98750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5368576,
      "curriculum_late_phase": 1.0,
      "epoch": 8.38e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5391466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.9932571907070943,
      "rewrite6_full_passes": 0.0,
      "step": 98760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5393152,
      "curriculum_late_phase": 1.0,
      "epoch": 8.385e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0667,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5394533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.470931562916125,
      "rewrite6_current_pass_fraction": 0.9939290374092034,
      "rewrite6_full_passes": 0.0,
      "step": 98770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5417728,
      "curriculum_late_phase": 1.0,
      "epoch": 8.39e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0773,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5398266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.699733947773797,
      "rewrite6_current_pass_fraction": 0.9944304882086036,
      "rewrite6_full_passes": 0.0,
      "step": 98780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5442304,
      "curriculum_late_phase": 1.0,
      "epoch": 8.395e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0696,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5400133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.53328763705487,
      "rewrite6_current_pass_fraction": 0.9950682557301708,
      "rewrite6_full_passes": 0.0,
      "step": 98790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.546688,
      "curriculum_late_phase": 1.0,
      "epoch": 8.4e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0256,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5404,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.606364821685087,
      "rewrite6_current_pass_fraction": 0.9956573387081069,
      "rewrite6_full_passes": 0.0,
      "step": 98800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5491456,
      "curriculum_late_phase": 1.0,
      "epoch": 8.405e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.54068,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.9962707639578586,
      "rewrite6_full_passes": 0.0,
      "step": 98810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5516032,
      "curriculum_late_phase": 1.0,
      "epoch": 8.41e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0584,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5409733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.29346035528703,
      "rewrite6_current_pass_fraction": 0.9968452415727055,
      "rewrite6_full_passes": 0.0,
      "step": 98820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5540608,
      "curriculum_late_phase": 1.0,
      "epoch": 8.415e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0319,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.54136,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.736594713485793,
      "rewrite6_current_pass_fraction": 0.9973369554633795,
      "rewrite6_full_passes": 0.0,
      "step": 98830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5565184,
      "curriculum_late_phase": 1.0,
      "epoch": 8.42e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.54168,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.662077184548252,
      "rewrite6_current_pass_fraction": 0.9980818289809351,
      "rewrite6_full_passes": 0.0,
      "step": 98840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.558976,
      "curriculum_late_phase": 1.0,
      "epoch": 8.425e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.54204,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.24241732656946,
      "rewrite6_current_pass_fraction": 0.998656306595782,
      "rewrite6_full_passes": 0.0,
      "step": 98850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5614336,
      "curriculum_late_phase": 1.0,
      "epoch": 8.43e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0749,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5423866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.647717031566703,
      "rewrite6_current_pass_fraction": 0.9991528889408191,
      "rewrite6_full_passes": 0.0,
      "step": 98860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5638912,
      "curriculum_late_phase": 1.0,
      "epoch": 8.435e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0589,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5426,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.304109747590893,
      "rewrite6_current_pass_fraction": 0.9995813129247726,
      "rewrite6_full_passes": 0.0,
      "step": 98870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5663488,
      "curriculum_late_phase": 1.0,
      "epoch": 8.44e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5429333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.911515748893304,
      "rewrite6_current_pass_fraction": 0.000219080446339896,
      "rewrite6_full_passes": 1.0,
      "step": 98880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5688064,
      "curriculum_late_phase": 1.0,
      "epoch": 8.445e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0315,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5432133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.728301734306807,
      "rewrite6_current_pass_fraction": 0.0008568479679071488,
      "rewrite6_full_passes": 1.0,
      "step": 98890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.571264,
      "curriculum_late_phase": 1.0,
      "epoch": 8.45e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0705,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5434533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.552676319526597,
      "rewrite6_current_pass_fraction": 0.0014118517653015522,
      "rewrite6_full_passes": 1.0,
      "step": 98900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5737216,
      "curriculum_late_phase": 1.0,
      "epoch": 8.455e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.068,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5437866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.498861924749583,
      "rewrite6_current_pass_fraction": 0.0019376448365173026,
      "rewrite6_full_passes": 1.0,
      "step": 98910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5761792,
      "curriculum_late_phase": 1.0,
      "epoch": 8.46e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5442133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.002536464723179685,
      "rewrite6_full_passes": 1.0,
      "step": 98920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5786368,
      "curriculum_late_phase": 1.0,
      "epoch": 8.465e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0364,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5445733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.83011966300961,
      "rewrite6_current_pass_fraction": 0.003174232244746938,
      "rewrite6_full_passes": 1.0,
      "step": 98930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5810944,
      "curriculum_late_phase": 1.0,
      "epoch": 8.47e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0637,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5449066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.406615290872633,
      "rewrite6_current_pass_fraction": 0.0036659461354209264,
      "rewrite6_full_passes": 1.0,
      "step": 98940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.583552,
      "curriculum_late_phase": 1.0,
      "epoch": 8.475e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.282821655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0745,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5452266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.639059676340555,
      "rewrite6_current_pass_fraction": 0.0043085821113512885,
      "rewrite6_full_passes": 1.0,
      "step": 98950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5860096,
      "curriculum_late_phase": 1.0,
      "epoch": 8.48e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.283870697021484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0466,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5457066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.043674159992648,
      "rewrite6_current_pass_fraction": 0.004917138906739889,
      "rewrite6_full_passes": 1.0,
      "step": 98960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5884672,
      "curriculum_late_phase": 1.0,
      "epoch": 8.485e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.283870697021484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.54612,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23604555718468,
      "rewrite6_current_pass_fraction": 0.005496484975949835,
      "rewrite6_full_passes": 1.0,
      "step": 98970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5909248,
      "curriculum_late_phase": 1.0,
      "epoch": 8.49e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.928401947021484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0366,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.54644,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.83428610357238,
      "rewrite6_current_pass_fraction": 0.006100173316975327,
      "rewrite6_full_passes": 1.0,
      "step": 98980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5933824,
      "curriculum_late_phase": 1.0,
      "epoch": 8.495e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.928401947021484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5467866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14704120581873,
      "rewrite6_current_pass_fraction": 0.00661622947946486,
      "rewrite6_full_passes": 1.0,
      "step": 98990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.59584,
      "curriculum_late_phase": 1.0,
      "epoch": 8.5e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.928401947021484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.54712,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.007103074915775739,
      "rewrite6_full_passes": 1.0,
      "step": 99000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 8.5e-06,
      "eval_loss": 3.083064317703247,
      "eval_perplexity": 21.82517931444733,
      "eval_runtime": 0.0501,
      "eval_samples_per_second": 638.448,
      "eval_steps_per_second": 19.951,
      "step": 99000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.5982976,
      "curriculum_late_phase": 1.0,
      "epoch": 8.505e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.9306640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0306,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5474666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.709654655190217,
      "rewrite6_current_pass_fraction": 0.0077603162547954276,
      "rewrite6_full_passes": 1.0,
      "step": 99010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6007552,
      "curriculum_late_phase": 1.0,
      "epoch": 8.51e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.9306640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0495,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5477866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.104789389307548,
      "rewrite6_current_pass_fraction": 0.008344530778368484,
      "rewrite6_full_passes": 1.0,
      "step": 99020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6032128,
      "curriculum_late_phase": 1.0,
      "epoch": 8.515e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0612,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5481733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35316559260667,
      "rewrite6_current_pass_fraction": 0.00892387684757843,
      "rewrite6_full_passes": 1.0,
      "step": 99030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6056704,
      "curriculum_late_phase": 1.0,
      "epoch": 8.52e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5484,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.009508091371151486,
      "rewrite6_full_passes": 1.0,
      "step": 99040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.608128,
      "curriculum_late_phase": 1.0,
      "epoch": 8.525e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0344,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5486533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.788501056163312,
      "rewrite6_current_pass_fraction": 0.010238359525617807,
      "rewrite6_full_passes": 1.0,
      "step": 99050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6105856,
      "curriculum_late_phase": 1.0,
      "epoch": 8.53e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0317,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5490133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.732447809247343,
      "rewrite6_current_pass_fraction": 0.010817705594827753,
      "rewrite6_full_passes": 1.0,
      "step": 99060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6130432,
      "curriculum_late_phase": 1.0,
      "epoch": 8.535e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0663,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.54932,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.462344907736483,
      "rewrite6_current_pass_fraction": 0.011348367120406613,
      "rewrite6_full_passes": 1.0,
      "step": 99070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6155008,
      "curriculum_late_phase": 1.0,
      "epoch": 8.54e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.54956,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.221185526912038,
      "rewrite6_current_pass_fraction": 0.011937450098342779,
      "rewrite6_full_passes": 1.0,
      "step": 99080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6179584,
      "curriculum_late_phase": 1.0,
      "epoch": 8.545e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0567,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.54988,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.257292224304877,
      "rewrite6_current_pass_fraction": 0.012516796167552725,
      "rewrite6_full_passes": 1.0,
      "step": 99090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.620416,
      "curriculum_late_phase": 1.0,
      "epoch": 8.55e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0289,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5501866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.67447815077682,
      "rewrite6_current_pass_fraction": 0.013110747599851999,
      "rewrite6_full_passes": 1.0,
      "step": 99100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6228736,
      "curriculum_late_phase": 1.0,
      "epoch": 8.555e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0556,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5505333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.233922058805657,
      "rewrite6_current_pass_fraction": 0.01358785612743666,
      "rewrite6_full_passes": 1.0,
      "step": 99110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6253312,
      "curriculum_late_phase": 1.0,
      "epoch": 8.56e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5508666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20845663464543,
      "rewrite6_current_pass_fraction": 0.014172070651009717,
      "rewrite6_full_passes": 1.0,
      "step": 99120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6277888,
      "curriculum_late_phase": 1.0,
      "epoch": 8.565e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0278,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.55112,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.651748728285217,
      "rewrite6_current_pass_fraction": 0.014766022083308992,
      "rewrite6_full_passes": 1.0,
      "step": 99130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6302464,
      "curriculum_late_phase": 1.0,
      "epoch": 8.57e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0682,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5516133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.503162127140435,
      "rewrite6_current_pass_fraction": 0.015408658059239353,
      "rewrite6_full_passes": 1.0,
      "step": 99140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.632704,
      "curriculum_late_phase": 1.0,
      "epoch": 8.575e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0556,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.55188,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.233922058805657,
      "rewrite6_current_pass_fraction": 0.01606103094389593,
      "rewrite6_full_passes": 1.0,
      "step": 99150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6351616,
      "curriculum_late_phase": 1.0,
      "epoch": 8.58e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.07,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.55216,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.54190267500241,
      "rewrite6_current_pass_fraction": 0.016659850830558313,
      "rewrite6_full_passes": 1.0,
      "step": 99160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6376192,
      "curriculum_late_phase": 1.0,
      "epoch": 8.585e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5524933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21482012611182,
      "rewrite6_current_pass_fraction": 0.017190512356137173,
      "rewrite6_full_passes": 1.0,
      "step": 99170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6400768,
      "curriculum_late_phase": 1.0,
      "epoch": 8.59e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5527066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.909424701872513,
      "rewrite6_current_pass_fraction": 0.017823411423341318,
      "rewrite6_full_passes": 1.0,
      "step": 99180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6425344,
      "curriculum_late_phase": 1.0,
      "epoch": 8.595e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.553,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14704120581873,
      "rewrite6_current_pass_fraction": 0.01833946758583085,
      "rewrite6_full_passes": 1.0,
      "step": 99190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.644992,
      "curriculum_late_phase": 1.0,
      "epoch": 8.6e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0421,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5533733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.949190374231975,
      "rewrite6_current_pass_fraction": 0.01895289283558256,
      "rewrite6_full_passes": 1.0,
      "step": 99200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6474496,
      "curriculum_late_phase": 1.0,
      "epoch": 8.605e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.55376,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.32969002438444,
      "rewrite6_current_pass_fraction": 0.019468948998072094,
      "rewrite6_full_passes": 1.0,
      "step": 99210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6499072,
      "curriculum_late_phase": 1.0,
      "epoch": 8.61e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0555,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.554,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.231798772765842,
      "rewrite6_current_pass_fraction": 0.020028821249829604,
      "rewrite6_full_passes": 1.0,
      "step": 99220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6523648,
      "curriculum_late_phase": 1.0,
      "epoch": 8.615e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.55448,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.020583825047224006,
      "rewrite6_full_passes": 1.0,
      "step": 99230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6548224,
      "curriculum_late_phase": 1.0,
      "epoch": 8.62e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0658,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5547466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45161641762865,
      "rewrite6_current_pass_fraction": 0.021250803294969912,
      "rewrite6_full_passes": 1.0,
      "step": 99240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.65728,
      "curriculum_late_phase": 1.0,
      "epoch": 8.625e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.07,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5550266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.54190267500241,
      "rewrite6_current_pass_fraction": 0.021937255360168252,
      "rewrite6_full_passes": 1.0,
      "step": 99250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6597376,
      "curriculum_late_phase": 1.0,
      "epoch": 8.63e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5552666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.022472785340110223,
      "rewrite6_full_passes": 1.0,
      "step": 99260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6621952,
      "curriculum_late_phase": 1.0,
      "epoch": 8.635e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0395,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5555733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.89479322619514,
      "rewrite6_current_pass_fraction": 0.02316897431403478,
      "rewrite6_full_passes": 1.0,
      "step": 99270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6646528,
      "curriculum_late_phase": 1.0,
      "epoch": 8.64e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.55596,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21482012611182,
      "rewrite6_current_pass_fraction": 0.02377753110942338,
      "rewrite6_full_passes": 1.0,
      "step": 99280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6671104,
      "curriculum_late_phase": 1.0,
      "epoch": 8.645e-06,
      "grad_norm": 0.4296875,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0344,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5561866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.788501056163312,
      "rewrite6_current_pass_fraction": 0.02434714026990711,
      "rewrite6_full_passes": 1.0,
      "step": 99290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.669568,
      "curriculum_late_phase": 1.0,
      "epoch": 8.65e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.930984497070312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5565466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.153386269892504,
      "rewrite6_current_pass_fraction": 0.024863196432396643,
      "rewrite6_full_passes": 1.0,
      "step": 99300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6720256,
      "curriculum_late_phase": 1.0,
      "epoch": 8.655e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.56075668334961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0649,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5568266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.432318648151654,
      "rewrite6_current_pass_fraction": 0.025471753227785242,
      "rewrite6_full_passes": 1.0,
      "step": 99310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6744832,
      "curriculum_late_phase": 1.0,
      "epoch": 8.66e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.563335418701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0706,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5571333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.55483169492553,
      "rewrite6_current_pass_fraction": 0.026075441568810732,
      "rewrite6_full_passes": 1.0,
      "step": 99320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6769408,
      "curriculum_late_phase": 1.0,
      "epoch": 8.665e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.565662384033203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.55748,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.026674261455473115,
      "rewrite6_full_passes": 1.0,
      "step": 99330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6793984,
      "curriculum_late_phase": 1.0,
      "epoch": 8.67e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.565670013427734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.55776,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.027287686705224825,
      "rewrite6_full_passes": 1.0,
      "step": 99340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.681856,
      "curriculum_late_phase": 1.0,
      "epoch": 8.675e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.565673828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.558,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.678613459924104,
      "rewrite6_current_pass_fraction": 0.02784755895698234,
      "rewrite6_full_passes": 1.0,
      "step": 99350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6843136,
      "curriculum_late_phase": 1.0,
      "epoch": 8.68e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.565673828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.062,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5582266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.37025495991624,
      "rewrite6_current_pass_fraction": 0.02845124729800783,
      "rewrite6_full_passes": 1.0,
      "step": 99360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6867712,
      "curriculum_late_phase": 1.0,
      "epoch": 8.685e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.565673828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5586,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01213231105346,
      "rewrite6_current_pass_fraction": 0.028977040369223578,
      "rewrite6_full_passes": 1.0,
      "step": 99370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6892288,
      "curriculum_late_phase": 1.0,
      "epoch": 8.69e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.565673828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0594,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.55892,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3147644659223,
      "rewrite6_current_pass_fraction": 0.029566123347159745,
      "rewrite6_full_passes": 1.0,
      "step": 99380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6916864,
      "curriculum_late_phase": 1.0,
      "epoch": 8.695e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.565677642822266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5592133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.934531072324265,
      "rewrite6_current_pass_fraction": 0.030213627777453213,
      "rewrite6_full_passes": 1.0,
      "step": 99390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.694144,
      "curriculum_late_phase": 1.0,
      "epoch": 8.7e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.573490142822266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0756,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.55948,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.66287573841722,
      "rewrite6_current_pass_fraction": 0.030754026211758292,
      "rewrite6_full_passes": 1.0,
      "step": 99400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6966016,
      "curriculum_late_phase": 1.0,
      "epoch": 8.705e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.574024200439453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0771,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5598133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.695394434949993,
      "rewrite6_current_pass_fraction": 0.031309030009152694,
      "rewrite6_full_passes": 1.0,
      "step": 99410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.6990592,
      "curriculum_late_phase": 1.0,
      "epoch": 8.71e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.574024200439453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5601466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.738668476643575,
      "rewrite6_current_pass_fraction": 0.0319224552589044,
      "rewrite6_full_passes": 1.0,
      "step": 99420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7015168,
      "curriculum_late_phase": 1.0,
      "epoch": 8.715e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 13.912025451660156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5605733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.032569959689197875,
      "rewrite6_full_passes": 1.0,
      "step": 99430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7039744,
      "curriculum_late_phase": 1.0,
      "epoch": 8.72e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.043403625488281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.082,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5608,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.801962746819974,
      "rewrite6_current_pass_fraction": 0.03311522657786606,
      "rewrite6_full_passes": 1.0,
      "step": 99440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.706432,
      "curriculum_late_phase": 1.0,
      "epoch": 8.725e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.648590087890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0674,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5612133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.48596647661604,
      "rewrite6_current_pass_fraction": 0.03376273100815953,
      "rewrite6_full_passes": 1.0,
      "step": 99450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7088896,
      "curriculum_late_phase": 1.0,
      "epoch": 8.73e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.649028778076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56144,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.295589807793416,
      "rewrite6_current_pass_fraction": 0.03433720862300637,
      "rewrite6_full_passes": 1.0,
      "step": 99460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7113472,
      "curriculum_late_phase": 1.0,
      "epoch": 8.735e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.649436950683594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0466,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5618266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.043674159992648,
      "rewrite6_current_pass_fraction": 0.03497984459893673,
      "rewrite6_full_passes": 1.0,
      "step": 99470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7138048,
      "curriculum_late_phase": 1.0,
      "epoch": 8.74e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.649688720703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0566,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5622133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25516660136536,
      "rewrite6_current_pass_fraction": 0.03545208467215828,
      "rewrite6_full_passes": 1.0,
      "step": 99480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7162624,
      "curriculum_late_phase": 1.0,
      "epoch": 8.745e-06,
      "grad_norm": 0.5234375,
      "host_rss_gib": 15.64971923828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0686,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5624933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.51176511247365,
      "rewrite6_current_pass_fraction": 0.03617261591789839,
      "rewrite6_full_passes": 1.0,
      "step": 99490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.71872,
      "curriculum_late_phase": 1.0,
      "epoch": 8.75e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.64971923828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0466,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5628,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.043674159992648,
      "rewrite6_current_pass_fraction": 0.036781172713286986,
      "rewrite6_full_passes": 1.0,
      "step": 99500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 8.75e-06,
      "eval_loss": 3.082996129989624,
      "eval_perplexity": 21.823691156108097,
      "eval_runtime": 0.049,
      "eval_samples_per_second": 653.646,
      "eval_steps_per_second": 20.426,
      "step": 99500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7211776,
      "curriculum_late_phase": 1.0,
      "epoch": 8.755e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.724472045898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.065,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5629866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.434461987181628,
      "rewrite6_current_pass_fraction": 0.037443282506669784,
      "rewrite6_full_passes": 1.0,
      "step": 99510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7236352,
      "curriculum_late_phase": 1.0,
      "epoch": 8.76e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.724472045898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0703,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.548366215287484,
      "rewrite6_current_pass_fraction": 0.037920391034254444,
      "rewrite6_full_passes": 1.0,
      "step": 99520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7260928,
      "curriculum_late_phase": 1.0,
      "epoch": 8.765e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.724472045898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56372,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.144926607429838,
      "rewrite6_current_pass_fraction": 0.03854842164709548,
      "rewrite6_full_passes": 1.0,
      "step": 99530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7285504,
      "curriculum_late_phase": 1.0,
      "epoch": 8.77e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.724761962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.079,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5640533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.73665486937659,
      "rewrite6_current_pass_fraction": 0.03911803080757921,
      "rewrite6_full_passes": 1.0,
      "step": 99540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.731008,
      "curriculum_late_phase": 1.0,
      "epoch": 8.775e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.72476577758789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0656,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5644,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.44732652334885,
      "rewrite6_current_pass_fraction": 0.03960487624389009,
      "rewrite6_full_passes": 1.0,
      "step": 99550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7334656,
      "curriculum_late_phase": 1.0,
      "epoch": 8.78e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.72476577758789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0648,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5646266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.430175523444856,
      "rewrite6_current_pass_fraction": 0.04016961695001071,
      "rewrite6_full_passes": 1.0,
      "step": 99560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7359232,
      "curriculum_late_phase": 1.0,
      "epoch": 8.785e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.72476577758789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.06,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56492,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.327557162026903,
      "rewrite6_current_pass_fraction": 0.040719752293042005,
      "rewrite6_full_passes": 1.0,
      "step": 99570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7383808,
      "curriculum_late_phase": 1.0,
      "epoch": 8.79e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.72476577758789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56516,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.26367036864598,
      "rewrite6_current_pass_fraction": 0.04131857217970439,
      "rewrite6_full_passes": 1.0,
      "step": 99580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7408384,
      "curriculum_late_phase": 1.0,
      "epoch": 8.795e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.72476577758789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0703,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56556,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.548366215287484,
      "rewrite6_current_pass_fraction": 0.04200989269926584,
      "rewrite6_full_passes": 1.0,
      "step": 99590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.743296,
      "curriculum_late_phase": 1.0,
      "epoch": 8.8e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.72476577758789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5658266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06894172649243,
      "rewrite6_current_pass_fraction": 0.04261358104029133,
      "rewrite6_full_passes": 1.0,
      "step": 99600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7457536,
      "curriculum_late_phase": 1.0,
      "epoch": 8.805e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.72476577758789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0551,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56612,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.22330775157419,
      "rewrite6_current_pass_fraction": 0.04316371638332262,
      "rewrite6_full_passes": 1.0,
      "step": 99610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7482112,
      "curriculum_late_phase": 1.0,
      "epoch": 8.81e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.725112915039062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5664533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35530101593532,
      "rewrite6_current_pass_fraction": 0.04374793090689568,
      "rewrite6_full_passes": 1.0,
      "step": 99620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7506688,
      "curriculum_late_phase": 1.0,
      "epoch": 8.815e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.72573471069336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56688,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.905243235092758,
      "rewrite6_current_pass_fraction": 0.04436135615664739,
      "rewrite6_full_passes": 1.0,
      "step": 99630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7531264,
      "curriculum_late_phase": 1.0,
      "epoch": 8.82e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.72573471069336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56716,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.04496504449767288,
      "rewrite6_full_passes": 1.0,
      "step": 99640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.755584,
      "curriculum_late_phase": 1.0,
      "epoch": 8.825e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.72573471069336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56748,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.961763660065063,
      "rewrite6_current_pass_fraction": 0.04557846974742459,
      "rewrite6_full_passes": 1.0,
      "step": 99650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7580416,
      "curriculum_late_phase": 1.0,
      "epoch": 8.83e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5677466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.064728359497874,
      "rewrite6_current_pass_fraction": 0.04620650036026562,
      "rewrite6_full_passes": 1.0,
      "step": 99660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7604992,
      "curriculum_late_phase": 1.0,
      "epoch": 8.835e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5679733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.046780977975112464,
      "rewrite6_full_passes": 1.0,
      "step": 99670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7629568,
      "curriculum_late_phase": 1.0,
      "epoch": 8.84e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0614,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5683333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.357436652816972,
      "rewrite6_current_pass_fraction": 0.04736032404432241,
      "rewrite6_full_passes": 1.0,
      "step": 99680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7654144,
      "curriculum_late_phase": 1.0,
      "epoch": 8.845e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0679,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5686666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.49671214604783,
      "rewrite6_current_pass_fraction": 0.047900722478627485,
      "rewrite6_full_passes": 1.0,
      "step": 99690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.767872,
      "curriculum_late_phase": 1.0,
      "epoch": 8.85e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0657,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56896,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.449471363241397,
      "rewrite6_current_pass_fraction": 0.04848980545656365,
      "rewrite6_full_passes": 1.0,
      "step": 99700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7703296,
      "curriculum_late_phase": 1.0,
      "epoch": 8.855e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56916,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.04896691398414831,
      "rewrite6_full_passes": 1.0,
      "step": 99710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7727872,
      "curriculum_late_phase": 1.0,
      "epoch": 8.86e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.058,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5696,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.284944674394637,
      "rewrite6_current_pass_fraction": 0.049619286868804895,
      "rewrite6_full_passes": 1.0,
      "step": 99720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7752448,
      "curriculum_late_phase": 1.0,
      "epoch": 8.865e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.56984,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.117456062563107,
      "rewrite6_current_pass_fraction": 0.05033981811454499,
      "rewrite6_full_passes": 1.0,
      "step": 99730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7777024,
      "curriculum_late_phase": 1.0,
      "epoch": 8.87e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5702133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.049988209300682,
      "rewrite6_current_pass_fraction": 0.05089969036630251,
      "rewrite6_full_passes": 1.0,
      "step": 99740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.78016,
      "curriculum_late_phase": 1.0,
      "epoch": 8.875e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5705466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.921974121142934,
      "rewrite6_current_pass_fraction": 0.05155206325095908,
      "rewrite6_full_passes": 1.0,
      "step": 99750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7826176,
      "curriculum_late_phase": 1.0,
      "epoch": 8.88e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5709333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.02684595284501,
      "rewrite6_current_pass_fraction": 0.052204436135615664,
      "rewrite6_full_passes": 1.0,
      "step": 99760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7850752,
      "curriculum_late_phase": 1.0,
      "epoch": 8.885e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.57128,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.197855056943403,
      "rewrite6_current_pass_fraction": 0.052715623843742086,
      "rewrite6_full_passes": 1.0,
      "step": 99770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7875328,
      "curriculum_late_phase": 1.0,
      "epoch": 8.89e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.726070404052734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0737,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5715466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.621755351252418,
      "rewrite6_current_pass_fraction": 0.05336799672839867,
      "rewrite6_full_passes": 1.0,
      "step": 99780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7899904,
      "curriculum_late_phase": 1.0,
      "epoch": 8.895e-06,
      "grad_norm": 0.546875,
      "host_rss_gib": 15.726192474365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0645,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5720133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.423747435049286,
      "rewrite6_current_pass_fraction": 0.05402523806741836,
      "rewrite6_full_passes": 1.0,
      "step": 99790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.792448,
      "curriculum_late_phase": 1.0,
      "epoch": 8.9e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.726192474365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0891,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5722666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.957307503632205,
      "rewrite6_current_pass_fraction": 0.054599715682265196,
      "rewrite6_full_passes": 1.0,
      "step": 99800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7949056,
      "curriculum_late_phase": 1.0,
      "epoch": 8.905e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.726192474365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5725866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.018436896387286,
      "rewrite6_current_pass_fraction": 0.055135245662207163,
      "rewrite6_full_passes": 1.0,
      "step": 99810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7973632,
      "curriculum_late_phase": 1.0,
      "epoch": 8.91e-06,
      "grad_norm": 0.431640625,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.57288,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.05570485482269089,
      "rewrite6_full_passes": 1.0,
      "step": 99820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.7998208,
      "curriculum_late_phase": 1.0,
      "epoch": 8.915e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5732666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.05624525325699597,
      "rewrite6_full_passes": 1.0,
      "step": 99830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8022784,
      "curriculum_late_phase": 1.0,
      "epoch": 8.92e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5736266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.05680999396311659,
      "rewrite6_full_passes": 1.0,
      "step": 99840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.804736,
      "curriculum_late_phase": 1.0,
      "epoch": 8.925e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5740666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.05734552394305856,
      "rewrite6_full_passes": 1.0,
      "step": 99850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8071936,
      "curriculum_late_phase": 1.0,
      "epoch": 8.93e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.57444,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.259418059817307,
      "rewrite6_current_pass_fraction": 0.057968686101536486,
      "rewrite6_full_passes": 1.0,
      "step": 99860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8096512,
      "curriculum_late_phase": 1.0,
      "epoch": 8.935e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5748,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.125904734609893,
      "rewrite6_current_pass_fraction": 0.05851395299020467,
      "rewrite6_full_passes": 1.0,
      "step": 99870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8121088,
      "curriculum_late_phase": 1.0,
      "epoch": 8.94e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0706,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.57512,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.55483169492553,
      "rewrite6_current_pass_fraction": 0.059088430605051506,
      "rewrite6_full_passes": 1.0,
      "step": 99880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8145664,
      "curriculum_late_phase": 1.0,
      "epoch": 8.945e-06,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0246,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5754266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.58576875661228,
      "rewrite6_current_pass_fraction": 0.05968725049171389,
      "rewrite6_full_passes": 1.0,
      "step": 99890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.817024,
      "curriculum_late_phase": 1.0,
      "epoch": 8.95e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0408,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5758666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.921974121142934,
      "rewrite6_current_pass_fraction": 0.060354228739459795,
      "rewrite6_full_passes": 1.0,
      "step": 99900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8194816,
      "curriculum_late_phase": 1.0,
      "epoch": 8.955e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5762666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.0610212069872057,
      "rewrite6_full_passes": 1.0,
      "step": 99910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8219392,
      "curriculum_late_phase": 1.0,
      "epoch": 8.96e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0763,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5765333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.678045060077274,
      "rewrite6_current_pass_fraction": 0.061571342330237,
      "rewrite6_full_passes": 1.0,
      "step": 99920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8243968,
      "curriculum_late_phase": 1.0,
      "epoch": 8.965e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0758,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5768266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.667208746851305,
      "rewrite6_current_pass_fraction": 0.0621798991256256,
      "rewrite6_full_passes": 1.0,
      "step": 99930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8268544,
      "curriculum_late_phase": 1.0,
      "epoch": 8.97e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0481,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5771733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.075263357207568,
      "rewrite6_current_pass_fraction": 0.06275437674047243,
      "rewrite6_full_passes": 1.0,
      "step": 99940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.829312,
      "curriculum_late_phase": 1.0,
      "epoch": 8.975e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0836,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.57748,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.836873808616637,
      "rewrite6_current_pass_fraction": 0.0633970127164028,
      "rewrite6_full_passes": 1.0,
      "step": 99950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8317696,
      "curriculum_late_phase": 1.0,
      "epoch": 8.98e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5777333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.153386269892504,
      "rewrite6_current_pass_fraction": 0.06392767424198166,
      "rewrite6_full_passes": 1.0,
      "step": 99960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8342272,
      "curriculum_late_phase": 1.0,
      "epoch": 8.985e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0328,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5779066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.755266049568853,
      "rewrite6_current_pass_fraction": 0.06451675721991781,
      "rewrite6_full_passes": 1.0,
      "step": 99970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8366848,
      "curriculum_late_phase": 1.0,
      "epoch": 8.99e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0556,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5781733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.233922058805657,
      "rewrite6_current_pass_fraction": 0.06518373546766372,
      "rewrite6_full_passes": 1.0,
      "step": 99980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8391424,
      "curriculum_late_phase": 1.0,
      "epoch": 8.995e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0736,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5784933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.61959328382246,
      "rewrite6_current_pass_fraction": 0.06578742380868921,
      "rewrite6_full_passes": 1.0,
      "step": 99990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8416,
      "curriculum_late_phase": 1.0,
      "epoch": 9e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.7261962890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0751,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5788133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.65204700795622,
      "rewrite6_current_pass_fraction": 0.06644953360207201,
      "rewrite6_full_passes": 1.0,
      "step": 100000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 9e-06,
      "eval_loss": 3.082587718963623,
      "eval_perplexity": 21.814779939855256,
      "eval_runtime": 0.0498,
      "eval_samples_per_second": 643.116,
      "eval_steps_per_second": 20.097,
      "step": 100000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8440576,
      "curriculum_late_phase": 1.0,
      "epoch": 9.005e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.85842514038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0556,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5792666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.233922058805657,
      "rewrite6_current_pass_fraction": 0.06700940585382953,
      "rewrite6_full_passes": 1.0,
      "step": 100010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8465152,
      "curriculum_late_phase": 1.0,
      "epoch": 9.01e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0744,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5796266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.63689587856461,
      "rewrite6_current_pass_fraction": 0.0676033572861288,
      "rewrite6_full_passes": 1.0,
      "step": 100020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8489728,
      "curriculum_late_phase": 1.0,
      "epoch": 9.015e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.57988,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.0682070456271543,
      "rewrite6_full_passes": 1.0,
      "step": 100030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8514304,
      "curriculum_late_phase": 1.0,
      "epoch": 9.02e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0773,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5801866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.699733947773797,
      "rewrite6_current_pass_fraction": 0.06883020778563222,
      "rewrite6_full_passes": 1.0,
      "step": 100040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.853888,
      "curriculum_late_phase": 1.0,
      "epoch": 9.025e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0578,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5804666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.280688111130274,
      "rewrite6_current_pass_fraction": 0.06943389612665771,
      "rewrite6_full_passes": 1.0,
      "step": 100050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8563456,
      "curriculum_late_phase": 1.0,
      "epoch": 9.03e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0639,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5808,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.41089704209166,
      "rewrite6_current_pass_fraction": 0.07010574282876672,
      "rewrite6_full_passes": 1.0,
      "step": 100060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8588032,
      "curriculum_late_phase": 1.0,
      "epoch": 9.035e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5810666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.07066561508052424,
      "rewrite6_full_passes": 1.0,
      "step": 100070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8612608,
      "curriculum_late_phase": 1.0,
      "epoch": 9.04e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0533,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.58148,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18514015876014,
      "rewrite6_current_pass_fraction": 0.07130338260209149,
      "rewrite6_full_passes": 1.0,
      "step": 100080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8637184,
      "curriculum_late_phase": 1.0,
      "epoch": 9.045e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0693,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.58176,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.52682861966481,
      "rewrite6_current_pass_fraction": 0.07182430721894413,
      "rewrite6_full_passes": 1.0,
      "step": 100090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.866176,
      "curriculum_late_phase": 1.0,
      "epoch": 9.05e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0834,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.58204,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.832506870563275,
      "rewrite6_current_pass_fraction": 0.07245233783178516,
      "rewrite6_full_passes": 1.0,
      "step": 100100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8686336,
      "curriculum_late_phase": 1.0,
      "epoch": 9.055e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5823466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.151271037028916,
      "rewrite6_current_pass_fraction": 0.07306089462717376,
      "rewrite6_full_passes": 1.0,
      "step": 100110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8710912,
      "curriculum_late_phase": 1.0,
      "epoch": 9.06e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0645,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.58264,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.423747435049286,
      "rewrite6_current_pass_fraction": 0.07365484605947303,
      "rewrite6_full_passes": 1.0,
      "step": 100120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8735488,
      "curriculum_late_phase": 1.0,
      "epoch": 9.065e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0814,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.58296,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.788885492740427,
      "rewrite6_current_pass_fraction": 0.07416116531323635,
      "rewrite6_full_passes": 1.0,
      "step": 100130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8760064,
      "curriculum_late_phase": 1.0,
      "epoch": 9.07e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0708,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5833866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.55914309238989,
      "rewrite6_current_pass_fraction": 0.07471616911063075,
      "rewrite6_full_passes": 1.0,
      "step": 100140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.878464,
      "curriculum_late_phase": 1.0,
      "epoch": 9.075e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0346,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5838133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.792659172172286,
      "rewrite6_current_pass_fraction": 0.07530525208856692,
      "rewrite6_full_passes": 1.0,
      "step": 100150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8809216,
      "curriculum_late_phase": 1.0,
      "epoch": 9.08e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5842133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11534442254061,
      "rewrite6_current_pass_fraction": 0.07588459815777687,
      "rewrite6_full_passes": 1.0,
      "step": 100160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8833792,
      "curriculum_late_phase": 1.0,
      "epoch": 9.085e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.85272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0687,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.58452,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.513916396547316,
      "rewrite6_current_pass_fraction": 0.07646394422698681,
      "rewrite6_full_passes": 1.0,
      "step": 100170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8858368,
      "curriculum_late_phase": 1.0,
      "epoch": 9.09e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.14337158203125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0774,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5848266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.701904029670857,
      "rewrite6_current_pass_fraction": 0.07705302720492298,
      "rewrite6_full_passes": 1.0,
      "step": 100180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8882944,
      "curriculum_late_phase": 1.0,
      "epoch": 9.095e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.147514343261719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.09,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5852,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.977077975763418,
      "rewrite6_current_pass_fraction": 0.0775593464586863,
      "rewrite6_full_passes": 1.0,
      "step": 100190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.890752,
      "curriculum_late_phase": 1.0,
      "epoch": 9.1e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.147514343261719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5855066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.090021206202152,
      "rewrite6_current_pass_fraction": 0.0781679032540749,
      "rewrite6_full_passes": 1.0,
      "step": 100200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8932096,
      "curriculum_late_phase": 1.0,
      "epoch": 9.105e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.149662017822266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0609,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.58576,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34676060372526,
      "rewrite6_current_pass_fraction": 0.07882514459309459,
      "rewrite6_full_passes": 1.0,
      "step": 100210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8956672,
      "curriculum_late_phase": 1.0,
      "epoch": 9.11e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.151195526123047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.58616,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.014233629348723,
      "rewrite6_current_pass_fraction": 0.07938014839048899,
      "rewrite6_full_passes": 1.0,
      "step": 100220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.8981248,
      "curriculum_late_phase": 1.0,
      "epoch": 9.115e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.151203155517578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0513,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.58648,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.1428122204902,
      "rewrite6_current_pass_fraction": 0.07993028373352028,
      "rewrite6_full_passes": 1.0,
      "step": 100230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9005824,
      "curriculum_late_phase": 1.0,
      "epoch": 9.12e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.158210754394531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0743,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5868666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.63473229715763,
      "rewrite6_current_pass_fraction": 0.0806654203423497,
      "rewrite6_full_passes": 1.0,
      "step": 100240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.90304,
      "curriculum_late_phase": 1.0,
      "epoch": 9.125e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.158214569091797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0598,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5870933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.323292077117205,
      "rewrite6_current_pass_fraction": 0.08112305505248193,
      "rewrite6_full_passes": 1.0,
      "step": 100250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9054976,
      "curriculum_late_phase": 1.0,
      "epoch": 9.13e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.158214569091797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0807,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5873733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.77363860992708,
      "rewrite6_current_pass_fraction": 0.0816050320344297,
      "rewrite6_full_passes": 1.0,
      "step": 100260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9079552,
      "curriculum_late_phase": 1.0,
      "epoch": 9.135e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.159725189208984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5878,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35530101593532,
      "rewrite6_current_pass_fraction": 0.08218437810363965,
      "rewrite6_full_passes": 1.0,
      "step": 100270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9104128,
      "curriculum_late_phase": 1.0,
      "epoch": 9.14e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.95444107055664,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.588,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.71586848361443,
      "rewrite6_current_pass_fraction": 0.08276859262721271,
      "rewrite6_full_passes": 1.0,
      "step": 100280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9128704,
      "curriculum_late_phase": 1.0,
      "epoch": 9.145e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.885074615478516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0556,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5883733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.233922058805657,
      "rewrite6_current_pass_fraction": 0.08335280715078577,
      "rewrite6_full_passes": 1.0,
      "step": 100290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.915328,
      "curriculum_late_phase": 1.0,
      "epoch": 9.15e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.195491790771484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5887733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10267901588904,
      "rewrite6_current_pass_fraction": 0.08390294249381706,
      "rewrite6_full_passes": 1.0,
      "step": 100300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9177856,
      "curriculum_late_phase": 1.0,
      "epoch": 9.155e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.195491790771484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0263,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5890933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.620794326797853,
      "rewrite6_current_pass_fraction": 0.08459913146774162,
      "rewrite6_full_passes": 1.0,
      "step": 100310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9202432,
      "curriculum_late_phase": 1.0,
      "epoch": 9.16e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.195491790771484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5894533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.08516874062822535,
      "rewrite6_full_passes": 1.0,
      "step": 100320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9227008,
      "curriculum_late_phase": 1.0,
      "epoch": 9.165e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.19549560546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5897733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.08581624505851883,
      "rewrite6_full_passes": 1.0,
      "step": 100330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9251584,
      "curriculum_late_phase": 1.0,
      "epoch": 9.17e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.19549560546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5900666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.855130810292252,
      "rewrite6_current_pass_fraction": 0.08643940721699675,
      "rewrite6_full_passes": 1.0,
      "step": 100340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.927616,
      "curriculum_late_phase": 1.0,
      "epoch": 9.175e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.195716857910156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0244,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.59028,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.581652014548887,
      "rewrite6_current_pass_fraction": 0.08704796401238535,
      "rewrite6_full_passes": 1.0,
      "step": 100350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9300736,
      "curriculum_late_phase": 1.0,
      "epoch": 9.18e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.195716857910156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5905466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.216941714202072,
      "rewrite6_current_pass_fraction": 0.08777823216685167,
      "rewrite6_full_passes": 1.0,
      "step": 100360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9325312,
      "curriculum_late_phase": 1.0,
      "epoch": 9.185e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.195716857910156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0402,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5908933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.909424701872513,
      "rewrite6_current_pass_fraction": 0.08831863060115674,
      "rewrite6_full_passes": 1.0,
      "step": 100370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9349888,
      "curriculum_late_phase": 1.0,
      "epoch": 9.19e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.195716857910156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5912133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.976442031457733,
      "rewrite6_current_pass_fraction": 0.08894179275963467,
      "rewrite6_full_passes": 1.0,
      "step": 100380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9374464,
      "curriculum_late_phase": 1.0,
      "epoch": 9.195e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0438,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5915866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.98483428660945,
      "rewrite6_current_pass_fraction": 0.08953574419193394,
      "rewrite6_full_passes": 1.0,
      "step": 100390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.939904,
      "curriculum_late_phase": 1.0,
      "epoch": 9.2e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0551,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5918266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.22330775157419,
      "rewrite6_current_pass_fraction": 0.0901150902611439,
      "rewrite6_full_passes": 1.0,
      "step": 100400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9423616,
      "curriculum_late_phase": 1.0,
      "epoch": 9.205e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5921866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.132243456791066,
      "rewrite6_current_pass_fraction": 0.09067983096726451,
      "rewrite6_full_passes": 1.0,
      "step": 100410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9448192,
      "curriculum_late_phase": 1.0,
      "epoch": 9.21e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0833,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5924266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.83032372903511,
      "rewrite6_current_pass_fraction": 0.09123970321902203,
      "rewrite6_full_passes": 1.0,
      "step": 100420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9472768,
      "curriculum_late_phase": 1.0,
      "epoch": 9.215e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0448,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5927066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.005829616811546,
      "rewrite6_current_pass_fraction": 0.09188233919495238,
      "rewrite6_full_passes": 1.0,
      "step": 100430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9497344,
      "curriculum_late_phase": 1.0,
      "epoch": 9.22e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0611,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.59288,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.351030382809682,
      "rewrite6_current_pass_fraction": 0.09248115908161476,
      "rewrite6_full_passes": 1.0,
      "step": 100440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.952192,
      "curriculum_late_phase": 1.0,
      "epoch": 9.225e-06,
      "grad_norm": 0.52734375,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0617,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5931866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.363844844993586,
      "rewrite6_current_pass_fraction": 0.09308484742264025,
      "rewrite6_full_passes": 1.0,
      "step": 100450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9546496,
      "curriculum_late_phase": 1.0,
      "epoch": 9.23e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0502,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.59344,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11956791376016,
      "rewrite6_current_pass_fraction": 0.09369340421802887,
      "rewrite6_full_passes": 1.0,
      "step": 100460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9571072,
      "curriculum_late_phase": 1.0,
      "epoch": 9.235e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0512,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5938133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14069804497869,
      "rewrite6_current_pass_fraction": 0.09423867110669705,
      "rewrite6_full_passes": 1.0,
      "step": 100470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9595648,
      "curriculum_late_phase": 1.0,
      "epoch": 9.24e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0331,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5941333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.761493563464107,
      "rewrite6_current_pass_fraction": 0.09481801717590699,
      "rewrite6_full_passes": 1.0,
      "step": 100480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9620224,
      "curriculum_late_phase": 1.0,
      "epoch": 9.245e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0354,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5944666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.809299954935618,
      "rewrite6_current_pass_fraction": 0.09538762633639072,
      "rewrite6_full_passes": 1.0,
      "step": 100490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.96448,
      "curriculum_late_phase": 1.0,
      "epoch": 9.25e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.195846557617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.59484,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.09601565694923175,
      "rewrite6_full_passes": 1.0,
      "step": 100500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 9.25e-06,
      "eval_loss": 3.0829901695251465,
      "eval_perplexity": 21.82356107715986,
      "eval_runtime": 0.0503,
      "eval_samples_per_second": 636.691,
      "eval_steps_per_second": 19.897,
      "step": 100500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9669376,
      "curriculum_late_phase": 1.0,
      "epoch": 9.255e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.230998992919922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0543,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5952,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.206335895020718,
      "rewrite6_current_pass_fraction": 0.0965901345640786,
      "rewrite6_full_passes": 1.0,
      "step": 100510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9693952,
      "curriculum_late_phase": 1.0,
      "epoch": 9.26e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.230998992919922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0525,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5955866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.16819882407055,
      "rewrite6_current_pass_fraction": 0.09723277054000896,
      "rewrite6_full_passes": 1.0,
      "step": 100520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9718528,
      "curriculum_late_phase": 1.0,
      "epoch": 9.265e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.230998992919922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0399,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5958933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.903152815291975,
      "rewrite6_current_pass_fraction": 0.09789974878775487,
      "rewrite6_full_passes": 1.0,
      "step": 100530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9743104,
      "curriculum_late_phase": 1.0,
      "epoch": 9.27e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.231128692626953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0631,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5961466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.393775174118343,
      "rewrite6_current_pass_fraction": 0.09851317403750658,
      "rewrite6_full_passes": 1.0,
      "step": 100540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.976768,
      "curriculum_late_phase": 1.0,
      "epoch": 9.275e-06,
      "grad_norm": 0.75,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0659,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.59656,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.453761686532072,
      "rewrite6_current_pass_fraction": 0.09907791474362719,
      "rewrite6_full_passes": 1.0,
      "step": 100550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9792256,
      "curriculum_late_phase": 1.0,
      "epoch": 9.28e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5969066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.968053132536788,
      "rewrite6_current_pass_fraction": 0.0996329185410216,
      "rewrite6_full_passes": 1.0,
      "step": 100560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9816832,
      "curriculum_late_phase": 1.0,
      "epoch": 9.285e-06,
      "grad_norm": 0.53515625,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0595,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5972,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.316896048946262,
      "rewrite6_current_pass_fraction": 0.1002366068820471,
      "rewrite6_full_passes": 1.0,
      "step": 100570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9841408,
      "curriculum_late_phase": 1.0,
      "epoch": 9.29e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5973866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.240293191045346,
      "rewrite6_current_pass_fraction": 0.10089384822106677,
      "rewrite6_full_passes": 1.0,
      "step": 100580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9865984,
      "curriculum_late_phase": 1.0,
      "epoch": 9.295e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.057,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5977066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.26367036864598,
      "rewrite6_current_pass_fraction": 0.1016338532842593,
      "rewrite6_full_passes": 1.0,
      "step": 100590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.989056,
      "curriculum_late_phase": 1.0,
      "epoch": 9.3e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.59808,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.959667588504388,
      "rewrite6_current_pass_fraction": 0.1023446476212732,
      "rewrite6_full_passes": 1.0,
      "step": 100600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9915136,
      "curriculum_late_phase": 1.0,
      "epoch": 9.305e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0236,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5984133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.565193277309945,
      "rewrite6_current_pass_fraction": 0.10290938832739382,
      "rewrite6_full_passes": 1.0,
      "step": 100610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9939712,
      "curriculum_late_phase": 1.0,
      "epoch": 9.31e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0357,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5987733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.81554368143424,
      "rewrite6_current_pass_fraction": 0.10343518139860956,
      "rewrite6_full_passes": 1.0,
      "step": 100620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9964288,
      "curriculum_late_phase": 1.0,
      "epoch": 9.315e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.59904,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.10406808046581371,
      "rewrite6_full_passes": 1.0,
      "step": 100630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 18.9988864,
      "curriculum_late_phase": 1.0,
      "epoch": 9.32e-06,
      "grad_norm": 0.53515625,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0666,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5993466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.468784577110917,
      "rewrite6_current_pass_fraction": 0.10459387353702947,
      "rewrite6_full_passes": 1.0,
      "step": 100640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.001344,
      "curriculum_late_phase": 1.0,
      "epoch": 9.325e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0874,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.5996666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.920011791213636,
      "rewrite6_current_pass_fraction": 0.10515861424315008,
      "rewrite6_full_passes": 1.0,
      "step": 100650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0038016,
      "curriculum_late_phase": 1.0,
      "epoch": 9.33e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6000933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86347453124931,
      "rewrite6_current_pass_fraction": 0.10577690794726491,
      "rewrite6_full_passes": 1.0,
      "step": 100660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0062592,
      "curriculum_late_phase": 1.0,
      "epoch": 9.335e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.231849670410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6003733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.197855056943403,
      "rewrite6_current_pass_fraction": 0.10641954392319526,
      "rewrite6_full_passes": 1.0,
      "step": 100670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0087168,
      "curriculum_late_phase": 1.0,
      "epoch": 9.34e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.827552795410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0861,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6006266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.89153429027125,
      "rewrite6_current_pass_fraction": 0.10707678526221495,
      "rewrite6_full_passes": 1.0,
      "step": 100680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0111744,
      "curriculum_late_phase": 1.0,
      "epoch": 9.345e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.827552795410156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0709,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6010533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.561299114498432,
      "rewrite6_current_pass_fraction": 0.10772915814687153,
      "rewrite6_full_passes": 1.0,
      "step": 100690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.013632,
      "curriculum_late_phase": 1.0,
      "epoch": 9.35e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.828472137451172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6013466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.99532932729427,
      "rewrite6_current_pass_fraction": 0.10829876730735526,
      "rewrite6_full_passes": 1.0,
      "step": 100700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0160896,
      "curriculum_late_phase": 1.0,
      "epoch": 9.355e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.828472137451172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0761,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.60164,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.67370988459726,
      "rewrite6_current_pass_fraction": 0.10885863955911278,
      "rewrite6_full_passes": 1.0,
      "step": 100710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0185472,
      "curriculum_late_phase": 1.0,
      "epoch": 9.36e-06,
      "grad_norm": 0.431640625,
      "host_rss_gib": 15.828472137451172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0357,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6020133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.81554368143424,
      "rewrite6_current_pass_fraction": 0.10952074935249557,
      "rewrite6_full_passes": 1.0,
      "step": 100720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0210048,
      "curriculum_late_phase": 1.0,
      "epoch": 9.365e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.828472137451172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6021733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.259418059817307,
      "rewrite6_current_pass_fraction": 0.11006601624116376,
      "rewrite6_full_passes": 1.0,
      "step": 100730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0234624,
      "curriculum_late_phase": 1.0,
      "epoch": 9.37e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.828472137451172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.062,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6025466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.37025495991624,
      "rewrite6_current_pass_fraction": 0.11063562540164748,
      "rewrite6_full_passes": 1.0,
      "step": 100740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.02592,
      "curriculum_late_phase": 1.0,
      "epoch": 9.375e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.828472137451172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0648,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6027733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.430175523444856,
      "rewrite6_current_pass_fraction": 0.11121983992522054,
      "rewrite6_full_passes": 1.0,
      "step": 100750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0283776,
      "curriculum_late_phase": 1.0,
      "epoch": 9.38e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0377,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6029466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.857216427652418,
      "rewrite6_current_pass_fraction": 0.11178944908570428,
      "rewrite6_full_passes": 1.0,
      "step": 100760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0308352,
      "curriculum_late_phase": 1.0,
      "epoch": 9.385e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0694,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6031333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.5289814101645,
      "rewrite6_current_pass_fraction": 0.11227142606765204,
      "rewrite6_full_passes": 1.0,
      "step": 100770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0332928,
      "curriculum_late_phase": 1.0,
      "epoch": 9.39e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.60336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.17878556994849,
      "rewrite6_current_pass_fraction": 0.11278261377577846,
      "rewrite6_full_passes": 1.0,
      "step": 100780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0357504,
      "curriculum_late_phase": 1.0,
      "epoch": 9.395e-06,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0634,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6036666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.400194269486743,
      "rewrite6_current_pass_fraction": 0.11339117057116707,
      "rewrite6_full_passes": 1.0,
      "step": 100790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.038208,
      "curriculum_late_phase": 1.0,
      "epoch": 9.4e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0627,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6040533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.38521937532253,
      "rewrite6_current_pass_fraction": 0.11394130591419836,
      "rewrite6_full_passes": 1.0,
      "step": 100800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0406656,
      "curriculum_late_phase": 1.0,
      "epoch": 9.405e-06,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0606,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.60436,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.340357536052313,
      "rewrite6_current_pass_fraction": 0.1145742049814025,
      "rewrite6_full_passes": 1.0,
      "step": 100810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0431232,
      "curriculum_late_phase": 1.0,
      "epoch": 9.41e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6046,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.11517302486806488,
      "rewrite6_full_passes": 1.0,
      "step": 100820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0455808,
      "curriculum_late_phase": 1.0,
      "epoch": 9.415e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0825,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6049333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.812866453892997,
      "rewrite6_current_pass_fraction": 0.1157328971198224,
      "rewrite6_full_passes": 1.0,
      "step": 100830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0480384,
      "curriculum_late_phase": 1.0,
      "epoch": 9.42e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.077,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6051866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.693225003979848,
      "rewrite6_current_pass_fraction": 0.11631224318903234,
      "rewrite6_full_passes": 1.0,
      "step": 100840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.050496,
      "curriculum_late_phase": 1.0,
      "epoch": 9.425e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0627,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6054266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.38521937532253,
      "rewrite6_current_pass_fraction": 0.1169500107105996,
      "rewrite6_full_passes": 1.0,
      "step": 100850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0529536,
      "curriculum_late_phase": 1.0,
      "epoch": 9.43e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0586,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6058,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.297719473255697,
      "rewrite6_current_pass_fraction": 0.1176121205039824,
      "rewrite6_full_passes": 1.0,
      "step": 100860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0554112,
      "curriculum_late_phase": 1.0,
      "epoch": 9.435e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6061066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.216941714202072,
      "rewrite6_current_pass_fraction": 0.11807949212284084,
      "rewrite6_full_passes": 1.0,
      "step": 100870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0578688,
      "curriculum_late_phase": 1.0,
      "epoch": 9.44e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.075,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6064133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.649881911512058,
      "rewrite6_current_pass_fraction": 0.11868318046386633,
      "rewrite6_full_passes": 1.0,
      "step": 100880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0603264,
      "curriculum_late_phase": 1.0,
      "epoch": 9.445e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0383,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.60664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.869734512558935,
      "rewrite6_current_pass_fraction": 0.11913107826527235,
      "rewrite6_full_passes": 1.0,
      "step": 100890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.062784,
      "curriculum_late_phase": 1.0,
      "epoch": 9.45e-06,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0554,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6069333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.229675699044027,
      "rewrite6_current_pass_fraction": 0.11971042433448228,
      "rewrite6_full_passes": 1.0,
      "step": 100900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0652416,
      "curriculum_late_phase": 1.0,
      "epoch": 9.455e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0598,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.60724,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.323292077117205,
      "rewrite6_current_pass_fraction": 0.12027516504060291,
      "rewrite6_full_passes": 1.0,
      "step": 100910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0676992,
      "curriculum_late_phase": 1.0,
      "epoch": 9.46e-06,
      "grad_norm": 0.5703125,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0735,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.60748,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.617431432588447,
      "rewrite6_current_pass_fraction": 0.12077661584000311,
      "rewrite6_full_passes": 1.0,
      "step": 100920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0701568,
      "curriculum_late_phase": 1.0,
      "epoch": 9.465e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6078266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.1213754357266655,
      "rewrite6_full_passes": 1.0,
      "step": 100930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0726144,
      "curriculum_late_phase": 1.0,
      "epoch": 9.47e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6081066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13435678680148,
      "rewrite6_current_pass_fraction": 0.12202780861132208,
      "rewrite6_full_passes": 1.0,
      "step": 100940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.075072,
      "curriculum_late_phase": 1.0,
      "epoch": 9.475e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6083466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.69516296961328,
      "rewrite6_current_pass_fraction": 0.12263149695234757,
      "rewrite6_full_passes": 1.0,
      "step": 100950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0775296,
      "curriculum_late_phase": 1.0,
      "epoch": 9.48e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6086533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.035258373598133,
      "rewrite6_current_pass_fraction": 0.12324005374773617,
      "rewrite6_full_passes": 1.0,
      "step": 100960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0799872,
      "curriculum_late_phase": 1.0,
      "epoch": 9.485e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6089066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.936624630107648,
      "rewrite6_current_pass_fraction": 0.12384374208876166,
      "rewrite6_full_passes": 1.0,
      "step": 100970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0824448,
      "curriculum_late_phase": 1.0,
      "epoch": 9.49e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.071,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6091733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.563455352219975,
      "rewrite6_current_pass_fraction": 0.12434519288816187,
      "rewrite6_full_passes": 1.0,
      "step": 100980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0849024,
      "curriculum_late_phase": 1.0,
      "epoch": 9.495e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0695,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6094666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.531134415954018,
      "rewrite6_current_pass_fraction": 0.12496835504663979,
      "rewrite6_full_passes": 1.0,
      "step": 100990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.08736,
      "curriculum_late_phase": 1.0,
      "epoch": 9.5e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.828487396240234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0705,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6098666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.552676319526597,
      "rewrite6_current_pass_fraction": 0.12560125411384393,
      "rewrite6_full_passes": 1.0,
      "step": 101000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 9.5e-06,
      "eval_loss": 3.0830399990081787,
      "eval_perplexity": 21.82464856102041,
      "eval_runtime": 0.0521,
      "eval_samples_per_second": 613.819,
      "eval_steps_per_second": 19.182,
      "step": 101000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0898176,
      "curriculum_late_phase": 1.0,
      "epoch": 9.505e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.832412719726562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0793,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6102133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.743176844084687,
      "rewrite6_current_pass_fraction": 0.1261757317286908,
      "rewrite6_full_passes": 1.0,
      "step": 101010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0922752,
      "curriculum_late_phase": 1.0,
      "epoch": 9.51e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.832412719726562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0383,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6106533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.869734512558935,
      "rewrite6_current_pass_fraction": 0.12677942006971626,
      "rewrite6_full_passes": 1.0,
      "step": 101020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0947328,
      "curriculum_late_phase": 1.0,
      "epoch": 9.515e-06,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.832412719726562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0612,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.611,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35316559260667,
      "rewrite6_current_pass_fraction": 0.12745126677182528,
      "rewrite6_full_passes": 1.0,
      "step": 101030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.0971904,
      "curriculum_late_phase": 1.0,
      "epoch": 9.52e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.262069702148438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0632,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6113066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3959146586082,
      "rewrite6_current_pass_fraction": 0.1280744289303032,
      "rewrite6_full_passes": 1.0,
      "step": 101040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.099648,
      "curriculum_late_phase": 1.0,
      "epoch": 9.525e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.264152526855469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0695,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6115866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.531134415954018,
      "rewrite6_current_pass_fraction": 0.1286781172713287,
      "rewrite6_full_passes": 1.0,
      "step": 101050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1021056,
      "curriculum_late_phase": 1.0,
      "epoch": 9.53e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.264724731445312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0499,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.61188,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.113232993671563,
      "rewrite6_current_pass_fraction": 0.12915522579891336,
      "rewrite6_full_passes": 1.0,
      "step": 101060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1045632,
      "curriculum_late_phase": 1.0,
      "epoch": 9.535e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.267845153808594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.61212,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.974344492133298,
      "rewrite6_current_pass_fraction": 0.12975404568557575,
      "rewrite6_full_passes": 1.0,
      "step": 101070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1070208,
      "curriculum_late_phase": 1.0,
      "epoch": 9.54e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.268489837646484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0719,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6124933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.58287119785693,
      "rewrite6_current_pass_fraction": 0.1302262857587973,
      "rewrite6_full_passes": 1.0,
      "step": 101080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1094784,
      "curriculum_late_phase": 1.0,
      "epoch": 9.545e-06,
      "grad_norm": 0.53515625,
      "host_rss_gib": 15.268741607666016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0657,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6128,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.449471363241397,
      "rewrite6_current_pass_fraction": 0.13080563182800725,
      "rewrite6_full_passes": 1.0,
      "step": 101090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.111936,
      "curriculum_late_phase": 1.0,
      "epoch": 9.55e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.269248962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0647,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6131466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.42803261303982,
      "rewrite6_current_pass_fraction": 0.1314433993495745,
      "rewrite6_full_passes": 1.0,
      "step": 101100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1143936,
      "curriculum_late_phase": 1.0,
      "epoch": 9.555e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.269248962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0574,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6135466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.2721775381139,
      "rewrite6_current_pass_fraction": 0.13197406087515334,
      "rewrite6_full_passes": 1.0,
      "step": 101110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1168512,
      "curriculum_late_phase": 1.0,
      "epoch": 9.56e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.269603729248047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6138666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.221185526912038,
      "rewrite6_current_pass_fraction": 0.13253393312691086,
      "rewrite6_full_passes": 1.0,
      "step": 101120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1193088,
      "curriculum_late_phase": 1.0,
      "epoch": 9.565e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.269607543945312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6141733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.961763660065063,
      "rewrite6_current_pass_fraction": 0.1331084107417577,
      "rewrite6_full_passes": 1.0,
      "step": 101130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1217664,
      "curriculum_late_phase": 1.0,
      "epoch": 9.57e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.871650695800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.61448,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10689997377394,
      "rewrite6_current_pass_fraction": 0.1337753889895036,
      "rewrite6_full_passes": 1.0,
      "step": 101140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.124224,
      "curriculum_late_phase": 1.0,
      "epoch": 9.575e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.57955551147461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.61488,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.13427197133454072,
      "rewrite6_full_passes": 1.0,
      "step": 101150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1266816,
      "curriculum_late_phase": 1.0,
      "epoch": 9.58e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.57955551147461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0477,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6152133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.066834937660975,
      "rewrite6_current_pass_fraction": 0.13490000194738175,
      "rewrite6_full_passes": 1.0,
      "step": 101160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1291392,
      "curriculum_late_phase": 1.0,
      "epoch": 9.585e-06,
      "grad_norm": 0.55859375,
      "host_rss_gib": 16.57956314086914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0681,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6155066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.501011918439943,
      "rewrite6_current_pass_fraction": 0.135493953379681,
      "rewrite6_full_passes": 1.0,
      "step": 101170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1315968,
      "curriculum_late_phase": 1.0,
      "epoch": 9.59e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.57956314086914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6158666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91988202833721,
      "rewrite6_current_pass_fraction": 0.13606356254016475,
      "rewrite6_full_passes": 1.0,
      "step": 101180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1340544,
      "curriculum_late_phase": 1.0,
      "epoch": 9.595e-06,
      "grad_norm": 9.5625,
      "host_rss_gib": 16.57956314086914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0788,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6162,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.73230797310683,
      "rewrite6_current_pass_fraction": 0.1367110669704582,
      "rewrite6_full_passes": 1.0,
      "step": 101190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.136512,
      "curriculum_late_phase": 1.0,
      "epoch": 9.6e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 16.57956314086914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0282,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.61656,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.660011080136737,
      "rewrite6_current_pass_fraction": 0.1373147553114837,
      "rewrite6_full_passes": 1.0,
      "step": 101200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1389696,
      "curriculum_late_phase": 1.0,
      "epoch": 9.605e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.57956314086914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6167733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20845663464543,
      "rewrite6_current_pass_fraction": 0.13801094428540828,
      "rewrite6_full_passes": 1.0,
      "step": 101210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1414272,
      "curriculum_late_phase": 1.0,
      "epoch": 9.61e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.57956314086914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6170533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.064728359497874,
      "rewrite6_current_pass_fraction": 0.13859515880898132,
      "rewrite6_full_passes": 1.0,
      "step": 101220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1438848,
      "curriculum_late_phase": 1.0,
      "epoch": 9.615e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.57980728149414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6175066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.970150042693795,
      "rewrite6_current_pass_fraction": 0.13923292633054857,
      "rewrite6_full_passes": 1.0,
      "step": 101230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1463424,
      "curriculum_late_phase": 1.0,
      "epoch": 9.62e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.57980728149414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0192,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.61788,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.474905206310687,
      "rewrite6_current_pass_fraction": 0.13976845631049054,
      "rewrite6_full_passes": 1.0,
      "step": 101240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1488,
      "curriculum_late_phase": 1.0,
      "epoch": 9.625e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.58053970336914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0391,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.61824,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.886436980265266,
      "rewrite6_current_pass_fraction": 0.1403526708340636,
      "rewrite6_full_passes": 1.0,
      "step": 101250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1512576,
      "curriculum_late_phase": 1.0,
      "epoch": 9.63e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.58053970336914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0594,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6185333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3147644659223,
      "rewrite6_current_pass_fraction": 0.14099043835563085,
      "rewrite6_full_passes": 1.0,
      "step": 101260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1537152,
      "curriculum_late_phase": 1.0,
      "epoch": 9.635e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.58053970336914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.61888,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.153386269892504,
      "rewrite6_current_pass_fraction": 0.14163307433156122,
      "rewrite6_full_passes": 1.0,
      "step": 101270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1561728,
      "curriculum_late_phase": 1.0,
      "epoch": 9.64e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.58053970336914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6192533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.959667588504388,
      "rewrite6_current_pass_fraction": 0.14213939358532454,
      "rewrite6_full_passes": 1.0,
      "step": 101280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1586304,
      "curriculum_late_phase": 1.0,
      "epoch": 9.645e-06,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.58053970336914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.072,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6197066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.585029592894678,
      "rewrite6_current_pass_fraction": 0.14266031820217717,
      "rewrite6_full_passes": 1.0,
      "step": 101290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.161088,
      "curriculum_late_phase": 1.0,
      "epoch": 9.65e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.58053970336914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0818,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6198933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.7976027902808,
      "rewrite6_current_pass_fraction": 0.1432201904539347,
      "rewrite6_full_passes": 1.0,
      "step": 101300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1635456,
      "curriculum_late_phase": 1.0,
      "epoch": 9.655e-06,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.58053970336914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0713,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6202533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.56992535927817,
      "rewrite6_current_pass_fraction": 0.14387256333859127,
      "rewrite6_full_passes": 1.0,
      "step": 101310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1660032,
      "curriculum_late_phase": 1.0,
      "epoch": 9.66e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.58053970336914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6206133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.295589807793416,
      "rewrite6_current_pass_fraction": 0.14444217249907498,
      "rewrite6_full_passes": 1.0,
      "step": 101320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1684608,
      "curriculum_late_phase": 1.0,
      "epoch": 9.665e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.58053970336914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0763,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.62108,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.678045060077274,
      "rewrite6_current_pass_fraction": 0.14496309711592764,
      "rewrite6_full_passes": 1.0,
      "step": 101330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1709184,
      "curriculum_late_phase": 1.0,
      "epoch": 9.67e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.58102035522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0619,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.62136,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.36811804126797,
      "rewrite6_current_pass_fraction": 0.1455959961831318,
      "rewrite6_full_passes": 1.0,
      "step": 101340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.173376,
      "curriculum_late_phase": 1.0,
      "epoch": 9.675e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.58102035522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0802,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6217866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.76275451187338,
      "rewrite6_current_pass_fraction": 0.14619968452415727,
      "rewrite6_full_passes": 1.0,
      "step": 101350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1758336,
      "curriculum_late_phase": 1.0,
      "epoch": 9.68e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.58102035522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0448,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6220533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.005829616811546,
      "rewrite6_current_pass_fraction": 0.1467595567759148,
      "rewrite6_full_passes": 1.0,
      "step": 101360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1782912,
      "curriculum_late_phase": 1.0,
      "epoch": 9.685e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.58102035522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0409,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6223066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.924066423168412,
      "rewrite6_current_pass_fraction": 0.14739732429748204,
      "rewrite6_full_passes": 1.0,
      "step": 101370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1807488,
      "curriculum_late_phase": 1.0,
      "epoch": 9.69e-06,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.58102035522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6225066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.71586848361443,
      "rewrite6_current_pass_fraction": 0.14803509181904928,
      "rewrite6_full_passes": 1.0,
      "step": 101380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1832064,
      "curriculum_late_phase": 1.0,
      "epoch": 9.695e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.58102035522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6227733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.02684595284501,
      "rewrite6_current_pass_fraction": 0.14858035870771746,
      "rewrite6_full_passes": 1.0,
      "step": 101390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.185664,
      "curriculum_late_phase": 1.0,
      "epoch": 9.7e-06,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.58102035522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0365,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6231066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.832202779129986,
      "rewrite6_current_pass_fraction": 0.14923760004673717,
      "rewrite6_full_passes": 1.0,
      "step": 101400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1881216,
      "curriculum_late_phase": 1.0,
      "epoch": 9.705e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.58102035522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0422,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.62344,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.951285398018836,
      "rewrite6_current_pass_fraction": 0.14983641993339955,
      "rewrite6_full_passes": 1.0,
      "step": 101410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1905792,
      "curriculum_late_phase": 1.0,
      "epoch": 9.71e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.58102035522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6237733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.151271037028916,
      "rewrite6_current_pass_fraction": 0.15037194991334152,
      "rewrite6_full_passes": 1.0,
      "step": 101420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1930368,
      "curriculum_late_phase": 1.0,
      "epoch": 9.715e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.58102035522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0251,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6241866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.596064214640606,
      "rewrite6_current_pass_fraction": 0.15098050670873012,
      "rewrite6_full_passes": 1.0,
      "step": 101430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.1954944,
      "curriculum_late_phase": 1.0,
      "epoch": 9.72e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.581024169921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0561,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.62444,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.244541674517738,
      "rewrite6_current_pass_fraction": 0.15162314268466048,
      "rewrite6_full_passes": 1.0,
      "step": 101440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.197952,
      "curriculum_late_phase": 1.0,
      "epoch": 9.725e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.581024169921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0701,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.62472,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.54405697298302,
      "rewrite6_current_pass_fraction": 0.15221709411695974,
      "rewrite6_full_passes": 1.0,
      "step": 101450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2004096,
      "curriculum_late_phase": 1.0,
      "epoch": 9.73e-06,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.58191680908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0684,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6252266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50746318965778,
      "rewrite6_current_pass_fraction": 0.1527964401861697,
      "rewrite6_full_passes": 1.0,
      "step": 101460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2028672,
      "curriculum_late_phase": 1.0,
      "epoch": 9.735e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.58191680908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0344,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6256266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.788501056163312,
      "rewrite6_current_pass_fraction": 0.1534147338902845,
      "rewrite6_full_passes": 1.0,
      "step": 101470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2053248,
      "curriculum_late_phase": 1.0,
      "epoch": 9.74e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.58191680908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.1012,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.62592,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.224604811899212,
      "rewrite6_current_pass_fraction": 0.15396486923331582,
      "rewrite6_full_passes": 1.0,
      "step": 101480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2077824,
      "curriculum_late_phase": 1.0,
      "epoch": 9.745e-06,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.58191680908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.62616,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.240293191045346,
      "rewrite6_current_pass_fraction": 0.15467079511596657,
      "rewrite6_full_passes": 1.0,
      "step": 101490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.21024,
      "curriculum_late_phase": 1.0,
      "epoch": 9.75e-06,
      "grad_norm": 0.64453125,
      "host_rss_gib": 16.58191680908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0827,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6264533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.817229463470188,
      "rewrite6_current_pass_fraction": 0.1551868512784561,
      "rewrite6_full_passes": 1.0,
      "step": 101500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 9.75e-06,
      "eval_loss": 3.082852363586426,
      "eval_perplexity": 21.82055386804979,
      "eval_runtime": 0.051,
      "eval_samples_per_second": 626.855,
      "eval_steps_per_second": 19.589,
      "step": 101500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2126976,
      "curriculum_late_phase": 1.0,
      "epoch": 9.755e-06,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.625972747802734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0559,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6268133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.240293191045346,
      "rewrite6_current_pass_fraction": 0.15572238125839807,
      "rewrite6_full_passes": 1.0,
      "step": 101510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2151552,
      "curriculum_late_phase": 1.0,
      "epoch": 9.76e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.625972747802734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0405,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6271733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91569847030129,
      "rewrite6_current_pass_fraction": 0.15640883332359642,
      "rewrite6_full_passes": 1.0,
      "step": 101520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2176128,
      "curriculum_late_phase": 1.0,
      "epoch": 9.765e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.625972747802734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6274133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.859302253584733,
      "rewrite6_current_pass_fraction": 0.15702712702771124,
      "rewrite6_full_passes": 1.0,
      "step": 101530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2200704,
      "curriculum_late_phase": 1.0,
      "epoch": 9.77e-06,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.625972747802734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0487,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6278533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08791230952813,
      "rewrite6_current_pass_fraction": 0.15766489454927848,
      "rewrite6_full_passes": 1.0,
      "step": 101540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.222528,
      "curriculum_late_phase": 1.0,
      "epoch": 9.775e-06,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.625972747802734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.62828,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.859302253584733,
      "rewrite6_current_pass_fraction": 0.15821016143794667,
      "rewrite6_full_passes": 1.0,
      "step": 101550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2249856,
      "curriculum_late_phase": 1.0,
      "epoch": 9.78e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.625972747802734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.062,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6286533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.37025495991624,
      "rewrite6_current_pass_fraction": 0.15879437596151974,
      "rewrite6_full_passes": 1.0,
      "step": 101560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2274432,
      "curriculum_late_phase": 1.0,
      "epoch": 9.785e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.62689208984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6291466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.15942240657436077,
      "rewrite6_full_passes": 1.0,
      "step": 101570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2299008,
      "curriculum_late_phase": 1.0,
      "epoch": 9.79e-06,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.62689208984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0223,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6294,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.538475896109915,
      "rewrite6_current_pass_fraction": 0.16002122646102315,
      "rewrite6_full_passes": 1.0,
      "step": 101580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2323584,
      "curriculum_late_phase": 1.0,
      "epoch": 9.795e-06,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.62689208984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0627,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6296533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.38521937532253,
      "rewrite6_current_pass_fraction": 0.1606443886195011,
      "rewrite6_full_passes": 1.0,
      "step": 101590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.234816,
      "curriculum_late_phase": 1.0,
      "epoch": 9.8e-06,
      "grad_norm": 0.4375,
      "host_rss_gib": 16.62689208984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0649,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.62996,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.432318648151654,
      "rewrite6_current_pass_fraction": 0.1612139977799848,
      "rewrite6_full_passes": 1.0,
      "step": 101600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2372736,
      "curriculum_late_phase": 1.0,
      "epoch": 9.805e-06,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.62689208984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0648,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6301733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.430175523444856,
      "rewrite6_current_pass_fraction": 0.16184202839282585,
      "rewrite6_full_passes": 1.0,
      "step": 101610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2397312,
      "curriculum_late_phase": 1.0,
      "epoch": 9.81e-06,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0747,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6305866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.643387921085868,
      "rewrite6_current_pass_fraction": 0.16243111137076202,
      "rewrite6_full_passes": 1.0,
      "step": 101620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2421888,
      "curriculum_late_phase": 1.0,
      "epoch": 9.815e-06,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6308933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.225430188469428,
      "rewrite6_current_pass_fraction": 0.16305914198360305,
      "rewrite6_full_passes": 1.0,
      "step": 101630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2446464,
      "curriculum_late_phase": 1.0,
      "epoch": 9.82e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0685,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6312133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50961404351764,
      "rewrite6_current_pass_fraction": 0.1636530934159023,
      "rewrite6_full_passes": 1.0,
      "step": 101640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.247104,
      "curriculum_late_phase": 1.0,
      "epoch": 9.825e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0373,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6314133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.848875209436216,
      "rewrite6_current_pass_fraction": 0.16424217639383848,
      "rewrite6_full_passes": 1.0,
      "step": 101650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2495616,
      "curriculum_late_phase": 1.0,
      "epoch": 9.83e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6317466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.986932874965774,
      "rewrite6_current_pass_fraction": 0.16485073318922708,
      "rewrite6_full_passes": 1.0,
      "step": 101660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2520192,
      "curriculum_late_phase": 1.0,
      "epoch": 9.835e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0562,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6320266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.246666234911444,
      "rewrite6_current_pass_fraction": 0.16550310607388366,
      "rewrite6_full_passes": 1.0,
      "step": 101670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2544768,
      "curriculum_late_phase": 1.0,
      "epoch": 9.84e-06,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0562,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6324266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.246666234911444,
      "rewrite6_current_pass_fraction": 0.16609705750618295,
      "rewrite6_full_passes": 1.0,
      "step": 101680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2569344,
      "curriculum_late_phase": 1.0,
      "epoch": 9.845e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.078,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6327333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.71492907921277,
      "rewrite6_current_pass_fraction": 0.16670561430157155,
      "rewrite6_full_passes": 1.0,
      "step": 101690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.259392,
      "curriculum_late_phase": 1.0,
      "epoch": 9.85e-06,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0523,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.63308,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.16396560764148,
      "rewrite6_current_pass_fraction": 0.1674066717298592,
      "rewrite6_full_passes": 1.0,
      "step": 101700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2618496,
      "curriculum_late_phase": 1.0,
      "epoch": 9.855e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6334133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.197855056943403,
      "rewrite6_current_pass_fraction": 0.16807851843196822,
      "rewrite6_full_passes": 1.0,
      "step": 101710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2643072,
      "curriculum_late_phase": 1.0,
      "epoch": 9.86e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6337733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21482012611182,
      "rewrite6_current_pass_fraction": 0.16868707522735682,
      "rewrite6_full_passes": 1.0,
      "step": 101720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2667648,
      "curriculum_late_phase": 1.0,
      "epoch": 9.865e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0618,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6340933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.365981336300866,
      "rewrite6_current_pass_fraction": 0.16936379038382895,
      "rewrite6_full_passes": 1.0,
      "step": 101730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2692224,
      "curriculum_late_phase": 1.0,
      "epoch": 9.87e-06,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0561,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6343466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.244541674517738,
      "rewrite6_current_pass_fraction": 0.17001129481412242,
      "rewrite6_full_passes": 1.0,
      "step": 101740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.27168,
      "curriculum_late_phase": 1.0,
      "epoch": 9.875e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0522,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6347733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.161849316897023,
      "rewrite6_current_pass_fraction": 0.1705565617027906,
      "rewrite6_full_passes": 1.0,
      "step": 101750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2741376,
      "curriculum_late_phase": 1.0,
      "epoch": 9.88e-06,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6350933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.058409888808015,
      "rewrite6_current_pass_fraction": 0.1711213024089112,
      "rewrite6_full_passes": 1.0,
      "step": 101760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2765952,
      "curriculum_late_phase": 1.0,
      "epoch": 9.885e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6354133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.759417517911764,
      "rewrite6_current_pass_fraction": 0.17175420147611536,
      "rewrite6_full_passes": 1.0,
      "step": 101770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2790528,
      "curriculum_late_phase": 1.0,
      "epoch": 9.89e-06,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.63576,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.934531072324265,
      "rewrite6_current_pass_fraction": 0.17231894218223598,
      "rewrite6_full_passes": 1.0,
      "step": 101780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2815104,
      "curriculum_late_phase": 1.0,
      "epoch": 9.895e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0526,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.63604,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.170315749797474,
      "rewrite6_current_pass_fraction": 0.17296644661252947,
      "rewrite6_full_passes": 1.0,
      "step": 101790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.283968,
      "curriculum_late_phase": 1.0,
      "epoch": 9.9e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0813,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.63632,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.786706713131952,
      "rewrite6_current_pass_fraction": 0.17356039804482873,
      "rewrite6_full_passes": 1.0,
      "step": 101800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2864256,
      "curriculum_late_phase": 1.0,
      "epoch": 9.905e-06,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0731,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6366666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.608786189179362,
      "rewrite6_current_pass_fraction": 0.1741056649334969,
      "rewrite6_full_passes": 1.0,
      "step": 101810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2888832,
      "curriculum_late_phase": 1.0,
      "epoch": 9.91e-06,
      "grad_norm": 0.58984375,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0502,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6370666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11956791376016,
      "rewrite6_current_pass_fraction": 0.17473369554633794,
      "rewrite6_full_passes": 1.0,
      "step": 101820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2913408,
      "curriculum_late_phase": 1.0,
      "epoch": 9.915e-06,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.63736,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.014233629348723,
      "rewrite6_current_pass_fraction": 0.1753665946135421,
      "rewrite6_full_passes": 1.0,
      "step": 101830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2937984,
      "curriculum_late_phase": 1.0,
      "epoch": 9.92e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.058,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6377066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.284944674394637,
      "rewrite6_current_pass_fraction": 0.1759702829545676,
      "rewrite6_full_passes": 1.0,
      "step": 101840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.296256,
      "curriculum_late_phase": 1.0,
      "epoch": 9.925e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6380933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.686886559817744,
      "rewrite6_current_pass_fraction": 0.1765301552063251,
      "rewrite6_full_passes": 1.0,
      "step": 101850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.2987136,
      "curriculum_late_phase": 1.0,
      "epoch": 9.93e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6383333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.191496654234413,
      "rewrite6_current_pass_fraction": 0.17709976436680883,
      "rewrite6_full_passes": 1.0,
      "step": 101860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3011712,
      "curriculum_late_phase": 1.0,
      "epoch": 9.935e-06,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6387333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.17785437479309069,
      "rewrite6_full_passes": 1.0,
      "step": 101870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3036288,
      "curriculum_late_phase": 1.0,
      "epoch": 9.94e-06,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6391733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.17854569531265213,
      "rewrite6_full_passes": 1.0,
      "step": 101880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3060864,
      "curriculum_late_phase": 1.0,
      "epoch": 9.945e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0801,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6395866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.760578345232332,
      "rewrite6_current_pass_fraction": 0.1791737259254932,
      "rewrite6_full_passes": 1.0,
      "step": 101890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.308544,
      "curriculum_late_phase": 1.0,
      "epoch": 9.95e-06,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.626903533935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0331,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.63992,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.761493563464107,
      "rewrite6_current_pass_fraction": 0.1798601779906915,
      "rewrite6_full_passes": 1.0,
      "step": 101900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3110016,
      "curriculum_late_phase": 1.0,
      "epoch": 9.955e-06,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.083736419677734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0607,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6402133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34249167851127,
      "rewrite6_current_pass_fraction": 0.18044926096862768,
      "rewrite6_full_passes": 1.0,
      "step": 101910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3134592,
      "curriculum_late_phase": 1.0,
      "epoch": 9.96e-06,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.087696075439453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6404933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.1810140016747483,
      "rewrite6_full_passes": 1.0,
      "step": 101920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3159168,
      "curriculum_late_phase": 1.0,
      "epoch": 9.965e-06,
      "grad_norm": 0.51953125,
      "host_rss_gib": 16.087696075439453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0541,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6408,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20209505194016,
      "rewrite6_current_pass_fraction": 0.18161282156141068,
      "rewrite6_full_passes": 1.0,
      "step": 101930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3183744,
      "curriculum_late_phase": 1.0,
      "epoch": 9.97e-06,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.0882568359375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0605,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.64108,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.338223606996944,
      "rewrite6_current_pass_fraction": 0.18215808845007886,
      "rewrite6_full_passes": 1.0,
      "step": 101940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.320832,
      "curriculum_late_phase": 1.0,
      "epoch": 9.975e-06,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.090133666992188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.64136,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.892703851343008,
      "rewrite6_current_pass_fraction": 0.18280072442600923,
      "rewrite6_full_passes": 1.0,
      "step": 101950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3232896,
      "curriculum_late_phase": 1.0,
      "epoch": 9.98e-06,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.09032440185547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0463,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6417333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.037362004615296,
      "rewrite6_current_pass_fraction": 0.18342388658448716,
      "rewrite6_full_passes": 1.0,
      "step": 101960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3257472,
      "curriculum_late_phase": 1.0,
      "epoch": 9.985e-06,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.091285705566406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0318,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.64204,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.734521157693965,
      "rewrite6_current_pass_fraction": 0.18397889038188156,
      "rewrite6_full_passes": 1.0,
      "step": 101970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3282048,
      "curriculum_late_phase": 1.0,
      "epoch": 9.99e-06,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.672096252441406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6424133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.090021206202152,
      "rewrite6_current_pass_fraction": 0.18453876263363908,
      "rewrite6_full_passes": 1.0,
      "step": 101980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3306624,
      "curriculum_late_phase": 1.0,
      "epoch": 9.995e-06,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.160892486572266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6427066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.185161924792117,
      "rewrite6_full_passes": 1.0,
      "step": 101990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.33312,
      "curriculum_late_phase": 1.0,
      "epoch": 1e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.161029815673828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.64304,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.125904734609893,
      "rewrite6_current_pass_fraction": 0.185653638682791,
      "rewrite6_full_passes": 1.0,
      "step": 102000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1e-05,
      "eval_loss": 3.083085060119629,
      "eval_perplexity": 21.825632026099427,
      "eval_runtime": 0.049,
      "eval_samples_per_second": 653.376,
      "eval_steps_per_second": 20.418,
      "step": 102000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3355776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0005e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.32880401611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0588,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6433066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.301979443133135,
      "rewrite6_current_pass_fraction": 0.18628653774999512,
      "rewrite6_full_passes": 1.0,
      "step": 102010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3380352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.001e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.323455810546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.052,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6436266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.157617370242413,
      "rewrite6_current_pass_fraction": 0.18686101536484198,
      "rewrite6_full_passes": 1.0,
      "step": 102020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3404928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0015e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.323455810546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0374,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6439466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.850960201205005,
      "rewrite6_current_pass_fraction": 0.1874403614340519,
      "rewrite6_full_passes": 1.0,
      "step": 102030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3429504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.002e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.323455810546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6444,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.062621991982052,
      "rewrite6_current_pass_fraction": 0.187980759868357,
      "rewrite6_full_passes": 1.0,
      "step": 102040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.345408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0025e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.323455810546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0328,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6446666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.755266049568853,
      "rewrite6_current_pass_fraction": 0.18856497439193004,
      "rewrite6_full_passes": 1.0,
      "step": 102050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3478656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.003e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.323455810546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0211,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6450533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.513844506823926,
      "rewrite6_current_pass_fraction": 0.18923682109403905,
      "rewrite6_full_passes": 1.0,
      "step": 102060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3503232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0035e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.323455810546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0308,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6453333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.713797000341962,
      "rewrite6_current_pass_fraction": 0.1897626141652548,
      "rewrite6_full_passes": 1.0,
      "step": 102070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3527808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.004e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.32345962524414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0707,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.64568,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.55698728587277,
      "rewrite6_current_pass_fraction": 0.1903030125995599,
      "rewrite6_full_passes": 1.0,
      "step": 102080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3552384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0045e-05,
      "grad_norm": 0.5546875,
      "host_rss_gib": 16.32345962524414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0658,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6460533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45161641762865,
      "rewrite6_current_pass_fraction": 0.190799594944597,
      "rewrite6_full_passes": 1.0,
      "step": 102090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.357696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.005e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.32345962524414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0681,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6462933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.501011918439943,
      "rewrite6_current_pass_fraction": 0.19137407255944383,
      "rewrite6_full_passes": 1.0,
      "step": 102100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3601536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0055e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.324081420898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0541,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6466533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20209505194016,
      "rewrite6_current_pass_fraction": 0.19208973535082083,
      "rewrite6_full_passes": 1.0,
      "step": 102110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3626112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.006e-05,
      "grad_norm": 0.5859375,
      "host_rss_gib": 16.324081420898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0693,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6470666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.52682861966481,
      "rewrite6_current_pass_fraction": 0.1927469766898405,
      "rewrite6_full_passes": 1.0,
      "step": 102120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3650688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0065e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.324081420898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0673,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.64748,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.483817987394623,
      "rewrite6_current_pass_fraction": 0.19325816439796695,
      "rewrite6_full_passes": 1.0,
      "step": 102130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3675264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.007e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.324081420898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.64776,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.60662541860077,
      "rewrite6_current_pass_fraction": 0.19382290510408756,
      "rewrite6_full_passes": 1.0,
      "step": 102140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.369984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0075e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.324081420898438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6480933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.24241732656946,
      "rewrite6_current_pass_fraction": 0.1943535666296664,
      "rewrite6_full_passes": 1.0,
      "step": 102150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3724416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.008e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.324726104736328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0691,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6483066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.52252368444875,
      "rewrite6_current_pass_fraction": 0.1950156764230492,
      "rewrite6_full_passes": 1.0,
      "step": 102160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3748992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0085e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.324726104736328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0495,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6486,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.104789389307548,
      "rewrite6_current_pass_fraction": 0.1956096278553485,
      "rewrite6_full_passes": 1.0,
      "step": 102170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3773568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.009e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.324726104736328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6489866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.957571726540376,
      "rewrite6_current_pass_fraction": 0.19618410547019532,
      "rewrite6_full_passes": 1.0,
      "step": 102180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3798144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0095e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.324726104736328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6493733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.216941714202072,
      "rewrite6_current_pass_fraction": 0.19673424081322663,
      "rewrite6_full_passes": 1.0,
      "step": 102190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.382272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.01e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.324726104736328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0696,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.64972,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.53328763705487,
      "rewrite6_current_pass_fraction": 0.19744016669587738,
      "rewrite6_full_passes": 1.0,
      "step": 102200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3847296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0105e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0995,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.65004,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.18685508008242,
      "rewrite6_current_pass_fraction": 0.19809253958053397,
      "rewrite6_full_passes": 1.0,
      "step": 102210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3871872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.011e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.061,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6503333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34889538652299,
      "rewrite6_current_pass_fraction": 0.19869135946719635,
      "rewrite6_full_passes": 1.0,
      "step": 102220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3896448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0115e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0686,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6505866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.51176511247365,
      "rewrite6_current_pass_fraction": 0.19923175790150144,
      "rewrite6_full_passes": 1.0,
      "step": 102230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3921024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.012e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6508533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.60662541860077,
      "rewrite6_current_pass_fraction": 0.20003505287141438,
      "rewrite6_full_passes": 1.0,
      "step": 102240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.39456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0125e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0659,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6512666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.453761686532072,
      "rewrite6_current_pass_fraction": 0.20050242449027283,
      "rewrite6_full_passes": 1.0,
      "step": 102250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3970176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.013e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0698,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6514533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.537594725276744,
      "rewrite6_current_pass_fraction": 0.2011061128312983,
      "rewrite6_full_passes": 1.0,
      "step": 102260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.3994752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0135e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.06,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6517466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.327557162026903,
      "rewrite6_current_pass_fraction": 0.20167085353741895,
      "rewrite6_full_passes": 1.0,
      "step": 102270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4019328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.014e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.65216,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.970150042693795,
      "rewrite6_current_pass_fraction": 0.20224533115226578,
      "rewrite6_full_passes": 1.0,
      "step": 102280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4043904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0145e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0456,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6525733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.022641004163326,
      "rewrite6_current_pass_fraction": 0.2028733617651068,
      "rewrite6_full_passes": 1.0,
      "step": 102290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.406848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.015e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0596,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6528666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.319027845139193,
      "rewrite6_current_pass_fraction": 0.2035306031041265,
      "rewrite6_full_passes": 1.0,
      "step": 102300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4093056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0155e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0886,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6531733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.946331594086434,
      "rewrite6_current_pass_fraction": 0.20410508071897335,
      "rewrite6_full_passes": 1.0,
      "step": 102310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4117632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.016e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6535333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23604555718468,
      "rewrite6_current_pass_fraction": 0.20466008451636775,
      "rewrite6_full_passes": 1.0,
      "step": 102320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4142208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0165e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0602,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6537733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.33182310003889,
      "rewrite6_current_pass_fraction": 0.20520535140503593,
      "rewrite6_full_passes": 1.0,
      "step": 102330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4166784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.017e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0611,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.65416,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.351030382809682,
      "rewrite6_current_pass_fraction": 0.20575548674806723,
      "rewrite6_full_passes": 1.0,
      "step": 102340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.419136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0175e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0398,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6545333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.901062604522732,
      "rewrite6_current_pass_fraction": 0.2063494381803665,
      "rewrite6_full_passes": 1.0,
      "step": 102350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4215936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.018e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.324729919433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.65488,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.144926607429838,
      "rewrite6_current_pass_fraction": 0.20700181106502308,
      "rewrite6_full_passes": 1.0,
      "step": 102360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4240512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0185e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.754417419433594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.65528,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.828036755190443,
      "rewrite6_current_pass_fraction": 0.2075762886798699,
      "rewrite6_full_passes": 1.0,
      "step": 102370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4265088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.019e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6556133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.911515748893304,
      "rewrite6_current_pass_fraction": 0.20825787229070514,
      "rewrite6_full_passes": 1.0,
      "step": 102380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4289664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0195e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.035,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6559733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.800977899475697,
      "rewrite6_current_pass_fraction": 0.20882748145118887,
      "rewrite6_full_passes": 1.0,
      "step": 102390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.431424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.02e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0367,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6563333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.836369636357645,
      "rewrite6_current_pass_fraction": 0.20942143288348816,
      "rewrite6_full_passes": 1.0,
      "step": 102400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4338816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0205e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0281,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6565333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.65794518232533,
      "rewrite6_current_pass_fraction": 0.2100007789526981,
      "rewrite6_full_passes": 1.0,
      "step": 102410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4363392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.021e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.072,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6567866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.585029592894678,
      "rewrite6_current_pass_fraction": 0.21067749410917022,
      "rewrite6_full_passes": 1.0,
      "step": 102420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4387968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0215e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0695,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6572,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.531134415954018,
      "rewrite6_current_pass_fraction": 0.21130065626764816,
      "rewrite6_full_passes": 1.0,
      "step": 102430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4412544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.022e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0562,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6575066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.246666234911444,
      "rewrite6_current_pass_fraction": 0.21190434460867363,
      "rewrite6_full_passes": 1.0,
      "step": 102440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.443712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0225e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0664,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.65784,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.464491249542554,
      "rewrite6_current_pass_fraction": 0.21250803294969914,
      "rewrite6_full_passes": 1.0,
      "step": 102450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4461696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.023e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0543,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.65808,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.206335895020718,
      "rewrite6_current_pass_fraction": 0.21307764211018285,
      "rewrite6_full_passes": 1.0,
      "step": 102460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4486272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0235e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6583733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.191496654234413,
      "rewrite6_current_pass_fraction": 0.21374462035792877,
      "rewrite6_full_passes": 1.0,
      "step": 102470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4510848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.024e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6587066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.926158934434543,
      "rewrite6_current_pass_fraction": 0.21437751942513292,
      "rewrite6_full_passes": 1.0,
      "step": 102480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4535424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0245e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.062,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6590266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.37025495991624,
      "rewrite6_current_pass_fraction": 0.21490818095071176,
      "rewrite6_full_passes": 1.0,
      "step": 102490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.025e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.754528045654297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0522,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6595066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.161849316897023,
      "rewrite6_current_pass_fraction": 0.21549239547428484,
      "rewrite6_full_passes": 1.0,
      "step": 102500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.025e-05,
      "eval_loss": 3.0828821659088135,
      "eval_perplexity": 21.82120418092121,
      "eval_runtime": 0.0499,
      "eval_samples_per_second": 641.693,
      "eval_steps_per_second": 20.053,
      "step": 102500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4584576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0255e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.754692077636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0365,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6599066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.832202779129986,
      "rewrite6_current_pass_fraction": 0.21610095226967344,
      "rewrite6_full_passes": 1.0,
      "step": 102510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4609152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.026e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.754692077636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0598,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6601333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.323292077117205,
      "rewrite6_current_pass_fraction": 0.21672898288251446,
      "rewrite6_full_passes": 1.0,
      "step": 102520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4633728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0265e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.754692077636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0614,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6604266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.357436652816972,
      "rewrite6_current_pass_fraction": 0.21731806586045063,
      "rewrite6_full_passes": 1.0,
      "step": 102530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4658304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.027e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.754692077636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0531,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.66076,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18090355440295,
      "rewrite6_current_pass_fraction": 0.21783899047730326,
      "rewrite6_full_passes": 1.0,
      "step": 102540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.468288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0275e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.754692077636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6611466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.383080960307538,
      "rewrite6_current_pass_fraction": 0.21843781036396565,
      "rewrite6_full_passes": 1.0,
      "step": 102550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4707456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.028e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.754692077636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0493,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6615066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.100568853497336,
      "rewrite6_current_pass_fraction": 0.21904636715935424,
      "rewrite6_full_passes": 1.0,
      "step": 102560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4732032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0285e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.754692077636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0753,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.66168,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.656377850427624,
      "rewrite6_current_pass_fraction": 0.21967439777219527,
      "rewrite6_full_passes": 1.0,
      "step": 102570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4756608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.029e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.754692077636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0857,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.66196,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.882779427644397,
      "rewrite6_current_pass_fraction": 0.22037058674611984,
      "rewrite6_full_passes": 1.0,
      "step": 102580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4781184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0295e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.754692077636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6623333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.22098401199587156,
      "rewrite6_full_passes": 1.0,
      "step": 102590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.480576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.03e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.754692077636719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6627333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.81346223114035,
      "rewrite6_current_pass_fraction": 0.2215682265194446,
      "rewrite6_full_passes": 1.0,
      "step": 102600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4830336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0305e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.015758514404297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6629466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.32969002438444,
      "rewrite6_current_pass_fraction": 0.22215244104301765,
      "rewrite6_full_passes": 1.0,
      "step": 102610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4854912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.031e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.494293212890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6634,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.383080960307538,
      "rewrite6_current_pass_fraction": 0.22268797102295962,
      "rewrite6_full_passes": 1.0,
      "step": 102620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4879488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0315e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.992240905761719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6637733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.22324297482035405,
      "rewrite6_full_passes": 1.0,
      "step": 102630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4904064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.032e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6641466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.2238320577982902,
      "rewrite6_full_passes": 1.0,
      "step": 102640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.492864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0325e-05,
      "grad_norm": 0.427734375,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0276,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6645066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.647618791547,
      "rewrite6_current_pass_fraction": 0.2244990360460361,
      "rewrite6_full_passes": 1.0,
      "step": 102650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4953216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.033e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0374,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6648666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.850960201205005,
      "rewrite6_current_pass_fraction": 0.22512219820451404,
      "rewrite6_full_passes": 1.0,
      "step": 102660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.4977792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0335e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6652533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.926158934434543,
      "rewrite6_current_pass_fraction": 0.2257161496368133,
      "rewrite6_full_passes": 1.0,
      "step": 102670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5002368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.034e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6656133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.945000955113006,
      "rewrite6_current_pass_fraction": 0.22632957488656502,
      "rewrite6_full_passes": 1.0,
      "step": 102680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5026944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0345e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0749,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6659466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.647717031566703,
      "rewrite6_current_pass_fraction": 0.22686997332087008,
      "rewrite6_full_passes": 1.0,
      "step": 102690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.505152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.035e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6663466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.383080960307538,
      "rewrite6_current_pass_fraction": 0.2274882670249849,
      "rewrite6_full_passes": 1.0,
      "step": 102700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5076096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0355e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0759,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.66668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.66937557606564,
      "rewrite6_current_pass_fraction": 0.22811142918346283,
      "rewrite6_full_passes": 1.0,
      "step": 102710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5100672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.036e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.045778632630523,
      "rewrite6_current_pass_fraction": 0.22873945979630386,
      "rewrite6_full_passes": 1.0,
      "step": 102720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5125248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0365e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.66736,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.391635903566247,
      "rewrite6_current_pass_fraction": 0.22931880586551381,
      "rewrite6_full_passes": 1.0,
      "step": 102730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5149824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.037e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0683,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6676266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.50531255087254,
      "rewrite6_current_pass_fraction": 0.22998578411325973,
      "rewrite6_full_passes": 1.0,
      "step": 102740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.51744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0375e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.66796,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.23065276236100563,
      "rewrite6_full_passes": 1.0,
      "step": 102750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5198976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.038e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.193927764892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.66824,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01213231105346,
      "rewrite6_current_pass_fraction": 0.23115908161476895,
      "rewrite6_full_passes": 1.0,
      "step": 102760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5223552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0385e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.560760498046875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0666,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6684533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.468784577110917,
      "rewrite6_current_pass_fraction": 0.2316848746859847,
      "rewrite6_full_passes": 1.0,
      "step": 102770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5248128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.039e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.561229705810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0487,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.66876,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08791230952813,
      "rewrite6_current_pass_fraction": 0.23227882611828396,
      "rewrite6_full_passes": 1.0,
      "step": 102780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5272704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0395e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.561946868896484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0501,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6691333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.117456062563107,
      "rewrite6_current_pass_fraction": 0.23286304064185703,
      "rewrite6_full_passes": 1.0,
      "step": 102790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.529728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.04e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.56341552734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0596,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6694666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.319027845139193,
      "rewrite6_current_pass_fraction": 0.23336935989562033,
      "rewrite6_full_passes": 1.0,
      "step": 102800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5321856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0405e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.563430786132812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0584,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6699333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.29346035528703,
      "rewrite6_current_pass_fraction": 0.23401686432591381,
      "rewrite6_full_passes": 1.0,
      "step": 102810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5346432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.041e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.563430786132812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0497,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6702933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10901076930934,
      "rewrite6_current_pass_fraction": 0.23454265739712957,
      "rewrite6_full_passes": 1.0,
      "step": 102820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5371008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0415e-05,
      "grad_norm": 0.65234375,
      "host_rss_gib": 14.565807342529297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.05,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6705866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.11534442254061,
      "rewrite6_current_pass_fraction": 0.2349954236528987,
      "rewrite6_full_passes": 1.0,
      "step": 102830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5395584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.042e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.566810607910156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0922,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6709066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.025480770862252,
      "rewrite6_current_pass_fraction": 0.23559424353956107,
      "rewrite6_full_passes": 1.0,
      "step": 102840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.542016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0425e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 14.566810607910156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0637,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6712533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.406615290872633,
      "rewrite6_current_pass_fraction": 0.23617845806313412,
      "rewrite6_full_passes": 1.0,
      "step": 102850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5444736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.043e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.566810607910156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0632,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6715733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3959146586082,
      "rewrite6_current_pass_fraction": 0.23681622558470136,
      "rewrite6_full_passes": 1.0,
      "step": 102860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5469312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0435e-05,
      "grad_norm": 0.427734375,
      "host_rss_gib": 14.566810607910156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0583,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6720133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.29133111571526,
      "rewrite6_current_pass_fraction": 0.23741991392572687,
      "rewrite6_full_passes": 1.0,
      "step": 102870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5493888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.044e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.566810607910156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0529,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6722133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.176667797281898,
      "rewrite6_current_pass_fraction": 0.23808202371910966,
      "rewrite6_full_passes": 1.0,
      "step": 102880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5518464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0445e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.56711196899414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0636,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6726,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.40447473637306,
      "rewrite6_current_pass_fraction": 0.23867597515140893,
      "rewrite6_full_passes": 1.0,
      "step": 102890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.554304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.045e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.56711196899414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0543,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.67288,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.206335895020718,
      "rewrite6_current_pass_fraction": 0.23925045276625576,
      "rewrite6_full_passes": 1.0,
      "step": 102900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5567616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0455e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.56723403930664,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0622,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6732666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.37452943834182,
      "rewrite6_current_pass_fraction": 0.2398249303811026,
      "rewrite6_full_passes": 1.0,
      "step": 102910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5592192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.046e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.56723403930664,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.67356,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.24035072345231837,
      "rewrite6_full_passes": 1.0,
      "step": 102920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5616768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0465e-05,
      "grad_norm": 0.546875,
      "host_rss_gib": 14.568309783935547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0637,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6740266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.406615290872633,
      "rewrite6_current_pass_fraction": 0.24105178088060603,
      "rewrite6_full_passes": 1.0,
      "step": 102930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5641344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.047e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.574172973632812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0558,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6743066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23816926792417,
      "rewrite6_current_pass_fraction": 0.2416457323129053,
      "rewrite6_full_passes": 1.0,
      "step": 102940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.566592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0475e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.574764251708984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6746,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.242259157562657,
      "rewrite6_full_passes": 1.0,
      "step": 102950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5690496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.048e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 14.574764251708984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0662,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6749733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.460198780553853,
      "rewrite6_current_pass_fraction": 0.2428141613600514,
      "rewrite6_full_passes": 1.0,
      "step": 102960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5715072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0485e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 14.57476806640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0606,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6752266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.340357536052313,
      "rewrite6_current_pass_fraction": 0.24329613834199917,
      "rewrite6_full_passes": 1.0,
      "step": 102970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5739648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.049e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.57476806640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0494,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6754533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10267901588904,
      "rewrite6_current_pass_fraction": 0.24389008977429846,
      "rewrite6_full_passes": 1.0,
      "step": 102980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5764224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0495e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.57476806640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6757333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.125904734609893,
      "rewrite6_current_pass_fraction": 0.24453759420459192,
      "rewrite6_full_passes": 1.0,
      "step": 102990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.57888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.05e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.574943542480469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.676,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.189377610522946,
      "rewrite6_current_pass_fraction": 0.2450293080952659,
      "rewrite6_full_passes": 1.0,
      "step": 103000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.05e-05,
      "eval_loss": 3.0829718112945557,
      "eval_perplexity": 21.823160438870808,
      "eval_runtime": 0.0502,
      "eval_samples_per_second": 638.002,
      "eval_steps_per_second": 19.938,
      "step": 103000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5813376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0505e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.645858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0692,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6764666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.524676044433395,
      "rewrite6_current_pass_fraction": 0.2456914178886487,
      "rewrite6_full_passes": 1.0,
      "step": 103010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5837952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.051e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.645858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0456,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6768,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.022641004163326,
      "rewrite6_current_pass_fraction": 0.24635839613639463,
      "rewrite6_full_passes": 1.0,
      "step": 103020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5862528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0515e-05,
      "grad_norm": 0.54296875,
      "host_rss_gib": 14.645858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0687,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6770666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.513916396547316,
      "rewrite6_current_pass_fraction": 0.24693287375124146,
      "rewrite6_full_passes": 1.0,
      "step": 103030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5887104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.052e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.645858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0579,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.67736,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28281628634838,
      "rewrite6_current_pass_fraction": 0.24758524663589804,
      "rewrite6_full_passes": 1.0,
      "step": 103040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.591168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0525e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.645858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0688,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6775733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.51606789576013,
      "rewrite6_current_pass_fraction": 0.24814025043329244,
      "rewrite6_full_passes": 1.0,
      "step": 103050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5936256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.053e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 14.645858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0605,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6778933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.338223606996944,
      "rewrite6_current_pass_fraction": 0.24868551732196062,
      "rewrite6_full_passes": 1.0,
      "step": 103060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5960832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0535e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.645858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0533,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.67816,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18514015876014,
      "rewrite6_current_pass_fraction": 0.249284337208623,
      "rewrite6_full_passes": 1.0,
      "step": 103070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.5985408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.054e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 14.645858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6784666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.02684595284501,
      "rewrite6_current_pass_fraction": 0.2498247356429281,
      "rewrite6_full_passes": 1.0,
      "step": 103080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6009984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0545e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.645858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6788,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.976442031457733,
      "rewrite6_current_pass_fraction": 0.2504625031644953,
      "rewrite6_full_passes": 1.0,
      "step": 103090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.603456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.055e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.645858764648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0721,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6791466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.587188203782706,
      "rewrite6_current_pass_fraction": 0.2509736908726218,
      "rewrite6_full_passes": 1.0,
      "step": 103100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6059136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0555e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0531,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6795066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18090355440295,
      "rewrite6_current_pass_fraction": 0.2516455375747308,
      "rewrite6_full_passes": 1.0,
      "step": 103110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6083712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.056e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0481,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.67988,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.075263357207568,
      "rewrite6_current_pass_fraction": 0.25223462055266693,
      "rewrite6_full_passes": 1.0,
      "step": 103120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6108288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0565e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0567,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.68024,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.257292224304877,
      "rewrite6_current_pass_fraction": 0.25285778271114484,
      "rewrite6_full_passes": 1.0,
      "step": 103130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6132864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.057e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0742,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6805066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.632568932097964,
      "rewrite6_current_pass_fraction": 0.2534079180541762,
      "rewrite6_full_passes": 1.0,
      "step": 103140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.615744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0575e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0368,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6807466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.838453377506596,
      "rewrite6_current_pass_fraction": 0.2539239742166657,
      "rewrite6_full_passes": 1.0,
      "step": 103150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6182016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.058e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.68108,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.99532932729427,
      "rewrite6_current_pass_fraction": 0.2544351619247921,
      "rewrite6_full_passes": 1.0,
      "step": 103160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6206592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0585e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6814266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.84679042615617,
      "rewrite6_current_pass_fraction": 0.2551508247161691,
      "rewrite6_full_passes": 1.0,
      "step": 103170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6231168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.059e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6818133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.88017198896687,
      "rewrite6_current_pass_fraction": 0.2557058285135635,
      "rewrite6_full_passes": 1.0,
      "step": 103180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6255744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0595e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0526,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6822533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.170315749797474,
      "rewrite6_current_pass_fraction": 0.25633385912640455,
      "rewrite6_full_passes": 1.0,
      "step": 103190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.628032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.06e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0479,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6827066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.071048726013295,
      "rewrite6_current_pass_fraction": 0.25701057428287666,
      "rewrite6_full_passes": 1.0,
      "step": 103200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6304896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0605e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0734,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.68308,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.615269797528736,
      "rewrite6_current_pass_fraction": 0.25762886798699147,
      "rewrite6_full_passes": 1.0,
      "step": 103210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6329472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.061e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6834666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.189377610522946,
      "rewrite6_current_pass_fraction": 0.2581936086931121,
      "rewrite6_full_passes": 1.0,
      "step": 103220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6354048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0615e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6838133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91988202833721,
      "rewrite6_current_pass_fraction": 0.2587875601254114,
      "rewrite6_full_passes": 1.0,
      "step": 103230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6378624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.062e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0652,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6840933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.438749308296885,
      "rewrite6_current_pass_fraction": 0.25948374909933597,
      "rewrite6_full_passes": 1.0,
      "step": 103240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.64032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0625e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0745,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.68448,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.639059676340555,
      "rewrite6_current_pass_fraction": 0.2600046737161886,
      "rewrite6_full_passes": 1.0,
      "step": 103250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6427776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.063e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0625,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6848133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.380942759123343,
      "rewrite6_current_pass_fraction": 0.26052072987867814,
      "rewrite6_full_passes": 1.0,
      "step": 103260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6452352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0635e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6850933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.17878556994849,
      "rewrite6_current_pass_fraction": 0.26111468131097737,
      "rewrite6_full_passes": 1.0,
      "step": 103270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6476928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.064e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0525,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.68548,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.16819882407055,
      "rewrite6_current_pass_fraction": 0.2616745535627349,
      "rewrite6_full_passes": 1.0,
      "step": 103280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6501504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0645e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6859066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.911515748893304,
      "rewrite6_current_pass_fraction": 0.2622831103581235,
      "rewrite6_full_passes": 1.0,
      "step": 103290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.652608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.065e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.101,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.68616,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.2201603353993,
      "rewrite6_current_pass_fraction": 0.26284785106424413,
      "rewrite6_full_passes": 1.0,
      "step": 103300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6550656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0655e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0668,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.68652,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.473078763430657,
      "rewrite6_current_pass_fraction": 0.2633931179529123,
      "rewrite6_full_passes": 1.0,
      "step": 103310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6575232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.066e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6868533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.295589807793416,
      "rewrite6_current_pass_fraction": 0.2638361472999552,
      "rewrite6_full_passes": 1.0,
      "step": 103320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6599808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0665e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0751,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6871866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.65204700795622,
      "rewrite6_current_pass_fraction": 0.2643911510973496,
      "rewrite6_full_passes": 1.0,
      "step": 103330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6624384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.067e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0689,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6874933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.518219610133638,
      "rewrite6_current_pass_fraction": 0.2650386555276431,
      "rewrite6_full_passes": 1.0,
      "step": 103340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.664896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0675e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.68792,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.7801873185992,
      "rewrite6_current_pass_fraction": 0.26562773850557925,
      "rewrite6_full_passes": 1.0,
      "step": 103350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6673536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.068e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0729,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.68824,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.604464864088442,
      "rewrite6_current_pass_fraction": 0.2661243208506163,
      "rewrite6_full_passes": 1.0,
      "step": 103360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6698112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0685e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6886133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.828036755190443,
      "rewrite6_current_pass_fraction": 0.2667669568265467,
      "rewrite6_full_passes": 1.0,
      "step": 103370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6722688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.069e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0579,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6889733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28281628634838,
      "rewrite6_current_pass_fraction": 0.2674874880722868,
      "rewrite6_full_passes": 1.0,
      "step": 103380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6747264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0695e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0593,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6893466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.312633096045975,
      "rewrite6_current_pass_fraction": 0.2680960448676754,
      "rewrite6_full_passes": 1.0,
      "step": 103390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.677184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.07e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0794,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6898,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.74535127048861,
      "rewrite6_current_pass_fraction": 0.26870460166306404,
      "rewrite6_full_passes": 1.0,
      "step": 103400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6796416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0705e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69024,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21482012611182,
      "rewrite6_current_pass_fraction": 0.26944460672625653,
      "rewrite6_full_passes": 1.0,
      "step": 103410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6820992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.071e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0752,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6905466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.654212320920866,
      "rewrite6_current_pass_fraction": 0.2700385581585558,
      "rewrite6_full_passes": 1.0,
      "step": 103420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6845568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0715e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0286,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6908266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.668276737590073,
      "rewrite6_current_pass_fraction": 0.2705789565928609,
      "rewrite6_full_passes": 1.0,
      "step": 103430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6870144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.072e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0599,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6911333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.325424512944927,
      "rewrite6_current_pass_fraction": 0.27122646102315434,
      "rewrite6_full_passes": 1.0,
      "step": 103440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.689472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0725e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0438,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6915333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.98483428660945,
      "rewrite6_current_pass_fraction": 0.2717960701836381,
      "rewrite6_full_passes": 1.0,
      "step": 103450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6919296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.073e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.646316528320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0673,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69184,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.483817987394623,
      "rewrite6_current_pass_fraction": 0.2724046269790267,
      "rewrite6_full_passes": 1.0,
      "step": 103460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6943872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0735e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.188518524169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0574,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69232,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.2721775381139,
      "rewrite6_current_pass_fraction": 0.2730618683180464,
      "rewrite6_full_passes": 1.0,
      "step": 103470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6968448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.074e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.057109832763672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6925733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.216941714202072,
      "rewrite6_current_pass_fraction": 0.2735779244805359,
      "rewrite6_full_passes": 1.0,
      "step": 103480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.6993024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0745e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.515716552734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0753,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6928133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.656377850427624,
      "rewrite6_current_pass_fraction": 0.2741816128215614,
      "rewrite6_full_passes": 1.0,
      "step": 103490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.70176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.075e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.515716552734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0486,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69312,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.085803623733216,
      "rewrite6_current_pass_fraction": 0.2747317481645927,
      "rewrite6_full_passes": 1.0,
      "step": 103500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.075e-05,
      "eval_loss": 3.0825281143188477,
      "eval_perplexity": 21.813479716396152,
      "eval_runtime": 0.0502,
      "eval_samples_per_second": 637.996,
      "eval_steps_per_second": 19.937,
      "step": 103500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7042176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0755e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0677,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6934133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.492413233524204,
      "rewrite6_current_pass_fraction": 0.2752867519619871,
      "rewrite6_full_passes": 1.0,
      "step": 103510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7066752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.076e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0646,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6938666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.425889916915104,
      "rewrite6_current_pass_fraction": 0.27583688730501843,
      "rewrite6_full_passes": 1.0,
      "step": 103520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7091328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0765e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69424,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.2764697863722225,
      "rewrite6_full_passes": 1.0,
      "step": 103530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7115904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.077e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0511,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69452,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.138584080874157,
      "rewrite6_current_pass_fraction": 0.2770053163521645,
      "rewrite6_full_passes": 1.0,
      "step": 103540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.714048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0775e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0517,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69472,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.151271037028916,
      "rewrite6_current_pass_fraction": 0.27757005705828514,
      "rewrite6_full_passes": 1.0,
      "step": 103550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7165056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.078e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0437,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6949733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.982735908101457,
      "rewrite6_current_pass_fraction": 0.2781980876711262,
      "rewrite6_full_passes": 1.0,
      "step": 103560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7189632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0785e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6952666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.028948742578034,
      "rewrite6_current_pass_fraction": 0.27872874919670504,
      "rewrite6_full_passes": 1.0,
      "step": 103570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7214208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.079e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0639,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69572,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.41089704209166,
      "rewrite6_current_pass_fraction": 0.2793178321746412,
      "rewrite6_full_passes": 1.0,
      "step": 103580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7238784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0795e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6960666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.27983388833713074,
      "rewrite6_full_passes": 1.0,
      "step": 103590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.726336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.08e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6964666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.265796842004743,
      "rewrite6_current_pass_fraction": 0.28051547194796594,
      "rewrite6_full_passes": 1.0,
      "step": 103600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7287936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0805e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69676,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06894172649243,
      "rewrite6_current_pass_fraction": 0.2811483710151701,
      "rewrite6_full_passes": 1.0,
      "step": 103610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7312512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.081e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0738,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69708,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.623917634899918,
      "rewrite6_current_pass_fraction": 0.2817569278105587,
      "rewrite6_full_passes": 1.0,
      "step": 103620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7337088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0815e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.581748962402344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0469,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69736,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.049988209300682,
      "rewrite6_current_pass_fraction": 0.2824141691495784,
      "rewrite6_full_passes": 1.0,
      "step": 103630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7361664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.082e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.980655670166016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0677,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6977466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.492413233524204,
      "rewrite6_current_pass_fraction": 0.28299351521878835,
      "rewrite6_full_passes": 1.0,
      "step": 103640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.738624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0825e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.98095703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6980133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.2837091780101653,
      "rewrite6_full_passes": 1.0,
      "step": 103650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7410816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.083e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.982471466064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0738,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6982666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.623917634899918,
      "rewrite6_current_pass_fraction": 0.28426418180755975,
      "rewrite6_full_passes": 1.0,
      "step": 103660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7435392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0835e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.983146667480469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0427,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6986933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.961763660065063,
      "rewrite6_current_pass_fraction": 0.28488734396603765,
      "rewrite6_full_passes": 1.0,
      "step": 103670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7459968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.084e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.985210418701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.075,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6990666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.649881911512058,
      "rewrite6_current_pass_fraction": 0.28548129539833694,
      "rewrite6_full_passes": 1.0,
      "step": 103680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7484544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0845e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.986885070800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.084,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.69936,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.84561030532294,
      "rewrite6_current_pass_fraction": 0.28604603610445756,
      "rewrite6_full_passes": 1.0,
      "step": 103690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.750912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.085e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.987438201904297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.6997733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08369514879635,
      "rewrite6_current_pass_fraction": 0.28661077681057817,
      "rewrite6_full_passes": 1.0,
      "step": 103700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7533696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0855e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.987945556640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0564,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7001466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25091599312008,
      "rewrite6_current_pass_fraction": 0.2871901228797881,
      "rewrite6_full_passes": 1.0,
      "step": 103710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7558272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.086e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.98843002319336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7004933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.882260110570105,
      "rewrite6_current_pass_fraction": 0.287676968316099,
      "rewrite6_full_passes": 1.0,
      "step": 103720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7582848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0865e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.98843002319336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0334,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7007733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.767722945893784,
      "rewrite6_current_pass_fraction": 0.288168682206773,
      "rewrite6_full_passes": 1.0,
      "step": 103730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7607424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.087e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.98843002319336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7011466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01003120287951,
      "rewrite6_current_pass_fraction": 0.2888161866370665,
      "rewrite6_full_passes": 1.0,
      "step": 103740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0875e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.98843002319336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.70152,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23604555718468,
      "rewrite6_current_pass_fraction": 0.2893809273431871,
      "rewrite6_full_passes": 1.0,
      "step": 103750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7656576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.088e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.98843765258789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0446,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7018266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.00162887097677,
      "rewrite6_current_pass_fraction": 0.2899213257774922,
      "rewrite6_full_passes": 1.0,
      "step": 103760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7681152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0885e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.98843765258789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0792,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7022266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.741002635112544,
      "rewrite6_current_pass_fraction": 0.29056396175342253,
      "rewrite6_full_passes": 1.0,
      "step": 103770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7705728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.089e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 14.989154815673828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7024266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.221185526912038,
      "rewrite6_current_pass_fraction": 0.2910897548246383,
      "rewrite6_full_passes": 1.0,
      "step": 103780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7730304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0895e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.989154815673828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7027466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.74904040364757,
      "rewrite6_current_pass_fraction": 0.29166423243948514,
      "rewrite6_full_passes": 1.0,
      "step": 103790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.775488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.09e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.989154815673828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0604,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.703,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3360898913238,
      "rewrite6_current_pass_fraction": 0.29223384159996885,
      "rewrite6_full_passes": 1.0,
      "step": 103800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7779456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0905e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.989154815673828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0317,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7033733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.732447809247343,
      "rewrite6_current_pass_fraction": 0.2927352923993691,
      "rewrite6_full_passes": 1.0,
      "step": 103810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7804032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.091e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.70372,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.29334384919475764,
      "rewrite6_full_passes": 1.0,
      "step": 103820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7828608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0915e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0671,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.704,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.47952165344486,
      "rewrite6_current_pass_fraction": 0.2940351697143191,
      "rewrite6_full_passes": 1.0,
      "step": 103830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7853184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.092e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0782,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7043466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.719272499356148,
      "rewrite6_current_pass_fraction": 0.29456583123989794,
      "rewrite6_full_passes": 1.0,
      "step": 103840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.787776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0925e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0689,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7048266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.518219610133638,
      "rewrite6_current_pass_fraction": 0.29518412494401275,
      "rewrite6_full_passes": 1.0,
      "step": 103850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7902336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.093e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0621,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7052133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.372392092267074,
      "rewrite6_current_pass_fraction": 0.29578294483067513,
      "rewrite6_full_passes": 1.0,
      "step": 103860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7926912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0935e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0719,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.70556,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.58287119785693,
      "rewrite6_current_pass_fraction": 0.2964693968958735,
      "rewrite6_full_passes": 1.0,
      "step": 103870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7951488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.094e-05,
      "grad_norm": 0.58984375,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0786,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7059866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.727961946129383,
      "rewrite6_current_pass_fraction": 0.29710229596307763,
      "rewrite6_full_passes": 1.0,
      "step": 103880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.7976064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0945e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0713,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7062533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.56992535927817,
      "rewrite6_current_pass_fraction": 0.2977303265759187,
      "rewrite6_full_passes": 1.0,
      "step": 103890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.800064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.095e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0603,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.70648,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.333956389011558,
      "rewrite6_current_pass_fraction": 0.29831940955385483,
      "rewrite6_full_passes": 1.0,
      "step": 103900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8025216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0955e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.70672,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.84679042615617,
      "rewrite6_current_pass_fraction": 0.2989620455297852,
      "rewrite6_full_passes": 1.0,
      "step": 103910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8049792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.096e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7070266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.2996290237775311,
      "rewrite6_full_passes": 1.0,
      "step": 103920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8074368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0965e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.053,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.70716,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.17878556994849,
      "rewrite6_current_pass_fraction": 0.3002181067554673,
      "rewrite6_full_passes": 1.0,
      "step": 103930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8098944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.097e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7074,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.225430188469428,
      "rewrite6_current_pass_fraction": 0.3007925843703141,
      "rewrite6_full_passes": 1.0,
      "step": 103940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.812352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0975e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0317,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.70768,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.732447809247343,
      "rewrite6_current_pass_fraction": 0.3013865358026134,
      "rewrite6_full_passes": 1.0,
      "step": 103950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8148096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.098e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.70804,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.3019464080543709,
      "rewrite6_full_passes": 1.0,
      "step": 103960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8172672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0985e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7083466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.191496654234413,
      "rewrite6_current_pass_fraction": 0.3024527273081342,
      "rewrite6_full_passes": 1.0,
      "step": 103970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8197248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.099e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0661,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7088266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45805286797322,
      "rewrite6_current_pass_fraction": 0.3030028626511655,
      "rewrite6_full_passes": 1.0,
      "step": 103980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8221824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.0995e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0732,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7092533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.61094717584581,
      "rewrite6_current_pass_fraction": 0.3035481295398337,
      "rewrite6_full_passes": 1.0,
      "step": 103990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.82464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 14.990283966064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0583,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7096533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.29133111571526,
      "rewrite6_current_pass_fraction": 0.3040593172479601,
      "rewrite6_full_passes": 1.0,
      "step": 104000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.1e-05,
      "eval_loss": 3.083242177963257,
      "eval_perplexity": 21.829061491747254,
      "eval_runtime": 0.0521,
      "eval_samples_per_second": 614.581,
      "eval_steps_per_second": 19.206,
      "step": 104000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8270976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1005e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.173599243164062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7100266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.89061468541791,
      "rewrite6_current_pass_fraction": 0.3047555062218847,
      "rewrite6_full_passes": 1.0,
      "step": 104010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8295552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.101e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0592,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7102133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.310501939295985,
      "rewrite6_current_pass_fraction": 0.3054273529239937,
      "rewrite6_full_passes": 1.0,
      "step": 104020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8320128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1015e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0451,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.71052,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01213231105346,
      "rewrite6_current_pass_fraction": 0.3061040680804658,
      "rewrite6_full_passes": 1.0,
      "step": 104030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8344704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.102e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0638,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.71096,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.40875605943837,
      "rewrite6_current_pass_fraction": 0.30671262487585443,
      "rewrite6_full_passes": 1.0,
      "step": 104040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.836928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1025e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7112,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.88017198896687,
      "rewrite6_current_pass_fraction": 0.307184864949076,
      "rewrite6_full_passes": 1.0,
      "step": 104050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8393856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.103e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.71152,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.32969002438444,
      "rewrite6_current_pass_fraction": 0.30773986874647036,
      "rewrite6_full_passes": 1.0,
      "step": 104060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8418432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1035e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.043,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7118933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.968053132536788,
      "rewrite6_current_pass_fraction": 0.30833382017876965,
      "rewrite6_full_passes": 1.0,
      "step": 104070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8443008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.104e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0535,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7122266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.189377610522946,
      "rewrite6_current_pass_fraction": 0.30887908706743783,
      "rewrite6_full_passes": 1.0,
      "step": 104080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8467584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1045e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7124933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.028948742578034,
      "rewrite6_current_pass_fraction": 0.3094730384997371,
      "rewrite6_full_passes": 1.0,
      "step": 104090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.849216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.105e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7129066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.31014488520184613,
      "rewrite6_full_passes": 1.0,
      "step": 104100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8516736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1055e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7132,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.957571726540376,
      "rewrite6_current_pass_fraction": 0.31084107417577067,
      "rewrite6_full_passes": 1.0,
      "step": 104110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8541312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.106e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7134933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13435678680148,
      "rewrite6_current_pass_fraction": 0.3114544994255224,
      "rewrite6_full_passes": 1.0,
      "step": 104120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8565888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1065e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0569,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7138133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.26154410792393,
      "rewrite6_current_pass_fraction": 0.31198516095110124,
      "rewrite6_full_passes": 1.0,
      "step": 104130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8590464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.107e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7140133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.3125888492921267,
      "rewrite6_full_passes": 1.0,
      "step": 104140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.861504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1075e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7142,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.3132655644485989,
      "rewrite6_full_passes": 1.0,
      "step": 104150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8639616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.108e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0712,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.71444,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.567768474588277,
      "rewrite6_current_pass_fraction": 0.3137426729761835,
      "rewrite6_full_passes": 1.0,
      "step": 104160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8664192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1085e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7147066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.31420030768631574,
      "rewrite6_full_passes": 1.0,
      "step": 104170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8688768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.109e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0231,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.71504,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.554913250892056,
      "rewrite6_current_pass_fraction": 0.3148332067535199,
      "rewrite6_full_passes": 1.0,
      "step": 104180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8713344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1095e-05,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0242,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.71536,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.577536095751576,
      "rewrite6_current_pass_fraction": 0.31555373799926,
      "rewrite6_full_passes": 1.0,
      "step": 104190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.873792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.11e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0532,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7156666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18302175066643,
      "rewrite6_current_pass_fraction": 0.3161184787053806,
      "rewrite6_full_passes": 1.0,
      "step": 104200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8762496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1105e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.71596,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.31675624622694787,
      "rewrite6_full_passes": 1.0,
      "step": 104210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8787072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.111e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7164133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.730374668125208,
      "rewrite6_current_pass_fraction": 0.3173940137485151,
      "rewrite6_full_passes": 1.0,
      "step": 104220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8811648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1115e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0577,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7167066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.27856014871906,
      "rewrite6_current_pass_fraction": 0.3179928336351775,
      "rewrite6_full_passes": 1.0,
      "step": 104230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8836224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.112e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.71708,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.976442031457733,
      "rewrite6_current_pass_fraction": 0.3186938910634652,
      "rewrite6_full_passes": 1.0,
      "step": 104240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.88608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1125e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0497,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7173333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.10901076930934,
      "rewrite6_current_pass_fraction": 0.31933165858503243,
      "rewrite6_full_passes": 1.0,
      "step": 104250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8885376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.113e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0627,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7176933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.38521937532253,
      "rewrite6_current_pass_fraction": 0.3199353469260579,
      "rewrite6_full_passes": 1.0,
      "step": 104260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8909952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1135e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7181066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.32051956144963095,
      "rewrite6_full_passes": 1.0,
      "step": 104270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8934528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.114e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7184266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.759417517911764,
      "rewrite6_current_pass_fraction": 0.320967459251037,
      "rewrite6_full_passes": 1.0,
      "step": 104280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.8959104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1145e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0709,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.71864,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.561299114498432,
      "rewrite6_current_pass_fraction": 0.3214835154135265,
      "rewrite6_full_passes": 1.0,
      "step": 104290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.898368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.115e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.1654052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.1072,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7190133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.35835328494457,
      "rewrite6_current_pass_fraction": 0.32200444003037915,
      "rewrite6_full_passes": 1.0,
      "step": 104300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9008256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1155e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.165794372558594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0689,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7193466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.518219610133638,
      "rewrite6_current_pass_fraction": 0.3225253646472318,
      "rewrite6_full_passes": 1.0,
      "step": 104310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9032832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.116e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.165794372558594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0267,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7196,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.629044294412097,
      "rewrite6_current_pass_fraction": 0.3231485268057097,
      "rewrite6_full_passes": 1.0,
      "step": 104320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9057408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1165e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.449115753173828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.71988,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.24241732656946,
      "rewrite6_current_pass_fraction": 0.32373274132928276,
      "rewrite6_full_passes": 1.0,
      "step": 104330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9081984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.117e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0726,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7201866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.597984496732916,
      "rewrite6_current_pass_fraction": 0.3243461665790345,
      "rewrite6_full_passes": 1.0,
      "step": 104340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.910656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1175e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7204533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.248791007771803,
      "rewrite6_current_pass_fraction": 0.32499367100932797,
      "rewrite6_full_passes": 1.0,
      "step": 104350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9131136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.118e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0713,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7208533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.56992535927817,
      "rewrite6_current_pass_fraction": 0.32569959689197875,
      "rewrite6_full_passes": 1.0,
      "step": 104360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9155712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1185e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.069,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7211733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.520371539689332,
      "rewrite6_current_pass_fraction": 0.3262789429611887,
      "rewrite6_full_passes": 1.0,
      "step": 104370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9180288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.119e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0963,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.72156,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.11597061945123,
      "rewrite6_current_pass_fraction": 0.32678039376058887,
      "rewrite6_full_passes": 1.0,
      "step": 104380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9204864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1195e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0612,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7219866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35316559260667,
      "rewrite6_current_pass_fraction": 0.32729644992307844,
      "rewrite6_full_passes": 1.0,
      "step": 104390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.922944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.12e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0666,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7223466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.468784577110917,
      "rewrite6_current_pass_fraction": 0.32787092753792524,
      "rewrite6_full_passes": 1.0,
      "step": 104400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9254016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1205e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0551,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.72272,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.22330775157419,
      "rewrite6_current_pass_fraction": 0.32843566824404585,
      "rewrite6_full_passes": 1.0,
      "step": 104410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9278592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.121e-05,
      "grad_norm": 0.54296875,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0902,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7229733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.981473830929435,
      "rewrite6_current_pass_fraction": 0.3290101458588927,
      "rewrite6_full_passes": 1.0,
      "step": 104420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9303168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1215e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0513,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7232266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.1428122204902,
      "rewrite6_current_pass_fraction": 0.329604097291192,
      "rewrite6_full_passes": 1.0,
      "step": 104430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9327744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.122e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0558,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7235866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23816926792417,
      "rewrite6_current_pass_fraction": 0.3302077856322175,
      "rewrite6_full_passes": 1.0,
      "step": 104440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.935232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1225e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.72384,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.3308406846994216,
      "rewrite6_full_passes": 1.0,
      "step": 104450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9376896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.123e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0898,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7241333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.972682999680526,
      "rewrite6_current_pass_fraction": 0.33140542540554224,
      "rewrite6_full_passes": 1.0,
      "step": 104460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9401472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1235e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7244133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.71586848361443,
      "rewrite6_current_pass_fraction": 0.3320626667445619,
      "rewrite6_full_passes": 1.0,
      "step": 104470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9426048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.124e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7246533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.197855056943403,
      "rewrite6_current_pass_fraction": 0.33261280208759325,
      "rewrite6_full_passes": 1.0,
      "step": 104480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9450624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1245e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.00653076171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7249733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.974344492133298,
      "rewrite6_current_pass_fraction": 0.33317754279371387,
      "rewrite6_full_passes": 1.0,
      "step": 104490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.94752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.125e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.40628433227539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7254,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35530101593532,
      "rewrite6_current_pass_fraction": 0.33380557340655487,
      "rewrite6_full_passes": 1.0,
      "step": 104500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.125e-05,
      "eval_loss": 3.0832529067993164,
      "eval_perplexity": 21.82929569342569,
      "eval_runtime": 0.0512,
      "eval_samples_per_second": 625.301,
      "eval_steps_per_second": 19.541,
      "step": 104500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9499776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1255e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.465744018554688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0529,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7258133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.176667797281898,
      "rewrite6_current_pass_fraction": 0.33442873556503283,
      "rewrite6_full_passes": 1.0,
      "step": 104510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9524352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.126e-05,
      "grad_norm": 0.53515625,
      "host_rss_gib": 15.467838287353516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0492,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.72612,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09845890211131,
      "rewrite6_current_pass_fraction": 0.3350129500886059,
      "rewrite6_full_passes": 1.0,
      "step": 104520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9548928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1265e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.46822738647461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.72644,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.945000955113006,
      "rewrite6_current_pass_fraction": 0.3355728223403634,
      "rewrite6_full_passes": 1.0,
      "step": 104530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9573504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.127e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.46823501586914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0603,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7268,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.333956389011558,
      "rewrite6_current_pass_fraction": 0.33621058986193064,
      "rewrite6_full_passes": 1.0,
      "step": 104540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.959808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1275e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.468563079833984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0649,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7271866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.432318648151654,
      "rewrite6_current_pass_fraction": 0.3368142782029561,
      "rewrite6_full_passes": 1.0,
      "step": 104550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9622656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.128e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.468563079833984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7274666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.007930304805882,
      "rewrite6_current_pass_fraction": 0.3374033611808923,
      "rewrite6_full_passes": 1.0,
      "step": 104560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9647232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1285e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.469078063964844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0351,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7277466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.803058101273997,
      "rewrite6_current_pass_fraction": 0.33804112870245956,
      "rewrite6_full_passes": 1.0,
      "step": 104570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9671808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.129e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.469078063964844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0688,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7280533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.51606789576013,
      "rewrite6_current_pass_fraction": 0.33859613249985393,
      "rewrite6_full_passes": 1.0,
      "step": 104580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9696384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1295e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.470550537109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0579,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.72836,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28281628634838,
      "rewrite6_current_pass_fraction": 0.3391657416603377,
      "rewrite6_full_passes": 1.0,
      "step": 104590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.972096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.13e-05,
      "grad_norm": 0.53125,
      "host_rss_gib": 15.470752716064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0664,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7287866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.464491249542554,
      "rewrite6_current_pass_fraction": 0.33974995618391074,
      "rewrite6_full_passes": 1.0,
      "step": 104600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9745536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1305e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.47076416015625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7292666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.3403536445249362,
      "rewrite6_full_passes": 1.0,
      "step": 104610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9770112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.131e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.471244812011719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0649,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.72956,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.432318648151654,
      "rewrite6_current_pass_fraction": 0.34102062277268214,
      "rewrite6_full_passes": 1.0,
      "step": 104620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9794688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1315e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.471244812011719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0462,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7298666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.035258373598133,
      "rewrite6_current_pass_fraction": 0.34159023193316584,
      "rewrite6_full_passes": 1.0,
      "step": 104630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9819264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.132e-05,
      "grad_norm": 0.54296875,
      "host_rss_gib": 15.471244812011719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0732,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.73032,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.61094717584581,
      "rewrite6_current_pass_fraction": 0.342266947089638,
      "rewrite6_full_passes": 1.0,
      "step": 104640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.984384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1325e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.73064,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.125904734609893,
      "rewrite6_current_pass_fraction": 0.3430020836984674,
      "rewrite6_full_passes": 1.0,
      "step": 104650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9868416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.133e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0543,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7309066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.206335895020718,
      "rewrite6_current_pass_fraction": 0.34359116667640355,
      "rewrite6_full_passes": 1.0,
      "step": 104660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9892992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1335e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0657,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7312,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.449471363241397,
      "rewrite6_current_pass_fraction": 0.3443019610134175,
      "rewrite6_full_passes": 1.0,
      "step": 104670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9917568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.134e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0673,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7315466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.483817987394623,
      "rewrite6_current_pass_fraction": 0.3448082802671808,
      "rewrite6_full_passes": 1.0,
      "step": 104680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9942144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1345e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0626,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.73188,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.383080960307538,
      "rewrite6_current_pass_fraction": 0.34544604778874805,
      "rewrite6_full_passes": 1.0,
      "step": 104690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.996672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.135e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7322,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.221185526912038,
      "rewrite6_current_pass_fraction": 0.3460594730384997,
      "rewrite6_full_passes": 1.0,
      "step": 104700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 19.9991296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1355e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0617,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7325333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.363844844993586,
      "rewrite6_current_pass_fraction": 0.3467361881949719,
      "rewrite6_full_passes": 1.0,
      "step": 104710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0015872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.136e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0753,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7328933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.656377850427624,
      "rewrite6_current_pass_fraction": 0.3473057973554556,
      "rewrite6_full_passes": 1.0,
      "step": 104720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0040448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1365e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7331866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.225430188469428,
      "rewrite6_current_pass_fraction": 0.34784619578976067,
      "rewrite6_full_passes": 1.0,
      "step": 104730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0065024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.137e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7335333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.193615909860856,
      "rewrite6_current_pass_fraction": 0.3485131740375066,
      "rewrite6_full_passes": 1.0,
      "step": 104740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.00896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1375e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0618,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7338933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.365981336300866,
      "rewrite6_current_pass_fraction": 0.34906330938053787,
      "rewrite6_full_passes": 1.0,
      "step": 104750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0114176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.138e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.472286224365234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7341866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.81346223114035,
      "rewrite6_current_pass_fraction": 0.3496767346302896,
      "rewrite6_full_passes": 1.0,
      "step": 104760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0138752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1385e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7344533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.018436896387286,
      "rewrite6_current_pass_fraction": 0.3502171330645947,
      "rewrite6_full_passes": 1.0,
      "step": 104770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0163328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.139e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0521,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.73476,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.159733237771047,
      "rewrite6_current_pass_fraction": 0.35082568985998325,
      "rewrite6_full_passes": 1.0,
      "step": 104780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0187904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1395e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.73528,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.896882809995212,
      "rewrite6_current_pass_fraction": 0.3514585889271874,
      "rewrite6_full_passes": 1.0,
      "step": 104790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.021248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.14e-05,
      "grad_norm": 0.64453125,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0522,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7355333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.161849316897023,
      "rewrite6_current_pass_fraction": 0.3520720141769391,
      "rewrite6_full_passes": 1.0,
      "step": 104800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0237056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1405e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7357733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.844705851344035,
      "rewrite6_current_pass_fraction": 0.3526221495199704,
      "rewrite6_full_passes": 1.0,
      "step": 104810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0261632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.141e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7360133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.221185526912038,
      "rewrite6_current_pass_fraction": 0.35330860158516875,
      "rewrite6_full_passes": 1.0,
      "step": 104820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0286208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1415e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0547,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.73628,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21482012611182,
      "rewrite6_current_pass_fraction": 0.35382465774765826,
      "rewrite6_full_passes": 1.0,
      "step": 104830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0310784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.142e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0724,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7365466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.593665331764466,
      "rewrite6_current_pass_fraction": 0.35433584545578467,
      "rewrite6_full_passes": 1.0,
      "step": 104840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.033536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1425e-05,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0647,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7368133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.42803261303982,
      "rewrite6_current_pass_fraction": 0.3549151915249946,
      "rewrite6_full_passes": 1.0,
      "step": 104850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0359936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.143e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.73716,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.3554994060485677,
      "rewrite6_full_passes": 1.0,
      "step": 104860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0384512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1435e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0344,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.73756,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.788501056163312,
      "rewrite6_current_pass_fraction": 0.35606901520905143,
      "rewrite6_full_passes": 1.0,
      "step": 104870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0409088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.144e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0272,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7378666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.639361395619666,
      "rewrite6_current_pass_fraction": 0.35662401900644586,
      "rewrite6_full_passes": 1.0,
      "step": 104880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0433664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1445e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0499,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7382,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.113232993671563,
      "rewrite6_current_pass_fraction": 0.35724231271056067,
      "rewrite6_full_passes": 1.0,
      "step": 104890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.045824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.145e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0485,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7385866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.08369514879635,
      "rewrite6_current_pass_fraction": 0.3577778426905026,
      "rewrite6_full_passes": 1.0,
      "step": 104900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0482816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1455e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0762,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.73884,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.675877363957884,
      "rewrite6_current_pass_fraction": 0.3583425833966232,
      "rewrite6_full_passes": 1.0,
      "step": 104910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0507392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.146e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7390533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.216941714202072,
      "rewrite6_current_pass_fraction": 0.35879534965239235,
      "rewrite6_full_passes": 1.0,
      "step": 104920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0531968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1465e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7393333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.041569897791508,
      "rewrite6_current_pass_fraction": 0.359452590991412,
      "rewrite6_full_passes": 1.0,
      "step": 104930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0556544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.147e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0606,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7397333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.340357536052313,
      "rewrite6_current_pass_fraction": 0.3599783840626278,
      "rewrite6_full_passes": 1.0,
      "step": 104940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.058112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1475e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7401066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91988202833721,
      "rewrite6_current_pass_fraction": 0.36056746704056397,
      "rewrite6_full_passes": 1.0,
      "step": 104950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0605696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.148e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.069,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.74048,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.520371539689332,
      "rewrite6_current_pass_fraction": 0.36117602383595254,
      "rewrite6_full_passes": 1.0,
      "step": 104960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0630272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1485e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0531,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.74068,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.18090355440295,
      "rewrite6_current_pass_fraction": 0.3616336585460848,
      "rewrite6_full_passes": 1.0,
      "step": 104970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0654848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.149e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0642,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.74096,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.417321274791014,
      "rewrite6_current_pass_fraction": 0.3622860314307414,
      "rewrite6_full_passes": 1.0,
      "step": 104980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0679424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1495e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0591,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7412933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.30837099565101,
      "rewrite6_current_pass_fraction": 0.3628605090455882,
      "rewrite6_full_passes": 1.0,
      "step": 104990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.15e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.472545623779297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0478,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7415733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06894172649243,
      "rewrite6_current_pass_fraction": 0.36344472356916124,
      "rewrite6_full_passes": 1.0,
      "step": 105000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.15e-05,
      "eval_loss": 3.083348035812378,
      "eval_perplexity": 21.831372391556386,
      "eval_runtime": 0.0492,
      "eval_samples_per_second": 649.807,
      "eval_steps_per_second": 20.306,
      "step": 105000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0728576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1505e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0704,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7419466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.55052115965443,
      "rewrite6_current_pass_fraction": 0.36406301727327606,
      "rewrite6_full_passes": 1.0,
      "step": 105010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0753152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.151e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0564,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.74216,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25091599312008,
      "rewrite6_current_pass_fraction": 0.36471052170356955,
      "rewrite6_full_passes": 1.0,
      "step": 105020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0777728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1515e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.048,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7427333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.073155936244657,
      "rewrite6_current_pass_fraction": 0.36519249868551734,
      "rewrite6_full_passes": 1.0,
      "step": 105030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0802304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.152e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0633,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7431066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.398054357057195,
      "rewrite6_current_pass_fraction": 0.3658253977527215,
      "rewrite6_full_passes": 1.0,
      "step": 105040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.082688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1525e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0631,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.74348,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.393775174118343,
      "rewrite6_current_pass_fraction": 0.36641448073065763,
      "rewrite6_full_passes": 1.0,
      "step": 105050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0851456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.153e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7438533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.039465846006077,
      "rewrite6_current_pass_fraction": 0.3670473797978618,
      "rewrite6_full_passes": 1.0,
      "step": 105060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0876032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1535e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.067,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7442266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.477373808673548,
      "rewrite6_current_pass_fraction": 0.36768514731942903,
      "rewrite6_full_passes": 1.0,
      "step": 105070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0900608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.154e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0655,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.74448,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.44518189792958,
      "rewrite6_current_pass_fraction": 0.3682888356604545,
      "rewrite6_full_passes": 1.0,
      "step": 105080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0925184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1545e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0795,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7448,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.747525914346035,
      "rewrite6_current_pass_fraction": 0.36884383945784893,
      "rewrite6_full_passes": 1.0,
      "step": 105090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.094976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.155e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0548,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7451066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.216941714202072,
      "rewrite6_current_pass_fraction": 0.3693452902572491,
      "rewrite6_full_passes": 1.0,
      "step": 105100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0974336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1555e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0579,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7454933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28281628634838,
      "rewrite6_current_pass_fraction": 0.36997818932445325,
      "rewrite6_full_passes": 1.0,
      "step": 105110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.0998912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.156e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7458266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.945000955113006,
      "rewrite6_current_pass_fraction": 0.3706451675721992,
      "rewrite6_full_passes": 1.0,
      "step": 105120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1023488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1565e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0323,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.74608,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.744891010519986,
      "rewrite6_current_pass_fraction": 0.3713170142743082,
      "rewrite6_full_passes": 1.0,
      "step": 105130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1048064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.157e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7465733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23604555718468,
      "rewrite6_current_pass_fraction": 0.37192070261533366,
      "rewrite6_full_passes": 1.0,
      "step": 105140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.107264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1575e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.74692,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.37245136414091257,
      "rewrite6_full_passes": 1.0,
      "step": 105150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1097216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.158e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0505,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.74728,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.125904734609893,
      "rewrite6_current_pass_fraction": 0.3729771572121283,
      "rewrite6_full_passes": 1.0,
      "step": 105160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1121792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1585e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.477996826171875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0912,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7475733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.003466299161783,
      "rewrite6_current_pass_fraction": 0.37361979318805866,
      "rewrite6_full_passes": 1.0,
      "step": 105170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1146368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.159e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.485946655273438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0583,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7478,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.29133111571526,
      "rewrite6_current_pass_fraction": 0.37424295534653657,
      "rewrite6_full_passes": 1.0,
      "step": 105180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1170944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1595e-05,
      "grad_norm": 0.61328125,
      "host_rss_gib": 16.242916107177734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0864,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7481866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.898102735775886,
      "rewrite6_current_pass_fraction": 0.3748515121419252,
      "rewrite6_full_passes": 1.0,
      "step": 105190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.119552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.16e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.242916107177734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7486133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.3753432260325992,
      "rewrite6_full_passes": 1.0,
      "step": 105200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1220096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1605e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.242916107177734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0802,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7489733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.76275451187338,
      "rewrite6_current_pass_fraction": 0.37581059765145763,
      "rewrite6_full_passes": 1.0,
      "step": 105210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1244672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.161e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0717,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7494133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.57855505524601,
      "rewrite6_current_pass_fraction": 0.37640941753812,
      "rewrite6_full_passes": 1.0,
      "step": 105220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1269248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1615e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7497066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.132243456791066,
      "rewrite6_current_pass_fraction": 0.37697415824424063,
      "rewrite6_full_passes": 1.0,
      "step": 105230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1293824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.162e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.75012,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.007930304805882,
      "rewrite6_current_pass_fraction": 0.3775194251329088,
      "rewrite6_full_passes": 1.0,
      "step": 105240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.13184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1625e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0769,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7504266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.691055789941966,
      "rewrite6_current_pass_fraction": 0.3780939027477556,
      "rewrite6_full_passes": 1.0,
      "step": 105250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1342976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.163e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0596,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7508266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.319027845139193,
      "rewrite6_current_pass_fraction": 0.37870245954314424,
      "rewrite6_full_passes": 1.0,
      "step": 105260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1367552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1635e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7510933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.60662541860077,
      "rewrite6_current_pass_fraction": 0.37918930497945513,
      "rewrite6_full_passes": 1.0,
      "step": 105270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1392128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.164e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7514133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.299849351695165,
      "rewrite6_current_pass_fraction": 0.37974917723121265,
      "rewrite6_full_passes": 1.0,
      "step": 105280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1416704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1645e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0459,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7517466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.028948742578034,
      "rewrite6_current_pass_fraction": 0.38031878639169636,
      "rewrite6_full_passes": 1.0,
      "step": 105290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.144128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.165e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7520666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.132243456791066,
      "rewrite6_current_pass_fraction": 0.3808445794629121,
      "rewrite6_full_passes": 1.0,
      "step": 105300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1465856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1655e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0524,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.75244,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.166082110025602,
      "rewrite6_current_pass_fraction": 0.38145313625830074,
      "rewrite6_full_passes": 1.0,
      "step": 105310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1490432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.166e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.242919921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0503,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.75268,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.1216799761529,
      "rewrite6_current_pass_fraction": 0.38208603532550484,
      "rewrite6_full_passes": 1.0,
      "step": 105320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1515008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1665e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.242923736572266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0738,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7530533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.623917634899918,
      "rewrite6_current_pass_fraction": 0.38272867130143523,
      "rewrite6_full_passes": 1.0,
      "step": 105330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1539584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.167e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.242923736572266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.046,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7533466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.031051742600535,
      "rewrite6_current_pass_fraction": 0.3833080173706452,
      "rewrite6_full_passes": 1.0,
      "step": 105340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.156416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1675e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.242923736572266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0406,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7536266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.917790144730294,
      "rewrite6_current_pass_fraction": 0.38391170571167066,
      "rewrite6_full_passes": 1.0,
      "step": 105350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1588736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.168e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.242923736572266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7539333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14704120581873,
      "rewrite6_current_pass_fraction": 0.3844861833265175,
      "rewrite6_full_passes": 1.0,
      "step": 105360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1613312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1685e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.64791488647461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0471,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7542933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.054198627970376,
      "rewrite6_current_pass_fraction": 0.3851482931199003,
      "rewrite6_full_passes": 1.0,
      "step": 105370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1637888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.169e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.649417877197266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0686,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7546,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.51176511247365,
      "rewrite6_current_pass_fraction": 0.38569356000856847,
      "rewrite6_full_passes": 1.0,
      "step": 105380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1662464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1695e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.652080535888672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0436,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.75496,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.980637739420835,
      "rewrite6_current_pass_fraction": 0.3863215906214095,
      "rewrite6_full_passes": 1.0,
      "step": 105390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.168704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.17e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.652088165283203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7554,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.3867548830597262,
      "rewrite6_full_passes": 1.0,
      "step": 105400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1711616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1705e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.652091979980469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7556666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.69516296961328,
      "rewrite6_current_pass_fraction": 0.3872709392222157,
      "rewrite6_full_passes": 1.0,
      "step": 105410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1736192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.171e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.652091979980469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.75596,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.38790870674378297,
      "rewrite6_full_passes": 1.0,
      "step": 105420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1760768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1715e-05,
      "grad_norm": 0.58203125,
      "host_rss_gib": 15.65262222290039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0738,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7563066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.623917634899918,
      "rewrite6_current_pass_fraction": 0.3885172635391716,
      "rewrite6_full_passes": 1.0,
      "step": 105430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1785344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.172e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.653423309326172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0707,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7565866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.55698728587277,
      "rewrite6_current_pass_fraction": 0.3891696364238282,
      "rewrite6_full_passes": 1.0,
      "step": 105440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.180992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1725e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.653434753417969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0305,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7569466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.707583793269517,
      "rewrite6_current_pass_fraction": 0.38989016766956824,
      "rewrite6_full_passes": 1.0,
      "step": 105450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1834496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.173e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.654563903808594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7573333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.390547409008588,
      "rewrite6_full_passes": 1.0,
      "step": 105460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1859072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1735e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 15.654563903808594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0566,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7576266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25516660136536,
      "rewrite6_current_pass_fraction": 0.39115596580397655,
      "rewrite6_full_passes": 1.0,
      "step": 105470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1883648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.174e-05,
      "grad_norm": 0.99609375,
      "host_rss_gib": 15.654563903808594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0657,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.75796,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.449471363241397,
      "rewrite6_current_pass_fraction": 0.39172070651009716,
      "rewrite6_full_passes": 1.0,
      "step": 105480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1908224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1745e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.654563903808594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7582266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.738668476643575,
      "rewrite6_current_pass_fraction": 0.3922318942182236,
      "rewrite6_full_passes": 1.0,
      "step": 105490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.19328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.175e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.654747009277344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0613,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7585733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.35530101593532,
      "rewrite6_current_pass_fraction": 0.39271387120017137,
      "rewrite6_full_passes": 1.0,
      "step": 105500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.175e-05,
      "eval_loss": 3.083340644836426,
      "eval_perplexity": 21.831211037004323,
      "eval_runtime": 0.0501,
      "eval_samples_per_second": 638.706,
      "eval_steps_per_second": 19.96,
      "step": 105500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1957376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1755e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.710670471191406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0777,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7588533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.70841557756311,
      "rewrite6_current_pass_fraction": 0.39331755954119685,
      "rewrite6_full_passes": 1.0,
      "step": 105510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.1981952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.176e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.710670471191406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0544,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7591333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20845663464543,
      "rewrite6_current_pass_fraction": 0.3939212478822223,
      "rewrite6_full_passes": 1.0,
      "step": 105520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2006528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1765e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.710670471191406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0637,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.75948,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.406615290872633,
      "rewrite6_current_pass_fraction": 0.3945930945843314,
      "rewrite6_full_passes": 1.0,
      "step": 105530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2031104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.177e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.711040496826172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0539,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7598266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.197855056943403,
      "rewrite6_current_pass_fraction": 0.39508967692936847,
      "rewrite6_full_passes": 1.0,
      "step": 105540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.205568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1775e-05,
      "grad_norm": 0.6328125,
      "host_rss_gib": 15.711040496826172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0582,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7602533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.28920208905679,
      "rewrite6_current_pass_fraction": 0.3956544176354891,
      "rewrite6_full_passes": 1.0,
      "step": 105550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2080256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.178e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.711429595947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0374,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.76052,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.850960201205005,
      "rewrite6_current_pass_fraction": 0.3962727113396039,
      "rewrite6_full_passes": 1.0,
      "step": 105560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2104832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1785e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.711429595947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0881,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.76072,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.93536117112368,
      "rewrite6_current_pass_fraction": 0.396813109773909,
      "rewrite6_full_passes": 1.0,
      "step": 105570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2129408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.179e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.711429595947266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7610666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.3973827189343927,
      "rewrite6_full_passes": 1.0,
      "step": 105580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2153984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1795e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.711433410644531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0418,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7613066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.942906559739004,
      "rewrite6_current_pass_fraction": 0.39807403945395414,
      "rewrite6_full_passes": 1.0,
      "step": 105590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.217856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.18e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.711437225341797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0526,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7616933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.170315749797474,
      "rewrite6_current_pass_fraction": 0.3986728593406165,
      "rewrite6_full_passes": 1.0,
      "step": 105600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2203136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1805e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.711456298828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0816,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7618666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.793243705645736,
      "rewrite6_current_pass_fraction": 0.3993057584078207,
      "rewrite6_full_passes": 1.0,
      "step": 105610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2227712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.181e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.71316909790039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0589,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.76212,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.304109747590893,
      "rewrite6_current_pass_fraction": 0.3999143152032093,
      "rewrite6_full_passes": 1.0,
      "step": 105620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2252288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1815e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.713172912597656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0645,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7623733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.423747435049286,
      "rewrite6_current_pass_fraction": 0.4005666880878659,
      "rewrite6_full_passes": 1.0,
      "step": 105630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2276864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.182e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.713172912597656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0518,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7627333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.153386269892504,
      "rewrite6_current_pass_fraction": 0.4011411657027127,
      "rewrite6_full_passes": 1.0,
      "step": 105640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.230144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1825e-05,
      "grad_norm": 0.55078125,
      "host_rss_gib": 15.713172912597656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0624,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7631066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.378804771748577,
      "rewrite6_current_pass_fraction": 0.4017302486806489,
      "rewrite6_full_passes": 1.0,
      "step": 105650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2326016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.183e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.713642120361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.76352,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.97853978054658,
      "rewrite6_current_pass_fraction": 0.4023582792934899,
      "rewrite6_full_passes": 1.0,
      "step": 105660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2350592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1835e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.713642120361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.1014,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7638266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.229050177383325,
      "rewrite6_current_pass_fraction": 0.40299117836069404,
      "rewrite6_full_passes": 1.0,
      "step": 105670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2375168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.184e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.713642120361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7640533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.14704120581873,
      "rewrite6_current_pass_fraction": 0.40360460361044576,
      "rewrite6_full_passes": 1.0,
      "step": 105680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2399744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1845e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.713642120361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0442,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7642933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.993229899334697,
      "rewrite6_current_pass_fraction": 0.4043007925843703,
      "rewrite6_full_passes": 1.0,
      "step": 105690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.242432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.185e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.713642120361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.067,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7646266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.477373808673548,
      "rewrite6_current_pass_fraction": 0.40488013865358025,
      "rewrite6_full_passes": 1.0,
      "step": 105700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2448896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1855e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.713642120361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.76496,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.986932874965774,
      "rewrite6_current_pass_fraction": 0.4055714591731417,
      "rewrite6_full_passes": 1.0,
      "step": 105710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2473472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.186e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0458,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.76528,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.02684595284501,
      "rewrite6_current_pass_fraction": 0.4061702790598041,
      "rewrite6_full_passes": 1.0,
      "step": 105720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2498048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1865e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0702,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7657066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.54621148640419,
      "rewrite6_current_pass_fraction": 0.4068421257619131,
      "rewrite6_full_passes": 1.0,
      "step": 105730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2522624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.187e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0563,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.766,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.248791007771803,
      "rewrite6_current_pass_fraction": 0.4073825241962182,
      "rewrite6_full_passes": 1.0,
      "step": 105740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.25472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1875e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0552,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7662533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.225430188469428,
      "rewrite6_current_pass_fraction": 0.40797647562851747,
      "rewrite6_full_passes": 1.0,
      "step": 105750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2571776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.188e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0508,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7665866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.132243456791066,
      "rewrite6_current_pass_fraction": 0.40857042706081675,
      "rewrite6_full_passes": 1.0,
      "step": 105760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2596352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1885e-05,
      "grad_norm": 0.53125,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0194,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.76688,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.479000596877352,
      "rewrite6_current_pass_fraction": 0.40916924694747914,
      "rewrite6_full_passes": 1.0,
      "step": 105770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2620928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.189e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0537,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7671466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.193615909860856,
      "rewrite6_current_pass_fraction": 0.40968530310996865,
      "rewrite6_full_passes": 1.0,
      "step": 105780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2645504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1895e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7674,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23604555718468,
      "rewrite6_current_pass_fraction": 0.41024517536172617,
      "rewrite6_full_passes": 1.0,
      "step": 105790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.267008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.19e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0349,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7677333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.798897905687166,
      "rewrite6_current_pass_fraction": 0.4108586006114779,
      "rewrite6_full_passes": 1.0,
      "step": 105800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2694656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1905e-05,
      "grad_norm": 0.5625,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0686,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7681066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.51176511247365,
      "rewrite6_current_pass_fraction": 0.4114525520437771,
      "rewrite6_full_passes": 1.0,
      "step": 105810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2719232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.191e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7684266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.4120951880197075,
      "rewrite6_full_passes": 1.0,
      "step": 105820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2743808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1915e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0655,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7687466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.44518189792958,
      "rewrite6_current_pass_fraction": 0.4126501918171019,
      "rewrite6_full_passes": 1.0,
      "step": 105830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2768384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.192e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0405,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7690666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.91569847030129,
      "rewrite6_current_pass_fraction": 0.4132051956144963,
      "rewrite6_full_passes": 1.0,
      "step": 105840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.279296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1925e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.066,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7693733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.455907169973106,
      "rewrite6_current_pass_fraction": 0.4138040155011587,
      "rewrite6_full_passes": 1.0,
      "step": 105850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2817536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.193e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0608,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7696933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34462603439513,
      "rewrite6_current_pass_fraction": 0.41444178302272594,
      "rewrite6_full_passes": 1.0,
      "step": 105860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2842112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1935e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7699333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.782265441235463,
      "rewrite6_current_pass_fraction": 0.4149870499113941,
      "rewrite6_full_passes": 1.0,
      "step": 105870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2866688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.194e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0454,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7702533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.018436896387286,
      "rewrite6_current_pass_fraction": 0.4155810013436934,
      "rewrite6_full_passes": 1.0,
      "step": 105880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2891264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1945e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0739,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7705466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.626080134786605,
      "rewrite6_current_pass_fraction": 0.4161506105041771,
      "rewrite6_full_passes": 1.0,
      "step": 105890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.291584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.195e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0631,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7710266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.393775174118343,
      "rewrite6_current_pass_fraction": 0.4166764035753929,
      "rewrite6_full_passes": 1.0,
      "step": 105900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2940416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1955e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0353,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7714,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.80721912898315,
      "rewrite6_current_pass_fraction": 0.41719732819224553,
      "rewrite6_full_passes": 1.0,
      "step": 105910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2964992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.196e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0606,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.77172,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.340357536052313,
      "rewrite6_current_pass_fraction": 0.4177718058070924,
      "rewrite6_full_passes": 1.0,
      "step": 105920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.2989568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1965e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0526,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.77208,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.170315749797474,
      "rewrite6_current_pass_fraction": 0.4182878619695819,
      "rewrite6_full_passes": 1.0,
      "step": 105930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3014144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.197e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0287,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.77252,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.670343668608663,
      "rewrite6_current_pass_fraction": 0.4188428657669763,
      "rewrite6_full_passes": 1.0,
      "step": 105940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.303872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1975e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7727866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.024743373380456,
      "rewrite6_current_pass_fraction": 0.4194708963798173,
      "rewrite6_full_passes": 1.0,
      "step": 105950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3063296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.198e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0546,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7730266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.21269875016978,
      "rewrite6_current_pass_fraction": 0.4199431364530389,
      "rewrite6_full_passes": 1.0,
      "step": 105960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3087872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1985e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0633,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7735066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.398054357057195,
      "rewrite6_current_pass_fraction": 0.4205030087047964,
      "rewrite6_full_passes": 1.0,
      "step": 105970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3112448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.199e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7738666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.045778632630523,
      "rewrite6_current_pass_fraction": 0.42102393332164906,
      "rewrite6_full_passes": 1.0,
      "step": 105980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3137024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.1995e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0514,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7742933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.144926607429838,
      "rewrite6_current_pass_fraction": 0.42154485793850166,
      "rewrite6_full_passes": 1.0,
      "step": 105990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.31616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.713977813720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0841,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7744933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.84779497558516,
      "rewrite6_current_pass_fraction": 0.4221242040077116,
      "rewrite6_full_passes": 1.0,
      "step": 106000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.2e-05,
      "eval_loss": 3.0831007957458496,
      "eval_perplexity": 21.825975468789157,
      "eval_runtime": 0.05,
      "eval_samples_per_second": 639.763,
      "eval_steps_per_second": 19.993,
      "step": 106000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3186176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2005e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.840400695800781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.061,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7747466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.34889538652299,
      "rewrite6_current_pass_fraction": 0.42262565480711184,
      "rewrite6_full_passes": 1.0,
      "step": 106010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3210752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.201e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.605396270751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7750533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.041569897791508,
      "rewrite6_current_pass_fraction": 0.42323908005686356,
      "rewrite6_full_passes": 1.0,
      "step": 106020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3235328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2015e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0777,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7752933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.70841557756311,
      "rewrite6_current_pass_fraction": 0.4238232945804366,
      "rewrite6_full_passes": 1.0,
      "step": 106030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3259904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.202e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.77564,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.210577586354717,
      "rewrite6_current_pass_fraction": 0.42451461509999805,
      "rewrite6_full_passes": 1.0,
      "step": 106040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.328448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2025e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0498,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7758533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.111121775934837,
      "rewrite6_current_pass_fraction": 0.4251961987108333,
      "rewrite6_full_passes": 1.0,
      "step": 106050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3309056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.203e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.067,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7761866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.477373808673548,
      "rewrite6_current_pass_fraction": 0.42573659714513834,
      "rewrite6_full_passes": 1.0,
      "step": 106060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3333632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2035e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.77644,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13435678680148,
      "rewrite6_current_pass_fraction": 0.4263110747599852,
      "rewrite6_full_passes": 1.0,
      "step": 106070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3358208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.204e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.052,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7767066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.157617370242413,
      "rewrite6_current_pass_fraction": 0.4269877899164573,
      "rewrite6_full_passes": 1.0,
      "step": 106080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3382784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2045e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0738,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7769866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.623917634899918,
      "rewrite6_current_pass_fraction": 0.4275281883507624,
      "rewrite6_full_passes": 1.0,
      "step": 106090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.340736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.205e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.076,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7773466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.67154262197374,
      "rewrite6_current_pass_fraction": 0.42805398142197815,
      "rewrite6_full_passes": 1.0,
      "step": 106100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3431936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2055e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0717,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7776933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.57855505524601,
      "rewrite6_current_pass_fraction": 0.42859924831064633,
      "rewrite6_full_passes": 1.0,
      "step": 106110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3456512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.206e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.77804,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.210577586354717,
      "rewrite6_current_pass_fraction": 0.42921267356039805,
      "rewrite6_full_passes": 1.0,
      "step": 106120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3481088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2065e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0571,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7783733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.265796842004743,
      "rewrite6_current_pass_fraction": 0.4298747833537808,
      "rewrite6_full_passes": 1.0,
      "step": 106130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3505664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.207e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0587,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7787066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.299849351695165,
      "rewrite6_current_pass_fraction": 0.4304151817880859,
      "rewrite6_full_passes": 1.0,
      "step": 106140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.353024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2075e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0382,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.77908,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86764764345287,
      "rewrite6_current_pass_fraction": 0.43097505403984343,
      "rewrite6_full_passes": 1.0,
      "step": 106150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3554816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.208e-05,
      "grad_norm": 0.42578125,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.058,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.77956,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.284944674394637,
      "rewrite6_current_pass_fraction": 0.4315787423808689,
      "rewrite6_full_passes": 1.0,
      "step": 106160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3579392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2085e-05,
      "grad_norm": 0.53125,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0756,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.77992,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.66287573841722,
      "rewrite6_current_pass_fraction": 0.4321921676306206,
      "rewrite6_full_passes": 1.0,
      "step": 106170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3603968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.209e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0651,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7803733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.436605540556233,
      "rewrite6_current_pass_fraction": 0.4327276976105626,
      "rewrite6_full_passes": 1.0,
      "step": 106180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3628544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2095e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0564,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7807733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.25091599312008,
      "rewrite6_current_pass_fraction": 0.43337033358649296,
      "rewrite6_full_passes": 1.0,
      "step": 106190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.365312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.21e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0581,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7810933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.287073275290354,
      "rewrite6_current_pass_fraction": 0.43396915347315534,
      "rewrite6_full_passes": 1.0,
      "step": 106200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3677696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2105e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0791,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7813466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.73882864355042,
      "rewrite6_current_pass_fraction": 0.4345046834530973,
      "rewrite6_full_passes": 1.0,
      "step": 106210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3702272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.211e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.60147476196289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0551,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7816533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.22330775157419,
      "rewrite6_current_pass_fraction": 0.4352154777901112,
      "rewrite6_full_passes": 1.0,
      "step": 106220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3726848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2115e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.037609100341797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7819333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.259418059817307,
      "rewrite6_current_pass_fraction": 0.4357315339526007,
      "rewrite6_full_passes": 1.0,
      "step": 106230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3751424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.212e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.039154052734375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0242,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7823466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.577536095751576,
      "rewrite6_current_pass_fraction": 0.43629140620435825,
      "rewrite6_full_passes": 1.0,
      "step": 106240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2125e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0227,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.78276,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.54669292976553,
      "rewrite6_current_pass_fraction": 0.4368415415473895,
      "rewrite6_full_passes": 1.0,
      "step": 106250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3800576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.213e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.039794921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0713,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7831466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.56992535927817,
      "rewrite6_current_pass_fraction": 0.43734786080115284,
      "rewrite6_full_passes": 1.0,
      "step": 106260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3825152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2135e-05,
      "grad_norm": 0.5546875,
      "host_rss_gib": 15.04031753540039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.066,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7834533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.455907169973106,
      "rewrite6_current_pass_fraction": 0.4379466806878152,
      "rewrite6_full_passes": 1.0,
      "step": 106270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3849728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.214e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.04031753540039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7837866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.871821590362348,
      "rewrite6_current_pass_fraction": 0.4385747113006563,
      "rewrite6_full_passes": 1.0,
      "step": 106280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3874304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2145e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.040752410888672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0637,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7842,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.406615290872633,
      "rewrite6_current_pass_fraction": 0.43917353118731867,
      "rewrite6_full_passes": 1.0,
      "step": 106290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.389888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.215e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.040756225585938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0235,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7845333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.563136860804747,
      "rewrite6_current_pass_fraction": 0.43978208798270724,
      "rewrite6_full_passes": 1.0,
      "step": 106300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3923456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2155e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.040763854980469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7849066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.4403565655975541,
      "rewrite6_full_passes": 1.0,
      "step": 106310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3948032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.216e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.040763854980469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0338,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7853466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.77603169671152,
      "rewrite6_current_pass_fraction": 0.4409505170298534,
      "rewrite6_full_passes": 1.0,
      "step": 106320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3972608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2165e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.040763854980469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.78568,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13647032815547,
      "rewrite6_current_pass_fraction": 0.4414617047379798,
      "rewrite6_full_passes": 1.0,
      "step": 106330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.3997184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.217e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.04172134399414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0658,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7860133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45161641762865,
      "rewrite6_current_pass_fraction": 0.4421189460769995,
      "rewrite6_full_passes": 1.0,
      "step": 106340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.402176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2175e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.04172134399414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0468,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.78632,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.04788331572619,
      "rewrite6_current_pass_fraction": 0.4427518451442036,
      "rewrite6_full_passes": 1.0,
      "step": 106350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4046336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.218e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.042366027832031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7866666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.738668476643575,
      "rewrite6_current_pass_fraction": 0.44326303285233004,
      "rewrite6_full_passes": 1.0,
      "step": 106360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4070912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2185e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.042366027832031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.055,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.78708,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.221185526912038,
      "rewrite6_current_pass_fraction": 0.44381803664972447,
      "rewrite6_full_passes": 1.0,
      "step": 106370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4095488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.219e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.042366027832031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0486,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.78744,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.085803623733216,
      "rewrite6_current_pass_fraction": 0.4444071196276606,
      "rewrite6_full_passes": 1.0,
      "step": 106380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4120064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2195e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.042617797851562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0357,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7877466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.81554368143424,
      "rewrite6_current_pass_fraction": 0.4451179139646745,
      "rewrite6_full_passes": 1.0,
      "step": 106390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.414464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.22e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.043006896972656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0658,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7880933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.45161641762865,
      "rewrite6_current_pass_fraction": 0.4456583123989796,
      "rewrite6_full_passes": 1.0,
      "step": 106400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4169216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2205e-05,
      "grad_norm": 0.546875,
      "host_rss_gib": 15.043006896972656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0892,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7884133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.95950334417276,
      "rewrite6_current_pass_fraction": 0.4462279215594633,
      "rewrite6_full_passes": 1.0,
      "step": 106410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4193792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.221e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.043006896972656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0635,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7887066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.402334395918224,
      "rewrite6_current_pass_fraction": 0.44674884617631594,
      "rewrite6_full_passes": 1.0,
      "step": 106420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4218368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2215e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.043018341064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.1011,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.78904,
      "no_prompt_full_passes": 0.0,
      "perplexity": 22.22238246253735,
      "rewrite6_current_pass_fraction": 0.4472941130649841,
      "rewrite6_full_passes": 1.0,
      "step": 106430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4242944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.222e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.043018341064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0591,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.78948,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.30837099565101,
      "rewrite6_current_pass_fraction": 0.4479221436778252,
      "rewrite6_full_passes": 1.0,
      "step": 106440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.426752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2225e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.043018341064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0491,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7898133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.09634916170988,
      "rewrite6_current_pass_fraction": 0.4485453058363031,
      "rewrite6_full_passes": 1.0,
      "step": 106450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4292096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.223e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.043018341064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0641,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7902,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.415179649746566,
      "rewrite6_current_pass_fraction": 0.44912952035987613,
      "rewrite6_full_passes": 1.0,
      "step": 106460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4316672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2235e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.043018341064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0557,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.79064,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.23604555718468,
      "rewrite6_current_pass_fraction": 0.4496212342505501,
      "rewrite6_full_passes": 1.0,
      "step": 106470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4341248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.224e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.043033599853516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0755,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7910133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.660709559154142,
      "rewrite6_current_pass_fraction": 0.45028334404393294,
      "rewrite6_full_passes": 1.0,
      "step": 106480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4365824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2245e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.043033599853516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0778,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7913733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.710586527666557,
      "rewrite6_current_pass_fraction": 0.4508821639305953,
      "rewrite6_full_passes": 1.0,
      "step": 106490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.43904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.225e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.043312072753906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0614,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7916,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.357436652816972,
      "rewrite6_current_pass_fraction": 0.45154914217834125,
      "rewrite6_full_passes": 1.0,
      "step": 106500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.225e-05,
      "eval_loss": 3.083246946334839,
      "eval_perplexity": 21.8291655810719,
      "eval_runtime": 0.0498,
      "eval_samples_per_second": 642.984,
      "eval_steps_per_second": 20.093,
      "step": 106500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4414976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2255e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.096759796142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0629,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7918933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.3894968469305,
      "rewrite6_current_pass_fraction": 0.4520895406126463,
      "rewrite6_full_passes": 1.0,
      "step": 106510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4439552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.226e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.096763610839844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7921866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.391635903566247,
      "rewrite6_current_pass_fraction": 0.4526786235905825,
      "rewrite6_full_passes": 1.0,
      "step": 106520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4464128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2265e-05,
      "grad_norm": 0.578125,
      "host_rss_gib": 15.097099304199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0504,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7925333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.12379224976243,
      "rewrite6_current_pass_fraction": 0.4532579696597924,
      "rewrite6_full_passes": 1.0,
      "step": 106530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4488704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.227e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.097099304199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0714,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7928133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.572082459667325,
      "rewrite6_current_pass_fraction": 0.45385192109209166,
      "rewrite6_full_passes": 1.0,
      "step": 106540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.451328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2275e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.097099304199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0783,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7932666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.72144453520607,
      "rewrite6_current_pass_fraction": 0.4544069248894861,
      "rewrite6_full_passes": 1.0,
      "step": 106550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4537856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.228e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.097099304199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7935066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.4550252185936009,
      "rewrite6_full_passes": 1.0,
      "step": 106560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4562432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2285e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.097099304199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7938266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.4555753539366322,
      "rewrite6_full_passes": 1.0,
      "step": 106570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4587008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.229e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.097099304199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7942533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.934531072324265,
      "rewrite6_current_pass_fraction": 0.45625206909310434,
      "rewrite6_full_passes": 1.0,
      "step": 106580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4611584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2295e-05,
      "grad_norm": 0.546875,
      "host_rss_gib": 15.097099304199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0734,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7944533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.615269797528736,
      "rewrite6_current_pass_fraction": 0.45682654670795114,
      "rewrite6_full_passes": 1.0,
      "step": 106590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.463616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.23e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.097099304199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.06,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7947466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.327557162026903,
      "rewrite6_current_pass_fraction": 0.45732799750735137,
      "rewrite6_full_passes": 1.0,
      "step": 106600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4660736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2305e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.097099304199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0456,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7951866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.022641004163326,
      "rewrite6_current_pass_fraction": 0.45798523884637105,
      "rewrite6_full_passes": 1.0,
      "step": 106610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4685312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.231e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.097099304199219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0333,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.79552,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.765646277434353,
      "rewrite6_current_pass_fraction": 0.4585110319175868,
      "rewrite6_full_passes": 1.0,
      "step": 106620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4709888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2315e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.231815338134766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0523,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7958533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.16396560764148,
      "rewrite6_current_pass_fraction": 0.45914879943915404,
      "rewrite6_full_passes": 1.0,
      "step": 106630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4734464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.232e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.65871810913086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9559,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7962533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.219012049312248,
      "rewrite6_current_pass_fraction": 0.4596989347821854,
      "rewrite6_full_passes": 1.0,
      "step": 106640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.475904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2325e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.31003189086914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9851,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.79664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.788480985529784,
      "rewrite6_current_pass_fraction": 0.4603951237561099,
      "rewrite6_full_passes": 1.0,
      "step": 106650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4783616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.233e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9881,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7969866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.8479355657658,
      "rewrite6_current_pass_fraction": 0.4610182859145878,
      "rewrite6_full_passes": 1.0,
      "step": 106660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4808192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2335e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0059,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7973866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.2043918683445,
      "rewrite6_current_pass_fraction": 0.46160250043816087,
      "rewrite6_full_passes": 1.0,
      "step": 106670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4832768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.234e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9909,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7977066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.903587661925197,
      "rewrite6_current_pass_fraction": 0.4622110572335495,
      "rewrite6_full_passes": 1.0,
      "step": 106680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4857344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2345e-05,
      "grad_norm": 0.5703125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9875,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.798,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.836030376340315,
      "rewrite6_current_pass_fraction": 0.462858561663843,
      "rewrite6_full_passes": 1.0,
      "step": 106690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.488192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.235e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9942,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7982933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.969377995555327,
      "rewrite6_current_pass_fraction": 0.4634525130961422,
      "rewrite6_full_passes": 1.0,
      "step": 106700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4906496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2355e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9533,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7985733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.169107521982443,
      "rewrite6_current_pass_fraction": 0.4640367276197153,
      "rewrite6_full_passes": 1.0,
      "step": 106710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4931072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.236e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9963,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7988,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.011357752663397,
      "rewrite6_current_pass_fraction": 0.46462094214328836,
      "rewrite6_full_passes": 1.0,
      "step": 106720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4955648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2365e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9936,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7991066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.957399962527237,
      "rewrite6_current_pass_fraction": 0.4652002882124983,
      "rewrite6_full_passes": 1.0,
      "step": 106730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.4980224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.237e-05,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9779,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.7995066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.6465156110719,
      "rewrite6_current_pass_fraction": 0.46581858191661313,
      "rewrite6_full_passes": 1.0,
      "step": 106740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.50048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2375e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9995,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.79996,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.07549666499979,
      "rewrite6_current_pass_fraction": 0.4664174018032755,
      "rewrite6_full_passes": 1.0,
      "step": 106750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5029376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.238e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9759,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.80028,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.607261846698723,
      "rewrite6_current_pass_fraction": 0.4669724056006699,
      "rewrite6_full_passes": 1.0,
      "step": 106760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5053952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2385e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0023,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8006133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.131786825109714,
      "rewrite6_current_pass_fraction": 0.46751767248933807,
      "rewrite6_full_passes": 1.0,
      "step": 106770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5078528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.239e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9936,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8009466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.957399962527237,
      "rewrite6_current_pass_fraction": 0.46805320246928006,
      "rewrite6_full_passes": 1.0,
      "step": 106780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5103104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2395e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0181,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8012133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.452395193360626,
      "rewrite6_current_pass_fraction": 0.4686130747210376,
      "rewrite6_full_passes": 1.0,
      "step": 106790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.512768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.24e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9816,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8016133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.719342365244916,
      "rewrite6_current_pass_fraction": 0.46927031606005726,
      "rewrite6_full_passes": 1.0,
      "step": 106800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5152256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2405e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.015,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8018,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.389090940549213,
      "rewrite6_current_pass_fraction": 0.46992268894471384,
      "rewrite6_full_passes": 1.0,
      "step": 106810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5176832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.241e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9947,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8020133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.979365181141432,
      "rewrite6_current_pass_fraction": 0.47046308737901893,
      "rewrite6_full_passes": 1.0,
      "step": 106820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5201408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2415e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9756,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8023733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.60138055038327,
      "rewrite6_current_pass_fraction": 0.471047301902592,
      "rewrite6_full_passes": 1.0,
      "step": 106830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5225984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.242e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9762,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8026,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.61314490766775,
      "rewrite6_current_pass_fraction": 0.47148546279527176,
      "rewrite6_full_passes": 1.0,
      "step": 106840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.525056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2425e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9837,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.80296,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.760796495814635,
      "rewrite6_current_pass_fraction": 0.4720696773188448,
      "rewrite6_full_passes": 1.0,
      "step": 106850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5275136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.243e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9823,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8032266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.733150737266957,
      "rewrite6_current_pass_fraction": 0.47258573348133437,
      "rewrite6_full_passes": 1.0,
      "step": 106860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5299712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2435e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9947,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.80356,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.979365181141432,
      "rewrite6_current_pass_fraction": 0.47319915873108603,
      "rewrite6_full_passes": 1.0,
      "step": 106870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5324288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.244e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9776,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8038666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.640622540393377,
      "rewrite6_current_pass_fraction": 0.4737736363459329,
      "rewrite6_full_passes": 1.0,
      "step": 106880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5348864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2445e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.979,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.80416,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.668138668745478,
      "rewrite6_current_pass_fraction": 0.4743286401433273,
      "rewrite6_full_passes": 1.0,
      "step": 106890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.537344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.245e-05,
      "grad_norm": 0.55859375,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0098,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8044133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.28334285097683,
      "rewrite6_current_pass_fraction": 0.4749274600299897,
      "rewrite6_full_passes": 1.0,
      "step": 106900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5398016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2455e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9949,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8048266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.983361453791602,
      "rewrite6_current_pass_fraction": 0.47556522755155695,
      "rewrite6_full_passes": 1.0,
      "step": 106910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5422592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.246e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9926,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8051733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.93745253793929,
      "rewrite6_current_pass_fraction": 0.47617865280130867,
      "rewrite6_full_passes": 1.0,
      "step": 106920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5447168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2465e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0081,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.80548,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.248890460958975,
      "rewrite6_current_pass_fraction": 0.47673365659870304,
      "rewrite6_full_passes": 1.0,
      "step": 106930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5471744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.247e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0292,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8056933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.680681424666606,
      "rewrite6_current_pass_fraction": 0.47728866039609746,
      "rewrite6_full_passes": 1.0,
      "step": 106940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.549632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2475e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9857,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8061333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.800357636760168,
      "rewrite6_current_pass_fraction": 0.47786313801094427,
      "rewrite6_full_passes": 1.0,
      "step": 106950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5520896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.248e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9989,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8064933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.063454979867583,
      "rewrite6_current_pass_fraction": 0.47837919417343383,
      "rewrite6_full_passes": 1.0,
      "step": 106960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5545472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2485e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9857,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.80672,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.800357636760168,
      "rewrite6_current_pass_fraction": 0.47900235633191174,
      "rewrite6_full_passes": 1.0,
      "step": 106970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5570048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.249e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9872,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8069733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.83008045975953,
      "rewrite6_current_pass_fraction": 0.47948433331385953,
      "rewrite6_full_passes": 1.0,
      "step": 106980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5594624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2495e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0147,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.80732,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.382975130684393,
      "rewrite6_current_pass_fraction": 0.48005394247434324,
      "rewrite6_full_passes": 1.0,
      "step": 106990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.56192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.25e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.56435775756836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0042,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8076666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.170073580977565,
      "rewrite6_current_pass_fraction": 0.48056999863683275,
      "rewrite6_full_passes": 1.0,
      "step": 107000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.25e-05,
      "eval_loss": 3.0823168754577637,
      "eval_perplexity": 21.808872348429293,
      "eval_runtime": 0.0509,
      "eval_samples_per_second": 628.128,
      "eval_steps_per_second": 19.629,
      "step": 107000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5643776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2505e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.564388275146484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0106,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.80804,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.29957601765852,
      "rewrite6_current_pass_fraction": 0.4811201339798641,
      "rewrite6_full_passes": 1.0,
      "step": 107010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5668352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.251e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.564388275146484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0166,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.80832,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.421739598015016,
      "rewrite6_current_pass_fraction": 0.4816361901423536,
      "rewrite6_full_passes": 1.0,
      "step": 107020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5692928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2515e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.564388275146484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0189,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8086,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.468763656027395,
      "rewrite6_current_pass_fraction": 0.48222040466592664,
      "rewrite6_full_passes": 1.0,
      "step": 107030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5717504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.252e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.564388275146484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9918,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8088133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.921508954192763,
      "rewrite6_current_pass_fraction": 0.4827851453720473,
      "rewrite6_full_passes": 1.0,
      "step": 107040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.574208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2525e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.564388275146484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.80916,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.180161139447517,
      "rewrite6_current_pass_fraction": 0.48337422834998345,
      "rewrite6_full_passes": 1.0,
      "step": 107050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5766656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.253e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.564388275146484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9762,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8094,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.61314490766775,
      "rewrite6_current_pass_fraction": 0.48394383751046716,
      "rewrite6_full_passes": 1.0,
      "step": 107060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5791232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2535e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.564388275146484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9879,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8096266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.843966375584888,
      "rewrite6_current_pass_fraction": 0.4844939728534985,
      "rewrite6_full_passes": 1.0,
      "step": 107070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5815808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.254e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.564388275146484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0261,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8099466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.61667058032089,
      "rewrite6_current_pass_fraction": 0.48509766119452397,
      "rewrite6_full_passes": 1.0,
      "step": 107080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5840384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2545e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 15.564388275146484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0019,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8102266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.1237357207079,
      "rewrite6_current_pass_fraction": 0.48570621798991254,
      "rewrite6_full_passes": 1.0,
      "step": 107090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.586496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.255e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.081016540527344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.01,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8104933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.287399925240926,
      "rewrite6_current_pass_fraction": 0.48624174796985453,
      "rewrite6_full_passes": 1.0,
      "step": 107100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5889536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2555e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.081550598144531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0151,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8107866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.391129951592116,
      "rewrite6_current_pass_fraction": 0.4868746470370587,
      "rewrite6_full_passes": 1.0,
      "step": 107110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5914112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.256e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.081554412841797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9773,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81116,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.634731237370904,
      "rewrite6_current_pass_fraction": 0.48741991392572687,
      "rewrite6_full_passes": 1.0,
      "step": 107120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5938688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2565e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.081554412841797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9726,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81148,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.542664526804494,
      "rewrite6_current_pass_fraction": 0.4879457069969426,
      "rewrite6_full_passes": 1.0,
      "step": 107130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.5963264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.257e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.08187484741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9994,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8118133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.073489215707433,
      "rewrite6_current_pass_fraction": 0.48851044770306323,
      "rewrite6_full_passes": 1.0,
      "step": 107140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.598784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2575e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.08187484741211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9993,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8120933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.071481967149957,
      "rewrite6_current_pass_fraction": 0.4892212420400771,
      "rewrite6_full_passes": 1.0,
      "step": 107150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6012416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.258e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 14.082786560058594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0106,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81232,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.29957601765852,
      "rewrite6_current_pass_fraction": 0.4898151934723764,
      "rewrite6_full_passes": 1.0,
      "step": 107160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6036992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2585e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.084098815917969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0086,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8126,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.25901743772266,
      "rewrite6_current_pass_fraction": 0.49037019726977077,
      "rewrite6_full_passes": 1.0,
      "step": 107170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6061568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.259e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 14.085193634033203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9895,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8128133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.875742135615024,
      "rewrite6_current_pass_fraction": 0.4908619111604448,
      "rewrite6_full_passes": 1.0,
      "step": 107180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6086144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2595e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.08575439453125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9926,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8130266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.93745253793929,
      "rewrite6_current_pass_fraction": 0.4913779673229343,
      "rewrite6_full_passes": 1.0,
      "step": 107190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.611072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.26e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 14.085758209228516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0007,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81328,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.09960172113887,
      "rewrite6_current_pass_fraction": 0.49190376039415007,
      "rewrite6_full_passes": 1.0,
      "step": 107200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6135296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2605e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.085758209228516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9951,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8137333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.987358525776234,
      "rewrite6_current_pass_fraction": 0.4924685011002707,
      "rewrite6_full_passes": 1.0,
      "step": 107210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6159872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.261e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.085941314697266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8139733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.174107999122125,
      "rewrite6_current_pass_fraction": 0.49311600553056417,
      "rewrite6_full_passes": 1.0,
      "step": 107220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6184448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2615e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.086715698242188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0224,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81432,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.540529846395334,
      "rewrite6_current_pass_fraction": 0.4936564039648692,
      "rewrite6_full_passes": 1.0,
      "step": 107230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6209024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.262e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 14.086715698242188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9977,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8145733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.039393273802787,
      "rewrite6_current_pass_fraction": 0.4942162762166267,
      "rewrite6_full_passes": 1.0,
      "step": 107240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.62336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2625e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.086715698242188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0112,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81492,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.311759417923692,
      "rewrite6_current_pass_fraction": 0.49477614846838425,
      "rewrite6_full_passes": 1.0,
      "step": 107250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6258176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.263e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.087474822998047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.004,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81528,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.16603996963595,
      "rewrite6_current_pass_fraction": 0.4952775992677845,
      "rewrite6_full_passes": 1.0,
      "step": 107260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6282752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2635e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.087810516357422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0028,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8156933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.14185523541509,
      "rewrite6_current_pass_fraction": 0.4957887869759109,
      "rewrite6_full_passes": 1.0,
      "step": 107270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6307328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.264e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.087821960449219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0148,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.816,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.38501353011574,
      "rewrite6_current_pass_fraction": 0.4962999746840373,
      "rewrite6_full_passes": 1.0,
      "step": 107280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6331904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2645e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.087821960449219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0026,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8162,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.137827267178256,
      "rewrite6_current_pass_fraction": 0.49683063620961615,
      "rewrite6_full_passes": 1.0,
      "step": 107290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.635648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.265e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.08782958984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9935,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8165333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.955404322314664,
      "rewrite6_current_pass_fraction": 0.49737103464392124,
      "rewrite6_full_passes": 1.0,
      "step": 107300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6381056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2655e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.08782958984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9786,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8168266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.660272986519303,
      "rewrite6_current_pass_fraction": 0.4978676169889584,
      "rewrite6_full_passes": 1.0,
      "step": 107310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6405632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.266e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.087936401367188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9873,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81716,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.83206356695921,
      "rewrite6_current_pass_fraction": 0.49852972678234114,
      "rewrite6_full_passes": 1.0,
      "step": 107320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6430208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2665e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.087936401367188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0197,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8175333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.4851452187036,
      "rewrite6_current_pass_fraction": 0.4991139413059142,
      "rewrite6_full_passes": 1.0,
      "step": 107330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6454784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.267e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.087936401367188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9911,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8178533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.907568777555873,
      "rewrite6_current_pass_fraction": 0.49969815582948723,
      "rewrite6_full_passes": 1.0,
      "step": 107340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.647936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2675e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 14.087936401367188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0109,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8182,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.30566680403609,
      "rewrite6_current_pass_fraction": 0.5001703959027088,
      "rewrite6_full_passes": 1.0,
      "step": 107350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6503936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.268e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 14.087936401367188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0074,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8184533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.234721197457105,
      "rewrite6_current_pass_fraction": 0.5006621097933828,
      "rewrite6_full_passes": 1.0,
      "step": 107360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6528512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2685e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 14.087936401367188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.006,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8188133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.20641240855666,
      "rewrite6_current_pass_fraction": 0.5011684290471461,
      "rewrite6_full_passes": 1.0,
      "step": 107370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6553088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.269e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.087940216064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0178,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81908,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.446260395068368,
      "rewrite6_current_pass_fraction": 0.501791591205624,
      "rewrite6_full_passes": 1.0,
      "step": 107380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6577664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2695e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 14.087940216064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9793,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81932,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.674039995500852,
      "rewrite6_current_pass_fraction": 0.5024244902728282,
      "rewrite6_full_passes": 1.0,
      "step": 107390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.660224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.27e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.087940216064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.029,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81956,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.67654570196773,
      "rewrite6_current_pass_fraction": 0.5029600202527702,
      "rewrite6_full_passes": 1.0,
      "step": 107400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6626816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2705e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.087940216064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0079,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.81992,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.244841087817587,
      "rewrite6_current_pass_fraction": 0.5036610776810578,
      "rewrite6_full_passes": 1.0,
      "step": 107410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6651392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.271e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 14.087940216064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9828,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8202666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.74301977969059,
      "rewrite6_current_pass_fraction": 0.5042891082938988,
      "rewrite6_full_passes": 1.0,
      "step": 107420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6675968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2715e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.087940216064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9943,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8205733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.971375033205096,
      "rewrite6_current_pass_fraction": 0.5048635859087457,
      "rewrite6_full_passes": 1.0,
      "step": 107430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6700544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.272e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.087940216064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9755,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8208266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.59942051033186,
      "rewrite6_current_pass_fraction": 0.5054234581605032,
      "rewrite6_full_passes": 1.0,
      "step": 107440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.672512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2725e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.087940216064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9781,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8212133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.650445307150626,
      "rewrite6_current_pass_fraction": 0.5059930673209869,
      "rewrite6_full_passes": 1.0,
      "step": 107450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6749696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.273e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 13.283252716064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9847,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8214533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.780567176002993,
      "rewrite6_current_pass_fraction": 0.5065042550291133,
      "rewrite6_full_passes": 1.0,
      "step": 107460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6774272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2735e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 13.283252716064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0127,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8218933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.342249919209575,
      "rewrite6_current_pass_fraction": 0.5070300481003291,
      "rewrite6_full_passes": 1.0,
      "step": 107470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6798848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.274e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 13.283252716064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9915,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8221733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.915533397884765,
      "rewrite6_current_pass_fraction": 0.507565578080271,
      "rewrite6_full_passes": 1.0,
      "step": 107480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6823424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2745e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 13.526248931884766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9697,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.82252,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.486072897200984,
      "rewrite6_current_pass_fraction": 0.508101108060213,
      "rewrite6_full_passes": 1.0,
      "step": 107490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.275e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.678150177001953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.992,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8228133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.925493654440345,
      "rewrite6_current_pass_fraction": 0.5086707172206968,
      "rewrite6_full_passes": 1.0,
      "step": 107500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.275e-05,
      "eval_loss": 3.0827438831329346,
      "eval_perplexity": 21.818186892858346,
      "eval_runtime": 0.05,
      "eval_samples_per_second": 639.857,
      "eval_steps_per_second": 19.996,
      "step": 107500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6872576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2755e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.216880798339844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9992,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8229866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.06947491930731,
      "rewrite6_current_pass_fraction": 0.5092451948355436,
      "rewrite6_full_passes": 1.0,
      "step": 107510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6897152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.276e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.216880798339844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.995,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8233066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.985359889857122,
      "rewrite6_current_pass_fraction": 0.5099219099920157,
      "rewrite6_full_passes": 1.0,
      "step": 107520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6921728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2765e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.217819213867188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0153,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.82364,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.395208585432226,
      "rewrite6_current_pass_fraction": 0.5104769137894102,
      "rewrite6_full_passes": 1.0,
      "step": 107530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6946304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.277e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0246,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8238933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.58576875661228,
      "rewrite6_current_pass_fraction": 0.5109588907713579,
      "rewrite6_full_passes": 1.0,
      "step": 107540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.697088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2775e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.82424,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.174107999122125,
      "rewrite6_current_pass_fraction": 0.511455473116395,
      "rewrite6_full_passes": 1.0,
      "step": 107550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.6995456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.278e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0136,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.82464,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.360566185220215,
      "rewrite6_current_pass_fraction": 0.5120542930030574,
      "rewrite6_full_passes": 1.0,
      "step": 107560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7020032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2785e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0097,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8249866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.281314618105068,
      "rewrite6_current_pass_fraction": 0.5126433759809935,
      "rewrite6_full_passes": 1.0,
      "step": 107570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7044608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.279e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0039,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8253066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.16402346646582,
      "rewrite6_current_pass_fraction": 0.5131837744152986,
      "rewrite6_full_passes": 1.0,
      "step": 107580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7069184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2795e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0212,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8255866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.515895993847245,
      "rewrite6_current_pass_fraction": 0.5137533835757824,
      "rewrite6_full_passes": 1.0,
      "step": 107590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.709376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.28e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0135,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8259066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.358530230401133,
      "rewrite6_current_pass_fraction": 0.5143522034624447,
      "rewrite6_full_passes": 1.0,
      "step": 107600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7118336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2805e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9885,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8263066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.855876328038683,
      "rewrite6_current_pass_fraction": 0.5149023388054761,
      "rewrite6_full_passes": 1.0,
      "step": 107610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7142912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.281e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.007,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.82664,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.226628927540002,
      "rewrite6_current_pass_fraction": 0.5154719479659597,
      "rewrite6_full_passes": 1.0,
      "step": 107620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7167488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2815e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0007,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8268133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.09960172113887,
      "rewrite6_current_pass_fraction": 0.5161340577593425,
      "rewrite6_full_passes": 1.0,
      "step": 107630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7192064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.282e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0243,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8272533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.579593952252267,
      "rewrite6_current_pass_fraction": 0.5166209031956535,
      "rewrite6_full_passes": 1.0,
      "step": 107640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.721664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2825e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.975,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8275066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.589623249595995,
      "rewrite6_current_pass_fraction": 0.5171807754474109,
      "rewrite6_full_passes": 1.0,
      "step": 107650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7241216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.283e-05,
      "grad_norm": 0.53515625,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0173,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8278,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.436039820227478,
      "rewrite6_current_pass_fraction": 0.5177795953340734,
      "rewrite6_full_passes": 1.0,
      "step": 107660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7265792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2835e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9751,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8280533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.59158230987233,
      "rewrite6_current_pass_fraction": 0.518349204494557,
      "rewrite6_full_passes": 1.0,
      "step": 107670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7290368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.284e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0094,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8284,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.275231136287534,
      "rewrite6_current_pass_fraction": 0.518972366653035,
      "rewrite6_full_passes": 1.0,
      "step": 107680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7314944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2845e-05,
      "grad_norm": 0.57421875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0211,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.82876,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.513844506823926,
      "rewrite6_current_pass_fraction": 0.5194786859067984,
      "rewrite6_full_passes": 1.0,
      "step": 107690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.733952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.285e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0086,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8291866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.25901743772266,
      "rewrite6_current_pass_fraction": 0.5200385581585558,
      "rewrite6_full_passes": 1.0,
      "step": 107700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7364096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2855e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0162,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8295466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.413572535697167,
      "rewrite6_current_pass_fraction": 0.5205107982317774,
      "rewrite6_full_passes": 1.0,
      "step": 107710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7388672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.286e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0135,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8299466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.358530230401133,
      "rewrite6_current_pass_fraction": 0.5210025121224514,
      "rewrite6_full_passes": 1.0,
      "step": 107720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7413248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2865e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9992,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.83028,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.06947491930731,
      "rewrite6_current_pass_fraction": 0.5215136998305778,
      "rewrite6_full_passes": 1.0,
      "step": 107730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7437824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.287e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9892,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.83064,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.869780307293297,
      "rewrite6_current_pass_fraction": 0.5220687036279722,
      "rewrite6_full_passes": 1.0,
      "step": 107740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.74624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2875e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0178,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8308933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.446260395068368,
      "rewrite6_current_pass_fraction": 0.5225604175186461,
      "rewrite6_full_passes": 1.0,
      "step": 107750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7486976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.288e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0011,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8311066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.107643170009883,
      "rewrite6_current_pass_fraction": 0.5231543689509455,
      "rewrite6_full_passes": 1.0,
      "step": 107760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7511552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2885e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0181,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.83148,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.452395193360626,
      "rewrite6_current_pass_fraction": 0.5237093727483398,
      "rewrite6_full_passes": 1.0,
      "step": 107770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7536128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.289e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8319333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.212475241658723,
      "rewrite6_current_pass_fraction": 0.5243033241806392,
      "rewrite6_full_passes": 1.0,
      "step": 107780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7560704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2895e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9842,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8322533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.770679364573844,
      "rewrite6_current_pass_fraction": 0.5249167494303908,
      "rewrite6_full_passes": 1.0,
      "step": 107790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.758528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.29e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8325333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.64348968071354,
      "rewrite6_current_pass_fraction": 0.5255350431345056,
      "rewrite6_full_passes": 1.0,
      "step": 107800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7609856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2905e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0034,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8328266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.153943974815494,
      "rewrite6_current_pass_fraction": 0.5260997838406263,
      "rewrite6_full_passes": 1.0,
      "step": 107810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7634432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.291e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9765,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8330933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.61902973381984,
      "rewrite6_current_pass_fraction": 0.526625576911842,
      "rewrite6_full_passes": 1.0,
      "step": 107820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7659008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2915e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0132,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.83328,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.352423587374265,
      "rewrite6_current_pass_fraction": 0.527297423613951,
      "rewrite6_full_passes": 1.0,
      "step": 107830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7683584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.292e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0144,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8336933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.376861155287358,
      "rewrite6_current_pass_fraction": 0.5279449280442445,
      "rewrite6_full_passes": 1.0,
      "step": 107840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.770816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2925e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9998,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8339733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.08152021748699,
      "rewrite6_current_pass_fraction": 0.5285291425678176,
      "rewrite6_full_passes": 1.0,
      "step": 107850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7732736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.293e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9967,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8344533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.01936389688656,
      "rewrite6_current_pass_fraction": 0.5291766469981111,
      "rewrite6_full_passes": 1.0,
      "step": 107860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7757312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2935e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9864,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8347466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.814222739325643,
      "rewrite6_current_pass_fraction": 0.5296927031606006,
      "rewrite6_full_passes": 1.0,
      "step": 107870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7781888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.294e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9938,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8351466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.96139184169435,
      "rewrite6_current_pass_fraction": 0.5302963915016261,
      "rewrite6_full_passes": 1.0,
      "step": 107880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7806464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2945e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0103,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8354533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.2934870582428,
      "rewrite6_current_pass_fraction": 0.5309049482970146,
      "rewrite6_full_passes": 1.0,
      "step": 107890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.783104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.295e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0065,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.83584,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.21651814098351,
      "rewrite6_current_pass_fraction": 0.5314745574574984,
      "rewrite6_full_passes": 1.0,
      "step": 107900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7855616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2955e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9955,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8361733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.995355068388445,
      "rewrite6_current_pass_fraction": 0.5319760082568986,
      "rewrite6_full_passes": 1.0,
      "step": 107910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7880192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.296e-05,
      "grad_norm": 0.56640625,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0099,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8365066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.285371286682025,
      "rewrite6_current_pass_fraction": 0.532487195965025,
      "rewrite6_full_passes": 1.0,
      "step": 107920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7904768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2965e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8367866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.198331459890714,
      "rewrite6_current_pass_fraction": 0.5330762789429612,
      "rewrite6_full_passes": 1.0,
      "step": 107930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7929344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.297e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9898,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.83712,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.881705752753547,
      "rewrite6_current_pass_fraction": 0.5336166773772663,
      "rewrite6_full_passes": 1.0,
      "step": 107940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.795392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2975e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.218631744384766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0166,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.83728,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.421739598015016,
      "rewrite6_current_pass_fraction": 0.5342008919008393,
      "rewrite6_full_passes": 1.0,
      "step": 107950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.7978496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.298e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.541099548339844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0053,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.83764,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.192272869286782,
      "rewrite6_current_pass_fraction": 0.5348435278767697,
      "rewrite6_full_passes": 1.0,
      "step": 107960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8003072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2985e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.544120788574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0039,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8379066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.16402346646582,
      "rewrite6_current_pass_fraction": 0.5353401102218068,
      "rewrite6_full_passes": 1.0,
      "step": 107970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8027648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.299e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.544269561767578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9982,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8382533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.04941547578139,
      "rewrite6_current_pass_fraction": 0.5359827461977371,
      "rewrite6_full_passes": 1.0,
      "step": 107980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8052224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.2995e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.544269561767578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9914,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8386,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.91354194411933,
      "rewrite6_current_pass_fraction": 0.5365913029931257,
      "rewrite6_full_passes": 1.0,
      "step": 107990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.80768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.545219421386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8388266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.212475241658723,
      "rewrite6_current_pass_fraction": 0.5371803859710619,
      "rewrite6_full_passes": 1.0,
      "step": 108000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.3e-05,
      "eval_loss": 3.0819668769836426,
      "eval_perplexity": 21.801240612011,
      "eval_runtime": 0.0487,
      "eval_samples_per_second": 657.575,
      "eval_steps_per_second": 20.549,
      "step": 108000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8101376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3005e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.756141662597656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0068,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8391333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.222584006260107,
      "rewrite6_current_pass_fraction": 0.5377499951315456,
      "rewrite6_full_passes": 1.0,
      "step": 108010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8125952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.301e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 14.751689910888672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0058,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8394133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.202371530176254,
      "rewrite6_current_pass_fraction": 0.538256314385309,
      "rewrite6_full_passes": 1.0,
      "step": 108020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8150528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3015e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.751689910888672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.008,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8397866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.246865673153952,
      "rewrite6_current_pass_fraction": 0.5387723705477985,
      "rewrite6_full_passes": 1.0,
      "step": 108030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8175104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.302e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.752792358398438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0234,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.84008,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.56108064993093,
      "rewrite6_current_pass_fraction": 0.5392738213471987,
      "rewrite6_full_passes": 1.0,
      "step": 108040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.819968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3025e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.752792358398438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0017,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8403333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.11971137601164,
      "rewrite6_current_pass_fraction": 0.5398629043251348,
      "rewrite6_full_passes": 1.0,
      "step": 108050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8224256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.303e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.752792358398438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0112,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8406666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.311759417923692,
      "rewrite6_current_pass_fraction": 0.5404422503943448,
      "rewrite6_full_passes": 1.0,
      "step": 108060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8248832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3035e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.753044128417969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0125,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.84084,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.33818187604361,
      "rewrite6_current_pass_fraction": 0.5410410702810072,
      "rewrite6_full_passes": 1.0,
      "step": 108070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8273408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.304e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.753044128417969,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0123,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8412133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.334114646404924,
      "rewrite6_current_pass_fraction": 0.5416398901676696,
      "rewrite6_full_passes": 1.0,
      "step": 108080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8297984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3045e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 14.7530517578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0068,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8415333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.222584006260107,
      "rewrite6_current_pass_fraction": 0.5421705516932485,
      "rewrite6_full_passes": 1.0,
      "step": 108090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.832256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.305e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 14.7530517578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0228,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8418533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.5487477017954,
      "rewrite6_current_pass_fraction": 0.5427547662168215,
      "rewrite6_full_passes": 1.0,
      "step": 108100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8347136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3055e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 14.7530517578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0137,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8421066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.362602343644966,
      "rewrite6_current_pass_fraction": 0.5433681914665732,
      "rewrite6_full_passes": 1.0,
      "step": 108110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8371712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.306e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.753105163574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0116,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8424,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.319885746848296,
      "rewrite6_current_pass_fraction": 0.5438842476290627,
      "rewrite6_full_passes": 1.0,
      "step": 108120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8396288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3065e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.753105163574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9979,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8427333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.043401553272133,
      "rewrite6_current_pass_fraction": 0.5444197776090047,
      "rewrite6_full_passes": 1.0,
      "step": 108130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8420864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.307e-05,
      "grad_norm": 0.53125,
      "host_rss_gib": 14.753105163574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0332,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8430133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.763569816631378,
      "rewrite6_current_pass_fraction": 0.5449796498607622,
      "rewrite6_full_passes": 1.0,
      "step": 108140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.844544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3075e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.753105163574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0222,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8433866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.536422151209266,
      "rewrite6_current_pass_fraction": 0.5456076804736032,
      "rewrite6_full_passes": 1.0,
      "step": 108150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8470016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.308e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.753105163574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8437066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.631107301990202,
      "rewrite6_current_pass_fraction": 0.5461480789079083,
      "rewrite6_full_passes": 1.0,
      "step": 108160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8494592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3085e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.753105163574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.016,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8441866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.409490229434265,
      "rewrite6_current_pass_fraction": 0.5467566357032969,
      "rewrite6_full_passes": 1.0,
      "step": 108170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8519168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.309e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.753105163574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9962,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8445733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.009356716941582,
      "rewrite6_current_pass_fraction": 0.5472824287745127,
      "rewrite6_full_passes": 1.0,
      "step": 108180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8543744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3095e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.753105163574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0009,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8450266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.103622043501936,
      "rewrite6_current_pass_fraction": 0.5479445385678955,
      "rewrite6_full_passes": 1.0,
      "step": 108190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.856832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.31e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 14.753105163574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9958,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8453066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.001354574789932,
      "rewrite6_current_pass_fraction": 0.5485190161827423,
      "rewrite6_full_passes": 1.0,
      "step": 108200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8592896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3105e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.753105163574219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8455866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.782265441235463,
      "rewrite6_current_pass_fraction": 0.5490983622519523,
      "rewrite6_full_passes": 1.0,
      "step": 108210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8617472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.311e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.753192901611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0223,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8459866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.538475896109915,
      "rewrite6_current_pass_fraction": 0.5497312613191564,
      "rewrite6_full_passes": 1.0,
      "step": 108220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8642048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3115e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 14.753192901611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9858,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8462933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.80233777152893,
      "rewrite6_current_pass_fraction": 0.5503008704796402,
      "rewrite6_full_passes": 1.0,
      "step": 108230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8666624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.312e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.753192901611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0251,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8465333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.596064214640606,
      "rewrite6_current_pass_fraction": 0.5508120581877666,
      "rewrite6_full_passes": 1.0,
      "step": 108240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.86912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3125e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.753192901611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0333,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8469333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.765646277434353,
      "rewrite6_current_pass_fraction": 0.551323245895893,
      "rewrite6_full_passes": 1.0,
      "step": 108250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8715776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.313e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.753192901611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8473466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.23269782650759,
      "rewrite6_current_pass_fraction": 0.5519269342369184,
      "rewrite6_full_passes": 1.0,
      "step": 108260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8740352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3135e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 14.753192901611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0039,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8476533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.16402346646582,
      "rewrite6_current_pass_fraction": 0.5525890440303013,
      "rewrite6_full_passes": 1.0,
      "step": 108270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8764928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.314e-05,
      "grad_norm": 0.55859375,
      "host_rss_gib": 14.753192901611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0182,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8478933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.45444053514535,
      "rewrite6_current_pass_fraction": 0.5531586531907849,
      "rewrite6_full_passes": 1.0,
      "step": 108280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8789504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3145e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.753192901611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0499,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8482266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.113232993671563,
      "rewrite6_current_pass_fraction": 0.5537136569881794,
      "rewrite6_full_passes": 1.0,
      "step": 108290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.881408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.315e-05,
      "grad_norm": 0.53125,
      "host_rss_gib": 14.753192901611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0089,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8484666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.26509605470094,
      "rewrite6_current_pass_fraction": 0.5541810286070379,
      "rewrite6_full_passes": 1.0,
      "step": 108300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8838656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3155e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.753192901611328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8487466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.23269782650759,
      "rewrite6_current_pass_fraction": 0.5548187961286051,
      "rewrite6_full_passes": 1.0,
      "step": 108310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8863232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.316e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 14.754642486572266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9936,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8490666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.957399962527237,
      "rewrite6_current_pass_fraction": 0.5552910362018266,
      "rewrite6_full_passes": 1.0,
      "step": 108320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8887808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3165e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.754642486572266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9831,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8493466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.748943574149234,
      "rewrite6_current_pass_fraction": 0.5558363030904948,
      "rewrite6_full_passes": 1.0,
      "step": 108330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8912384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.317e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.6954345703125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0184,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8496266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.458531832368465,
      "rewrite6_current_pass_fraction": 0.5563182800724426,
      "rewrite6_full_passes": 1.0,
      "step": 108340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.893696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3175e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.753223419189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0107,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8499066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.301606076761544,
      "rewrite6_current_pass_fraction": 0.5569268368678312,
      "rewrite6_full_passes": 1.0,
      "step": 108350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8961536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.318e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.824851989746094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0194,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.85016,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.479000596877352,
      "rewrite6_current_pass_fraction": 0.5576035520243033,
      "rewrite6_full_passes": 1.0,
      "step": 108360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.8986112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3185e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.824851989746094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0192,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8504666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.474905206310687,
      "rewrite6_current_pass_fraction": 0.5582023719109657,
      "rewrite6_full_passes": 1.0,
      "step": 108370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9010688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.319e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.824851989746094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.019,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8507733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.470810634740232,
      "rewrite6_current_pass_fraction": 0.5587768495258125,
      "rewrite6_full_passes": 1.0,
      "step": 108380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9035264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3195e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.824867248535156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0143,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8511066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.374823571052737,
      "rewrite6_current_pass_fraction": 0.5593269848688438,
      "rewrite6_full_passes": 1.0,
      "step": 108390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.905984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.32e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.824867248535156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0143,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8514933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.374823571052737,
      "rewrite6_current_pass_fraction": 0.5599209363011431,
      "rewrite6_full_passes": 1.0,
      "step": 108400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9084416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3205e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.824867248535156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8517466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.738668476643575,
      "rewrite6_current_pass_fraction": 0.5604223871005434,
      "rewrite6_full_passes": 1.0,
      "step": 108410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9108992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.321e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.824867248535156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8520266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.5610066016241164,
      "rewrite6_full_passes": 1.0,
      "step": 108420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9133568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3215e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.824867248535156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0088,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8523066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.263069646417566,
      "rewrite6_current_pass_fraction": 0.5616346322369574,
      "rewrite6_full_passes": 1.0,
      "step": 108430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9158144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.322e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.824867248535156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8525733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.78642230999673,
      "rewrite6_current_pass_fraction": 0.5621604253081731,
      "rewrite6_full_passes": 1.0,
      "step": 108440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.918272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3225e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.824867248535156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9871,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.85288,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.828097550860647,
      "rewrite6_current_pass_fraction": 0.5627300344686569,
      "rewrite6_full_passes": 1.0,
      "step": 108450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9207296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.323e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.824867248535156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0139,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8532666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.366675271392893,
      "rewrite6_current_pass_fraction": 0.5633093805378668,
      "rewrite6_full_passes": 1.0,
      "step": 108460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9231872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3235e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.824867248535156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0191,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8534666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.472857818161167,
      "rewrite6_current_pass_fraction": 0.5639958326030652,
      "rewrite6_full_passes": 1.0,
      "step": 108470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9256448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.324e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.825202941894531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0079,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8537333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.244841087817587,
      "rewrite6_current_pass_fraction": 0.5646092578528169,
      "rewrite6_full_passes": 1.0,
      "step": 108480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9281024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3245e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.825202941894531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0114,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.85416,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.31582217606955,
      "rewrite6_current_pass_fraction": 0.5651691301045744,
      "rewrite6_full_passes": 1.0,
      "step": 108490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.93056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.325e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.82659912109375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0168,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8544666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.425824354396642,
      "rewrite6_current_pass_fraction": 0.5657922922630523,
      "rewrite6_full_passes": 1.0,
      "step": 108500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.325e-05,
      "eval_loss": 3.0818240642547607,
      "eval_perplexity": 21.79812733965894,
      "eval_runtime": 0.0491,
      "eval_samples_per_second": 651.783,
      "eval_steps_per_second": 20.368,
      "step": 108500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9330176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3255e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0111,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.85484,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.30972834353731,
      "rewrite6_current_pass_fraction": 0.566405717512804,
      "rewrite6_full_passes": 1.0,
      "step": 108510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9354752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.326e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9912,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8551466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.909559633974794,
      "rewrite6_current_pass_fraction": 0.5669558528558353,
      "rewrite6_full_passes": 1.0,
      "step": 108520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9379328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3265e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.03,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8555066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.6972325893895,
      "rewrite6_current_pass_fraction": 0.5675790150143133,
      "rewrite6_full_passes": 1.0,
      "step": 108530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9403904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.327e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9996,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8558133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.077504315047122,
      "rewrite6_current_pass_fraction": 0.5682411248076961,
      "rewrite6_full_passes": 1.0,
      "step": 108540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.942848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3275e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0081,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.85612,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.248890460958975,
      "rewrite6_current_pass_fraction": 0.5688009970594535,
      "rewrite6_full_passes": 1.0,
      "step": 108550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9453056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.328e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0101,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8565466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.289428766673836,
      "rewrite6_current_pass_fraction": 0.5693949484917529,
      "rewrite6_full_passes": 1.0,
      "step": 108560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9477632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3285e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0146,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8567866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.380936935082808,
      "rewrite6_current_pass_fraction": 0.5699402153804211,
      "rewrite6_full_passes": 1.0,
      "step": 108570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9502208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.329e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0277,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.85704,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.649683656667687,
      "rewrite6_current_pass_fraction": 0.5704173239080057,
      "rewrite6_full_passes": 1.0,
      "step": 108580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9526784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3295e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0167,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8572666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.423781874086924,
      "rewrite6_current_pass_fraction": 0.5709431169792214,
      "rewrite6_full_passes": 1.0,
      "step": 108590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.955136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.33e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0258,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8576666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.610486506804197,
      "rewrite6_current_pass_fraction": 0.5714737785048003,
      "rewrite6_full_passes": 1.0,
      "step": 108600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9575936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3305e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.992,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8579466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.925493654440345,
      "rewrite6_current_pass_fraction": 0.5720336507565578,
      "rewrite6_full_passes": 1.0,
      "step": 108610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9600512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.331e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9927,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8583066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.939446382883673,
      "rewrite6_current_pass_fraction": 0.5725837860995892,
      "rewrite6_full_passes": 1.0,
      "step": 108620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9625088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3315e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9761,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8586133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.611183691239443,
      "rewrite6_current_pass_fraction": 0.5732118167124302,
      "rewrite6_full_passes": 1.0,
      "step": 108630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9649664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.332e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0133,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.85896,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.354458931498517,
      "rewrite6_current_pass_fraction": 0.5737522151467352,
      "rewrite6_full_passes": 1.0,
      "step": 108640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.967424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3325e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.855175018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0102,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8591866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.29145781100103,
      "rewrite6_current_pass_fraction": 0.5743705088508501,
      "rewrite6_full_passes": 1.0,
      "step": 108650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9698816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.333e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0107,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.85948,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.301606076761544,
      "rewrite6_current_pass_fraction": 0.5750034079180542,
      "rewrite6_full_passes": 1.0,
      "step": 108660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9723392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3335e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0072,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.85988,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.23067465788506,
      "rewrite6_current_pass_fraction": 0.5755827539872641,
      "rewrite6_full_passes": 1.0,
      "step": 108670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9747968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.334e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0039,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8602133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.16402346646582,
      "rewrite6_current_pass_fraction": 0.5761621000564741,
      "rewrite6_full_passes": 1.0,
      "step": 108680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9772544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3345e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0113,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86044,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.313790695427663,
      "rewrite6_current_pass_fraction": 0.5767171038538684,
      "rewrite6_full_passes": 1.0,
      "step": 108690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.979712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.335e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.011,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8609333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.307697472248215,
      "rewrite6_current_pass_fraction": 0.5772526338338104,
      "rewrite6_full_passes": 1.0,
      "step": 108700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9821696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3355e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0058,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8612266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.202371530176254,
      "rewrite6_current_pass_fraction": 0.5778173745399311,
      "rewrite6_full_passes": 1.0,
      "step": 108710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9846272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.336e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0194,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86152,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.479000596877352,
      "rewrite6_current_pass_fraction": 0.5784064575178672,
      "rewrite6_full_passes": 1.0,
      "step": 108720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9870848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3365e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8618933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.54258400208604,
      "rewrite6_current_pass_fraction": 0.5789079083172675,
      "rewrite6_full_passes": 1.0,
      "step": 108730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9895424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.337e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9921,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86224,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.927486303436584,
      "rewrite6_current_pass_fraction": 0.5795797550193764,
      "rewrite6_full_passes": 1.0,
      "step": 108740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3375e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9711,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86252,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.51337250452325,
      "rewrite6_current_pass_fraction": 0.5801444957254971,
      "rewrite6_full_passes": 1.0,
      "step": 108750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9944576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.338e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0145,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86284,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.378898943290586,
      "rewrite6_current_pass_fraction": 0.5806556834336235,
      "rewrite6_full_passes": 1.0,
      "step": 108760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9969152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3385e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0218,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8631333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.528209225043515,
      "rewrite6_current_pass_fraction": 0.5812252925941073,
      "rewrite6_full_passes": 1.0,
      "step": 108770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 20.9993728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.339e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0085,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86344,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.256991637270605,
      "rewrite6_current_pass_fraction": 0.5817559541196861,
      "rewrite6_full_passes": 1.0,
      "step": 108780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0018304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3395e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9979,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8637333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.043401553272133,
      "rewrite6_current_pass_fraction": 0.5822768787365388,
      "rewrite6_full_passes": 1.0,
      "step": 108790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.004288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.34e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9935,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86404,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.955404322314664,
      "rewrite6_current_pass_fraction": 0.5828562248057487,
      "rewrite6_full_passes": 1.0,
      "step": 108800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0067456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3405e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0079,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8643466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.244841087817587,
      "rewrite6_current_pass_fraction": 0.5834404393293218,
      "rewrite6_full_passes": 1.0,
      "step": 108810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0092032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.341e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.855567932128906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9995,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86472,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.07549666499979,
      "rewrite6_current_pass_fraction": 0.5840489961247103,
      "rewrite6_full_passes": 1.0,
      "step": 108820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0116608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3415e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.970909118652344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0208,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8650133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.507691276502573,
      "rewrite6_current_pass_fraction": 0.5845455784697474,
      "rewrite6_full_passes": 1.0,
      "step": 108830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0141184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.342e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 14.975692749023438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.003,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8652666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.14588400932613,
      "rewrite6_current_pass_fraction": 0.5850129500886059,
      "rewrite6_full_passes": 1.0,
      "step": 108840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.016576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3425e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.976112365722656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0288,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8655866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.672410806330685,
      "rewrite6_current_pass_fraction": 0.5855192693423692,
      "rewrite6_full_passes": 1.0,
      "step": 108850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0190336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.343e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.976116180419922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0239,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8658666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.57136376081938,
      "rewrite6_current_pass_fraction": 0.5860791415941267,
      "rewrite6_full_passes": 1.0,
      "step": 108860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0214912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3435e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.976116180419922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.023,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86608,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.55285786233811,
      "rewrite6_current_pass_fraction": 0.5866244084827948,
      "rewrite6_full_passes": 1.0,
      "step": 108870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0239488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.344e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.977169036865234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0016,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86644,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.11769950546924,
      "rewrite6_current_pass_fraction": 0.5871453330996476,
      "rewrite6_full_passes": 1.0,
      "step": 108880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0264064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3445e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 14.977169036865234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8668,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.198331459890714,
      "rewrite6_current_pass_fraction": 0.5877344160775837,
      "rewrite6_full_passes": 1.0,
      "step": 108890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.028864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.345e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.978328704833984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0041,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8672533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.168056674466477,
      "rewrite6_current_pass_fraction": 0.5883234990555198,
      "rewrite6_full_passes": 1.0,
      "step": 108900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0313216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3455e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.978519439697266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0265,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8675866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.624918898106596,
      "rewrite6_current_pass_fraction": 0.5888444236723724,
      "rewrite6_full_passes": 1.0,
      "step": 108910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0337792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.346e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.978519439697266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9841,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8678133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.76870239548749,
      "rewrite6_current_pass_fraction": 0.5893410060174096,
      "rewrite6_full_passes": 1.0,
      "step": 108920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0362368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3465e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.97921371459961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0123,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8680533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.334114646404924,
      "rewrite6_current_pass_fraction": 0.5899592997215244,
      "rewrite6_full_passes": 1.0,
      "step": 108930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0386944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.347e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.97921371459961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0082,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8684133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.250915451252894,
      "rewrite6_current_pass_fraction": 0.5905629880625499,
      "rewrite6_full_passes": 1.0,
      "step": 108940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.041152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3475e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.97921371459961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8688,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86347453124931,
      "rewrite6_current_pass_fraction": 0.5912348347646589,
      "rewrite6_full_passes": 1.0,
      "step": 108950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0436096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.348e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.97921371459961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0025,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8691866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.135813585137313,
      "rewrite6_current_pass_fraction": 0.5918287861969582,
      "rewrite6_full_passes": 1.0,
      "step": 108960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0460672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3485e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.97921371459961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0107,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.86956,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.301606076761544,
      "rewrite6_current_pass_fraction": 0.5923789215399895,
      "rewrite6_full_passes": 1.0,
      "step": 108970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0485248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.349e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.97921371459961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0187,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8698533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.46467031264417,
      "rewrite6_current_pass_fraction": 0.5929923467897412,
      "rewrite6_full_passes": 1.0,
      "step": 108980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0509824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3495e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.97921371459961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8702,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.23269782650759,
      "rewrite6_current_pass_fraction": 0.5936495881287609,
      "rewrite6_full_passes": 1.0,
      "step": 108990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.05344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.35e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.98116683959961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8704933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.01003120287951,
      "rewrite6_current_pass_fraction": 0.5942045919261553,
      "rewrite6_full_passes": 1.0,
      "step": 109000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.35e-05,
      "eval_loss": 3.08249568939209,
      "eval_perplexity": 21.812772427380974,
      "eval_runtime": 0.0492,
      "eval_samples_per_second": 650.283,
      "eval_steps_per_second": 20.321,
      "step": 109000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0558976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3505e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9965,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8708,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.01536042446777,
      "rewrite6_current_pass_fraction": 0.5947449903604604,
      "rewrite6_full_passes": 1.0,
      "step": 109010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0583552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.351e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0031,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8711466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.147898698459837,
      "rewrite6_current_pass_fraction": 0.5952707834316762,
      "rewrite6_full_passes": 1.0,
      "step": 109020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0608128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3515e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0084,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8714,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.254966039388457,
      "rewrite6_current_pass_fraction": 0.5958209187747074,
      "rewrite6_full_passes": 1.0,
      "step": 109030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0632704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.352e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9928,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8717733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.941440427222513,
      "rewrite6_current_pass_fraction": 0.5963759225721018,
      "rewrite6_full_passes": 1.0,
      "step": 109040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.065728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3525e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9967,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.87216,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.01936389688656,
      "rewrite6_current_pass_fraction": 0.5969260579151331,
      "rewrite6_full_passes": 1.0,
      "step": 109050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0681856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.353e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0178,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8725066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.446260395068368,
      "rewrite6_current_pass_fraction": 0.5976027730716053,
      "rewrite6_full_passes": 1.0,
      "step": 109060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0706432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3535e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0038,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.87284,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.162007164935936,
      "rewrite6_current_pass_fraction": 0.5982113298669939,
      "rewrite6_full_passes": 1.0,
      "step": 109070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0731008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.354e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0158,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.87312,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.40540873955097,
      "rewrite6_current_pass_fraction": 0.5987955443905669,
      "rewrite6_full_passes": 1.0,
      "step": 109080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0755584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3545e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0041,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8734533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.168056674466477,
      "rewrite6_current_pass_fraction": 0.5994041011859554,
      "rewrite6_full_passes": 1.0,
      "step": 109090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.078016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.355e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0154,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8738266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.397248208270213,
      "rewrite6_current_pass_fraction": 0.5999883157095285,
      "rewrite6_full_passes": 1.0,
      "step": 109100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0804736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3555e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0035,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.87412,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.15595946998605,
      "rewrite6_current_pass_fraction": 0.6005822671418278,
      "rewrite6_full_passes": 1.0,
      "step": 109110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0829312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.356e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.035453796386719,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9903,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.87444,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.8916490912574,
      "rewrite6_current_pass_fraction": 0.6010155595801445,
      "rewrite6_full_passes": 1.0,
      "step": 109120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0853888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3565e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.035457611083984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0468,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8745866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.04788331572619,
      "rewrite6_current_pass_fraction": 0.601531615742634,
      "rewrite6_full_passes": 1.0,
      "step": 109130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0878464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.357e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.035457611083984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.028,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8748666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.655879491093387,
      "rewrite6_current_pass_fraction": 0.6020963564487547,
      "rewrite6_full_passes": 1.0,
      "step": 109140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.090304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3575e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.035457611083984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.029,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8752666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.67654570196773,
      "rewrite6_current_pass_fraction": 0.6026513602461491,
      "rewrite6_full_passes": 1.0,
      "step": 109150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0927616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.358e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.035457611083984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0075,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8755333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.236744770753823,
      "rewrite6_current_pass_fraction": 0.603274522404627,
      "rewrite6_full_passes": 1.0,
      "step": 109160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0952192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3585e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.035457611083984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0235,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8758933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.563136860804747,
      "rewrite6_current_pass_fraction": 0.6038587369282,
      "rewrite6_full_passes": 1.0,
      "step": 109170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.0976768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.359e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.035457611083984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0148,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8762133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.38501353011574,
      "rewrite6_current_pass_fraction": 0.604394266908142,
      "rewrite6_full_passes": 1.0,
      "step": 109180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1001344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3595e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.542251586914062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9919,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8765733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.923501204699043,
      "rewrite6_current_pass_fraction": 0.6049979552491674,
      "rewrite6_full_passes": 1.0,
      "step": 109190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.102592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.36e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.602981567382812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9974,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.87688,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.033382357503168,
      "rewrite6_current_pass_fraction": 0.6055432221378357,
      "rewrite6_full_passes": 1.0,
      "step": 109200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1050496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3605e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9999,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8771333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.08352846991968,
      "rewrite6_current_pass_fraction": 0.606185858113766,
      "rewrite6_full_passes": 1.0,
      "step": 109210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1075072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.361e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0321,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8774933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.740742447188033,
      "rewrite6_current_pass_fraction": 0.6066824404588032,
      "rewrite6_full_passes": 1.0,
      "step": 109220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1099648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3615e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0181,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.87792,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.452395193360626,
      "rewrite6_current_pass_fraction": 0.6070962590796674,
      "rewrite6_full_passes": 1.0,
      "step": 109230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1124224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.362e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9925,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8782266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.935458892369443,
      "rewrite6_current_pass_fraction": 0.607612315242157,
      "rewrite6_full_passes": 1.0,
      "step": 109240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.11488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3625e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0048,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.87856,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.182179256465627,
      "rewrite6_current_pass_fraction": 0.6081819244026406,
      "rewrite6_full_passes": 1.0,
      "step": 109250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1173376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.363e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0181,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.87884,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.452395193360626,
      "rewrite6_current_pass_fraction": 0.60877587583494,
      "rewrite6_full_passes": 1.0,
      "step": 109260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1197952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3635e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.014,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8792133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.368712040756797,
      "rewrite6_current_pass_fraction": 0.6094233802652333,
      "rewrite6_full_passes": 1.0,
      "step": 109270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1222528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.364e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.02,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.87964,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.49129168419294,
      "rewrite6_current_pass_fraction": 0.6100368055149851,
      "rewrite6_full_passes": 1.0,
      "step": 109280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1247104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3645e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8798933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.701372449879624,
      "rewrite6_current_pass_fraction": 0.610703783762731,
      "rewrite6_full_passes": 1.0,
      "step": 109290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.127168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.365e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0219,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.88024,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.53026214861049,
      "rewrite6_current_pass_fraction": 0.6112539191057623,
      "rewrite6_full_passes": 1.0,
      "step": 109300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1296256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3655e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0223,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8805733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.538475896109915,
      "rewrite6_current_pass_fraction": 0.6117991859944305,
      "rewrite6_full_passes": 1.0,
      "step": 109310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1320832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.366e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0282,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8808666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.660011080136737,
      "rewrite6_current_pass_fraction": 0.6122811629763782,
      "rewrite6_full_passes": 1.0,
      "step": 109320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1345408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3665e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0009,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.88116,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.103622043501936,
      "rewrite6_current_pass_fraction": 0.6129968257677553,
      "rewrite6_full_passes": 1.0,
      "step": 109330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1369984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.367e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0165,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8814266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.419697526160515,
      "rewrite6_current_pass_fraction": 0.613610251017507,
      "rewrite6_full_passes": 1.0,
      "step": 109340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.139456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3675e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0039,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.88172,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.16402346646582,
      "rewrite6_current_pass_fraction": 0.6141311756343596,
      "rewrite6_full_passes": 1.0,
      "step": 109350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1419136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.368e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0146,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.88212,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.380936935082808,
      "rewrite6_current_pass_fraction": 0.6147640747015638,
      "rewrite6_full_passes": 1.0,
      "step": 109360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1443712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3685e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0154,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8825466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.397248208270213,
      "rewrite6_current_pass_fraction": 0.6153969737687679,
      "rewrite6_full_passes": 1.0,
      "step": 109370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1468288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.369e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0137,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8829466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.362602343644966,
      "rewrite6_current_pass_fraction": 0.6160347412903352,
      "rewrite6_full_passes": 1.0,
      "step": 109380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1492864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3695e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9907,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.88336,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.899607342438028,
      "rewrite6_current_pass_fraction": 0.6166092189051819,
      "rewrite6_full_passes": 1.0,
      "step": 109390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.151744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.37e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0189,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8836133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.468763656027395,
      "rewrite6_current_pass_fraction": 0.6172031703374813,
      "rewrite6_full_passes": 1.0,
      "step": 109400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1542016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3705e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8838666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.212475241658723,
      "rewrite6_current_pass_fraction": 0.617728963408697,
      "rewrite6_full_passes": 1.0,
      "step": 109410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1566592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.371e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8840933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.052093313375067,
      "rewrite6_current_pass_fraction": 0.6184154154738953,
      "rewrite6_full_passes": 1.0,
      "step": 109420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1591168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3715e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0236,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8844266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.565193277309945,
      "rewrite6_current_pass_fraction": 0.6189509454538373,
      "rewrite6_full_passes": 1.0,
      "step": 109430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1615744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.372e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.02,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.88468,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.49129168419294,
      "rewrite6_current_pass_fraction": 0.6194523962532376,
      "rewrite6_full_passes": 1.0,
      "step": 109440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.164032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3725e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.011,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8849866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.307697472248215,
      "rewrite6_current_pass_fraction": 0.6199587155070009,
      "rewrite6_full_passes": 1.0,
      "step": 109450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1664896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.373e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0008,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8853466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.10161178181234,
      "rewrite6_current_pass_fraction": 0.6205575353936632,
      "rewrite6_full_passes": 1.0,
      "step": 109460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1689472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3735e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8856133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.176125510795945,
      "rewrite6_current_pass_fraction": 0.6211320130085101,
      "rewrite6_full_passes": 1.0,
      "step": 109470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1714048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.374e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8858666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.6217746489844405,
      "rewrite6_full_passes": 1.0,
      "step": 109480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1738624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3745e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0048,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8862,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.182179256465627,
      "rewrite6_current_pass_fraction": 0.6223345212361979,
      "rewrite6_full_passes": 1.0,
      "step": 109490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.17632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.375e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.0218505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.007,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.88648,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.226628927540002,
      "rewrite6_current_pass_fraction": 0.6229382095772235,
      "rewrite6_full_passes": 1.0,
      "step": 109500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.375e-05,
      "eval_loss": 3.081403970718384,
      "eval_perplexity": 21.788972010440386,
      "eval_runtime": 0.0489,
      "eval_samples_per_second": 654.794,
      "eval_steps_per_second": 20.462,
      "step": 109500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1787776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3755e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8869066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.198331459890714,
      "rewrite6_current_pass_fraction": 0.6235224241007965,
      "rewrite6_full_passes": 1.0,
      "step": 109510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1812352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.376e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8872133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.15797516671621,
      "rewrite6_current_pass_fraction": 0.6241358493505482,
      "rewrite6_full_passes": 1.0,
      "step": 109520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1836928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3765e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9731,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.88736,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.552438302308154,
      "rewrite6_current_pass_fraction": 0.6246519055130377,
      "rewrite6_full_passes": 1.0,
      "step": 109530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1861504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.377e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0208,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8878666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.507691276502573,
      "rewrite6_current_pass_fraction": 0.6252117777647952,
      "rewrite6_full_passes": 1.0,
      "step": 109540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.188608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3775e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0095,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8880133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.2772587607807,
      "rewrite6_current_pass_fraction": 0.6258203345601838,
      "rewrite6_full_passes": 1.0,
      "step": 109550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1910656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.378e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9888,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8883066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.861833984540883,
      "rewrite6_current_pass_fraction": 0.6264629705361142,
      "rewrite6_full_passes": 1.0,
      "step": 109560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1935232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3785e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0188,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8885333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.466716882002203,
      "rewrite6_current_pass_fraction": 0.6269985005160562,
      "rewrite6_full_passes": 1.0,
      "step": 109570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1959808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.379e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8887733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.6275388989503612,
      "rewrite6_full_passes": 1.0,
      "step": 109580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.1984384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3795e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9994,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.88908,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.073489215707433,
      "rewrite6_current_pass_fraction": 0.6280695604759401,
      "rewrite6_full_passes": 1.0,
      "step": 109590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.200896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.38e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8893066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.159991065026112,
      "rewrite6_current_pass_fraction": 0.6285612743666141,
      "rewrite6_full_passes": 1.0,
      "step": 109600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2033536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3805e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8896266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.23269782650759,
      "rewrite6_current_pass_fraction": 0.6291357519814609,
      "rewrite6_full_passes": 1.0,
      "step": 109610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2058112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.381e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8899866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.821789281331817,
      "rewrite6_current_pass_fraction": 0.629719966505034,
      "rewrite6_full_passes": 1.0,
      "step": 109620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2082688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3815e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0328,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8903066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.755266049568853,
      "rewrite6_current_pass_fraction": 0.6302360226675235,
      "rewrite6_full_passes": 1.0,
      "step": 109630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2107264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.382e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9832,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.89072,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.750918567254665,
      "rewrite6_current_pass_fraction": 0.6309565539132637,
      "rewrite6_full_passes": 1.0,
      "step": 109640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.213184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3825e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.008,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8910266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.246865673153952,
      "rewrite6_current_pass_fraction": 0.6314823469844794,
      "rewrite6_full_passes": 1.0,
      "step": 109650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2156416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.383e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.89136,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.6321736675040408,
      "rewrite6_full_passes": 1.0,
      "step": 109660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2180992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3835e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0076,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8917066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.238768546418,
      "rewrite6_current_pass_fraction": 0.6326945921208934,
      "rewrite6_full_passes": 1.0,
      "step": 109670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2205568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.384e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.077075958251953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0333,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8920133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.765646277434353,
      "rewrite6_current_pass_fraction": 0.6332398590095616,
      "rewrite6_full_passes": 1.0,
      "step": 109680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2230144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3845e-05,
      "grad_norm": 0.5390625,
      "host_rss_gib": 15.421859741210938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0251,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8922133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.596064214640606,
      "rewrite6_current_pass_fraction": 0.6337218359915094,
      "rewrite6_full_passes": 1.0,
      "step": 109690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.225472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.385e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.42453384399414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0273,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8925466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.641425434959473,
      "rewrite6_current_pass_fraction": 0.6343206558781718,
      "rewrite6_full_passes": 1.0,
      "step": 109700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2279296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3855e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.424571990966797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0059,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8928133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.2043918683445,
      "rewrite6_current_pass_fraction": 0.6348367120406613,
      "rewrite6_full_passes": 1.0,
      "step": 109710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2303872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.386e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.424571990966797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0093,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8931466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.273203714546685,
      "rewrite6_current_pass_fraction": 0.6355231641058596,
      "rewrite6_full_passes": 1.0,
      "step": 109720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2328448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3865e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.424571990966797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9875,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8934666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.836030376340315,
      "rewrite6_current_pass_fraction": 0.6361560631730638,
      "rewrite6_full_passes": 1.0,
      "step": 109730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2353024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.387e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.424575805664062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0125,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8937866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.33818187604361,
      "rewrite6_current_pass_fraction": 0.6368084360577204,
      "rewrite6_full_passes": 1.0,
      "step": 109740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.23776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3875e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.425914764404297,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9944,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8942266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.973372270568625,
      "rewrite6_current_pass_fraction": 0.6373975190356566,
      "rewrite6_full_passes": 1.0,
      "step": 109750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2402176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.388e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.426097869873047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0071,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.89452,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.22865169156927,
      "rewrite6_current_pass_fraction": 0.6378989698350568,
      "rewrite6_full_passes": 1.0,
      "step": 109760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2426752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3885e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.426101684570312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0213,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.89484,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.517947686029533,
      "rewrite6_current_pass_fraction": 0.6385075266304454,
      "rewrite6_full_passes": 1.0,
      "step": 109770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2451328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.389e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.426101684570312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8950266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.78642230999673,
      "rewrite6_current_pass_fraction": 0.6389651613405776,
      "rewrite6_full_passes": 1.0,
      "step": 109780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2475904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3895e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.426101684570312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0338,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8953066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.77603169671152,
      "rewrite6_current_pass_fraction": 0.6395396389554244,
      "rewrite6_full_passes": 1.0,
      "step": 109790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.250048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.39e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.427173614501953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0237,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.89556,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.567249899467065,
      "rewrite6_current_pass_fraction": 0.6400897742984557,
      "rewrite6_full_passes": 1.0,
      "step": 109800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2525056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3905e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.42733383178711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0203,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8959066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.497439993898546,
      "rewrite6_current_pass_fraction": 0.6405863566434928,
      "rewrite6_full_passes": 1.0,
      "step": 109810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2549632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.391e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.42758560180664,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0248,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.89624,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.589886322106427,
      "rewrite6_current_pass_fraction": 0.6412387295281494,
      "rewrite6_full_passes": 1.0,
      "step": 109820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2574208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3915e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.428058624267578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0156,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.89648,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.40132806588403,
      "rewrite6_current_pass_fraction": 0.6417742595080914,
      "rewrite6_full_passes": 1.0,
      "step": 109830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2598784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.392e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.428058624267578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0165,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8968933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.419697526160515,
      "rewrite6_current_pass_fraction": 0.642295184124944,
      "rewrite6_full_passes": 1.0,
      "step": 109840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.262336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3925e-05,
      "grad_norm": 0.6640625,
      "host_rss_gib": 15.428123474121094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0084,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8971333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.254966039388457,
      "rewrite6_current_pass_fraction": 0.6429183462834219,
      "rewrite6_full_passes": 1.0,
      "step": 109850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2647936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.393e-05,
      "grad_norm": 0.4296875,
      "host_rss_gib": 15.428443908691406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0139,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.89752,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.366675271392893,
      "rewrite6_current_pass_fraction": 0.6435707191680785,
      "rewrite6_full_passes": 1.0,
      "step": 109860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2672512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3935e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.428443908691406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0176,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8977333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.442171551887302,
      "rewrite6_current_pass_fraction": 0.644086775330568,
      "rewrite6_full_passes": 1.0,
      "step": 109870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2697088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.394e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.428443908691406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0229,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8980266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.55080267931274,
      "rewrite6_current_pass_fraction": 0.6445541469494265,
      "rewrite6_full_passes": 1.0,
      "step": 109880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2721664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3945e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.429264068603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9905,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8983066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.895627818935157,
      "rewrite6_current_pass_fraction": 0.6450069132051957,
      "rewrite6_full_passes": 1.0,
      "step": 109890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.274624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.395e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.429264068603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0116,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8986666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.319885746848296,
      "rewrite6_current_pass_fraction": 0.6454986270958696,
      "rewrite6_full_passes": 1.0,
      "step": 109900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2770816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3955e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.429264068603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0174,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8989866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.4380835263931,
      "rewrite6_current_pass_fraction": 0.646053630893264,
      "rewrite6_full_passes": 1.0,
      "step": 109910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2795392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.396e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.429450988769531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8992,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.212475241658723,
      "rewrite6_current_pass_fraction": 0.6466427138712002,
      "rewrite6_full_passes": 1.0,
      "step": 109920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2819968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3965e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.429450988769531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9916,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.89952,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.917525050805544,
      "rewrite6_current_pass_fraction": 0.6472220599404102,
      "rewrite6_full_passes": 1.0,
      "step": 109930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2844544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.397e-05,
      "grad_norm": 0.431640625,
      "host_rss_gib": 15.429450988769531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9955,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.8997733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.995355068388445,
      "rewrite6_current_pass_fraction": 0.6477916691008938,
      "rewrite6_full_passes": 1.0,
      "step": 109940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.286912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3975e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.429450988769531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0342,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9000666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.784343771694374,
      "rewrite6_current_pass_fraction": 0.6483953574419193,
      "rewrite6_full_passes": 1.0,
      "step": 109950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2893696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.398e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.429450988769531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0069,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90032,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.224606365777017,
      "rewrite6_current_pass_fraction": 0.6490233880547603,
      "rewrite6_full_passes": 1.0,
      "step": 109960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2918272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3985e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.429450988769531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0151,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9005333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.391129951592116,
      "rewrite6_current_pass_fraction": 0.6497049716655956,
      "rewrite6_full_passes": 1.0,
      "step": 109970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2942848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.399e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 15.431411743164062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9008133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.701372449879624,
      "rewrite6_current_pass_fraction": 0.6503135284609842,
      "rewrite6_full_passes": 1.0,
      "step": 109980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2967424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.3995e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.43453598022461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0011,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9012,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.107643170009883,
      "rewrite6_current_pass_fraction": 0.6509172168020096,
      "rewrite6_full_passes": 1.0,
      "step": 109990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.2992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.43454360961914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0203,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9016666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.497439993898546,
      "rewrite6_current_pass_fraction": 0.6515403789604877,
      "rewrite6_full_passes": 1.0,
      "step": 110000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.4e-05,
      "eval_loss": 3.082461357116699,
      "eval_perplexity": 21.81202355812623,
      "eval_runtime": 0.0807,
      "eval_samples_per_second": 396.591,
      "eval_steps_per_second": 12.393,
      "step": 110000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3016576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4005e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.63125228881836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9019866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.174107999122125,
      "rewrite6_current_pass_fraction": 0.6520759089404295,
      "rewrite6_full_passes": 1.0,
      "step": 110010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3041152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.401e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.62518310546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0147,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9023066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.382975130684393,
      "rewrite6_current_pass_fraction": 0.6525822281941929,
      "rewrite6_full_passes": 1.0,
      "step": 110020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3065728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4015e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.62518310546875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0171,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9025866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.431953020956982,
      "rewrite6_current_pass_fraction": 0.6531761796264922,
      "rewrite6_full_passes": 1.0,
      "step": 110030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3090304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.402e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.144691467285156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0098,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90288,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.28334285097683,
      "rewrite6_current_pass_fraction": 0.653886973963506,
      "rewrite6_full_passes": 1.0,
      "step": 110040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.311488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4025e-05,
      "grad_norm": 0.55859375,
      "host_rss_gib": 15.691856384277344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9032,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.926158934434543,
      "rewrite6_current_pass_fraction": 0.6543932932172694,
      "rewrite6_full_passes": 1.0,
      "step": 110050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3139456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.403e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.114612579345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0201,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9036133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.49334091582123,
      "rewrite6_current_pass_fraction": 0.6549629023777531,
      "rewrite6_full_passes": 1.0,
      "step": 110060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3164032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4035e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.114612579345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0118,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90396,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.323950130422475,
      "rewrite6_current_pass_fraction": 0.6555617222644154,
      "rewrite6_full_passes": 1.0,
      "step": 110070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3188608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.404e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0006,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90428,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.097591861461414,
      "rewrite6_current_pass_fraction": 0.6560923837899943,
      "rewrite6_full_passes": 1.0,
      "step": 110080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3213184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4045e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0114,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90452,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.31582217606955,
      "rewrite6_current_pass_fraction": 0.6567398882202878,
      "rewrite6_full_passes": 1.0,
      "step": 110090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.323776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.405e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0164,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90472,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.417655658502976,
      "rewrite6_current_pass_fraction": 0.657285155108956,
      "rewrite6_full_passes": 1.0,
      "step": 110100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3262336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4055e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0126,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9050933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.34021579592551,
      "rewrite6_current_pass_fraction": 0.6578352904519873,
      "rewrite6_full_passes": 1.0,
      "step": 110110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3286912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.406e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0216,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90536,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.52410399373532,
      "rewrite6_current_pass_fraction": 0.6583610835232031,
      "rewrite6_full_passes": 1.0,
      "step": 110120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3311488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4065e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0048,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90568,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.182179256465627,
      "rewrite6_current_pass_fraction": 0.6589014819575081,
      "rewrite6_full_passes": 1.0,
      "step": 110130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3336064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.407e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9934,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9059866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.953408881656124,
      "rewrite6_current_pass_fraction": 0.6594613542092657,
      "rewrite6_full_passes": 1.0,
      "step": 110140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.336064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4075e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0109,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9062,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.30566680403609,
      "rewrite6_current_pass_fraction": 0.659880041284493,
      "rewrite6_full_passes": 1.0,
      "step": 110150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3385216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.408e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9984,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9063466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.05342575989159,
      "rewrite6_current_pass_fraction": 0.6604399135362505,
      "rewrite6_full_passes": 1.0,
      "step": 110160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3409792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4085e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0473,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9065333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.058409888808015,
      "rewrite6_current_pass_fraction": 0.6610241280598236,
      "rewrite6_full_passes": 1.0,
      "step": 110170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3434368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.409e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0183,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9068266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.456486081474473,
      "rewrite6_current_pass_fraction": 0.6615499211310393,
      "rewrite6_full_passes": 1.0,
      "step": 110180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3458944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4095e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9072133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.6620854511109813,
      "rewrite6_full_passes": 1.0,
      "step": 110190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.348352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.41e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.017,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9075333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.42990992781124,
      "rewrite6_current_pass_fraction": 0.6626063757278339,
      "rewrite6_full_passes": 1.0,
      "step": 110200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3508096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4105e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.005,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90784,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.186216095987415,
      "rewrite6_current_pass_fraction": 0.6632100640688594,
      "rewrite6_full_passes": 1.0,
      "step": 110210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3532672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.411e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0314,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90812,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.726229007771437,
      "rewrite6_current_pass_fraction": 0.6637650678662538,
      "rewrite6_full_passes": 1.0,
      "step": 110220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3557248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4115e-05,
      "grad_norm": 0.55078125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0345,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9084666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.790580010214892,
      "rewrite6_current_pass_fraction": 0.6643736246616424,
      "rewrite6_full_passes": 1.0,
      "step": 110230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3581824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.412e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0414,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9088933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.934531072324265,
      "rewrite6_current_pass_fraction": 0.6649042861872213,
      "rewrite6_full_passes": 1.0,
      "step": 110240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.36064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4125e-05,
      "grad_norm": 0.59375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0197,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.90932,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.4851452187036,
      "rewrite6_current_pass_fraction": 0.6654446846215264,
      "rewrite6_full_passes": 1.0,
      "step": 110250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3630976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.413e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.002,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9096133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.125748194902,
      "rewrite6_current_pass_fraction": 0.6659850830558315,
      "rewrite6_full_passes": 1.0,
      "step": 110260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3655552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4135e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9098,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.87808407616535,
      "rewrite6_current_pass_fraction": 0.6665157445814103,
      "rewrite6_full_passes": 1.0,
      "step": 110270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3680128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.414e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0178,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.91008,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.446260395068368,
      "rewrite6_current_pass_fraction": 0.6671243013767989,
      "rewrite6_full_passes": 1.0,
      "step": 110280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3704704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4145e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9993,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9102666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.071481967149957,
      "rewrite6_current_pass_fraction": 0.6677474635352768,
      "rewrite6_full_passes": 1.0,
      "step": 110290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.372928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.415e-05,
      "grad_norm": 0.5390625,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9106,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.23269782650759,
      "rewrite6_current_pass_fraction": 0.6683754941481178,
      "rewrite6_full_passes": 1.0,
      "step": 110300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3753856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4155e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0368,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.91096,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.838453377506596,
      "rewrite6_current_pass_fraction": 0.6689110241280598,
      "rewrite6_full_passes": 1.0,
      "step": 110310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3778432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.416e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0336,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9112266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.771876905865113,
      "rewrite6_current_pass_fraction": 0.6695049755603592,
      "rewrite6_full_passes": 1.0,
      "step": 110320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3803008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4165e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0022,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9116266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.129773747082783,
      "rewrite6_current_pass_fraction": 0.6701037954470215,
      "rewrite6_full_passes": 1.0,
      "step": 110330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3827584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.417e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0196,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9118533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.48309680660405,
      "rewrite6_current_pass_fraction": 0.6706052462464217,
      "rewrite6_full_passes": 1.0,
      "step": 110340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.385216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4175e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0113,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.91232,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.313790695427663,
      "rewrite6_current_pass_fraction": 0.671204066133084,
      "rewrite6_full_passes": 1.0,
      "step": 110350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3876736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.418e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9126533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.188234818531463,
      "rewrite6_current_pass_fraction": 0.6718467021090144,
      "rewrite6_full_passes": 1.0,
      "step": 110360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3901312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4185e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.91292,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.6723919689976826,
      "rewrite6_full_passes": 1.0,
      "step": 110370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3925888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.419e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0136,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9131733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.360566185220215,
      "rewrite6_current_pass_fraction": 0.6730589472454285,
      "rewrite6_full_passes": 1.0,
      "step": 110380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3950464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4195e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0231,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.91348,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.554913250892056,
      "rewrite6_current_pass_fraction": 0.6736821094039065,
      "rewrite6_full_passes": 1.0,
      "step": 110390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.397504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.42e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0311,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.91376,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.720012071656146,
      "rewrite6_current_pass_fraction": 0.674246850110027,
      "rewrite6_full_passes": 1.0,
      "step": 110400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.3999616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4205e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0069,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9141466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.224606365777017,
      "rewrite6_current_pass_fraction": 0.6748115908161477,
      "rewrite6_full_passes": 1.0,
      "step": 110410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4024192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.421e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9144666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.97853978054658,
      "rewrite6_current_pass_fraction": 0.6754834375182567,
      "rewrite6_full_passes": 1.0,
      "step": 110420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4048768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4215e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0127,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.91472,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.342249919209575,
      "rewrite6_current_pass_fraction": 0.6760627835874666,
      "rewrite6_full_passes": 1.0,
      "step": 110430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4073344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.422e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0108,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9150933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.30363633888064,
      "rewrite6_current_pass_fraction": 0.6766908142003076,
      "rewrite6_full_passes": 1.0,
      "step": 110440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.409792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4225e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9929,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.91532,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.943434670975765,
      "rewrite6_current_pass_fraction": 0.6771922649997079,
      "rewrite6_full_passes": 1.0,
      "step": 110450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4122496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.423e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0215,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9156533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.522051685953052,
      "rewrite6_current_pass_fraction": 0.6777862164320072,
      "rewrite6_full_passes": 1.0,
      "step": 110460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4147072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4235e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9879,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9159333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.843966375584888,
      "rewrite6_current_pass_fraction": 0.6783412202294016,
      "rewrite6_full_passes": 1.0,
      "step": 110470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4171648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.424e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0019,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9161866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.1237357207079,
      "rewrite6_current_pass_fraction": 0.6789497770247902,
      "rewrite6_full_passes": 1.0,
      "step": 110480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4196224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4245e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0057,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9165066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.20035139403173,
      "rewrite6_current_pass_fraction": 0.6794901754590953,
      "rewrite6_full_passes": 1.0,
      "step": 110490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.42208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.425e-05,
      "grad_norm": 0.431640625,
      "host_rss_gib": 16.11505889892578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9887,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9167733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.859847900448294,
      "rewrite6_current_pass_fraction": 0.6800354423477635,
      "rewrite6_full_passes": 1.0,
      "step": 110500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.425e-05,
      "eval_loss": 3.081953525543213,
      "eval_perplexity": 21.80094953598882,
      "eval_runtime": 0.0574,
      "eval_samples_per_second": 557.353,
      "eval_steps_per_second": 17.417,
      "step": 110500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4245376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4255e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 16.147544860839844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0269,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9170533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.63317051587937,
      "rewrite6_current_pass_fraction": 0.6806099199626102,
      "rewrite6_full_passes": 1.0,
      "step": 110510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4269952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.426e-05,
      "grad_norm": 0.5390625,
      "host_rss_gib": 16.147571563720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0292,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9172933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.680681424666606,
      "rewrite6_current_pass_fraction": 0.6812379505754513,
      "rewrite6_full_passes": 1.0,
      "step": 110520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4294528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4265e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 16.147571563720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0212,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.91768,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.515895993847245,
      "rewrite6_current_pass_fraction": 0.6817734805553932,
      "rewrite6_full_passes": 1.0,
      "step": 110530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4319104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.427e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.147571563720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0061,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9179466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.20843315083295,
      "rewrite6_current_pass_fraction": 0.6823966427138712,
      "rewrite6_full_passes": 1.0,
      "step": 110540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.434368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4275e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.147571563720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0143,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9183066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.374823571052737,
      "rewrite6_current_pass_fraction": 0.6829126988763607,
      "rewrite6_full_passes": 1.0,
      "step": 110550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4368256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.428e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.271446228027344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9984,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.91868,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.05342575989159,
      "rewrite6_current_pass_fraction": 0.6835017818542969,
      "rewrite6_full_passes": 1.0,
      "step": 110560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4392832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4285e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.27176284790039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9190666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.662077184548252,
      "rewrite6_current_pass_fraction": 0.6840665225604176,
      "rewrite6_full_passes": 1.0,
      "step": 110570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4417408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.429e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.275554656982422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0118,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9194,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.323950130422475,
      "rewrite6_current_pass_fraction": 0.6846069209947226,
      "rewrite6_full_passes": 1.0,
      "step": 110580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4441984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4295e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.275638580322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9197466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.631107301990202,
      "rewrite6_current_pass_fraction": 0.685205740881385,
      "rewrite6_full_passes": 1.0,
      "step": 110590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.446656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.43e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.275642395019531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0232,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.556968844995144,
      "rewrite6_current_pass_fraction": 0.685741270861327,
      "rewrite6_full_passes": 1.0,
      "step": 110600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4491136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4305e-05,
      "grad_norm": 0.55859375,
      "host_rss_gib": 15.275642395019531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0227,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92028,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.54669292976553,
      "rewrite6_current_pass_fraction": 0.6862427216607272,
      "rewrite6_full_passes": 1.0,
      "step": 110610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4515712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.431e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.275642395019531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0204,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9205333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.49948984038855,
      "rewrite6_current_pass_fraction": 0.6867393040057642,
      "rewrite6_full_passes": 1.0,
      "step": 110620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4540288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4315e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.276844024658203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0289,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92084,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.67447815077682,
      "rewrite6_current_pass_fraction": 0.6873381238924267,
      "rewrite6_full_passes": 1.0,
      "step": 110630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4564864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.432e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.276844024658203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0506,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9212533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.128017430716405,
      "rewrite6_current_pass_fraction": 0.68784444314619,
      "rewrite6_full_passes": 1.0,
      "step": 110640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.458944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4325e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.27707290649414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0027,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9215866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.139841150597462,
      "rewrite6_current_pass_fraction": 0.6883458939455902,
      "rewrite6_full_passes": 1.0,
      "step": 110650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4614016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.433e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.27707290649414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9975,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92192,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.035385795909175,
      "rewrite6_current_pass_fraction": 0.6889495822866156,
      "rewrite6_full_passes": 1.0,
      "step": 110660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4638592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4335e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.27707290649414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.01,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9222133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.287399925240926,
      "rewrite6_current_pass_fraction": 0.6894510330860159,
      "rewrite6_full_passes": 1.0,
      "step": 110670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4663168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.434e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.27707290649414,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0304,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92256,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.705513138424664,
      "rewrite6_current_pass_fraction": 0.69008393215322,
      "rewrite6_full_passes": 1.0,
      "step": 110680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4687744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4345e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.277938842773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0127,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9228933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.342249919209575,
      "rewrite6_current_pass_fraction": 0.6906389359506144,
      "rewrite6_full_passes": 1.0,
      "step": 110690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.471232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.435e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.277938842773438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0228,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92328,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.5487477017954,
      "rewrite6_current_pass_fraction": 0.6912426242916399,
      "rewrite6_full_passes": 1.0,
      "step": 110700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4736896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4355e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.28048324584961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9716,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9236133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.52313163035366,
      "rewrite6_current_pass_fraction": 0.6917830227259449,
      "rewrite6_full_passes": 1.0,
      "step": 110710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4761472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.436e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 15.28048324584961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0127,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9239466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.342249919209575,
      "rewrite6_current_pass_fraction": 0.6923623687951549,
      "rewrite6_full_passes": 1.0,
      "step": 110720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4786048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4365e-05,
      "grad_norm": 0.56640625,
      "host_rss_gib": 14.127616882324219,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0339,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9243733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.77810940376482,
      "rewrite6_current_pass_fraction": 0.6930682946778057,
      "rewrite6_full_passes": 1.0,
      "step": 110730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4810624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.437e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.913745880126953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92472,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.926158934434543,
      "rewrite6_current_pass_fraction": 0.6936379038382894,
      "rewrite6_full_passes": 1.0,
      "step": 110740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.48352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4375e-05,
      "grad_norm": 0.57421875,
      "host_rss_gib": 15.527976989746094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.986,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9251333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.806298635156402,
      "rewrite6_current_pass_fraction": 0.6942269868162256,
      "rewrite6_full_passes": 1.0,
      "step": 110750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4859776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.438e-05,
      "grad_norm": 0.6015625,
      "host_rss_gib": 15.678905487060547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0065,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9254,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.21651814098351,
      "rewrite6_current_pass_fraction": 0.6946748846176316,
      "rewrite6_full_passes": 1.0,
      "step": 110760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4884352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4385e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.679679870605469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0306,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92572,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.709654655190217,
      "rewrite6_current_pass_fraction": 0.6952493622324785,
      "rewrite6_full_passes": 1.0,
      "step": 110770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4908928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.439e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.679679870605469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0121,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92596,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.330048230130824,
      "rewrite6_current_pass_fraction": 0.6959066035714981,
      "rewrite6_full_passes": 1.0,
      "step": 110780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4933504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4395e-05,
      "grad_norm": 0.4296875,
      "host_rss_gib": 15.679679870605469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0163,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9262533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.415613995022007,
      "rewrite6_current_pass_fraction": 0.6964664758232556,
      "rewrite6_full_passes": 1.0,
      "step": 110790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.495808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.44e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.679679870605469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0242,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92652,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.577536095751576,
      "rewrite6_current_pass_fraction": 0.6970360849837394,
      "rewrite6_full_passes": 1.0,
      "step": 110800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.4982656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4405e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.679679870605469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0185,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9267733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.460577787847765,
      "rewrite6_current_pass_fraction": 0.6976592471422173,
      "rewrite6_full_passes": 1.0,
      "step": 110810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5007232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.441e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.679683685302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0238,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92708,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.56930672729669,
      "rewrite6_current_pass_fraction": 0.6981850402134331,
      "rewrite6_full_passes": 1.0,
      "step": 110820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5031808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4415e-05,
      "grad_norm": 0.53515625,
      "host_rss_gib": 15.679683685302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92728,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.15797516671621,
      "rewrite6_current_pass_fraction": 0.6988276761893634,
      "rewrite6_full_passes": 1.0,
      "step": 110830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5056384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.442e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.679683685302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.998,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9275866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.04540599364781,
      "rewrite6_current_pass_fraction": 0.6993972853498471,
      "rewrite6_full_passes": 1.0,
      "step": 110840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.508096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4425e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.679683685302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9961,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9279866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.00735588131334,
      "rewrite6_current_pass_fraction": 0.7000788689606824,
      "rewrite6_full_passes": 1.0,
      "step": 110850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5105536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.443e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.679683685302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0113,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9282133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.313790695427663,
      "rewrite6_current_pass_fraction": 0.7006533465755292,
      "rewrite6_full_passes": 1.0,
      "step": 110860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5130112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4435e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.679683685302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92852,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.817625339883573,
      "rewrite6_current_pass_fraction": 0.7011742711923818,
      "rewrite6_full_passes": 1.0,
      "step": 110870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5154688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.444e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.679683685302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0287,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9288266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.670343668608663,
      "rewrite6_current_pass_fraction": 0.7017341434441393,
      "rewrite6_full_passes": 1.0,
      "step": 110880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5179264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4445e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.685543060302734,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9841,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9290666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.76870239548749,
      "rewrite6_current_pass_fraction": 0.7022501996066289,
      "rewrite6_full_passes": 1.0,
      "step": 110890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.520384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.445e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.686328887939453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9294266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.7028879671281961,
      "rewrite6_full_passes": 1.0,
      "step": 110900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5228416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4455e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.686336517333984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0043,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.92972,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.172090689189396,
      "rewrite6_current_pass_fraction": 0.7034234971081381,
      "rewrite6_full_passes": 1.0,
      "step": 110910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5252992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.446e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.686336517333984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0243,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.93016,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.579593952252267,
      "rewrite6_current_pass_fraction": 0.7040077116317112,
      "rewrite6_full_passes": 1.0,
      "step": 110920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5277568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4465e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.686336517333984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0199,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9304266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.48924265747756,
      "rewrite6_current_pass_fraction": 0.7046065315183736,
      "rewrite6_full_passes": 1.0,
      "step": 110930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5302144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.447e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.68658447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0254,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9308,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.602243960820573,
      "rewrite6_current_pass_fraction": 0.7051177192265,
      "rewrite6_full_passes": 1.0,
      "step": 110940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.532672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4475e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.68658447265625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0074,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.93116,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.234721197457105,
      "rewrite6_current_pass_fraction": 0.7057798290198828,
      "rewrite6_full_passes": 1.0,
      "step": 110950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5351296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.448e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.49713134765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.011,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9314666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.307697472248215,
      "rewrite6_current_pass_fraction": 0.706325095908551,
      "rewrite6_full_passes": 1.0,
      "step": 110960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5375872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4485e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.49713134765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0186,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.93176,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.46262394793285,
      "rewrite6_current_pass_fraction": 0.7068265467079512,
      "rewrite6_full_passes": 1.0,
      "step": 110970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5400448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.449e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.49713134765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0143,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9320666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.374823571052737,
      "rewrite6_current_pass_fraction": 0.7074010243227979,
      "rewrite6_full_passes": 1.0,
      "step": 110980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5425024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4495e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.49713134765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0238,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.93232,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.56930672729669,
      "rewrite6_current_pass_fraction": 0.7079560281201924,
      "rewrite6_full_passes": 1.0,
      "step": 110990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.54496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.45e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.49713134765625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0264,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9326666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.622856509337947,
      "rewrite6_current_pass_fraction": 0.708520768826313,
      "rewrite6_full_passes": 1.0,
      "step": 111000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.45e-05,
      "eval_loss": 3.0820095539093018,
      "eval_perplexity": 21.802171041789677,
      "eval_runtime": 0.0957,
      "eval_samples_per_second": 334.432,
      "eval_steps_per_second": 10.451,
      "step": 111000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5474176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4505e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.502830505371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0019,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9330266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.1237357207079,
      "rewrite6_current_pass_fraction": 0.7091828786196959,
      "rewrite6_full_passes": 1.0,
      "step": 111010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5498752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.451e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.502830505371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0052,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9332533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.19025374295785,
      "rewrite6_current_pass_fraction": 0.7097816985063582,
      "rewrite6_full_passes": 1.0,
      "step": 111020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5523328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4515e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.502830505371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.988,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9334933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.845950871445588,
      "rewrite6_current_pass_fraction": 0.7103756499386574,
      "rewrite6_full_passes": 1.0,
      "step": 111030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5547904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.452e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.502830505371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0147,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9338933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.382975130684393,
      "rewrite6_current_pass_fraction": 0.710847890011879,
      "rewrite6_full_passes": 1.0,
      "step": 111040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.557248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4525e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.502830505371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0251,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9342666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.596064214640606,
      "rewrite6_current_pass_fraction": 0.7114661837159938,
      "rewrite6_full_passes": 1.0,
      "step": 111050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5597056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.453e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.502830505371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0133,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9346666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.354458931498517,
      "rewrite6_current_pass_fraction": 0.7118946076999474,
      "rewrite6_full_passes": 1.0,
      "step": 111060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5621632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4535e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.502830505371094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.93492,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.174107999122125,
      "rewrite6_current_pass_fraction": 0.7124106638624369,
      "rewrite6_full_passes": 1.0,
      "step": 111070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5646208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.454e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.503273010253906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0095,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9351466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.2772587607807,
      "rewrite6_current_pass_fraction": 0.7129997468403731,
      "rewrite6_full_passes": 1.0,
      "step": 111080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5670784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4545e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9915,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.93536,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.915533397884765,
      "rewrite6_current_pass_fraction": 0.7135109345484996,
      "rewrite6_full_passes": 1.0,
      "step": 111090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.569536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.455e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.999,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.93568,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.06546142568619,
      "rewrite6_current_pass_fraction": 0.7141876497049716,
      "rewrite6_full_passes": 1.0,
      "step": 111100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5719936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4555e-05,
      "grad_norm": 0.5390625,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0171,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9359333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.431953020956982,
      "rewrite6_current_pass_fraction": 0.7147767326829079,
      "rewrite6_full_passes": 1.0,
      "step": 111110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5744512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.456e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0481,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9362266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.075263357207568,
      "rewrite6_current_pass_fraction": 0.715458316293743,
      "rewrite6_full_passes": 1.0,
      "step": 111120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5769088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4565e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9843,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9365466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.772656531366998,
      "rewrite6_current_pass_fraction": 0.7160522677260424,
      "rewrite6_full_passes": 1.0,
      "step": 111130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5793664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.457e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0209,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9368,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.5097421481721,
      "rewrite6_current_pass_fraction": 0.716621876886526,
      "rewrite6_full_passes": 1.0,
      "step": 111140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.581824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4575e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0156,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9370666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.40132806588403,
      "rewrite6_current_pass_fraction": 0.7172255652275515,
      "rewrite6_full_passes": 1.0,
      "step": 111150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5842816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.458e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.93732,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.989031673191437,
      "rewrite6_current_pass_fraction": 0.7178292535685771,
      "rewrite6_full_passes": 1.0,
      "step": 111160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5867392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4585e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0234,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9376,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.56108064993093,
      "rewrite6_current_pass_fraction": 0.718452415727055,
      "rewrite6_full_passes": 1.0,
      "step": 111170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5891968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.459e-05,
      "grad_norm": 0.62109375,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0328,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9379466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.755266049568853,
      "rewrite6_current_pass_fraction": 0.7190171564331755,
      "rewrite6_full_passes": 1.0,
      "step": 111180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5916544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4595e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0077,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.93808,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.240792524469853,
      "rewrite6_current_pass_fraction": 0.7195770286849331,
      "rewrite6_full_passes": 1.0,
      "step": 111190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.594112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.46e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0249,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9382933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.591945413691505,
      "rewrite6_current_pass_fraction": 0.720063874121244,
      "rewrite6_full_passes": 1.0,
      "step": 111200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5965696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4605e-05,
      "grad_norm": 0.55859375,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9385333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.7206821678253588,
      "rewrite6_full_passes": 1.0,
      "step": 111210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.5990272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.461e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.038,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9388266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.86347453124931,
      "rewrite6_current_pass_fraction": 0.7212663823489318,
      "rewrite6_full_passes": 1.0,
      "step": 111220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6014848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4615e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9935,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9391733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.955404322314664,
      "rewrite6_current_pass_fraction": 0.7218019123288738,
      "rewrite6_full_passes": 1.0,
      "step": 111230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6039424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.462e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.503276824951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.025,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9397066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.59400471119603,
      "rewrite6_current_pass_fraction": 0.7223812583980838,
      "rewrite6_full_passes": 1.0,
      "step": 111240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4625e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0424,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9401333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.955476074152095,
      "rewrite6_current_pass_fraction": 0.7230238943740142,
      "rewrite6_full_passes": 1.0,
      "step": 111250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6088576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.463e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0234,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.94048,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.56108064993093,
      "rewrite6_current_pass_fraction": 0.7236032404432241,
      "rewrite6_full_passes": 1.0,
      "step": 111260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6113152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4635e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.028,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9408,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.655879491093387,
      "rewrite6_current_pass_fraction": 0.7241241650600767,
      "rewrite6_full_passes": 1.0,
      "step": 111270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6137728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.464e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0043,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.94104,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.172090689189396,
      "rewrite6_current_pass_fraction": 0.7246840373118343,
      "rewrite6_full_passes": 1.0,
      "step": 111280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6162304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4645e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9413066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.69516296961328,
      "rewrite6_current_pass_fraction": 0.7252925941072228,
      "rewrite6_full_passes": 1.0,
      "step": 111290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.618688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.465e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0355,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9417466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.811380988981075,
      "rewrite6_current_pass_fraction": 0.725837860995891,
      "rewrite6_full_passes": 1.0,
      "step": 111300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6211456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4655e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0105,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.942,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.29754616155125,
      "rewrite6_current_pass_fraction": 0.7265583922416311,
      "rewrite6_full_passes": 1.0,
      "step": 111310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6236032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.466e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9995,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9423733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.07549666499979,
      "rewrite6_current_pass_fraction": 0.7271085275846624,
      "rewrite6_full_passes": 1.0,
      "step": 111320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6260608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4665e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0147,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9426933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.382975130684393,
      "rewrite6_current_pass_fraction": 0.7277706373780453,
      "rewrite6_full_passes": 1.0,
      "step": 111330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6285184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.467e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0145,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9430666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.378898943290586,
      "rewrite6_current_pass_fraction": 0.7283645888103445,
      "rewrite6_full_passes": 1.0,
      "step": 111340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.630976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4675e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0105,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9432666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.29754616155125,
      "rewrite6_current_pass_fraction": 0.728924461062102,
      "rewrite6_full_passes": 1.0,
      "step": 111350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6334336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.468e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9976,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9434666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.037389434669027,
      "rewrite6_current_pass_fraction": 0.7295232809487644,
      "rewrite6_full_passes": 1.0,
      "step": 111360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6358912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4685e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9437333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.039465846006077,
      "rewrite6_current_pass_fraction": 0.7301999961052366,
      "rewrite6_full_passes": 1.0,
      "step": 111370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6383488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.469e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9972,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.94408,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.029376081672606,
      "rewrite6_current_pass_fraction": 0.730711183813363,
      "rewrite6_full_passes": 1.0,
      "step": 111380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6408064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4695e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0266,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9444133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.626981493124443,
      "rewrite6_current_pass_fraction": 0.7311931607953107,
      "rewrite6_full_passes": 1.0,
      "step": 111390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.643264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.47e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0189,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9447733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.468763656027395,
      "rewrite6_current_pass_fraction": 0.7318942182235983,
      "rewrite6_full_passes": 1.0,
      "step": 111400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6457216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4705e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.504100799560547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0076,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9450666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.238768546418,
      "rewrite6_current_pass_fraction": 0.7324784327471714,
      "rewrite6_full_passes": 1.0,
      "step": 111410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6481792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.471e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 14.054935455322266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.048,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9453466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.073155936244657,
      "rewrite6_current_pass_fraction": 0.733130805631828,
      "rewrite6_full_passes": 1.0,
      "step": 111420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6506368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4715e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 14.05697250366211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9456533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.631107301990202,
      "rewrite6_current_pass_fraction": 0.7337004147923117,
      "rewrite6_full_passes": 1.0,
      "step": 111430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6530944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.472e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.060157775878906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9971,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.94596,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.027373244207986,
      "rewrite6_current_pass_fraction": 0.734250550135343,
      "rewrite6_full_passes": 1.0,
      "step": 111440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.655552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4725e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.060474395751953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.94632,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.64348968071354,
      "rewrite6_current_pass_fraction": 0.7347081848454753,
      "rewrite6_full_passes": 1.0,
      "step": 111450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6580096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.473e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.060558319091797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9999,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9465866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.08352846991968,
      "rewrite6_current_pass_fraction": 0.735365426184495,
      "rewrite6_full_passes": 1.0,
      "step": 111460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6604672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4735e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.062572479248047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0087,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9468266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.2610434407649,
      "rewrite6_current_pass_fraction": 0.7359447722537049,
      "rewrite6_full_passes": 1.0,
      "step": 111470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6629248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.474e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 14.06298828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0162,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9470666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.413572535697167,
      "rewrite6_current_pass_fraction": 0.7365192498685518,
      "rewrite6_full_passes": 1.0,
      "step": 111480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6653824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4745e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 14.06298828125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0281,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9473733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.65794518232533,
      "rewrite6_current_pass_fraction": 0.7370207006679519,
      "rewrite6_full_passes": 1.0,
      "step": 111490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.66784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.475e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 14.063201904296875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0461,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9475066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.033154952933565,
      "rewrite6_current_pass_fraction": 0.7374783353780842,
      "rewrite6_full_passes": 1.0,
      "step": 111500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.475e-05,
      "eval_loss": 3.0818963050842285,
      "eval_perplexity": 21.799702111339524,
      "eval_runtime": 0.082,
      "eval_samples_per_second": 390.327,
      "eval_steps_per_second": 12.198,
      "step": 111500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6702976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4755e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 14.145843505859375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9478133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.738096629082199,
      "rewrite6_full_passes": 1.0,
      "step": 111510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6727552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.476e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.145957946777344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0144,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9482666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.376861155287358,
      "rewrite6_current_pass_fraction": 0.7386418959708672,
      "rewrite6_full_passes": 1.0,
      "step": 111520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6752128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4765e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 14.145957946777344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0114,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9485466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.31582217606955,
      "rewrite6_current_pass_fraction": 0.739260189674982,
      "rewrite6_full_passes": 1.0,
      "step": 111530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6776704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.477e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 14.146125793457031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.016,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9489066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.409490229434265,
      "rewrite6_current_pass_fraction": 0.7398687464703706,
      "rewrite6_full_passes": 1.0,
      "step": 111540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.680128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4775e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 14.146125793457031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9490933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.7404529609939436,
      "rewrite6_full_passes": 1.0,
      "step": 111550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6825856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.478e-05,
      "grad_norm": 0.5625,
      "host_rss_gib": 14.146209716796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0043,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9493733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.172090689189396,
      "rewrite6_current_pass_fraction": 0.7409690171564332,
      "rewrite6_full_passes": 1.0,
      "step": 111560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6850432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4785e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.146209716796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0313,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9496666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.724156488498345,
      "rewrite6_current_pass_fraction": 0.7414315203209285,
      "rewrite6_full_passes": 1.0,
      "step": 111570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6875008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.479e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.146209716796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0209,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9500266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.5097421481721,
      "rewrite6_current_pass_fraction": 0.7420546824794064,
      "rewrite6_full_passes": 1.0,
      "step": 111580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6899584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4795e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.05941390991211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9503866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.7426194231855271,
      "rewrite6_full_passes": 1.0,
      "step": 111590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.692416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.48e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.028579711914062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0237,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9506933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.567249899467065,
      "rewrite6_current_pass_fraction": 0.7432133746178263,
      "rewrite6_full_passes": 1.0,
      "step": 111600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6948736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4805e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.261974334716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9994,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9510666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.073489215707433,
      "rewrite6_current_pass_fraction": 0.7437391676890421,
      "rewrite6_full_passes": 1.0,
      "step": 111610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6973312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.481e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.261974334716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9943,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.95136,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.971375033205096,
      "rewrite6_current_pass_fraction": 0.7442503553971686,
      "rewrite6_full_passes": 1.0,
      "step": 111620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.6997888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4815e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.261974334716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0282,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.95164,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.660011080136737,
      "rewrite6_current_pass_fraction": 0.7448199645576522,
      "rewrite6_full_passes": 1.0,
      "step": 111630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7022464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.482e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.261974334716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0267,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9520533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.629044294412097,
      "rewrite6_current_pass_fraction": 0.7453554945375942,
      "rewrite6_full_passes": 1.0,
      "step": 111640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.704704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4825e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.261974334716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0071,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9522666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.22865169156927,
      "rewrite6_current_pass_fraction": 0.7459104983349886,
      "rewrite6_full_passes": 1.0,
      "step": 111650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7071616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.483e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.262401580810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.999,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9525866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.06546142568619,
      "rewrite6_current_pass_fraction": 0.7464070806800257,
      "rewrite6_full_passes": 1.0,
      "step": 111660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7096192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4835e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.262401580810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0222,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.95296,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.536422151209266,
      "rewrite6_current_pass_fraction": 0.746864715390158,
      "rewrite6_full_passes": 1.0,
      "step": 111670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7120768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.484e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.262401580810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0121,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9533866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.330048230130824,
      "rewrite6_current_pass_fraction": 0.7474684037311834,
      "rewrite6_full_passes": 1.0,
      "step": 111680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7145344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4845e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.262401580810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.005,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9537466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.186216095987415,
      "rewrite6_current_pass_fraction": 0.747945512258768,
      "rewrite6_full_passes": 1.0,
      "step": 111690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.716992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.485e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.262401580810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0308,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9540266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.713797000341962,
      "rewrite6_current_pass_fraction": 0.74848104223871,
      "rewrite6_full_passes": 1.0,
      "step": 111700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7194496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4855e-05,
      "grad_norm": 0.57421875,
      "host_rss_gib": 15.262401580810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0166,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9543066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.421739598015016,
      "rewrite6_current_pass_fraction": 0.7490506513991938,
      "rewrite6_full_passes": 1.0,
      "step": 111710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7219072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.486e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9545733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.738668476643575,
      "rewrite6_current_pass_fraction": 0.7496348659227668,
      "rewrite6_full_passes": 1.0,
      "step": 111720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7243648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4865e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9957,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.95488,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.999354539335886,
      "rewrite6_current_pass_fraction": 0.7502921072617865,
      "rewrite6_full_passes": 1.0,
      "step": 111730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7268224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.487e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0106,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9550533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.29957601765852,
      "rewrite6_current_pass_fraction": 0.7509298747833538,
      "rewrite6_full_passes": 1.0,
      "step": 111740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.72928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4875e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0149,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9552666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.387052133397212,
      "rewrite6_current_pass_fraction": 0.7514994839438375,
      "rewrite6_full_passes": 1.0,
      "step": 111750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7317376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.488e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0158,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.95568,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.40540873955097,
      "rewrite6_current_pass_fraction": 0.7521615937372204,
      "rewrite6_full_passes": 1.0,
      "step": 111760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7341952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4885e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0237,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9559866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.567249899467065,
      "rewrite6_current_pass_fraction": 0.7527263344433409,
      "rewrite6_full_passes": 1.0,
      "step": 111770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7366528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.489e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.028,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9562266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.655879491093387,
      "rewrite6_current_pass_fraction": 0.7532862066950985,
      "rewrite6_full_passes": 1.0,
      "step": 111780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7391104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4895e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9988,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9564933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.06144873468353,
      "rewrite6_current_pass_fraction": 0.7537681836770462,
      "rewrite6_full_passes": 1.0,
      "step": 111790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.741568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.49e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0276,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.95676,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.647618791547,
      "rewrite6_current_pass_fraction": 0.7543037136569882,
      "rewrite6_full_passes": 1.0,
      "step": 111800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7440256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4905e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0332,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9571733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.763569816631378,
      "rewrite6_current_pass_fraction": 0.7549560865416448,
      "rewrite6_full_passes": 1.0,
      "step": 111810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7464832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.491e-05,
      "grad_norm": 0.5859375,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9575466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.54258400208604,
      "rewrite6_current_pass_fraction": 0.7555256957021285,
      "rewrite6_full_passes": 1.0,
      "step": 111820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7489408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4915e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0272,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.95776,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.639361395619666,
      "rewrite6_current_pass_fraction": 0.7560758310451597,
      "rewrite6_full_passes": 1.0,
      "step": 111830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7513984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.492e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0029,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9579466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.143869521651258,
      "rewrite6_current_pass_fraction": 0.7566600455687328,
      "rewrite6_full_passes": 1.0,
      "step": 111840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.753856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4925e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0493,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9583333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.100568853497336,
      "rewrite6_current_pass_fraction": 0.7571566279137699,
      "rewrite6_full_passes": 1.0,
      "step": 111850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7563136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.493e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0138,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9586266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.36463870569573,
      "rewrite6_current_pass_fraction": 0.7577846585266109,
      "rewrite6_full_passes": 1.0,
      "step": 111860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7587712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4935e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0472,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9589333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.05630415310768,
      "rewrite6_current_pass_fraction": 0.7583007146891005,
      "rewrite6_full_passes": 1.0,
      "step": 111870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7612288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.494e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9592933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.511793224939044,
      "rewrite6_current_pass_fraction": 0.7588216393059531,
      "rewrite6_full_passes": 1.0,
      "step": 111880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7636864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4945e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0028,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9596133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.14185523541509,
      "rewrite6_current_pass_fraction": 0.7593669061946213,
      "rewrite6_full_passes": 1.0,
      "step": 111890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.766144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.495e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.009,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9599866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.267122665635263,
      "rewrite6_current_pass_fraction": 0.75997546299001,
      "rewrite6_full_passes": 1.0,
      "step": 111900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7686016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4955e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0191,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9602933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.472857818161167,
      "rewrite6_current_pass_fraction": 0.760515861424315,
      "rewrite6_full_passes": 1.0,
      "step": 111910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7710592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.496e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9802,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9606533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.691754601873935,
      "rewrite6_current_pass_fraction": 0.7611049444022512,
      "rewrite6_full_passes": 1.0,
      "step": 111920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7735168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4965e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.96096,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.759417517911764,
      "rewrite6_current_pass_fraction": 0.7617621857412709,
      "rewrite6_full_passes": 1.0,
      "step": 111930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7759744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.497e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0121,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9614133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.330048230130824,
      "rewrite6_current_pass_fraction": 0.7623366633561177,
      "rewrite6_full_passes": 1.0,
      "step": 111940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.778432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4975e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.96168,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.176125510795945,
      "rewrite6_current_pass_fraction": 0.7628721933360597,
      "rewrite6_full_passes": 1.0,
      "step": 111950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7808896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.498e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0204,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9619733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.49948984038855,
      "rewrite6_current_pass_fraction": 0.7634758816770851,
      "rewrite6_full_passes": 1.0,
      "step": 111960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7833472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4985e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0228,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9621733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.5487477017954,
      "rewrite6_current_pass_fraction": 0.7640600962006582,
      "rewrite6_full_passes": 1.0,
      "step": 111970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7858048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.499e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9625066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.54258400208604,
      "rewrite6_current_pass_fraction": 0.7645810208175109,
      "rewrite6_full_passes": 1.0,
      "step": 111980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7882624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.4995e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0406,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9627333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.917790144730294,
      "rewrite6_current_pass_fraction": 0.765345368152519,
      "rewrite6_full_passes": 1.0,
      "step": 111990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.79072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.26443099975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0171,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9630933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.431953020956982,
      "rewrite6_current_pass_fraction": 0.7659149773130026,
      "rewrite6_full_passes": 1.0,
      "step": 112000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.5e-05,
      "eval_loss": 3.081655263900757,
      "eval_perplexity": 21.79444811858303,
      "eval_runtime": 0.0797,
      "eval_samples_per_second": 401.752,
      "eval_steps_per_second": 12.555,
      "step": 112000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7931776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5005e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.357471466064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9633866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.188234818531463,
      "rewrite6_current_pass_fraction": 0.7663628751144087,
      "rewrite6_full_passes": 1.0,
      "step": 112010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7956352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.501e-05,
      "grad_norm": 0.5390625,
      "host_rss_gib": 15.357471466064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0103,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9637466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.2934870582428,
      "rewrite6_current_pass_fraction": 0.76691301045744,
      "rewrite6_full_passes": 1.0,
      "step": 112020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.7980928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5015e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.357471466064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0121,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9642,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.330048230130824,
      "rewrite6_current_pass_fraction": 0.7674241981655664,
      "rewrite6_full_passes": 1.0,
      "step": 112030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8005504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.502e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.357471466064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0048,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9644266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.182179256465627,
      "rewrite6_current_pass_fraction": 0.7679305174193297,
      "rewrite6_full_passes": 1.0,
      "step": 112040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.803008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5025e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.357471466064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0113,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.96484,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.313790695427663,
      "rewrite6_current_pass_fraction": 0.7685926272127125,
      "rewrite6_full_passes": 1.0,
      "step": 112050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8054656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.503e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.357471466064453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.009,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9652933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.267122665635263,
      "rewrite6_current_pass_fraction": 0.7692206578255535,
      "rewrite6_full_passes": 1.0,
      "step": 112060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8079232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5035e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9955,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.96556,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.995355068388445,
      "rewrite6_current_pass_fraction": 0.7697367139880431,
      "rewrite6_full_passes": 1.0,
      "step": 112070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8103808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.504e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0214,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9658666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.51999958339129,
      "rewrite6_current_pass_fraction": 0.7703939553270628,
      "rewrite6_full_passes": 1.0,
      "step": 112080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8128384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5045e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0287,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9661733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.670343668608663,
      "rewrite6_current_pass_fraction": 0.7710171174855407,
      "rewrite6_full_passes": 1.0,
      "step": 112090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.815296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.505e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0131,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.96652,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.350388446774257,
      "rewrite6_current_pass_fraction": 0.7715867266460245,
      "rewrite6_full_passes": 1.0,
      "step": 112100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8177536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5055e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0266,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9667733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.626981493124443,
      "rewrite6_current_pass_fraction": 0.7722585733481334,
      "rewrite6_full_passes": 1.0,
      "step": 112110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8202112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.506e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.042,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9670266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.947095559937008,
      "rewrite6_current_pass_fraction": 0.7728427878717065,
      "rewrite6_full_passes": 1.0,
      "step": 112120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8226688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5065e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0265,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9674,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.624918898106596,
      "rewrite6_current_pass_fraction": 0.7734610815758213,
      "rewrite6_full_passes": 1.0,
      "step": 112130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8251264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.507e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0046,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9676533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.17814322423101,
      "rewrite6_current_pass_fraction": 0.7740063484644895,
      "rewrite6_full_passes": 1.0,
      "step": 112140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.827584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5075e-05,
      "grad_norm": 0.58984375,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0045,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9679733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.176125510795945,
      "rewrite6_current_pass_fraction": 0.7745613522618839,
      "rewrite6_full_passes": 1.0,
      "step": 112150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8300416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.508e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.006,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9682666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.20641240855666,
      "rewrite6_current_pass_fraction": 0.7751504352398201,
      "rewrite6_full_passes": 1.0,
      "step": 112160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8324992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5085e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9685733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.15797516671621,
      "rewrite6_current_pass_fraction": 0.7756957021284883,
      "rewrite6_full_passes": 1.0,
      "step": 112170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8349568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.509e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9688533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.686886559817744,
      "rewrite6_current_pass_fraction": 0.7762701797433351,
      "rewrite6_full_passes": 1.0,
      "step": 112180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8374144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5095e-05,
      "grad_norm": 0.431640625,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9803,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9692,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.693723875796184,
      "rewrite6_current_pass_fraction": 0.7768836049930868,
      "rewrite6_full_passes": 1.0,
      "step": 112190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.839872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.51e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0012,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9694266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.109654034868445,
      "rewrite6_current_pass_fraction": 0.777428871881755,
      "rewrite6_full_passes": 1.0,
      "step": 112200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8423296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5105e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0716,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9698,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.57639730762967,
      "rewrite6_current_pass_fraction": 0.778008217950965,
      "rewrite6_full_passes": 1.0,
      "step": 112210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8447872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.511e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0098,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9700133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.28334285097683,
      "rewrite6_current_pass_fraction": 0.7785729586570855,
      "rewrite6_full_passes": 1.0,
      "step": 112220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8472448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5115e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.012,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9703066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.32801532695466,
      "rewrite6_current_pass_fraction": 0.7791376993632062,
      "rewrite6_full_passes": 1.0,
      "step": 112230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8497024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.512e-05,
      "grad_norm": 0.578125,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0097,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9705066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.281314618105068,
      "rewrite6_current_pass_fraction": 0.7796732293431481,
      "rewrite6_full_passes": 1.0,
      "step": 112240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.85216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5125e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9947,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9708933333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.979365181141432,
      "rewrite6_current_pass_fraction": 0.7801892855056377,
      "rewrite6_full_passes": 1.0,
      "step": 112250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8546176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.513e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0059,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9712133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.2043918683445,
      "rewrite6_current_pass_fraction": 0.7808124476641156,
      "rewrite6_full_passes": 1.0,
      "step": 112260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8570752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5135e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0226,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.97144,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.544638363202598,
      "rewrite6_current_pass_fraction": 0.7812603454655216,
      "rewrite6_full_passes": 1.0,
      "step": 112270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8595328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.514e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.358192443847656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0152,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9716666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.39316916654633,
      "rewrite6_current_pass_fraction": 0.781902981441452,
      "rewrite6_full_passes": 1.0,
      "step": 112280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8619904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5145e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 14.670845031738281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0334,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.972,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.767722945893784,
      "rewrite6_current_pass_fraction": 0.7824287745126677,
      "rewrite6_full_passes": 1.0,
      "step": 112290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.864448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.515e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 14.681583404541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9994,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9723466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.073489215707433,
      "rewrite6_current_pass_fraction": 0.7829643044926097,
      "rewrite6_full_passes": 1.0,
      "step": 112300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8669056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5155e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 14.682418823242188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.022,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9727066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.53231527748008,
      "rewrite6_current_pass_fraction": 0.7835679928336352,
      "rewrite6_full_passes": 1.0,
      "step": 112310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8693632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.516e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 14.683574676513672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0348,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.97308,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.796818119887625,
      "rewrite6_current_pass_fraction": 0.7841668127202975,
      "rewrite6_full_passes": 1.0,
      "step": 112320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8718208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5165e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 14.68454360961914,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0316,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9733333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.730374668125208,
      "rewrite6_current_pass_fraction": 0.7847510272438706,
      "rewrite6_full_passes": 1.0,
      "step": 112330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8742784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.517e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 14.684738159179688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0198,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9736533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.487193835654615,
      "rewrite6_current_pass_fraction": 0.7852086619540029,
      "rewrite6_full_passes": 1.0,
      "step": 112340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.876736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5175e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.68524169921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0222,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.974,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.536422151209266,
      "rewrite6_current_pass_fraction": 0.7856760335728613,
      "rewrite6_full_passes": 1.0,
      "step": 112350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8791936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.518e-05,
      "grad_norm": 0.5546875,
      "host_rss_gib": 14.68524169921875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0171,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9744933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.431953020956982,
      "rewrite6_current_pass_fraction": 0.786158010554809,
      "rewrite6_full_passes": 1.0,
      "step": 112360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8816512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5185e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 14.685310363769531,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0271,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9749466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.637297562673467,
      "rewrite6_current_pass_fraction": 0.7867568304414715,
      "rewrite6_full_passes": 1.0,
      "step": 112370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8841088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.519e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 14.685962677001953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0216,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9752666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.52410399373532,
      "rewrite6_current_pass_fraction": 0.7872972288757765,
      "rewrite6_full_passes": 1.0,
      "step": 112380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8865664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5195e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.685962677001953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9756266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.926158934434543,
      "rewrite6_current_pass_fraction": 0.7878765749449864,
      "rewrite6_full_passes": 1.0,
      "step": 112390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.889024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.52e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 14.685962677001953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0443,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9758133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.99532932729427,
      "rewrite6_current_pass_fraction": 0.7884705263772858,
      "rewrite6_full_passes": 1.0,
      "step": 112400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8914816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5205e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.685962677001953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9760533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.7890742147183112,
      "rewrite6_full_passes": 1.0,
      "step": 112410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8939392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.521e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 14.685962677001953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0391,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9763866666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.886436980265266,
      "rewrite6_current_pass_fraction": 0.7897168506942416,
      "rewrite6_full_passes": 1.0,
      "step": 112420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8963968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5215e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.687122344970703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9964,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9766666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.013358988498783,
      "rewrite6_current_pass_fraction": 0.790271854491636,
      "rewrite6_full_passes": 1.0,
      "step": 112430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.8988544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.522e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 14.687122344970703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.977,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.738668476643575,
      "rewrite6_current_pass_fraction": 0.7908268582890304,
      "rewrite6_full_passes": 1.0,
      "step": 112440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.901312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5225e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 14.994152069091797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9879,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9772933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.843966375584888,
      "rewrite6_current_pass_fraction": 0.7914451519931452,
      "rewrite6_full_passes": 1.0,
      "step": 112450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9037696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.523e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.711917877197266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.014,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9775466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.368712040756797,
      "rewrite6_current_pass_fraction": 0.7919709450643609,
      "rewrite6_full_passes": 1.0,
      "step": 112460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9062272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5235e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.712100982666016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0341,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9778533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.782265441235463,
      "rewrite6_current_pass_fraction": 0.7925454226792078,
      "rewrite6_full_passes": 1.0,
      "step": 112470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9086848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.524e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.712100982666016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0436,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.97816,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.980637739420835,
      "rewrite6_current_pass_fraction": 0.7930955580222391,
      "rewrite6_full_passes": 1.0,
      "step": 112480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9111424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5245e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.712100982666016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0243,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9784933333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.579593952252267,
      "rewrite6_current_pass_fraction": 0.7936700356370859,
      "rewrite6_full_passes": 1.0,
      "step": 112490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.525e-05,
      "grad_norm": 0.4296875,
      "host_rss_gib": 15.712100982666016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.017,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9788666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.42990992781124,
      "rewrite6_current_pass_fraction": 0.794254250160659,
      "rewrite6_full_passes": 1.0,
      "step": 112500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.525e-05,
      "eval_loss": 3.0819971561431885,
      "eval_perplexity": 21.801900745247877,
      "eval_runtime": 0.0734,
      "eval_samples_per_second": 436.187,
      "eval_steps_per_second": 13.631,
      "step": 112500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9160576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5255e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.746658325195312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9982,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9791466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.04941547578139,
      "rewrite6_current_pass_fraction": 0.7947703063231485,
      "rewrite6_full_passes": 1.0,
      "step": 112510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9185152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.526e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.746658325195312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0253,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9794533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.600183839432283,
      "rewrite6_current_pass_fraction": 0.795417810753442,
      "rewrite6_full_passes": 1.0,
      "step": 112520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9209728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5265e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.746906280517578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0065,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9797066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.21651814098351,
      "rewrite6_current_pass_fraction": 0.7959776830051996,
      "rewrite6_full_passes": 1.0,
      "step": 112530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9234304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.527e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.746906280517578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9799466666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.69516296961328,
      "rewrite6_current_pass_fraction": 0.796488870713326,
      "rewrite6_full_passes": 1.0,
      "step": 112540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.925888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5275e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.746906280517578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0247,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9804533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.58782743642022,
      "rewrite6_current_pass_fraction": 0.7971120328718039,
      "rewrite6_full_passes": 1.0,
      "step": 112550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9283456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.528e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.746906280517578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0092,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9807333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.271176495537865,
      "rewrite6_current_pass_fraction": 0.7977157212128293,
      "rewrite6_full_passes": 1.0,
      "step": 112560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9308032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5285e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.747241973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9943,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.98104,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.971375033205096,
      "rewrite6_current_pass_fraction": 0.7982755934645869,
      "rewrite6_full_passes": 1.0,
      "step": 112570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9332608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.529e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.747241973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9813733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.188234818531463,
      "rewrite6_current_pass_fraction": 0.7988549395337968,
      "rewrite6_full_passes": 1.0,
      "step": 112580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9357184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5295e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.747241973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0007,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.98164,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.09960172113887,
      "rewrite6_current_pass_fraction": 0.7994488909660961,
      "rewrite6_full_passes": 1.0,
      "step": 112590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.938176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.53e-05,
      "grad_norm": 0.5390625,
      "host_rss_gib": 15.747241973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9820266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.8000574477614847,
      "rewrite6_full_passes": 1.0,
      "step": 112600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9406336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5305e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.747241973876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0262,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9822133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.618732350465706,
      "rewrite6_current_pass_fraction": 0.8006660045568733,
      "rewrite6_full_passes": 1.0,
      "step": 112610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9430912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.531e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9825333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.54258400208604,
      "rewrite6_current_pass_fraction": 0.8012307452629939,
      "rewrite6_full_passes": 1.0,
      "step": 112620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9455488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5315e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0265,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9827733333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.624918898106596,
      "rewrite6_current_pass_fraction": 0.8017321960623941,
      "rewrite6_full_passes": 1.0,
      "step": 112630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9480064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.532e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0261,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.983,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.61667058032089,
      "rewrite6_current_pass_fraction": 0.8022725944966992,
      "rewrite6_full_passes": 1.0,
      "step": 112640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.950464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5325e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0344,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9833466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.788501056163312,
      "rewrite6_current_pass_fraction": 0.8027643083873732,
      "rewrite6_full_passes": 1.0,
      "step": 112650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9529216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.533e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0109,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9835733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.30566680403609,
      "rewrite6_current_pass_fraction": 0.8033679967283986,
      "rewrite6_full_passes": 1.0,
      "step": 112660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9553792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5335e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0379,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9838,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.861388288110085,
      "rewrite6_current_pass_fraction": 0.8039132636170668,
      "rewrite6_full_passes": 1.0,
      "step": 112670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9578368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.534e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0117,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.98404,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.32191783702579,
      "rewrite6_current_pass_fraction": 0.8044877412319137,
      "rewrite6_full_passes": 1.0,
      "step": 112680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9602944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5345e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0097,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9843733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.281314618105068,
      "rewrite6_current_pass_fraction": 0.8050135343031295,
      "rewrite6_full_passes": 1.0,
      "step": 112690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.962752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.535e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9846,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.54258400208604,
      "rewrite6_current_pass_fraction": 0.8055588011917977,
      "rewrite6_full_passes": 1.0,
      "step": 112700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9652096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5355e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0006,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.98496,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.097591861461414,
      "rewrite6_current_pass_fraction": 0.8060651204455609,
      "rewrite6_full_passes": 1.0,
      "step": 112710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9676672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.536e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.747295379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0212,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.98532,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.515895993847245,
      "rewrite6_current_pass_fraction": 0.8065957819711398,
      "rewrite6_full_passes": 1.0,
      "step": 112720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9701248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5365e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.481670379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9857333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.78642230999673,
      "rewrite6_current_pass_fraction": 0.8071507857685342,
      "rewrite6_full_passes": 1.0,
      "step": 112730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9725824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.537e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.481670379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0222,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9861066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.536422151209266,
      "rewrite6_current_pass_fraction": 0.8078080271075538,
      "rewrite6_full_passes": 1.0,
      "step": 112740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.97504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5375e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.481670379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9889,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9865333333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.86382026725182,
      "rewrite6_current_pass_fraction": 0.8084117154485794,
      "rewrite6_full_passes": 1.0,
      "step": 112750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9774976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.538e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.481670379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9868133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.8089813246090631,
      "rewrite6_full_passes": 1.0,
      "step": 112760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9799552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5385e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.481670379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0132,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.98708,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.352423587374265,
      "rewrite6_current_pass_fraction": 0.809604486767541,
      "rewrite6_full_passes": 1.0,
      "step": 112770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9824128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.539e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.481670379638672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0223,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.98732,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.538475896109915,
      "rewrite6_current_pass_fraction": 0.810183832836751,
      "rewrite6_full_passes": 1.0,
      "step": 112780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9848704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5395e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9876,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9876133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.83801407856141,
      "rewrite6_current_pass_fraction": 0.8109141009912173,
      "rewrite6_full_passes": 1.0,
      "step": 112790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.987328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.54e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0317,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9879333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.732447809247343,
      "rewrite6_current_pass_fraction": 0.8115129208778797,
      "rewrite6_full_passes": 1.0,
      "step": 112800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9897856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5405e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.004,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.98832,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.16603996963595,
      "rewrite6_current_pass_fraction": 0.8121117407645421,
      "rewrite6_full_passes": 1.0,
      "step": 112810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9922432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.541e-05,
      "grad_norm": 0.431640625,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0219,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9886266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.53026214861049,
      "rewrite6_current_pass_fraction": 0.812691086833752,
      "rewrite6_full_passes": 1.0,
      "step": 112820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9947008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5415e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0098,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9891733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.28334285097683,
      "rewrite6_current_pass_fraction": 0.8132753013573251,
      "rewrite6_full_passes": 1.0,
      "step": 112830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.9971584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.542e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0095,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9894266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.2772587607807,
      "rewrite6_current_pass_fraction": 0.8137816206110884,
      "rewrite6_full_passes": 1.0,
      "step": 112840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 21.999616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5425e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9898,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.13647032815547,
      "rewrite6_current_pass_fraction": 0.8143707035890245,
      "rewrite6_full_passes": 1.0,
      "step": 112850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0020736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.543e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0131,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9901866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.350388446774257,
      "rewrite6_current_pass_fraction": 0.8148234698447937,
      "rewrite6_full_passes": 1.0,
      "step": 112860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0045312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5435e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0038,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9905466666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.162007164935936,
      "rewrite6_current_pass_fraction": 0.815329789098557,
      "rewrite6_full_passes": 1.0,
      "step": 112870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0069888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.544e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9745,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9908266666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.57983088626603,
      "rewrite6_current_pass_fraction": 0.815909135167767,
      "rewrite6_full_passes": 1.0,
      "step": 112880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0094464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5445e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0284,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9912133333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.664143495580532,
      "rewrite6_current_pass_fraction": 0.8165030866000662,
      "rewrite6_full_passes": 1.0,
      "step": 112890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.011904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.545e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0212,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9915066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.515895993847245,
      "rewrite6_current_pass_fraction": 0.81707269576055,
      "rewrite6_full_passes": 1.0,
      "step": 112900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0143616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5455e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0297,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99184,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.691024350895024,
      "rewrite6_current_pass_fraction": 0.8176471733753968,
      "rewrite6_full_passes": 1.0,
      "step": 112910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0168192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.546e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9921066666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.64348968071354,
      "rewrite6_current_pass_fraction": 0.8182411248076961,
      "rewrite6_full_passes": 1.0,
      "step": 112920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0192768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5465e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0108,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9922666666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.30363633888064,
      "rewrite6_current_pass_fraction": 0.8187279702440069,
      "rewrite6_full_passes": 1.0,
      "step": 112930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0217344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.547e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9943,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9925066666666666,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.971375033205096,
      "rewrite6_current_pass_fraction": 0.8193121847675799,
      "rewrite6_full_passes": 1.0,
      "step": 112940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.024192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5475e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0354,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9928666666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.809299954935618,
      "rewrite6_current_pass_fraction": 0.8198282409300696,
      "rewrite6_full_passes": 1.0,
      "step": 112950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0266496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.548e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0094,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99316,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.275231136287534,
      "rewrite6_current_pass_fraction": 0.8203686393643747,
      "rewrite6_full_passes": 1.0,
      "step": 112960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0291072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5485e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0096,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99352,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.279286588046446,
      "rewrite6_current_pass_fraction": 0.820967459251037,
      "rewrite6_full_passes": 1.0,
      "step": 112970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0315648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.549e-05,
      "grad_norm": 0.59765625,
      "host_rss_gib": 15.482128143310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0086,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9938533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.25901743772266,
      "rewrite6_current_pass_fraction": 0.8215468053202469,
      "rewrite6_full_passes": 1.0,
      "step": 112980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0340224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5495e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.482425689697266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.052,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9940533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 21.157617370242413,
      "rewrite6_current_pass_fraction": 0.822179704387451,
      "rewrite6_full_passes": 1.0,
      "step": 112990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.03648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.55e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.482425689697266,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0384,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99436,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.871821590362348,
      "rewrite6_current_pass_fraction": 0.8229002356331911,
      "rewrite6_full_passes": 1.0,
      "step": 113000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.55e-05,
      "eval_loss": 3.0817556381225586,
      "eval_perplexity": 21.79663582914557,
      "eval_runtime": 0.0492,
      "eval_samples_per_second": 650.456,
      "eval_steps_per_second": 20.327,
      "step": 113000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0389376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5505e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0196,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9946533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.48309680660405,
      "rewrite6_current_pass_fraction": 0.8235087924285798,
      "rewrite6_full_passes": 1.0,
      "step": 113010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0413952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.551e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0233,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99488,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.559024644667907,
      "rewrite6_current_pass_fraction": 0.8241076123152422,
      "rewrite6_full_passes": 1.0,
      "step": 113020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0438528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5515e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0154,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9952533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.397248208270213,
      "rewrite6_current_pass_fraction": 0.8246480107495472,
      "rewrite6_full_passes": 1.0,
      "step": 113030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0463104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.552e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0114,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9955733333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.31582217606955,
      "rewrite6_current_pass_fraction": 0.8253588050865611,
      "rewrite6_full_passes": 1.0,
      "step": 113040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.048768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5525e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0006,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99588,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.097591861461414,
      "rewrite6_current_pass_fraction": 0.8258992035208662,
      "rewrite6_full_passes": 1.0,
      "step": 113050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0512256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.553e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0058,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9962,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.202371530176254,
      "rewrite6_current_pass_fraction": 0.8264347335008082,
      "rewrite6_full_passes": 1.0,
      "step": 113060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0536832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5535e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0223,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99652,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.538475896109915,
      "rewrite6_current_pass_fraction": 0.8270043426612919,
      "rewrite6_full_passes": 1.0,
      "step": 113070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0561408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.554e-05,
      "grad_norm": 0.4296875,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9923,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9968533333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 19.931472199273568,
      "rewrite6_current_pass_fraction": 0.8276518470915853,
      "rewrite6_full_passes": 1.0,
      "step": 113080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0585984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5545e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0292,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99708,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.680681424666606,
      "rewrite6_current_pass_fraction": 0.8282263247064322,
      "rewrite6_full_passes": 1.0,
      "step": 113090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.061056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.555e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0114,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9974133333333334,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.31582217606955,
      "rewrite6_current_pass_fraction": 0.8287618546863742,
      "rewrite6_full_passes": 1.0,
      "step": 113100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0635136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5555e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.027,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99768,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.635233936100253,
      "rewrite6_current_pass_fraction": 0.8293460692099472,
      "rewrite6_full_passes": 1.0,
      "step": 113110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0659712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.556e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0106,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99796,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.29957601765852,
      "rewrite6_current_pass_fraction": 0.8299448890966096,
      "rewrite6_full_passes": 1.0,
      "step": 113120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0684288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5565e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0195,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.99832,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.481048599335452,
      "rewrite6_current_pass_fraction": 0.8304804190765516,
      "rewrite6_full_passes": 1.0,
      "step": 113130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0708864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.557e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.482501983642578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0065,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9986,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.21651814098351,
      "rewrite6_current_pass_fraction": 0.8310208175108567,
      "rewrite6_full_passes": 1.0,
      "step": 113140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.073344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5575e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 14.887500762939453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0185,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9989333333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.460577787847765,
      "rewrite6_current_pass_fraction": 0.8316001635800666,
      "rewrite6_full_passes": 1.0,
      "step": 113150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0758016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.558e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 14.88875961303711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9978,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9992266666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.041397313330467,
      "rewrite6_current_pass_fraction": 0.8320431929271095,
      "rewrite6_full_passes": 1.0,
      "step": 113160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0782592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5585e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 14.88875961303711,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0152,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9994533333333333,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.39316916654633,
      "rewrite6_current_pass_fraction": 0.8326176705419563,
      "rewrite6_full_passes": 1.0,
      "step": 113170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0807168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.559e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.89215087890625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0029,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.9997866666666667,
      "no_prompt_full_passes": 0.0,
      "perplexity": 20.143869521651258,
      "rewrite6_current_pass_fraction": 0.8332067535198925,
      "rewrite6_full_passes": 1.0,
      "step": 113180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0831744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5595e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 14.89410400390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0118,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 4e-05,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.323950130422475,
      "rewrite6_current_pass_fraction": 0.8338834686763646,
      "rewrite6_full_passes": 1.0,
      "step": 113190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.085632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.56e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 14.894111633300781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0282,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0002933333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.660011080136737,
      "rewrite6_current_pass_fraction": 0.8343557087495862,
      "rewrite6_full_passes": 1.0,
      "step": 113200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0880896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5605e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 14.89571762084961,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0213,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0006,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.517947686029533,
      "rewrite6_current_pass_fraction": 0.8348815018208019,
      "rewrite6_full_passes": 1.0,
      "step": 113210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0905472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.561e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 14.896064758300781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0379,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0008533333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.861388288110085,
      "rewrite6_current_pass_fraction": 0.8353440049852973,
      "rewrite6_full_passes": 1.0,
      "step": 113220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0930048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5615e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 14.896064758300781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0135,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00124,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.358530230401133,
      "rewrite6_current_pass_fraction": 0.8357578236061615,
      "rewrite6_full_passes": 1.0,
      "step": 113230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.0954624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.562e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 14.896064758300781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0364,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00156,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.83011966300961,
      "rewrite6_current_pass_fraction": 0.8364345387626336,
      "rewrite6_full_passes": 1.0,
      "step": 113240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.09792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5625e-05,
      "grad_norm": 0.53515625,
      "host_rss_gib": 14.896797180175781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0258,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00196,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.610486506804197,
      "rewrite6_current_pass_fraction": 0.8370479640123853,
      "rewrite6_full_passes": 1.0,
      "step": 113250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1003776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.563e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.896797180175781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0251,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0022933333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.596064214640606,
      "rewrite6_current_pass_fraction": 0.8376273100815953,
      "rewrite6_full_passes": 1.0,
      "step": 113260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1028352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5635e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.896797180175781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0002,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0027866666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.089554432309825,
      "rewrite6_current_pass_fraction": 0.8381871823333528,
      "rewrite6_full_passes": 1.0,
      "step": 113270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1052928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.564e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 14.89859390258789,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0030533333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.926158934434543,
      "rewrite6_current_pass_fraction": 0.838688633132753,
      "rewrite6_full_passes": 1.0,
      "step": 113280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1077504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5645e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 14.899093627929688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0108,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0033066666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.30363633888064,
      "rewrite6_current_pass_fraction": 0.8393799536523144,
      "rewrite6_full_passes": 1.0,
      "step": 113290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.110208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.565e-05,
      "grad_norm": 0.56640625,
      "host_rss_gib": 14.33011245727539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0287,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0035066666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.670343668608663,
      "rewrite6_current_pass_fraction": 0.8399008782691672,
      "rewrite6_full_passes": 1.0,
      "step": 113300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1126656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5655e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.542888641357422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0205,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.003746666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.501539891873456,
      "rewrite6_current_pass_fraction": 0.8404510136121984,
      "rewrite6_full_passes": 1.0,
      "step": 113310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1151232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.566e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 16.114517211914062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0041466666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.511793224939044,
      "rewrite6_current_pass_fraction": 0.841015754318319,
      "rewrite6_full_passes": 1.0,
      "step": 113320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1175808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5665e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.114524841308594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0198,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.004506666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.487193835654615,
      "rewrite6_current_pass_fraction": 0.8415902319331658,
      "rewrite6_full_passes": 1.0,
      "step": 113330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1200384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.567e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.11453628540039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0266,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0048,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.626981493124443,
      "rewrite6_current_pass_fraction": 0.8421890518198283,
      "rewrite6_full_passes": 1.0,
      "step": 113340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.122496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5675e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.11453628540039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0321,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0051333333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.740742447188033,
      "rewrite6_current_pass_fraction": 0.8427586609803119,
      "rewrite6_full_passes": 1.0,
      "step": 113350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1249536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.568e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.11453628540039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0207,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00532,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.505640609909964,
      "rewrite6_current_pass_fraction": 0.8432844540515277,
      "rewrite6_full_passes": 1.0,
      "step": 113360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1274112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5685e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.11453628540039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0137,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00572,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.362602343644966,
      "rewrite6_current_pass_fraction": 0.8438540632120114,
      "rewrite6_full_passes": 1.0,
      "step": 113370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1298688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.569e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.11453628540039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.006093333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.819707206509154,
      "rewrite6_current_pass_fraction": 0.8444869622792156,
      "rewrite6_full_passes": 1.0,
      "step": 113380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1323264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5695e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.11453628540039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.006333333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.8449640708068002,
      "rewrite6_full_passes": 1.0,
      "step": 113390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.134784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.57e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 16.11648941040039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.819707206509154,
      "rewrite6_current_pass_fraction": 0.8454314424256587,
      "rewrite6_full_passes": 1.0,
      "step": 113400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1372416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5705e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.11648941040039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0179,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00696,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.44830512334259,
      "rewrite6_current_pass_fraction": 0.8459328932250589,
      "rewrite6_full_passes": 1.0,
      "step": 113410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1396992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.571e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.116493225097656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.022,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.007186666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53231527748008,
      "rewrite6_current_pass_fraction": 0.8465122392942689,
      "rewrite6_full_passes": 1.0,
      "step": 113420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1421568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5715e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.116493225097656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0206,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.007493333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.503590148373757,
      "rewrite6_current_pass_fraction": 0.8470721115460264,
      "rewrite6_full_passes": 1.0,
      "step": 113430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1446144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.572e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.116493225097656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0248,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.00776,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.589886322106427,
      "rewrite6_current_pass_fraction": 0.8475735623454266,
      "rewrite6_full_passes": 1.0,
      "step": 113440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.147072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5725e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.116493225097656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0074,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.008106666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.234721197457105,
      "rewrite6_current_pass_fraction": 0.8481480399602734,
      "rewrite6_full_passes": 1.0,
      "step": 113450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1495296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.573e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.116493225097656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0182,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.008493333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.45444053514535,
      "rewrite6_current_pass_fraction": 0.8486397538509474,
      "rewrite6_full_passes": 1.0,
      "step": 113460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1519872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5735e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.116493225097656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0311,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.008733333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.720012071656146,
      "rewrite6_current_pass_fraction": 0.8492726529181516,
      "rewrite6_full_passes": 1.0,
      "step": 113470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1544448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.574e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.116493225097656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.009093333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.180161139447517,
      "rewrite6_current_pass_fraction": 0.8498179198068198,
      "rewrite6_full_passes": 1.0,
      "step": 113480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1569024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5745e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.116493225097656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0146,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.009426666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.380936935082808,
      "rewrite6_current_pass_fraction": 0.8503388444236724,
      "rewrite6_full_passes": 1.0,
      "step": 113490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.15936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.575e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.116493225097656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.009666666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.844705851344035,
      "rewrite6_current_pass_fraction": 0.8509084535841561,
      "rewrite6_full_passes": 1.0,
      "step": 113500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.575e-05,
      "eval_loss": 3.0825860500335693,
      "eval_perplexity": 21.81474353254378,
      "eval_runtime": 0.0504,
      "eval_samples_per_second": 635.059,
      "eval_steps_per_second": 19.846,
      "step": 113500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1618176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5755e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0144,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.010106666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.376861155287358,
      "rewrite6_current_pass_fraction": 0.8514877996533661,
      "rewrite6_full_passes": 1.0,
      "step": 113510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1642752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.576e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0136,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.010373333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.360566185220215,
      "rewrite6_current_pass_fraction": 0.8520379349963974,
      "rewrite6_full_passes": 1.0,
      "step": 113520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1667328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5765e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0031,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.010666666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.147898698459837,
      "rewrite6_current_pass_fraction": 0.8526416233374229,
      "rewrite6_full_passes": 1.0,
      "step": 113530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1691904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.577e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.010853333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.8532112324979065,
      "rewrite6_full_passes": 1.0,
      "step": 113540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.171648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5775e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0148,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.011106666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.38501353011574,
      "rewrite6_current_pass_fraction": 0.8537613678409378,
      "rewrite6_full_passes": 1.0,
      "step": 113550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1741056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.578e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01144,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.819707206509154,
      "rewrite6_current_pass_fraction": 0.8543504508188741,
      "rewrite6_full_passes": 1.0,
      "step": 113560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1765632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5785e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.022,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.011693333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53231527748008,
      "rewrite6_current_pass_fraction": 0.8549882183404413,
      "rewrite6_full_passes": 1.0,
      "step": 113570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1790208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.579e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.011933333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.8555773013183774,
      "rewrite6_full_passes": 1.0,
      "step": 113580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1814784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5795e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0534,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.012226666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.187258778705246,
      "rewrite6_current_pass_fraction": 0.8561566473875873,
      "rewrite6_full_passes": 1.0,
      "step": 113590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.183936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.58e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01256,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.23269782650759,
      "rewrite6_current_pass_fraction": 0.8566532297326245,
      "rewrite6_full_passes": 1.0,
      "step": 113600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1863936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5805e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.016,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.012893333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.409490229434265,
      "rewrite6_current_pass_fraction": 0.8572228388931082,
      "rewrite6_full_passes": 1.0,
      "step": 113610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1888512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.581e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0149,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01316,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.387052133397212,
      "rewrite6_current_pass_fraction": 0.8578654748690385,
      "rewrite6_full_passes": 1.0,
      "step": 113620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1913088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5815e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0201,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.013546666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.49334091582123,
      "rewrite6_current_pass_fraction": 0.8584545578469748,
      "rewrite6_full_passes": 1.0,
      "step": 113630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1937664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.582e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0354,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.013933333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.809299954935618,
      "rewrite6_current_pass_fraction": 0.8590241670074584,
      "rewrite6_full_passes": 1.0,
      "step": 113640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.196224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5825e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0315,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0142,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.728301734306807,
      "rewrite6_current_pass_fraction": 0.8595596969874004,
      "rewrite6_full_passes": 1.0,
      "step": 113650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.1986816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.583e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0329,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.014626666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.757341679953605,
      "rewrite6_current_pass_fraction": 0.8601098323304317,
      "rewrite6_full_passes": 1.0,
      "step": 113660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2011392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5835e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0251,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01492,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.596064214640606,
      "rewrite6_current_pass_fraction": 0.860659967673463,
      "rewrite6_full_passes": 1.0,
      "step": 113670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2035968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.584e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0054,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.015266666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.194292197538438,
      "rewrite6_current_pass_fraction": 0.8612441821970361,
      "rewrite6_full_passes": 1.0,
      "step": 113680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2060544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5845e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0301,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01564,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.699302416138053,
      "rewrite6_current_pass_fraction": 0.8617845806313411,
      "rewrite6_full_passes": 1.0,
      "step": 113690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.208512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.585e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0364,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.015946666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.83011966300961,
      "rewrite6_current_pass_fraction": 0.8624369535159977,
      "rewrite6_full_passes": 1.0,
      "step": 113700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2109696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5855e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0134,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01632,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.35649447916735,
      "rewrite6_current_pass_fraction": 0.8630357734026601,
      "rewrite6_full_passes": 1.0,
      "step": 113710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2134272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.586e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0091,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01670666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.26914947924082,
      "rewrite6_current_pass_fraction": 0.8635859087456914,
      "rewrite6_full_passes": 1.0,
      "step": 113720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2158848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5865e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0064,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01688,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.214496590248636,
      "rewrite6_current_pass_fraction": 0.8641311756343596,
      "rewrite6_full_passes": 1.0,
      "step": 113730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2183424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.587e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0104,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.017226666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.295516508419446,
      "rewrite6_current_pass_fraction": 0.8646910478861172,
      "rewrite6_full_passes": 1.0,
      "step": 113740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5875e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9991,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.017546666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.067468072159407,
      "rewrite6_current_pass_fraction": 0.8652071040486067,
      "rewrite6_full_passes": 1.0,
      "step": 113750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2232576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.588e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0515,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01792,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.14704120581873,
      "rewrite6_current_pass_fraction": 0.865757239391638,
      "rewrite6_full_passes": 1.0,
      "step": 113760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2257152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5885e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.184688568115234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0182,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.976442031457733,
      "rewrite6_current_pass_fraction": 0.8663706646413897,
      "rewrite6_full_passes": 1.0,
      "step": 113770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2281728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.589e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0234,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.018626666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.56108064993093,
      "rewrite6_current_pass_fraction": 0.866969484528052,
      "rewrite6_full_passes": 1.0,
      "step": 113780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2306304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5895e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0574,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.018893333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.2721775381139,
      "rewrite6_current_pass_fraction": 0.8675001460536309,
      "rewrite6_full_passes": 1.0,
      "step": 113790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.233088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.59e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.019266666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.180161139447517,
      "rewrite6_current_pass_fraction": 0.8681135713033826,
      "rewrite6_full_passes": 1.0,
      "step": 113800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2355456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5905e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.01965333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.89061468541791,
      "rewrite6_current_pass_fraction": 0.8686539697376877,
      "rewrite6_full_passes": 1.0,
      "step": 113810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2380032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.591e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0251,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.019973333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.596064214640606,
      "rewrite6_current_pass_fraction": 0.8691748943545403,
      "rewrite6_full_passes": 1.0,
      "step": 113820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2404608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5915e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0453,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0202,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.016335157786333,
      "rewrite6_current_pass_fraction": 0.8697298981519347,
      "rewrite6_full_passes": 1.0,
      "step": 113830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2429184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.592e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0061,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02056,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.20843315083295,
      "rewrite6_current_pass_fraction": 0.8703287180385971,
      "rewrite6_full_passes": 1.0,
      "step": 113840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.245376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5925e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0239,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02084,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.57136376081938,
      "rewrite6_current_pass_fraction": 0.8708739849272653,
      "rewrite6_full_passes": 1.0,
      "step": 113850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2478336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.593e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0168,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.021093333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.425824354396642,
      "rewrite6_current_pass_fraction": 0.8715068839944694,
      "rewrite6_full_passes": 1.0,
      "step": 113860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2502912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5935e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0237,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.021453333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.567249899467065,
      "rewrite6_current_pass_fraction": 0.8721057038811318,
      "rewrite6_full_passes": 1.0,
      "step": 113870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2527488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.594e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.021706666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.511793224939044,
      "rewrite6_current_pass_fraction": 0.8727093922221573,
      "rewrite6_full_passes": 1.0,
      "step": 113880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2552064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5945e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0346,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.021866666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.792659172172286,
      "rewrite6_current_pass_fraction": 0.87323031683901,
      "rewrite6_full_passes": 1.0,
      "step": 113890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.257664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.595e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.022306666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.898972602764104,
      "rewrite6_current_pass_fraction": 0.8736976884578684,
      "rewrite6_full_passes": 1.0,
      "step": 113900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2601216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5955e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.022613333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.959667588504388,
      "rewrite6_current_pass_fraction": 0.8743159821619833,
      "rewrite6_full_passes": 1.0,
      "step": 113910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2625792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.596e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0006,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.023,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.097591861461414,
      "rewrite6_current_pass_fraction": 0.8750170395902709,
      "rewrite6_full_passes": 1.0,
      "step": 113920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2650368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5965e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0275,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.023306666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.645554132902497,
      "rewrite6_current_pass_fraction": 0.8756012541138439,
      "rewrite6_full_passes": 1.0,
      "step": 113930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2674944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.597e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0009,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0236,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.103622043501936,
      "rewrite6_current_pass_fraction": 0.8762779692703161,
      "rewrite6_full_passes": 1.0,
      "step": 113940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.269952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5975e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0169,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.023933333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.42786703896461,
      "rewrite6_current_pass_fraction": 0.8767940254328056,
      "rewrite6_full_passes": 1.0,
      "step": 113950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2724096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.598e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0104,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0242,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.295516508419446,
      "rewrite6_current_pass_fraction": 0.877256528597301,
      "rewrite6_full_passes": 1.0,
      "step": 113960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2748672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5985e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02444,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.8778115323946953,
      "rewrite6_full_passes": 1.0,
      "step": 113970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2773248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.599e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.024613333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.8783957469182684,
      "rewrite6_full_passes": 1.0,
      "step": 113980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2797824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.5995e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9984,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.024906666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.05342575989159,
      "rewrite6_current_pass_fraction": 0.8789994352592939,
      "rewrite6_full_passes": 1.0,
      "step": 113990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.28224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.185043334960938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0118,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.025173333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.323950130422475,
      "rewrite6_current_pass_fraction": 0.8795106229674203,
      "rewrite6_full_passes": 1.0,
      "step": 114000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.6e-05,
      "eval_loss": 3.0821659564971924,
      "eval_perplexity": 21.80558122443599,
      "eval_runtime": 0.0493,
      "eval_samples_per_second": 649.745,
      "eval_steps_per_second": 20.305,
      "step": 114000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2846976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6005e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.25707244873047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0108,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.025573333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.30363633888064,
      "rewrite6_current_pass_fraction": 0.8801337851258982,
      "rewrite6_full_passes": 1.0,
      "step": 114010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2871552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.601e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.600044250488281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0218,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02588,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.528209225043515,
      "rewrite6_current_pass_fraction": 0.8807374734669238,
      "rewrite6_full_passes": 1.0,
      "step": 114020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2896128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6015e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.600677490234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0272,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02616,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.639361395619666,
      "rewrite6_current_pass_fraction": 0.8813119510817705,
      "rewrite6_full_passes": 1.0,
      "step": 114030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2920704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.602e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.600749969482422,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0198,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02644,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.487193835654615,
      "rewrite6_current_pass_fraction": 0.8818426126073494,
      "rewrite6_full_passes": 1.0,
      "step": 114040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.294528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6025e-05,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.601066589355469,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0133,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.026773333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.354458931498517,
      "rewrite6_current_pass_fraction": 0.8824706432201904,
      "rewrite6_full_passes": 1.0,
      "step": 114050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2969856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.603e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.604045867919922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.993,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.027106666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.945429114163357,
      "rewrite6_current_pass_fraction": 0.8830159101088586,
      "rewrite6_full_passes": 1.0,
      "step": 114060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.2994432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6035e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.60751724243164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0264,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.027466666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.622856509337947,
      "rewrite6_current_pass_fraction": 0.8835417031800744,
      "rewrite6_full_passes": 1.0,
      "step": 114070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3019008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.604e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.60751724243164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0181,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.027853333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.452395193360626,
      "rewrite6_current_pass_fraction": 0.8841989445190941,
      "rewrite6_full_passes": 1.0,
      "step": 114080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3043584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6045e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.609874725341797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0128,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.028186666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.344284245916132,
      "rewrite6_current_pass_fraction": 0.88486592276684,
      "rewrite6_full_passes": 1.0,
      "step": 114090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.306816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.605e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.609954833984375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0008,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.028706666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.10161178181234,
      "rewrite6_current_pass_fraction": 0.8854793480165917,
      "rewrite6_full_passes": 1.0,
      "step": 114100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3092736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6055e-05,
      "grad_norm": 0.431640625,
      "host_rss_gib": 15.60995864868164,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.029026666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.188234818531463,
      "rewrite6_current_pass_fraction": 0.8860100095421706,
      "rewrite6_full_passes": 1.0,
      "step": 114110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3117312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.606e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.610443115234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0261,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02932,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.61667058032089,
      "rewrite6_current_pass_fraction": 0.8866234347919223,
      "rewrite6_full_passes": 1.0,
      "step": 114120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3141888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6065e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.612396240234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.029586666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.71586848361443,
      "rewrite6_current_pass_fraction": 0.887329360674573,
      "rewrite6_full_passes": 1.0,
      "step": 114130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3166464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.607e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.612396240234375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0211,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.02988,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.513844506823926,
      "rewrite6_current_pass_fraction": 0.887908706743783,
      "rewrite6_full_passes": 1.0,
      "step": 114140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.319104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6075e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 14.712074279785156,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.012,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.030186666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.32801532695466,
      "rewrite6_current_pass_fraction": 0.888492921267356,
      "rewrite6_full_passes": 1.0,
      "step": 114150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3215616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.608e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.573925018310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0042,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.030373333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.170073580977565,
      "rewrite6_current_pass_fraction": 0.889028451247298,
      "rewrite6_full_passes": 1.0,
      "step": 114160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3240192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6085e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.33905792236328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0393,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.030706666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.89061468541791,
      "rewrite6_current_pass_fraction": 0.8895347705010613,
      "rewrite6_full_passes": 1.0,
      "step": 114170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3264768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.609e-05,
      "grad_norm": 0.5546875,
      "host_rss_gib": 17.651649475097656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0176,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03092,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.442171551887302,
      "rewrite6_current_pass_fraction": 0.889968062939378,
      "rewrite6_full_passes": 1.0,
      "step": 114180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3289344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6095e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 17.718334197998047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0304,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.031186666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.705513138424664,
      "rewrite6_current_pass_fraction": 0.8904646452844152,
      "rewrite6_full_passes": 1.0,
      "step": 114190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.331392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.61e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 17.71837615966797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0396,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03152,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.896882809995212,
      "rewrite6_current_pass_fraction": 0.8910099121730833,
      "rewrite6_full_passes": 1.0,
      "step": 114200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3338496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6105e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 17.71837615966797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.025,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.031773333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.59400471119603,
      "rewrite6_current_pass_fraction": 0.891535705244299,
      "rewrite6_full_passes": 1.0,
      "step": 114210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3363072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.611e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 17.71837615966797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0332,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03202666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.763569816631378,
      "rewrite6_current_pass_fraction": 0.8921199197678721,
      "rewrite6_full_passes": 1.0,
      "step": 114220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3387648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6115e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 17.71837615966797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0207,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03238666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.505640609909964,
      "rewrite6_current_pass_fraction": 0.8926749235652665,
      "rewrite6_full_passes": 1.0,
      "step": 114230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3412224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.612e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 17.718402862548828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.022,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03276,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53231527748008,
      "rewrite6_current_pass_fraction": 0.8932299273626609,
      "rewrite6_full_passes": 1.0,
      "step": 114240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.34368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6125e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 17.718402862548828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0168,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.033093333333333336,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.425824354396642,
      "rewrite6_current_pass_fraction": 0.8938482210667757,
      "rewrite6_full_passes": 1.0,
      "step": 114250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3461376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.613e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 17.71979522705078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0178,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.033506666666666664,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.446260395068368,
      "rewrite6_current_pass_fraction": 0.8944713832252537,
      "rewrite6_full_passes": 1.0,
      "step": 114260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3485952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6135e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 17.71979522705078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0482,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.033826666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.07737098892312,
      "rewrite6_current_pass_fraction": 0.8949923078421063,
      "rewrite6_full_passes": 1.0,
      "step": 114270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3510528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.614e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 17.71979522705078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0452,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03409333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.014233629348723,
      "rewrite6_current_pass_fraction": 0.8955375747307744,
      "rewrite6_full_passes": 1.0,
      "step": 114280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3535104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6145e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 17.71979522705078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0029,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03442666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.143869521651258,
      "rewrite6_current_pass_fraction": 0.8961802107067048,
      "rewrite6_full_passes": 1.0,
      "step": 114290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.355968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.615e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 17.71979522705078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.034813333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.828036755190443,
      "rewrite6_current_pass_fraction": 0.8968131097739089,
      "rewrite6_full_passes": 1.0,
      "step": 114300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3584256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6155e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0163,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.035013333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.415613995022007,
      "rewrite6_current_pass_fraction": 0.897353508208214,
      "rewrite6_full_passes": 1.0,
      "step": 114310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3608832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.616e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0287,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03544,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.670343668608663,
      "rewrite6_current_pass_fraction": 0.8979669334579657,
      "rewrite6_full_passes": 1.0,
      "step": 114320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3633408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6165e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.999,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03577333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.06546142568619,
      "rewrite6_current_pass_fraction": 0.8986582539775272,
      "rewrite6_full_passes": 1.0,
      "step": 114330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3657984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.617e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0585,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.036026666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.295589807793416,
      "rewrite6_current_pass_fraction": 0.8992668107729158,
      "rewrite6_full_passes": 1.0,
      "step": 114340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.368256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6175e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0143,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03625333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.374823571052737,
      "rewrite6_current_pass_fraction": 0.8998510252964889,
      "rewrite6_full_passes": 1.0,
      "step": 114350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3707136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.618e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9839,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.036706666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.764749050356087,
      "rewrite6_current_pass_fraction": 0.9004303713656988,
      "rewrite6_full_passes": 1.0,
      "step": 114360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3731712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6185e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0218,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03708,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.528209225043515,
      "rewrite6_current_pass_fraction": 0.9010097174349088,
      "rewrite6_full_passes": 1.0,
      "step": 114370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3756288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.619e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0243,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03741333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.579593952252267,
      "rewrite6_current_pass_fraction": 0.9015355105061246,
      "rewrite6_full_passes": 1.0,
      "step": 114380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3780864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6195e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0093,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03784,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.273203714546685,
      "rewrite6_current_pass_fraction": 0.9021927518451442,
      "rewrite6_full_passes": 1.0,
      "step": 114390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.380544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.62e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0118,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03817333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.323950130422475,
      "rewrite6_current_pass_fraction": 0.902806177094896,
      "rewrite6_full_passes": 1.0,
      "step": 114400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3830016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6205e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0399,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.038533333333333336,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.903152815291975,
      "rewrite6_current_pass_fraction": 0.9033757862553796,
      "rewrite6_full_passes": 1.0,
      "step": 114410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3854592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.621e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0079,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03877333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.244841087817587,
      "rewrite6_current_pass_fraction": 0.9041255282272984,
      "rewrite6_full_passes": 1.0,
      "step": 114420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3879168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6215e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0176,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03917333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.442171551887302,
      "rewrite6_current_pass_fraction": 0.9047389534770501,
      "rewrite6_full_passes": 1.0,
      "step": 114430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3903744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.622e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0214,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03965333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.51999958339129,
      "rewrite6_current_pass_fraction": 0.9053523787268019,
      "rewrite6_full_passes": 1.0,
      "step": 114440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.392832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6225e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0143,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.03998666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.374823571052737,
      "rewrite6_current_pass_fraction": 0.9059950147027321,
      "rewrite6_full_passes": 1.0,
      "step": 114450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3952896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.623e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0399,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04022666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.903152815291975,
      "rewrite6_current_pass_fraction": 0.9065646238632159,
      "rewrite6_full_passes": 1.0,
      "step": 114460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.3977472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6235e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9865,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0406,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.816204260673988,
      "rewrite6_current_pass_fraction": 0.9071147592062472,
      "rewrite6_full_passes": 1.0,
      "step": 114470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4002048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.624e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0211,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04088,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.513844506823926,
      "rewrite6_current_pass_fraction": 0.9076746314580048,
      "rewrite6_full_passes": 1.0,
      "step": 114480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4026624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6245e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0054,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.041093333333333336,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.194292197538438,
      "rewrite6_current_pass_fraction": 0.9082977936164827,
      "rewrite6_full_passes": 1.0,
      "step": 114490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.40512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.625e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 17.719806671142578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0232,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04137333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.556968844995144,
      "rewrite6_current_pass_fraction": 0.9088625343226032,
      "rewrite6_full_passes": 1.0,
      "step": 114500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.625e-05,
      "eval_loss": 3.0815176963806152,
      "eval_perplexity": 21.79145011662108,
      "eval_runtime": 0.0509,
      "eval_samples_per_second": 628.557,
      "eval_steps_per_second": 19.642,
      "step": 114500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4075776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6255e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 17.77941131591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0164,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04154666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.417655658502976,
      "rewrite6_current_pass_fraction": 0.9093396428501879,
      "rewrite6_full_passes": 1.0,
      "step": 114510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4100352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.626e-05,
      "grad_norm": 0.431640625,
      "host_rss_gib": 17.77941131591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0265,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.041773333333333336,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.624918898106596,
      "rewrite6_current_pass_fraction": 0.9099774103717552,
      "rewrite6_full_passes": 1.0,
      "step": 114520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4124928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6265e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9871,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.042146666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.828097550860647,
      "rewrite6_current_pass_fraction": 0.9106930731631322,
      "rewrite6_full_passes": 1.0,
      "step": 114530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4149504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.627e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04236,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.701372449879624,
      "rewrite6_current_pass_fraction": 0.9112286031430742,
      "rewrite6_full_passes": 1.0,
      "step": 114540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.417408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6275e-05,
      "grad_norm": 0.59765625,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.029,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04288,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.67654570196773,
      "rewrite6_current_pass_fraction": 0.9117349223968374,
      "rewrite6_full_passes": 1.0,
      "step": 114550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4198656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.628e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0073,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.043186666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.23269782650759,
      "rewrite6_current_pass_fraction": 0.9123386107378629,
      "rewrite6_full_passes": 1.0,
      "step": 114560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4223232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6285e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0247,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04348,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.58782743642022,
      "rewrite6_current_pass_fraction": 0.9129666413507039,
      "rewrite6_full_passes": 1.0,
      "step": 114570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4247808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.629e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04378666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.9135119082393721,
      "rewrite6_full_passes": 1.0,
      "step": 114580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4272384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6295e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0143,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.044173333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.374823571052737,
      "rewrite6_current_pass_fraction": 0.9141107281260346,
      "rewrite6_full_passes": 1.0,
      "step": 114590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.429696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.63e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.044533333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.913607005029263,
      "rewrite6_current_pass_fraction": 0.9146511265603396,
      "rewrite6_full_passes": 1.0,
      "step": 114600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4321536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6305e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0111,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04484,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.30972834353731,
      "rewrite6_current_pass_fraction": 0.9152888940819068,
      "rewrite6_full_passes": 1.0,
      "step": 114610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4346112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.631e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0197,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04510666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.4851452187036,
      "rewrite6_current_pass_fraction": 0.9158731086054799,
      "rewrite6_full_passes": 1.0,
      "step": 114620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4370688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6315e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0024,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.045386666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.13380010445452,
      "rewrite6_current_pass_fraction": 0.9164281124028744,
      "rewrite6_full_passes": 1.0,
      "step": 114630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4395264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.632e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0367,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04572,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.836369636357645,
      "rewrite6_current_pass_fraction": 0.9170318007438998,
      "rewrite6_full_passes": 1.0,
      "step": 114640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.441984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6325e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0277,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.046253333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.649683656667687,
      "rewrite6_current_pass_fraction": 0.917620883721836,
      "rewrite6_full_passes": 1.0,
      "step": 114650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4444416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.633e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0339,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.046573333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.77810940376482,
      "rewrite6_current_pass_fraction": 0.9181369398843255,
      "rewrite6_full_passes": 1.0,
      "step": 114660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4468992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6335e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0257,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04692,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.60842556120252,
      "rewrite6_current_pass_fraction": 0.9187893127689821,
      "rewrite6_full_passes": 1.0,
      "step": 114670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4493568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.634e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 17.77989959716797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0243,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.047213333333333336,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.579593952252267,
      "rewrite6_current_pass_fraction": 0.9194319487449125,
      "rewrite6_full_passes": 1.0,
      "step": 114680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4518144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6345e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0615,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04744,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.359572503273004,
      "rewrite6_current_pass_fraction": 0.9199772156335807,
      "rewrite6_full_passes": 1.0,
      "step": 114690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.454272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.635e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.023,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04770666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.55285786233811,
      "rewrite6_current_pass_fraction": 0.9205614301571537,
      "rewrite6_full_passes": 1.0,
      "step": 114700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4567296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6355e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0136,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04805333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.360566185220215,
      "rewrite6_current_pass_fraction": 0.9210288017760122,
      "rewrite6_full_passes": 1.0,
      "step": 114710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4591872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.636e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0212,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04836,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.515895993847245,
      "rewrite6_current_pass_fraction": 0.921690911569395,
      "rewrite6_full_passes": 1.0,
      "step": 114720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4616448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6365e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04868,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.212475241658723,
      "rewrite6_current_pass_fraction": 0.9222215730949738,
      "rewrite6_full_passes": 1.0,
      "step": 114730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4641024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.637e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0092,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04905333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.271176495537865,
      "rewrite6_current_pass_fraction": 0.9227765768923682,
      "rewrite6_full_passes": 1.0,
      "step": 114740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.46656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6375e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0188,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04937333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.466716882002203,
      "rewrite6_current_pass_fraction": 0.9233169753266733,
      "rewrite6_full_passes": 1.0,
      "step": 114750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4690176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.638e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0229,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04968,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.55080267931274,
      "rewrite6_current_pass_fraction": 0.9238378999435259,
      "rewrite6_full_passes": 1.0,
      "step": 114760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4714752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6385e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0196,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.04992,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.48309680660405,
      "rewrite6_current_pass_fraction": 0.9243977721952834,
      "rewrite6_full_passes": 1.0,
      "step": 114770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4739328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.639e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0237,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05021333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.567249899467065,
      "rewrite6_current_pass_fraction": 0.9249965920819458,
      "rewrite6_full_passes": 1.0,
      "step": 114780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4763904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6395e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0176,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05048,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.442171551887302,
      "rewrite6_current_pass_fraction": 0.9256051488773345,
      "rewrite6_full_passes": 1.0,
      "step": 114790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.478848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.64e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0262,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05074666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.618732350465706,
      "rewrite6_current_pass_fraction": 0.9262234425814493,
      "rewrite6_full_passes": 1.0,
      "step": 114800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4813056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6405e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0166,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.051026666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.421739598015016,
      "rewrite6_current_pass_fraction": 0.926836867831201,
      "rewrite6_full_passes": 1.0,
      "step": 114810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4837632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.641e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0142,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05130666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.37278619056636,
      "rewrite6_current_pass_fraction": 0.9273918716285954,
      "rewrite6_full_passes": 1.0,
      "step": 114820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4862208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6415e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 17.780284881591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0516,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.9279663492434422,
      "rewrite6_full_passes": 1.0,
      "step": 114830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4886784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.642e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.551532745361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0168,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05197333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.425824354396642,
      "rewrite6_current_pass_fraction": 0.928540826858289,
      "rewrite6_full_passes": 1.0,
      "step": 114840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.491136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6425e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 17.119892120361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0164,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05224,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.417655658502976,
      "rewrite6_current_pass_fraction": 0.9291153044731358,
      "rewrite6_full_passes": 1.0,
      "step": 114850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4935936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.643e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 17.64334487915039,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0242,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.052533333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.577536095751576,
      "rewrite6_current_pass_fraction": 0.9296508344530778,
      "rewrite6_full_passes": 1.0,
      "step": 114860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4960512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6435e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 17.67461395263672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0144,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.052853333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.376861155287358,
      "rewrite6_current_pass_fraction": 0.9302301805222878,
      "rewrite6_full_passes": 1.0,
      "step": 114870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.4985088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.644e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.778671264648438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0067,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05316,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.220561848969027,
      "rewrite6_current_pass_fraction": 0.9308192635002239,
      "rewrite6_full_passes": 1.0,
      "step": 114880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5009664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6445e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.779129028320312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0201,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05352,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.49334091582123,
      "rewrite6_current_pass_fraction": 0.9314716363848805,
      "rewrite6_full_passes": 1.0,
      "step": 114890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.503424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.645e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.779914855957031,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.024,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05381333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.573421000055713,
      "rewrite6_current_pass_fraction": 0.932070456271543,
      "rewrite6_full_passes": 1.0,
      "step": 114900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5058816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6455e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.78005599975586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0181,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.054053333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.452395193360626,
      "rewrite6_current_pass_fraction": 0.9325280909816751,
      "rewrite6_full_passes": 1.0,
      "step": 114910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5083392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.646e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.780807495117188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0176,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.054426666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.442171551887302,
      "rewrite6_current_pass_fraction": 0.9331123055052482,
      "rewrite6_full_passes": 1.0,
      "step": 114920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5107968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6465e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.780811309814453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0209,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05481333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.5097421481721,
      "rewrite6_current_pass_fraction": 0.9337938891160834,
      "rewrite6_full_passes": 1.0,
      "step": 114930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5132544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.647e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.782764434814453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0307,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05514666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.711725724207458,
      "rewrite6_current_pass_fraction": 0.93444626200074,
      "rewrite6_full_passes": 1.0,
      "step": 114940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.515712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6475e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.783767700195312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0131,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.055426666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.350388446774257,
      "rewrite6_current_pass_fraction": 0.9350110027068607,
      "rewrite6_full_passes": 1.0,
      "step": 114950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5181696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.648e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.783767700195312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0223,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05574666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.538475896109915,
      "rewrite6_current_pass_fraction": 0.9356146910478861,
      "rewrite6_full_passes": 1.0,
      "step": 114960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5206272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6485e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.783767700195312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0273,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.055986666666666664,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.641425434959473,
      "rewrite6_current_pass_fraction": 0.9362183793889116,
      "rewrite6_full_passes": 1.0,
      "step": 114970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5230848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.649e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.783767700195312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0436,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.056253333333333336,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.980637739420835,
      "rewrite6_current_pass_fraction": 0.9368318046386633,
      "rewrite6_full_passes": 1.0,
      "step": 114980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5255424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6495e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.783767700195312,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0493,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05653333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.100568853497336,
      "rewrite6_current_pass_fraction": 0.9373137816206111,
      "rewrite6_full_passes": 1.0,
      "step": 114990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.65e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.78420639038086,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0582,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.056973333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.28920208905679,
      "rewrite6_current_pass_fraction": 0.9379077330529103,
      "rewrite6_full_passes": 1.0,
      "step": 115000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.65e-05,
      "eval_loss": 3.082225799560547,
      "eval_perplexity": 21.806886176260452,
      "eval_runtime": 0.0786,
      "eval_samples_per_second": 407.315,
      "eval_steps_per_second": 12.729,
      "step": 115000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5304576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6505e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.866352081298828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0197,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05717333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.4851452187036,
      "rewrite6_current_pass_fraction": 0.9384822106677572,
      "rewrite6_full_passes": 1.0,
      "step": 115010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5329152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.651e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.866352081298828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0479,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05753333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.071048726013295,
      "rewrite6_current_pass_fraction": 0.9390469513738778,
      "rewrite6_full_passes": 1.0,
      "step": 115020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5353728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6515e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.866352081298828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0315,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.057866666666666663,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.728301734306807,
      "rewrite6_current_pass_fraction": 0.9396068236256353,
      "rewrite6_full_passes": 1.0,
      "step": 115030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5378304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.652e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.866352081298828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0024,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.058186666666666664,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.13380010445452,
      "rewrite6_current_pass_fraction": 0.9402591965102919,
      "rewrite6_full_passes": 1.0,
      "step": 115040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.540288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6525e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.866352081298828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0303,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05862666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.70344269063493,
      "rewrite6_current_pass_fraction": 0.9408142003076864,
      "rewrite6_full_passes": 1.0,
      "step": 115050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5427456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.653e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.866352081298828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9994,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05892,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.073489215707433,
      "rewrite6_current_pass_fraction": 0.941427625557438,
      "rewrite6_full_passes": 1.0,
      "step": 115060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5452032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6535e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.042,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05926666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.947095559937008,
      "rewrite6_current_pass_fraction": 0.9419972347179217,
      "rewrite6_full_passes": 1.0,
      "step": 115070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5476608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.654e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0158,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05944,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.40540873955097,
      "rewrite6_current_pass_fraction": 0.9426155284220366,
      "rewrite6_full_passes": 1.0,
      "step": 115080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5501184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6545e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0082,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.059653333333333336,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.250915451252894,
      "rewrite6_current_pass_fraction": 0.94326303285233,
      "rewrite6_full_passes": 1.0,
      "step": 115090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.552576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.655e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9983,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.05998666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.051420517579384,
      "rewrite6_current_pass_fraction": 0.9438521158302662,
      "rewrite6_full_passes": 1.0,
      "step": 115100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5550336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6555e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0125,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.060346666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.33818187604361,
      "rewrite6_current_pass_fraction": 0.9444655410800179,
      "rewrite6_full_passes": 1.0,
      "step": 115110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5574912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.656e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0297,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06061333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.691024350895024,
      "rewrite6_current_pass_fraction": 0.9451130455103114,
      "rewrite6_full_passes": 1.0,
      "step": 115120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5599488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6565e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0206,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.060893333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.503590148373757,
      "rewrite6_current_pass_fraction": 0.9456680493077058,
      "rewrite6_full_passes": 1.0,
      "step": 115130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5624064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.657e-05,
      "grad_norm": 0.74609375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9949,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.061173333333333337,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.983361453791602,
      "rewrite6_current_pass_fraction": 0.9462327900138264,
      "rewrite6_full_passes": 1.0,
      "step": 115140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.564864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6575e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0013,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.061413333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.111665100823558,
      "rewrite6_current_pass_fraction": 0.9467780569024946,
      "rewrite6_full_passes": 1.0,
      "step": 115150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5673216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.658e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0106,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06170666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.29957601765852,
      "rewrite6_current_pass_fraction": 0.9474109559696987,
      "rewrite6_full_passes": 1.0,
      "step": 115160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5697792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6585e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0173,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06214666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.436039820227478,
      "rewrite6_current_pass_fraction": 0.9479270121321882,
      "rewrite6_full_passes": 1.0,
      "step": 115170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5722368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.659e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06248,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.15797516671621,
      "rewrite6_current_pass_fraction": 0.9485112266557614,
      "rewrite6_full_passes": 1.0,
      "step": 115180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5746944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6595e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0184,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06290666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.458531832368465,
      "rewrite6_current_pass_fraction": 0.9490857042706081,
      "rewrite6_full_passes": 1.0,
      "step": 115190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.577152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.66e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0061,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06324,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.20843315083295,
      "rewrite6_current_pass_fraction": 0.9497624194270803,
      "rewrite6_full_passes": 1.0,
      "step": 115200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5796096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6605e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0096,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06364,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.279286588046446,
      "rewrite6_current_pass_fraction": 0.9504050554030107,
      "rewrite6_full_passes": 1.0,
      "step": 115210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5820672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.661e-05,
      "grad_norm": 0.5546875,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0266,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06397333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.626981493124443,
      "rewrite6_current_pass_fraction": 0.950955190746042,
      "rewrite6_full_passes": 1.0,
      "step": 115220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5845248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6615e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0056,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0644,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.198331459890714,
      "rewrite6_current_pass_fraction": 0.9515588790870675,
      "rewrite6_full_passes": 1.0,
      "step": 115230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5869824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.662e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0133,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06470666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.354458931498517,
      "rewrite6_current_pass_fraction": 0.952074935249557,
      "rewrite6_full_passes": 1.0,
      "step": 115240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.58944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6625e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.016,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06505333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.409490229434265,
      "rewrite6_current_pass_fraction": 0.9526980974080349,
      "rewrite6_full_passes": 1.0,
      "step": 115250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5918976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.663e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0112,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06532,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.311759417923692,
      "rewrite6_current_pass_fraction": 0.9532287589336138,
      "rewrite6_full_passes": 1.0,
      "step": 115260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5943552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6635e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0417,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06553333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.940812373794078,
      "rewrite6_current_pass_fraction": 0.953866526455181,
      "rewrite6_full_passes": 1.0,
      "step": 115270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5968128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.664e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0161,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06577333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.411531280508054,
      "rewrite6_current_pass_fraction": 0.9543728457089443,
      "rewrite6_full_passes": 1.0,
      "step": 115280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.5992704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6645e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0001,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06614666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.087545577311023,
      "rewrite6_current_pass_fraction": 0.9550349555023271,
      "rewrite6_full_passes": 1.0,
      "step": 115290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.601728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.665e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.011,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06650666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.307697472248215,
      "rewrite6_current_pass_fraction": 0.9555558801191798,
      "rewrite6_full_passes": 1.0,
      "step": 115300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6041856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6655e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0084,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06673333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.254966039388457,
      "rewrite6_current_pass_fraction": 0.956101147007848,
      "rewrite6_full_passes": 1.0,
      "step": 115310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6066432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.666e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06690666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.759417517911764,
      "rewrite6_current_pass_fraction": 0.9567048353488734,
      "rewrite6_full_passes": 1.0,
      "step": 115320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6091008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6665e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.995,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06716,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.985359889857122,
      "rewrite6_current_pass_fraction": 0.957220891511363,
      "rewrite6_full_passes": 1.0,
      "step": 115330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6115584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.667e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0394,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06742666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.892703851343008,
      "rewrite6_current_pass_fraction": 0.9578197113980254,
      "rewrite6_full_passes": 1.0,
      "step": 115340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.614016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6675e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.866355895996094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0352,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06781333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.805138511102886,
      "rewrite6_current_pass_fraction": 0.9583503729236043,
      "rewrite6_full_passes": 1.0,
      "step": 115350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6164736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.668e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.019,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0682,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.470810634740232,
      "rewrite6_current_pass_fraction": 0.9589540612646297,
      "rewrite6_full_passes": 1.0,
      "step": 115360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6189312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6685e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0205,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06849333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.501539891873456,
      "rewrite6_current_pass_fraction": 0.959640513329828,
      "rewrite6_full_passes": 1.0,
      "step": 115370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6213888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.669e-05,
      "grad_norm": 0.498046875,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0227,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06886666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.54669292976553,
      "rewrite6_current_pass_fraction": 0.9601517010379544,
      "rewrite6_full_passes": 1.0,
      "step": 115380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6238464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6695e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06910666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.821789281331817,
      "rewrite6_current_pass_fraction": 0.9606823625635333,
      "rewrite6_full_passes": 1.0,
      "step": 115390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.626304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.67e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.027,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06953333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.635233936100253,
      "rewrite6_current_pass_fraction": 0.9612324979065646,
      "rewrite6_full_passes": 1.0,
      "step": 115400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6287616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6705e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.016,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.06984,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.409490229434265,
      "rewrite6_current_pass_fraction": 0.9616901326166969,
      "rewrite6_full_passes": 1.0,
      "step": 115410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6312192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.671e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0117,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07014666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.32191783702579,
      "rewrite6_current_pass_fraction": 0.9623035578664486,
      "rewrite6_full_passes": 1.0,
      "step": 115420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6336768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6715e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0203,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07044,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.497439993898546,
      "rewrite6_current_pass_fraction": 0.9628098771202118,
      "rewrite6_full_passes": 1.0,
      "step": 115430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6361344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.672e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07077333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.970150042693795,
      "rewrite6_current_pass_fraction": 0.9633600124632432,
      "rewrite6_full_passes": 1.0,
      "step": 115440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.638592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6725e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0181,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07094666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.452395193360626,
      "rewrite6_current_pass_fraction": 0.9639004108975483,
      "rewrite6_full_passes": 1.0,
      "step": 115450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6410496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.673e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0333,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07125333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.765646277434353,
      "rewrite6_current_pass_fraction": 0.964426203968764,
      "rewrite6_full_passes": 1.0,
      "step": 115460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6435072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6735e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0082,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07164,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.250915451252894,
      "rewrite6_current_pass_fraction": 0.9649276547681642,
      "rewrite6_full_passes": 1.0,
      "step": 115470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6459648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.674e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0074,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07186666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.234721197457105,
      "rewrite6_current_pass_fraction": 0.9655946330159101,
      "rewrite6_full_passes": 1.0,
      "step": 115480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6484224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6745e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0032,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0722,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.14991358907254,
      "rewrite6_current_pass_fraction": 0.9661301629958521,
      "rewrite6_full_passes": 1.0,
      "step": 115490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.65088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.675e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 15.867061614990234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07236,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.686886559817744,
      "rewrite6_current_pass_fraction": 0.9667241144281513,
      "rewrite6_full_passes": 1.0,
      "step": 115500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.675e-05,
      "eval_loss": 3.081629753112793,
      "eval_perplexity": 21.793892132130143,
      "eval_runtime": 0.0494,
      "eval_samples_per_second": 648.329,
      "eval_steps_per_second": 20.26,
      "step": 115500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6533376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6755e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0254,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07269333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.602243960820573,
      "rewrite6_current_pass_fraction": 0.9672742497711826,
      "rewrite6_full_passes": 1.0,
      "step": 115510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6557952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.676e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.073,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.9678049112967615,
      "rewrite6_full_passes": 1.0,
      "step": 115520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6582528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6765e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0133,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.354458931498517,
      "rewrite6_current_pass_fraction": 0.9683355728223404,
      "rewrite6_full_passes": 1.0,
      "step": 115530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6607104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.677e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0736,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.9688662343479192,
      "rewrite6_full_passes": 1.0,
      "step": 115540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.663168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6775e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0239,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07396,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.57136376081938,
      "rewrite6_current_pass_fraction": 0.9694553173258554,
      "rewrite6_full_passes": 1.0,
      "step": 115550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6656256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.678e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0351,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07424,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.803058101273997,
      "rewrite6_current_pass_fraction": 0.9700541372125178,
      "rewrite6_full_passes": 1.0,
      "step": 115560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6680832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6785e-05,
      "grad_norm": 0.435546875,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0055,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07444,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.196311727733022,
      "rewrite6_current_pass_fraction": 0.9706140094642752,
      "rewrite6_full_passes": 1.0,
      "step": 115570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6705408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.679e-05,
      "grad_norm": 0.4296875,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0064,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07482666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.214496590248636,
      "rewrite6_current_pass_fraction": 0.9712030924422115,
      "rewrite6_full_passes": 1.0,
      "step": 115580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6729984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6795e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9952,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07512,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.989357361568942,
      "rewrite6_current_pass_fraction": 0.9717142801503379,
      "rewrite6_full_passes": 1.0,
      "step": 115590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.675456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.68e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0238,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07546666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.56930672729669,
      "rewrite6_current_pass_fraction": 0.9723131000370002,
      "rewrite6_full_passes": 1.0,
      "step": 115600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6779136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6805e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0336,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07593333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.771876905865113,
      "rewrite6_current_pass_fraction": 0.9728534984713053,
      "rewrite6_full_passes": 1.0,
      "step": 115610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6803712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.681e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0177,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07633333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.44421587125676,
      "rewrite6_current_pass_fraction": 0.9734717921754201,
      "rewrite6_full_passes": 1.0,
      "step": 115620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6828288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6815e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07672,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.64348968071354,
      "rewrite6_current_pass_fraction": 0.9740073221553621,
      "rewrite6_full_passes": 1.0,
      "step": 115630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6852864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.682e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0344,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07718666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.788501056163312,
      "rewrite6_current_pass_fraction": 0.9745671944071196,
      "rewrite6_full_passes": 1.0,
      "step": 115640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.687744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6825e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0188,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07748,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.466716882002203,
      "rewrite6_current_pass_fraction": 0.975117329750151,
      "rewrite6_full_passes": 1.0,
      "step": 115650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6902016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.683e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.017,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0778,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.42990992781124,
      "rewrite6_current_pass_fraction": 0.9757697026348076,
      "rewrite6_full_passes": 1.0,
      "step": 115660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6926592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6835e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07801333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.78642230999673,
      "rewrite6_current_pass_fraction": 0.9764074701563747,
      "rewrite6_full_passes": 1.0,
      "step": 115670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6951168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.684e-05,
      "grad_norm": 0.4375,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0204,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07830666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.49948984038855,
      "rewrite6_current_pass_fraction": 0.9770062900430372,
      "rewrite6_full_passes": 1.0,
      "step": 115680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.6975744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6845e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 15.867130279541016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0134,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07854666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.35649447916735,
      "rewrite6_current_pass_fraction": 0.9774590562988063,
      "rewrite6_full_passes": 1.0,
      "step": 115690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.700032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.685e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 14.898494720458984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9965,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07877333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.01536042446777,
      "rewrite6_current_pass_fraction": 0.9780773500029211,
      "rewrite6_full_passes": 1.0,
      "step": 115700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7024896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6855e-05,
      "grad_norm": 0.57421875,
      "host_rss_gib": 15.953716278076172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0373,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07925333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.848875209436216,
      "rewrite6_current_pass_fraction": 0.9786761698895835,
      "rewrite6_full_passes": 1.0,
      "step": 115710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7049472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.686e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.52914047241211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0618,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07957333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.365981336300866,
      "rewrite6_current_pass_fraction": 0.9791581468715312,
      "rewrite6_full_passes": 1.0,
      "step": 115720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7074048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6865e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.52914047241211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9955,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.07989333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.995355068388445,
      "rewrite6_current_pass_fraction": 0.9797131506689256,
      "rewrite6_full_passes": 1.0,
      "step": 115730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7098624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.687e-05,
      "grad_norm": 0.578125,
      "host_rss_gib": 16.52914047241211,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0311,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08026666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.720012071656146,
      "rewrite6_current_pass_fraction": 0.9801707853790579,
      "rewrite6_full_passes": 1.0,
      "step": 115740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.71232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6875e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.909507751464844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0173,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08052,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.436039820227478,
      "rewrite6_current_pass_fraction": 0.9808085529006251,
      "rewrite6_full_passes": 1.0,
      "step": 115750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7147776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.688e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.911937713623047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0307,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08086666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.711725724207458,
      "rewrite6_current_pass_fraction": 0.9813148721543884,
      "rewrite6_full_passes": 1.0,
      "step": 115760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7172352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6885e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.913532257080078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08124,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.159991065026112,
      "rewrite6_current_pass_fraction": 0.9818650074974197,
      "rewrite6_full_passes": 1.0,
      "step": 115770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7196928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.689e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.914863586425781,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.039,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0814,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.884348440995947,
      "rewrite6_current_pass_fraction": 0.9824881696558977,
      "rewrite6_full_passes": 1.0,
      "step": 115780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7221504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6895e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.915786743164062,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0312,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08161333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.72208417646683,
      "rewrite6_current_pass_fraction": 0.9831162002687387,
      "rewrite6_full_passes": 1.0,
      "step": 115790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.724608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.69e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.91628646850586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0305,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08192,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.707583793269517,
      "rewrite6_current_pass_fraction": 0.9836371248855913,
      "rewrite6_full_passes": 1.0,
      "step": 115800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7270656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6905e-05,
      "grad_norm": 0.5625,
      "host_rss_gib": 15.91653823852539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0242,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0822,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.577536095751576,
      "rewrite6_current_pass_fraction": 0.9842651554984324,
      "rewrite6_full_passes": 1.0,
      "step": 115810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7295232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.691e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.91653823852539,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0129,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08246666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.346318776065537,
      "rewrite6_current_pass_fraction": 0.9849272652918152,
      "rewrite6_full_passes": 1.0,
      "step": 115820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7319808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6915e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.916927337646484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0185,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08278666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.460577787847765,
      "rewrite6_current_pass_fraction": 0.9854774006348465,
      "rewrite6_full_passes": 1.0,
      "step": 115830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7344384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.692e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.916927337646484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0256,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08310666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.606364821685087,
      "rewrite6_current_pass_fraction": 0.9860470097953302,
      "rewrite6_full_passes": 1.0,
      "step": 115840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.736896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6925e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.916927337646484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.993,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08348,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.945429114163357,
      "rewrite6_current_pass_fraction": 0.9866458296819925,
      "rewrite6_full_passes": 1.0,
      "step": 115850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7393536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.693e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.916927337646484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0264,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08373333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.622856509337947,
      "rewrite6_current_pass_fraction": 0.9871034643921248,
      "rewrite6_full_passes": 1.0,
      "step": 115860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7418112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6935e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.916927337646484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0488,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08396,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.090021206202152,
      "rewrite6_current_pass_fraction": 0.9876195205546143,
      "rewrite6_full_passes": 1.0,
      "step": 115870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7442688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.694e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.916927337646484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0204,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08434666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.49948984038855,
      "rewrite6_current_pass_fraction": 0.9881891297150981,
      "rewrite6_full_passes": 1.0,
      "step": 115880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7467264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6945e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 15.916927337646484,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0115,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08448,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.317853859869647,
      "rewrite6_current_pass_fraction": 0.9887051858775876,
      "rewrite6_full_passes": 1.0,
      "step": 115890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.749184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.695e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.91726303100586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08473333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.662077184548252,
      "rewrite6_current_pass_fraction": 0.9893283480360655,
      "rewrite6_full_passes": 1.0,
      "step": 115900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7516416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6955e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.91726303100586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0306,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08501333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.709654655190217,
      "rewrite6_current_pass_fraction": 0.9898930887421862,
      "rewrite6_full_passes": 1.0,
      "step": 115910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7540992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.696e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.91726303100586,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0528,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08528,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.174550236381975,
      "rewrite6_current_pass_fraction": 0.9904286187221281,
      "rewrite6_full_passes": 1.0,
      "step": 115920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7565568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6965e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.917633056640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0397,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08565333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.898972602764104,
      "rewrite6_current_pass_fraction": 0.9910371755175167,
      "rewrite6_full_passes": 1.0,
      "step": 115930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7590144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.697e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.917633056640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0468,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08588,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.04788331572619,
      "rewrite6_current_pass_fraction": 0.9916262584954528,
      "rewrite6_full_passes": 1.0,
      "step": 115940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.761472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6975e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 15.917633056640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08628,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.212475241658723,
      "rewrite6_current_pass_fraction": 0.9922104730190259,
      "rewrite6_full_passes": 1.0,
      "step": 115950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7639296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.698e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.917633056640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08649333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.905243235092758,
      "rewrite6_current_pass_fraction": 0.9928336351775039,
      "rewrite6_full_passes": 1.0,
      "step": 115960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7663872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6985e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.917633056640625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0076,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08688,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.238768546418,
      "rewrite6_current_pass_fraction": 0.9933691651574458,
      "rewrite6_full_passes": 1.0,
      "step": 115970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7688448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.699e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.91842269897461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08721333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.9939777219528344,
      "rewrite6_full_passes": 1.0,
      "step": 115980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7713024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.6995e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.918495178222656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.002,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08757333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.125748194902,
      "rewrite6_current_pass_fraction": 0.9945668049307705,
      "rewrite6_full_passes": 1.0,
      "step": 115990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.77376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 15.918495178222656,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0288,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08784,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.672410806330685,
      "rewrite6_current_pass_fraction": 0.9952289147241534,
      "rewrite6_full_passes": 1.0,
      "step": 116000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.7e-05,
      "eval_loss": 3.080976963043213,
      "eval_perplexity": 21.779669938327277,
      "eval_runtime": 0.0875,
      "eval_samples_per_second": 365.786,
      "eval_steps_per_second": 11.431,
      "step": 116000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7762176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7005e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0293,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08822666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.68274959621593,
      "rewrite6_current_pass_fraction": 0.9957595762497322,
      "rewrite6_full_passes": 1.0,
      "step": 116010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7786752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.701e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0176,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08856,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.442171551887302,
      "rewrite6_current_pass_fraction": 0.9963097115927635,
      "rewrite6_full_passes": 1.0,
      "step": 116020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7811328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7015e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08893333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.821789281331817,
      "rewrite6_current_pass_fraction": 0.9968257677552531,
      "rewrite6_full_passes": 1.0,
      "step": 116030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7835904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.702e-05,
      "grad_norm": 0.53515625,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0118,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08934666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.323950130422475,
      "rewrite6_current_pass_fraction": 0.9974148507331893,
      "rewrite6_full_passes": 1.0,
      "step": 116040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.786048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7025e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0149,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08969333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.387052133397212,
      "rewrite6_current_pass_fraction": 0.9979747229849467,
      "rewrite6_full_passes": 1.0,
      "step": 116050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7885056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.703e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0433,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.08994666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.974344492133298,
      "rewrite6_current_pass_fraction": 0.9984566999668945,
      "rewrite6_full_passes": 1.0,
      "step": 116060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7909632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7035e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09028,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.945000955113006,
      "rewrite6_current_pass_fraction": 0.9990117037642889,
      "rewrite6_full_passes": 1.0,
      "step": 116070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7934208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.704e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0159,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09070666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.407449382455365,
      "rewrite6_current_pass_fraction": 0.9995423652898677,
      "rewrite6_full_passes": 1.0,
      "step": 116080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.7958784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7045e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0239,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09096,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.57136376081938,
      "rewrite6_current_pass_fraction": 0.0002336858094292224,
      "rewrite6_full_passes": 2.0,
      "step": 116090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.798336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.705e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0372,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09132,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.84679042615617,
      "rewrite6_current_pass_fraction": 0.0009055325115382368,
      "rewrite6_full_passes": 2.0,
      "step": 116100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8007936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7055e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09165333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.0015481684874685984,
      "rewrite6_full_passes": 2.0,
      "step": 116110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8032512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.706e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0527,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09202666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.172432887227565,
      "rewrite6_current_pass_fraction": 0.0022394890070300483,
      "rewrite6_full_passes": 2.0,
      "step": 116120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8057088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7065e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09230666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.0029113357091390627,
      "rewrite6_full_passes": 2.0,
      "step": 116130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8081664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.707e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0249,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0926,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.591945413691505,
      "rewrite6_current_pass_fraction": 0.003466339506533466,
      "rewrite6_full_passes": 2.0,
      "step": 116140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.810624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7075e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0258,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09285333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.610486506804197,
      "rewrite6_current_pass_fraction": 0.004006737940838542,
      "rewrite6_full_passes": 2.0,
      "step": 116150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8130816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.708e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09317333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.188234818531463,
      "rewrite6_current_pass_fraction": 0.004576347101322272,
      "rewrite6_full_passes": 2.0,
      "step": 116160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8155392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7085e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09348,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.999528813094322,
      "rewrite6_current_pass_fraction": 0.005082666355085587,
      "rewrite6_full_passes": 2.0,
      "step": 116170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8179968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.709e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0165,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09373333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.419697526160515,
      "rewrite6_current_pass_fraction": 0.00559872251757512,
      "rewrite6_full_passes": 2.0,
      "step": 116180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8204544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7095e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9857,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09417333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.800357636760168,
      "rewrite6_current_pass_fraction": 0.006236490039142373,
      "rewrite6_full_passes": 2.0,
      "step": 116190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.822912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.71e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0095,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09456,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.2772587607807,
      "rewrite6_current_pass_fraction": 0.006772020019084341,
      "rewrite6_full_passes": 2.0,
      "step": 116200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8253696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7105e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.017,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09477333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.42990992781124,
      "rewrite6_current_pass_fraction": 0.007278339272847656,
      "rewrite6_full_passes": 2.0,
      "step": 116210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8278272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.711e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9949,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09510666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.983361453791602,
      "rewrite6_current_pass_fraction": 0.00787229070514693,
      "rewrite6_full_passes": 2.0,
      "step": 116220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8302848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7115e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0204,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09546666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.49948984038855,
      "rewrite6_current_pass_fraction": 0.008529532044166618,
      "rewrite6_full_passes": 2.0,
      "step": 116230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8327424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.712e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09578666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.64348968071354,
      "rewrite6_current_pass_fraction": 0.009167299565733871,
      "rewrite6_full_passes": 2.0,
      "step": 116240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7125e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09602666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.64348968071354,
      "rewrite6_current_pass_fraction": 0.009610328912776772,
      "rewrite6_full_passes": 2.0,
      "step": 116250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8376576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.713e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09637333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.010189674981986719,
      "rewrite6_full_passes": 2.0,
      "step": 116260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8401152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7135e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0278,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09661333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.651748728285217,
      "rewrite6_current_pass_fraction": 0.010827442503553972,
      "rewrite6_full_passes": 2.0,
      "step": 116270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8425728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.714e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0072,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09702666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.23067465788506,
      "rewrite6_current_pass_fraction": 0.01136297248349594,
      "rewrite6_full_passes": 2.0,
      "step": 116280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8450304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7145e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0189,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.0974,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.468763656027395,
      "rewrite6_current_pass_fraction": 0.011864423282896146,
      "rewrite6_full_passes": 2.0,
      "step": 116290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.847488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.715e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0249,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09772,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.591945413691505,
      "rewrite6_current_pass_fraction": 0.012492453895737181,
      "rewrite6_full_passes": 2.0,
      "step": 116300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8499456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7155e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0238,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09797333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.56930672729669,
      "rewrite6_current_pass_fraction": 0.013130221417304435,
      "rewrite6_full_passes": 2.0,
      "step": 116310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8524032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.716e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9932,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09832,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.949418598921365,
      "rewrite6_current_pass_fraction": 0.013646277579793966,
      "rewrite6_full_passes": 2.0,
      "step": 116320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8548608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7165e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0132,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09868,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.352423587374265,
      "rewrite6_current_pass_fraction": 0.014225623649003915,
      "rewrite6_full_passes": 2.0,
      "step": 116330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8573184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.717e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0348,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09909333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.796818119887625,
      "rewrite6_current_pass_fraction": 0.01480983817257697,
      "rewrite6_full_passes": 2.0,
      "step": 116340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.859776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7175e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0106,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09944,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.29957601765852,
      "rewrite6_current_pass_fraction": 0.01530155206325096,
      "rewrite6_full_passes": 2.0,
      "step": 116350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8622336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.718e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0376,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.09978666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.855130810292252,
      "rewrite6_current_pass_fraction": 0.015851687406282254,
      "rewrite6_full_passes": 2.0,
      "step": 116360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8646912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7185e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10004,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.957571726540376,
      "rewrite6_current_pass_fraction": 0.016465112656033964,
      "rewrite6_full_passes": 2.0,
      "step": 116370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8671488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.719e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0194,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10029333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.479000596877352,
      "rewrite6_current_pass_fraction": 0.017146696266869193,
      "rewrite6_full_passes": 2.0,
      "step": 116380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8696064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7195e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0248,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10068,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.589886322106427,
      "rewrite6_current_pass_fraction": 0.017755253062257795,
      "rewrite6_full_passes": 2.0,
      "step": 116390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.872064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.72e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0407,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10101333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.91988202833721,
      "rewrite6_current_pass_fraction": 0.01838328367509883,
      "rewrite6_full_passes": 2.0,
      "step": 116400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8745216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7205e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0349,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10132,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.798897905687166,
      "rewrite6_current_pass_fraction": 0.01887499756577282,
      "rewrite6_full_passes": 2.0,
      "step": 116410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8769792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.721e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0196,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10165333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.48309680660405,
      "rewrite6_current_pass_fraction": 0.019507896632976963,
      "rewrite6_full_passes": 2.0,
      "step": 116420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8794368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7215e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10182666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.999528813094322,
      "rewrite6_current_pass_fraction": 0.020072637339097584,
      "rewrite6_full_passes": 2.0,
      "step": 116430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8818944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.722e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0145,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10204,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.378898943290586,
      "rewrite6_current_pass_fraction": 0.02073961558684349,
      "rewrite6_full_passes": 2.0,
      "step": 116440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.884352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7225e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0323,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10245333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.744891010519986,
      "rewrite6_current_pass_fraction": 0.021280014021148565,
      "rewrite6_full_passes": 2.0,
      "step": 116450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8868096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.723e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0336,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10266666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.771876905865113,
      "rewrite6_current_pass_fraction": 0.02175712254873323,
      "rewrite6_full_passes": 2.0,
      "step": 116460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8892672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7235e-05,
      "grad_norm": 0.55859375,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0063,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.103,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.212475241658723,
      "rewrite6_current_pass_fraction": 0.022385153161574262,
      "rewrite6_full_passes": 2.0,
      "step": 116470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8917248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.724e-05,
      "grad_norm": 4.46875,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0259,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.61254765851075,
      "rewrite6_current_pass_fraction": 0.02296449923078421,
      "rewrite6_full_passes": 2.0,
      "step": 116480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8941824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7245e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0255,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10365333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.60430428823131,
      "rewrite6_current_pass_fraction": 0.02348542384763685,
      "rewrite6_full_passes": 2.0,
      "step": 116490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.89664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.725e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.150531768798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10406666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.024050164553757473,
      "rewrite6_full_passes": 2.0,
      "step": 116500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.725e-05,
      "eval_loss": 3.0810773372650146,
      "eval_perplexity": 21.781856165467026,
      "eval_runtime": 0.0749,
      "eval_samples_per_second": 427.01,
      "eval_steps_per_second": 13.344,
      "step": 116500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.8990976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7255e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.150619506835938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0304,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10433333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.705513138424664,
      "rewrite6_current_pass_fraction": 0.024639247531693637,
      "rewrite6_full_passes": 2.0,
      "step": 116510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9015552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.726e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 16.150619506835938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1046,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.74904040364757,
      "rewrite6_current_pass_fraction": 0.025223462055266693,
      "rewrite6_full_passes": 2.0,
      "step": 116520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9040128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7265e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.150619506835938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0205,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10506666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.501539891873456,
      "rewrite6_current_pass_fraction": 0.025943993301006798,
      "rewrite6_full_passes": 2.0,
      "step": 116530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9064704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.727e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.150619506835938,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0536,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10544,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.191496654234413,
      "rewrite6_current_pass_fraction": 0.026484391735311873,
      "rewrite6_full_passes": 2.0,
      "step": 116540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.908928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7275e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.160400390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0115,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10582666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.317853859869647,
      "rewrite6_current_pass_fraction": 0.027005316352164514,
      "rewrite6_full_passes": 2.0,
      "step": 116550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9113856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.728e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 17.05986785888672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0168,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10616,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.425824354396642,
      "rewrite6_current_pass_fraction": 0.027584662421374462,
      "rewrite6_full_passes": 2.0,
      "step": 116560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9138432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7285e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 17.141895294189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0387,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10645333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.87808407616535,
      "rewrite6_current_pass_fraction": 0.028188350762399953,
      "rewrite6_full_passes": 2.0,
      "step": 116570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9163008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.729e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 17.141895294189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0185,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10689333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.460577787847765,
      "rewrite6_current_pass_fraction": 0.028738486105431248,
      "rewrite6_full_passes": 2.0,
      "step": 116580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9187584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7295e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 17.141895294189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.006,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10718666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.20641240855666,
      "rewrite6_current_pass_fraction": 0.029283752994099434,
      "rewrite6_full_passes": 2.0,
      "step": 116590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.921216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.73e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 17.141895294189453,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0173,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10733333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.436039820227478,
      "rewrite6_current_pass_fraction": 0.029755993067320987,
      "rewrite6_full_passes": 2.0,
      "step": 116600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9236736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7305e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.368194580078125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10772,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.936624630107648,
      "rewrite6_current_pass_fraction": 0.03031099686471539,
      "rewrite6_full_passes": 2.0,
      "step": 116610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9261312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.731e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.370098114013672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0124,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10801333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.33614815954352,
      "rewrite6_current_pass_fraction": 0.0309244221144671,
      "rewrite6_full_passes": 2.0,
      "step": 116620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9285888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7315e-05,
      "grad_norm": 0.53515625,
      "host_rss_gib": 16.37024688720703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.03,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10826666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.6972325893895,
      "rewrite6_current_pass_fraction": 0.03151350509240326,
      "rewrite6_full_passes": 2.0,
      "step": 116630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9310464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.732e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.371047973632812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10850666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.64348968071354,
      "rewrite6_current_pass_fraction": 0.03210745652470254,
      "rewrite6_full_passes": 2.0,
      "step": 116640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.933504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7325e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.374393463134766,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9999,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10889333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.08352846991968,
      "rewrite6_current_pass_fraction": 0.03267219723082316,
      "rewrite6_full_passes": 2.0,
      "step": 116650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9359616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.733e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.374839782714844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0265,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10926666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.624918898106596,
      "rewrite6_current_pass_fraction": 0.033300227843664196,
      "rewrite6_full_passes": 2.0,
      "step": 116660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9384192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7335e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.374839782714844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0523,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10950666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.16396560764148,
      "rewrite6_current_pass_fraction": 0.0338601000954217,
      "rewrite6_full_passes": 2.0,
      "step": 116670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9408768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.734e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.374839782714844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.10981333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.631107301990202,
      "rewrite6_current_pass_fraction": 0.034458919982084085,
      "rewrite6_full_passes": 2.0,
      "step": 116680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9433344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7345e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.374839782714844,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0255,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11014666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.60430428823131,
      "rewrite6_current_pass_fraction": 0.035082082140562014,
      "rewrite6_full_passes": 2.0,
      "step": 116690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.945792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.735e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.37555694580078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0137,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11056,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.362602343644966,
      "rewrite6_current_pass_fraction": 0.03572958657085548,
      "rewrite6_full_passes": 2.0,
      "step": 116700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9482496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7355e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.37555694580078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11090666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.036304064185702326,
      "rewrite6_full_passes": 2.0,
      "step": 116710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9507072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.736e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.37555694580078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0355,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11128,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.811380988981075,
      "rewrite6_current_pass_fraction": 0.03699051625090066,
      "rewrite6_full_passes": 2.0,
      "step": 116720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9531648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7365e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.37555694580078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0392,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11168,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.888525728398964,
      "rewrite6_current_pass_fraction": 0.0375649938657475,
      "rewrite6_full_passes": 2.0,
      "step": 116730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9556224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.737e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.37555694580078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11193333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.936624630107648,
      "rewrite6_current_pass_fraction": 0.0381151292087788,
      "rewrite6_full_passes": 2.0,
      "step": 116740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.95808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7375e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.37555694580078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0349,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11230666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.798897905687166,
      "rewrite6_current_pass_fraction": 0.03869447527798874,
      "rewrite6_full_passes": 2.0,
      "step": 116750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9605376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.738e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.37555694580078,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11261333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.678613459924104,
      "rewrite6_current_pass_fraction": 0.0392251368035676,
      "rewrite6_full_passes": 2.0,
      "step": 116760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9629952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7385e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.375938415527344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0434,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11294666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.976442031457733,
      "rewrite6_current_pass_fraction": 0.0397752721465989,
      "rewrite6_full_passes": 2.0,
      "step": 116770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9654528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.739e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.375938415527344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0065,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11314666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.21651814098351,
      "rewrite6_current_pass_fraction": 0.04025724912854667,
      "rewrite6_full_passes": 2.0,
      "step": 116780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9679104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7395e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.375938415527344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0172,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11353333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.433996318422246,
      "rewrite6_current_pass_fraction": 0.040773305291036205,
      "rewrite6_full_passes": 2.0,
      "step": 116790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.970368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.74e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.376407623291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0246,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11386666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.58576875661228,
      "rewrite6_current_pass_fraction": 0.04125528227298397,
      "rewrite6_full_passes": 2.0,
      "step": 116800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9728256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7405e-05,
      "grad_norm": 0.43359375,
      "host_rss_gib": 16.376407623291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0147,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11424,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.382975130684393,
      "rewrite6_current_pass_fraction": 0.04191739206636677,
      "rewrite6_full_passes": 2.0,
      "step": 116810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9752832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.741e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.376407623291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.027,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11462666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.635233936100253,
      "rewrite6_current_pass_fraction": 0.04248213277248739,
      "rewrite6_full_passes": 2.0,
      "step": 116820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9777408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7415e-05,
      "grad_norm": 0.54296875,
      "host_rss_gib": 16.376407623291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11489333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.78642230999673,
      "rewrite6_current_pass_fraction": 0.04307121575042355,
      "rewrite6_full_passes": 2.0,
      "step": 116830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9801984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.742e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.376407623291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0215,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11522666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.522051685953052,
      "rewrite6_current_pass_fraction": 0.04355319273237133,
      "rewrite6_full_passes": 2.0,
      "step": 116840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.982656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7425e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.376407623291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0061,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11557333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.20843315083295,
      "rewrite6_current_pass_fraction": 0.044127670347218165,
      "rewrite6_full_passes": 2.0,
      "step": 116850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9851136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.743e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.376407623291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0409,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11577333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.924066423168412,
      "rewrite6_current_pass_fraction": 0.04462425269225526,
      "rewrite6_full_passes": 2.0,
      "step": 116860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9875712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7435e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.376407623291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0274,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.116,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.64348968071354,
      "rewrite6_current_pass_fraction": 0.04521333567019143,
      "rewrite6_full_passes": 2.0,
      "step": 116870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9900288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.744e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.376407623291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0339,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11641333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.77810940376482,
      "rewrite6_current_pass_fraction": 0.045846234737395575,
      "rewrite6_full_passes": 2.0,
      "step": 116880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9924864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7445e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.376407623291016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0475,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11676,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.062621991982052,
      "rewrite6_current_pass_fraction": 0.04642071235224241,
      "rewrite6_full_passes": 2.0,
      "step": 116890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.994944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.745e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0567,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11705333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.257292224304877,
      "rewrite6_current_pass_fraction": 0.047077953691262095,
      "rewrite6_full_passes": 2.0,
      "step": 116900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9974016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7455e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11733333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.81346223114035,
      "rewrite6_current_pass_fraction": 0.047642694397382716,
      "rewrite6_full_passes": 2.0,
      "step": 116910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 22.9998592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.746e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0275,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11752,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.645554132902497,
      "rewrite6_current_pass_fraction": 0.04816361901423536,
      "rewrite6_full_passes": 2.0,
      "step": 116920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0023168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7465e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0204,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11790666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.49948984038855,
      "rewrite6_current_pass_fraction": 0.048850071079433705,
      "rewrite6_full_passes": 2.0,
      "step": 116930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0047744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.747e-05,
      "grad_norm": 0.55078125,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0169,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11825333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.42786703896461,
      "rewrite6_current_pass_fraction": 0.049366127241923234,
      "rewrite6_full_passes": 2.0,
      "step": 116940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.007232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7475e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0222,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11849333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.536422151209266,
      "rewrite6_current_pass_fraction": 0.049891920313138986,
      "rewrite6_full_passes": 2.0,
      "step": 116950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0096896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.748e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0127,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1188,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.342249919209575,
      "rewrite6_current_pass_fraction": 0.050422581838717846,
      "rewrite6_full_passes": 2.0,
      "step": 116960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0121472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7485e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0028,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11908,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.14185523541509,
      "rewrite6_current_pass_fraction": 0.05101166481665401,
      "rewrite6_full_passes": 2.0,
      "step": 116970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0146048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.749e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9978,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11937333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.041397313330467,
      "rewrite6_current_pass_fraction": 0.05151311561605421,
      "rewrite6_full_passes": 2.0,
      "step": 116980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0170624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7495e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0363,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11968,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.828036755190443,
      "rewrite6_current_pass_fraction": 0.052072987867811726,
      "rewrite6_full_passes": 2.0,
      "step": 116990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.01952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.75e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.37789535522461,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0245,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.11988,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.583710282662036,
      "rewrite6_current_pass_fraction": 0.05269128157192655,
      "rewrite6_full_passes": 2.0,
      "step": 117000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.75e-05,
      "eval_loss": 3.0817835330963135,
      "eval_perplexity": 21.797243854210347,
      "eval_runtime": 0.0561,
      "eval_samples_per_second": 570.04,
      "eval_steps_per_second": 17.814,
      "step": 117000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0219776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7505e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0431,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12010666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.970150042693795,
      "rewrite6_current_pass_fraction": 0.05333391754785691,
      "rewrite6_full_passes": 2.0,
      "step": 117010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0244352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.751e-05,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12045333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.662077184548252,
      "rewrite6_current_pass_fraction": 0.053796420712352244,
      "rewrite6_full_passes": 2.0,
      "step": 117020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0268928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7515e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0496,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12081333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.10689997377394,
      "rewrite6_current_pass_fraction": 0.054385503690288405,
      "rewrite6_full_passes": 2.0,
      "step": 117030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0293504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.752e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0263,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12110666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.620794326797853,
      "rewrite6_current_pass_fraction": 0.05497458666822457,
      "rewrite6_full_passes": 2.0,
      "step": 117040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.031808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7525e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0196,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12149333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.48309680660405,
      "rewrite6_current_pass_fraction": 0.05557827500925006,
      "rewrite6_full_passes": 2.0,
      "step": 117050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0342656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.753e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0296,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12178666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.688955351911602,
      "rewrite6_current_pass_fraction": 0.05621117407645421,
      "rewrite6_full_passes": 2.0,
      "step": 117060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0367232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7535e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.122,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.7801873185992,
      "rewrite6_current_pass_fraction": 0.05675157251075928,
      "rewrite6_full_passes": 2.0,
      "step": 117070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0391808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.754e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0342,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12236,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.784343771694374,
      "rewrite6_current_pass_fraction": 0.05732605012560612,
      "rewrite6_full_passes": 2.0,
      "step": 117080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0416384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7545e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9977,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12277333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.039393273802787,
      "rewrite6_current_pass_fraction": 0.05798815991898892,
      "rewrite6_full_passes": 2.0,
      "step": 117090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.044096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.755e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12312,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.05856263753383576,
      "rewrite6_full_passes": 2.0,
      "step": 117100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0465536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7555e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.88017198896687,
      "rewrite6_current_pass_fraction": 0.0590835621506884,
      "rewrite6_full_passes": 2.0,
      "step": 117110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0490112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.756e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12352,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.059711592763529435,
      "rewrite6_full_passes": 2.0,
      "step": 117120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0514688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7565e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0196,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12384,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.48309680660405,
      "rewrite6_current_pass_fraction": 0.06037857101127534,
      "rewrite6_full_passes": 2.0,
      "step": 117130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0539264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.757e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0444,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12417333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.997428965207153,
      "rewrite6_current_pass_fraction": 0.06095304862612218,
      "rewrite6_full_passes": 2.0,
      "step": 117140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.056384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7575e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0237,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12453333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.567249899467065,
      "rewrite6_current_pass_fraction": 0.06148371015170104,
      "rewrite6_full_passes": 2.0,
      "step": 117150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0588416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.758e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0238,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12482666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.56930672729669,
      "rewrite6_current_pass_fraction": 0.062072793129637205,
      "rewrite6_full_passes": 2.0,
      "step": 117160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0612992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7585e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1252,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.631107301990202,
      "rewrite6_current_pass_fraction": 0.06264240229012093,
      "rewrite6_full_passes": 2.0,
      "step": 117170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0637568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.759e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0201,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12556,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.49334091582123,
      "rewrite6_current_pass_fraction": 0.0632850382660513,
      "rewrite6_full_passes": 2.0,
      "step": 117180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0662144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7595e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0252,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.126,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.59812392404582,
      "rewrite6_current_pass_fraction": 0.06383517360908258,
      "rewrite6_full_passes": 2.0,
      "step": 117190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.068672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.76e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0292,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12630666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.680681424666606,
      "rewrite6_current_pass_fraction": 0.06437557204338766,
      "rewrite6_full_passes": 2.0,
      "step": 117200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0711296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7605e-05,
      "grad_norm": 0.59765625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1266,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.13647032815547,
      "rewrite6_current_pass_fraction": 0.06484781211660921,
      "rewrite6_full_passes": 2.0,
      "step": 117210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0735872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.761e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0335,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12688,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.769799822030453,
      "rewrite6_current_pass_fraction": 0.06526163073747347,
      "rewrite6_full_passes": 2.0,
      "step": 117220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0760448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7615e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0123,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12712,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.334114646404924,
      "rewrite6_current_pass_fraction": 0.06587018753286207,
      "rewrite6_full_passes": 2.0,
      "step": 117230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0785024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.762e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0256,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12746666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.606364821685087,
      "rewrite6_current_pass_fraction": 0.06648361278261378,
      "rewrite6_full_passes": 2.0,
      "step": 117240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.08096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7625e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0509,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12777333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.13435678680148,
      "rewrite6_current_pass_fraction": 0.06701914276255574,
      "rewrite6_full_passes": 2.0,
      "step": 117250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0834176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.763e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0268,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12809333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.631107301990202,
      "rewrite6_current_pass_fraction": 0.06748651438141419,
      "rewrite6_full_passes": 2.0,
      "step": 117260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0858752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7635e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0184,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12848,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.458531832368465,
      "rewrite6_current_pass_fraction": 0.06823138789896983,
      "rewrite6_full_passes": 2.0,
      "step": 117270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0883328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.764e-05,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.04,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12876,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.905243235092758,
      "rewrite6_current_pass_fraction": 0.06870362797219139,
      "rewrite6_full_passes": 2.0,
      "step": 117280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0907904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7645e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0298,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12912,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.69309355678868,
      "rewrite6_current_pass_fraction": 0.06921481568031781,
      "rewrite6_full_passes": 2.0,
      "step": 117290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.093248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.765e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0281,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12942666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.65794518232533,
      "rewrite6_current_pass_fraction": 0.06983310938443263,
      "rewrite6_full_passes": 2.0,
      "step": 117300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0957056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7655e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0275,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.12981333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.645554132902497,
      "rewrite6_current_pass_fraction": 0.07043679772545812,
      "rewrite6_full_passes": 2.0,
      "step": 117310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.0981632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.766e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0033,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13002666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.15192868118437,
      "rewrite6_current_pass_fraction": 0.0710891706101147,
      "rewrite6_full_passes": 2.0,
      "step": 117320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1006208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7665e-05,
      "grad_norm": 0.515625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0016,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13034666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.11769950546924,
      "rewrite6_current_pass_fraction": 0.07161009522696735,
      "rewrite6_full_passes": 2.0,
      "step": 117330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1030784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.767e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0594,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13058666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.3147644659223,
      "rewrite6_current_pass_fraction": 0.07220404665926662,
      "rewrite6_full_passes": 2.0,
      "step": 117340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.105536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7675e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0148,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13084,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.38501353011574,
      "rewrite6_current_pass_fraction": 0.07276391891102413,
      "rewrite6_full_passes": 2.0,
      "step": 117350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1079936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.768e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13108,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.15797516671621,
      "rewrite6_current_pass_fraction": 0.07332379116278163,
      "rewrite6_full_passes": 2.0,
      "step": 117360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1104512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7685e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0307,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13141333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.711725724207458,
      "rewrite6_current_pass_fraction": 0.07386905805144983,
      "rewrite6_full_passes": 2.0,
      "step": 117370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1129088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.769e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13173333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.957571726540376,
      "rewrite6_current_pass_fraction": 0.07445327257502288,
      "rewrite6_full_passes": 2.0,
      "step": 117380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1153664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7695e-05,
      "grad_norm": 0.5625,
      "host_rss_gib": 16.431495666503906,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0286,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13206666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.668276737590073,
      "rewrite6_current_pass_fraction": 0.07496932873751241,
      "rewrite6_full_passes": 2.0,
      "step": 117390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.117824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.77e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.678752899169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1324,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.701372449879624,
      "rewrite6_current_pass_fraction": 0.07545130571946018,
      "rewrite6_full_passes": 2.0,
      "step": 117400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1202816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7705e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 17.437641143798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0329,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13269333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.757341679953605,
      "rewrite6_current_pass_fraction": 0.07604038869739635,
      "rewrite6_full_passes": 2.0,
      "step": 117410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1227392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.771e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 17.437641143798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0267,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13308,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.629044294412097,
      "rewrite6_current_pass_fraction": 0.07655644485988589,
      "rewrite6_full_passes": 2.0,
      "step": 117420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1251968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7715e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 17.437641143798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0447,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13337333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.003729138875517,
      "rewrite6_current_pass_fraction": 0.07705789565928609,
      "rewrite6_full_passes": 2.0,
      "step": 117430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1276544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.772e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 17.437641143798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0409,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13373333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.924066423168412,
      "rewrite6_current_pass_fraction": 0.07756908336741251,
      "rewrite6_full_passes": 2.0,
      "step": 117440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.130112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7725e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 17.437641143798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0413,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13408,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.932437723886203,
      "rewrite6_current_pass_fraction": 0.07810948180171759,
      "rewrite6_full_passes": 2.0,
      "step": 117450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1325696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.773e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 17.437641143798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0383,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13437333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.869734512558935,
      "rewrite6_current_pass_fraction": 0.07867909096220133,
      "rewrite6_full_passes": 2.0,
      "step": 117460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1350272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7735e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 17.437641143798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0258,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13472,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.610486506804197,
      "rewrite6_current_pass_fraction": 0.07924383166832194,
      "rewrite6_full_passes": 2.0,
      "step": 117470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1374848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.774e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.83525848388672,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0253,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13506666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.600183839432283,
      "rewrite6_current_pass_fraction": 0.07980370392007945,
      "rewrite6_full_passes": 2.0,
      "step": 117480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1399424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7745e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.835262298583984,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13544,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.511793224939044,
      "rewrite6_current_pass_fraction": 0.08033923390002141,
      "rewrite6_full_passes": 2.0,
      "step": 117490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.775e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 16.835586547851562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13574666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.936624630107648,
      "rewrite6_current_pass_fraction": 0.08093318533232069,
      "rewrite6_full_passes": 2.0,
      "step": 117500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.775e-05,
      "eval_loss": 3.081529140472412,
      "eval_perplexity": 21.79169950140359,
      "eval_runtime": 0.0509,
      "eval_samples_per_second": 629.276,
      "eval_steps_per_second": 19.665,
      "step": 117500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1448576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7755e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.89443588256836,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0483,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13605333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.079478831412377,
      "rewrite6_current_pass_fraction": 0.08154661058207241,
      "rewrite6_full_passes": 2.0,
      "step": 117510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1473152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.776e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.89507293701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.044,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13634666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.989031673191437,
      "rewrite6_current_pass_fraction": 0.08212595665128235,
      "rewrite6_full_passes": 2.0,
      "step": 117520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1497728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7765e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.89507293701172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0013,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.111665100823558,
      "rewrite6_current_pass_fraction": 0.08271990808358162,
      "rewrite6_full_passes": 2.0,
      "step": 117530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1522304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.777e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.895553588867188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.034,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13694666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.7801873185992,
      "rewrite6_current_pass_fraction": 0.08328464878970225,
      "rewrite6_full_passes": 2.0,
      "step": 117540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.154688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7775e-05,
      "grad_norm": 0.5625,
      "host_rss_gib": 16.897159576416016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0186,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13732,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.46262394793285,
      "rewrite6_current_pass_fraction": 0.08386399485891219,
      "rewrite6_full_passes": 2.0,
      "step": 117550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1571456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.778e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.897159576416016,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13756,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.08450663083484256,
      "rewrite6_full_passes": 2.0,
      "step": 117560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1596032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7785e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.897563934326172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0342,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13797333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.784343771694374,
      "rewrite6_current_pass_fraction": 0.08516874062822535,
      "rewrite6_full_passes": 2.0,
      "step": 117570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1620608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.779e-05,
      "grad_norm": 0.435546875,
      "host_rss_gib": 16.897567749023438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0195,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13822666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.481048599335452,
      "rewrite6_current_pass_fraction": 0.08584545578469747,
      "rewrite6_full_passes": 2.0,
      "step": 117580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1645184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7795e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 16.897567749023438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0099,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13853333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.285371286682025,
      "rewrite6_current_pass_fraction": 0.08638098576463944,
      "rewrite6_full_passes": 2.0,
      "step": 117590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.166976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.78e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.897567749023438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0426,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1388,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.959667588504388,
      "rewrite6_current_pass_fraction": 0.08702362174056981,
      "rewrite6_full_passes": 2.0,
      "step": 117600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1694336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7805e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.897567749023438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13913333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.08758836244669042,
      "rewrite6_full_passes": 2.0,
      "step": 117610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1718912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.781e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.897567749023438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0157,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13942666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.403368300700652,
      "rewrite6_current_pass_fraction": 0.08823099842262079,
      "rewrite6_full_passes": 2.0,
      "step": 117620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1743488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7815e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.897567749023438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0204,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.13970666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.49948984038855,
      "rewrite6_current_pass_fraction": 0.08877139685692587,
      "rewrite6_full_passes": 2.0,
      "step": 117630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1768064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.782e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.897567749023438,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.022,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14010666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53231527748008,
      "rewrite6_current_pass_fraction": 0.08946271737648731,
      "rewrite6_full_passes": 2.0,
      "step": 117640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.179264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7825e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.897808074951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0048,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14044,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.182179256465627,
      "rewrite6_current_pass_fraction": 0.09000311581079239,
      "rewrite6_full_passes": 2.0,
      "step": 117650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1817216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.783e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 16.897808074951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0197,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14070666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.4851452187036,
      "rewrite6_current_pass_fraction": 0.0905629880625499,
      "rewrite6_full_passes": 2.0,
      "step": 117660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1841792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7835e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 16.897808074951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0191,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14102666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.472857818161167,
      "rewrite6_current_pass_fraction": 0.09107417577067632,
      "rewrite6_full_passes": 2.0,
      "step": 117670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1866368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.784e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.897808074951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0507,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14121333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.13013033810308,
      "rewrite6_current_pass_fraction": 0.09155128429826098,
      "rewrite6_full_passes": 2.0,
      "step": 117680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1890944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7845e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.897808074951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0159,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14158666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.407449382455365,
      "rewrite6_current_pass_fraction": 0.09206734046075052,
      "rewrite6_full_passes": 2.0,
      "step": 117690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.191552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.785e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.897808074951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0369,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14188,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.840537327040092,
      "rewrite6_current_pass_fraction": 0.09264668652996047,
      "rewrite6_full_passes": 2.0,
      "step": 117700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1940096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7855e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.897808074951172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9979,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14232,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.043401553272133,
      "rewrite6_current_pass_fraction": 0.09319682187299176,
      "rewrite6_full_passes": 2.0,
      "step": 117710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1964672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.786e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.8978271484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14256,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.09373235185293373,
      "rewrite6_full_passes": 2.0,
      "step": 117720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.1989248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7865e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.8978271484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0255,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14296,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.60430428823131,
      "rewrite6_current_pass_fraction": 0.09424353956106016,
      "rewrite6_full_passes": 2.0,
      "step": 117730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2013824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.787e-05,
      "grad_norm": 0.52734375,
      "host_rss_gib": 16.8978271484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0309,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14330666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.71586848361443,
      "rewrite6_current_pass_fraction": 0.09479854335845456,
      "rewrite6_full_passes": 2.0,
      "step": 117740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.20384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7875e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 16.8978271484375,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0249,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1436,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.591945413691505,
      "rewrite6_current_pass_fraction": 0.09536328406457519,
      "rewrite6_full_passes": 2.0,
      "step": 117750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2062976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.788e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.89783477783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0049,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14388,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.184197575305536,
      "rewrite6_current_pass_fraction": 0.09591828786196958,
      "rewrite6_full_passes": 2.0,
      "step": 117760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2087552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7885e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.89783477783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0415,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14414666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.936624630107648,
      "rewrite6_current_pass_fraction": 0.09641973866136978,
      "rewrite6_full_passes": 2.0,
      "step": 117770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2112128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.789e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.89783477783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0126,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1446,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.34021579592551,
      "rewrite6_current_pass_fraction": 0.0969796109131273,
      "rewrite6_full_passes": 2.0,
      "step": 117780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2136704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7895e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.89783477783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0118,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14484,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.323950130422475,
      "rewrite6_current_pass_fraction": 0.09765145761523632,
      "rewrite6_full_passes": 2.0,
      "step": 117790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.216128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.79e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.89783477783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0129,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14525333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.346318776065537,
      "rewrite6_current_pass_fraction": 0.0983038304998929,
      "rewrite6_full_passes": 2.0,
      "step": 117800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2185856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7905e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.89783477783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.041,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14553333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.926158934434543,
      "rewrite6_current_pass_fraction": 0.09893186111273393,
      "rewrite6_full_passes": 2.0,
      "step": 117810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2210432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.791e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.89783477783203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0159,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14588,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.407449382455365,
      "rewrite6_current_pass_fraction": 0.0995793655430274,
      "rewrite6_full_passes": 2.0,
      "step": 117820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2235008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7915e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 16.898605346679688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0055,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14617333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.196311727733022,
      "rewrite6_current_pass_fraction": 0.10017331697532668,
      "rewrite6_full_passes": 2.0,
      "step": 117830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2259584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.792e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 16.898605346679688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0205,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14644,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.501539891873456,
      "rewrite6_current_pass_fraction": 0.10071858386399486,
      "rewrite6_full_passes": 2.0,
      "step": 117840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.228416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7925e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.898605346679688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0493,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14673333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.100568853497336,
      "rewrite6_current_pass_fraction": 0.10121029775466885,
      "rewrite6_full_passes": 2.0,
      "step": 117850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2308736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.793e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.898605346679688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0243,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14693333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.579593952252267,
      "rewrite6_current_pass_fraction": 0.10177503846078947,
      "rewrite6_full_passes": 2.0,
      "step": 117860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2333312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7935e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.898605346679688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0234,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14716,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.56108064993093,
      "rewrite6_current_pass_fraction": 0.10228135771455278,
      "rewrite6_full_passes": 2.0,
      "step": 117870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2357888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.794e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.898822784423828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14752,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.678613459924104,
      "rewrite6_current_pass_fraction": 0.10279254542267921,
      "rewrite6_full_passes": 2.0,
      "step": 117880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2382464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7945e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.898822784423828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.037,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14793333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.842621484978952,
      "rewrite6_current_pass_fraction": 0.10344491830733579,
      "rewrite6_full_passes": 2.0,
      "step": 117890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.240704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.795e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.898822784423828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14825333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.844705851344035,
      "rewrite6_current_pass_fraction": 0.10401939592218262,
      "rewrite6_full_passes": 2.0,
      "step": 117900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2431616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7955e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.89940643310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0198,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14852,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.487193835654615,
      "rewrite6_current_pass_fraction": 0.10463282117193433,
      "rewrite6_full_passes": 2.0,
      "step": 117910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2456192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.796e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.89940643310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0121,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14884,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.330048230130824,
      "rewrite6_current_pass_fraction": 0.10520729878678117,
      "rewrite6_full_passes": 2.0,
      "step": 117920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2480768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7965e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.89940643310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0206,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14914666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.503590148373757,
      "rewrite6_current_pass_fraction": 0.1056649334969134,
      "rewrite6_full_passes": 2.0,
      "step": 117930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2505344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.797e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.89940643310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0159,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14948,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.407449382455365,
      "rewrite6_current_pass_fraction": 0.10625888492921268,
      "rewrite6_full_passes": 2.0,
      "step": 117940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.252992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7975e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.89940643310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0216,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.14986666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.52410399373532,
      "rewrite6_current_pass_fraction": 0.1068333625440595,
      "rewrite6_full_passes": 2.0,
      "step": 117950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2554496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.798e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.89940643310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0371,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15018666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.844705851344035,
      "rewrite6_current_pass_fraction": 0.1073250764347335,
      "rewrite6_full_passes": 2.0,
      "step": 117960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2579072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7985e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.89940643310547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0115,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15050666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.317853859869647,
      "rewrite6_current_pass_fraction": 0.10783139568849681,
      "rewrite6_full_passes": 2.0,
      "step": 117970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2603648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.799e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 16.32323455810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0174,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15078666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.4380835263931,
      "rewrite6_current_pass_fraction": 0.10845455784697475,
      "rewrite6_full_passes": 2.0,
      "step": 117980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2628224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.7995e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.32323455810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0575,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15114666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.27430486223215,
      "rewrite6_current_pass_fraction": 0.10904850927927402,
      "rewrite6_full_passes": 2.0,
      "step": 117990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.26528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.32323455810547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0265,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1514,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.624918898106596,
      "rewrite6_current_pass_fraction": 0.10950614398940624,
      "rewrite6_full_passes": 2.0,
      "step": 118000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.8e-05,
      "eval_loss": 3.0819222927093506,
      "eval_perplexity": 21.800268641187117,
      "eval_runtime": 0.0504,
      "eval_samples_per_second": 634.299,
      "eval_steps_per_second": 19.822,
      "step": 118000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2677376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8005e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.422382354736328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0445,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15174666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.999528813094322,
      "rewrite6_current_pass_fraction": 0.10996864715390157,
      "rewrite6_full_passes": 2.0,
      "step": 118010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2701952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.801e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.416133880615234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0553,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15206666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.22755283761896,
      "rewrite6_current_pass_fraction": 0.11054799322311153,
      "rewrite6_full_passes": 2.0,
      "step": 118020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2726528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8015e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.416133880615234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0267,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15252,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.629044294412097,
      "rewrite6_current_pass_fraction": 0.11109326011177971,
      "rewrite6_full_passes": 2.0,
      "step": 118030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2751104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.802e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.416133880615234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0373,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15285333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.848875209436216,
      "rewrite6_current_pass_fraction": 0.11162392163735857,
      "rewrite6_full_passes": 2.0,
      "step": 118040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.277568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8025e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.416133880615234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0099,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15322666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.285371286682025,
      "rewrite6_current_pass_fraction": 0.11227629452201515,
      "rewrite6_full_passes": 2.0,
      "step": 118050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2800256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.803e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.416133880615234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0225,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15358666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.54258400208604,
      "rewrite6_current_pass_fraction": 0.11283616677377266,
      "rewrite6_full_passes": 2.0,
      "step": 118060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2824832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8035e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.416133880615234,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0242,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15374666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.577536095751576,
      "rewrite6_current_pass_fraction": 0.11343498666043504,
      "rewrite6_full_passes": 2.0,
      "step": 118070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2849408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.804e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0202,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15394666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.495390352382937,
      "rewrite6_current_pass_fraction": 0.11406788572763919,
      "rewrite6_full_passes": 2.0,
      "step": 118080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2873984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8045e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0436,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15429333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.980637739420835,
      "rewrite6_current_pass_fraction": 0.11460828416194427,
      "rewrite6_full_passes": 2.0,
      "step": 118090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.289856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.805e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0233,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15461333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.559024644667907,
      "rewrite6_current_pass_fraction": 0.1150707873264396,
      "rewrite6_full_passes": 2.0,
      "step": 118100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2923136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8055e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15505333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.686886559817744,
      "rewrite6_current_pass_fraction": 0.11569394948491753,
      "rewrite6_full_passes": 2.0,
      "step": 118110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2947712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.806e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.985,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15548,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.786502236370332,
      "rewrite6_current_pass_fraction": 0.11648263909174116,
      "rewrite6_full_passes": 2.0,
      "step": 118120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2972288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8065e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0261,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15589333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.61667058032089,
      "rewrite6_current_pass_fraction": 0.11707659052404043,
      "rewrite6_full_passes": 2.0,
      "step": 118130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.2996864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.807e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0601,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15608,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.32969002438444,
      "rewrite6_current_pass_fraction": 0.11764133123016104,
      "rewrite6_full_passes": 2.0,
      "step": 118140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.302144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8075e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0065,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15645333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.21651814098351,
      "rewrite6_current_pass_fraction": 0.11818659811882923,
      "rewrite6_full_passes": 2.0,
      "step": 118150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3046016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.808e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.022,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15676,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53231527748008,
      "rewrite6_current_pass_fraction": 0.11880002336858095,
      "rewrite6_full_passes": 2.0,
      "step": 118160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3070592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8085e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0089,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15706666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.26509605470094,
      "rewrite6_current_pass_fraction": 0.11948160697941618,
      "rewrite6_full_passes": 2.0,
      "step": 118170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3095168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.809e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.059,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15738666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.306240265089755,
      "rewrite6_current_pass_fraction": 0.12008042686607856,
      "rewrite6_full_passes": 2.0,
      "step": 118180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3119744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8095e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0102,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15774666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.29145781100103,
      "rewrite6_current_pass_fraction": 0.12081556347490799,
      "rewrite6_full_passes": 2.0,
      "step": 118190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.314432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.81e-05,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0286,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15802666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.668276737590073,
      "rewrite6_current_pass_fraction": 0.121487410177017,
      "rewrite6_full_passes": 2.0,
      "step": 118200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3168896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8105e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0304,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15838666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.705513138424664,
      "rewrite6_current_pass_fraction": 0.12206675624622695,
      "rewrite6_full_passes": 2.0,
      "step": 118210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3193472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.811e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0236,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15858666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.565193277309945,
      "rewrite6_current_pass_fraction": 0.12262176004362135,
      "rewrite6_full_passes": 2.0,
      "step": 118220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3218048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8115e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0344,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15894666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.788501056163312,
      "rewrite6_current_pass_fraction": 0.12327900138264104,
      "rewrite6_full_passes": 2.0,
      "step": 118230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3242624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.812e-05,
      "grad_norm": 0.55078125,
      "host_rss_gib": 16.417957305908203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0083,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15925333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.252940644055975,
      "rewrite6_current_pass_fraction": 0.12389729508675586,
      "rewrite6_full_passes": 2.0,
      "step": 118240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.32672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8125e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 16.898029327392578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15966666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.12447664115596581,
      "rewrite6_full_passes": 2.0,
      "step": 118250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3291776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.813e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.954654693603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0325,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.15996,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.74904040364757,
      "rewrite6_current_pass_fraction": 0.1250851979513544,
      "rewrite6_full_passes": 2.0,
      "step": 118260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3316352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8135e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.954654693603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0293,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16024,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.68274959621593,
      "rewrite6_current_pass_fraction": 0.12564020174874882,
      "rewrite6_full_passes": 2.0,
      "step": 118270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3340928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.814e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.954654693603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16046666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.210577586354717,
      "rewrite6_current_pass_fraction": 0.12615138945687523,
      "rewrite6_full_passes": 2.0,
      "step": 118280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3365504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8145e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.954654693603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0172,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1608,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.433996318422246,
      "rewrite6_current_pass_fraction": 0.1267015247999065,
      "rewrite6_full_passes": 2.0,
      "step": 118290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.339008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.815e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.954654693603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.02,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16102666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.49129168419294,
      "rewrite6_current_pass_fraction": 0.12726139705166403,
      "rewrite6_full_passes": 2.0,
      "step": 118300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3414656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8155e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.954654693603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9832,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1614,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.750918567254665,
      "rewrite6_current_pass_fraction": 0.12782126930342155,
      "rewrite6_full_passes": 2.0,
      "step": 118310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3439232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.816e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.954654693603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0227,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16170666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.54669292976553,
      "rewrite6_current_pass_fraction": 0.12836166773772661,
      "rewrite6_full_passes": 2.0,
      "step": 118320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3463808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8165e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.954654693603516,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0388,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16201333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.88017198896687,
      "rewrite6_current_pass_fraction": 0.12883390781094817,
      "rewrite6_full_passes": 2.0,
      "step": 118330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3488384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.817e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.350597381591797,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0157,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16232,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.403368300700652,
      "rewrite6_current_pass_fraction": 0.12937430624525326,
      "rewrite6_full_passes": 2.0,
      "step": 118340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.351296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8175e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.351795196533203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0183,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16258666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.456486081474473,
      "rewrite6_current_pass_fraction": 0.129900099316469,
      "rewrite6_full_passes": 2.0,
      "step": 118350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3537536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.818e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.35272216796875,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0145,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16286666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.378898943290586,
      "rewrite6_current_pass_fraction": 0.13043562929641098,
      "rewrite6_full_passes": 2.0,
      "step": 118360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3562112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8185e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.352825164794922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0257,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16317333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.60842556120252,
      "rewrite6_current_pass_fraction": 0.1310101069112578,
      "rewrite6_full_passes": 2.0,
      "step": 118370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3586688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.819e-05,
      "grad_norm": 0.54296875,
      "host_rss_gib": 16.354415893554688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0375,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16344,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.853045401483406,
      "rewrite6_current_pass_fraction": 0.13164300597846196,
      "rewrite6_full_passes": 2.0,
      "step": 118380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3611264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8195e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.354415893554688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16378666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.759417517911764,
      "rewrite6_current_pass_fraction": 0.13224182586512434,
      "rewrite6_full_passes": 2.0,
      "step": 118390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.363584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.82e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.354930877685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0127,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16416,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.342249919209575,
      "rewrite6_current_pass_fraction": 0.1328844618410547,
      "rewrite6_full_passes": 2.0,
      "step": 118400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3660416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8205e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.354930877685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0171,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16450666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.431953020956982,
      "rewrite6_current_pass_fraction": 0.13351249245389574,
      "rewrite6_full_passes": 2.0,
      "step": 118410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3684992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.821e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.354930877685547,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9964,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16474666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.013358988498783,
      "rewrite6_current_pass_fraction": 0.13412591770364746,
      "rewrite6_full_passes": 2.0,
      "step": 118420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3709568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8215e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 16.356338500976562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16496,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.188234818531463,
      "rewrite6_current_pass_fraction": 0.13471986913594672,
      "rewrite6_full_passes": 2.0,
      "step": 118430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3734144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.822e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.356338500976562,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0264,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16521333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.622856509337947,
      "rewrite6_current_pass_fraction": 0.13516289848298962,
      "rewrite6_full_passes": 2.0,
      "step": 118440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.375872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8225e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 16.356342315673828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16548,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.69516296961328,
      "rewrite6_current_pass_fraction": 0.1357665868240151,
      "rewrite6_full_passes": 2.0,
      "step": 118450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3783296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.823e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.356342315673828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.007,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16584,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.226628927540002,
      "rewrite6_current_pass_fraction": 0.13632159062140953,
      "rewrite6_full_passes": 2.0,
      "step": 118460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3807872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8235e-05,
      "grad_norm": 0.53125,
      "host_rss_gib": 16.356342315673828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0277,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16608,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.649683656667687,
      "rewrite6_current_pass_fraction": 0.136925278962435,
      "rewrite6_full_passes": 2.0,
      "step": 118470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3832448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.824e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.356342315673828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0338,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16652,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.77603169671152,
      "rewrite6_current_pass_fraction": 0.1375338357578236,
      "rewrite6_full_passes": 2.0,
      "step": 118480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3857024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8245e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.357654571533203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0385,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16682666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.873908876883966,
      "rewrite6_current_pass_fraction": 0.13817647173375397,
      "rewrite6_full_passes": 2.0,
      "step": 118490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.38816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.825e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.357654571533203,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0474,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16706666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.06051583509246,
      "rewrite6_current_pass_fraction": 0.13858055344589199,
      "rewrite6_full_passes": 2.0,
      "step": 118500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.825e-05,
      "eval_loss": 3.081754207611084,
      "eval_perplexity": 21.796604648830208,
      "eval_runtime": 0.0501,
      "eval_samples_per_second": 639.029,
      "eval_steps_per_second": 19.97,
      "step": 118500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3906176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8255e-05,
      "grad_norm": 0.546875,
      "host_rss_gib": 16.414337158203125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0521,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16734666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.159733237771047,
      "rewrite6_current_pass_fraction": 0.13921832096745926,
      "rewrite6_full_passes": 2.0,
      "step": 118510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3930752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.826e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 16.414337158203125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0359,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16764,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.819707206509154,
      "rewrite6_current_pass_fraction": 0.13970516640377012,
      "rewrite6_full_passes": 2.0,
      "step": 118520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3955328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8265e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.414337158203125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0187,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16784,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.46467031264417,
      "rewrite6_current_pass_fraction": 0.14023095947498587,
      "rewrite6_full_passes": 2.0,
      "step": 118530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.3979904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.827e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.414337158203125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0187,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16808,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.46467031264417,
      "rewrite6_current_pass_fraction": 0.14085899008782693,
      "rewrite6_full_passes": 2.0,
      "step": 118540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.400448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8275e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.414817810058594,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0288,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16838666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.672410806330685,
      "rewrite6_current_pass_fraction": 0.14148215224630484,
      "rewrite6_full_passes": 2.0,
      "step": 118550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4029056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.828e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.415245056152344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0263,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16868,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.620794326797853,
      "rewrite6_current_pass_fraction": 0.1420566298611517,
      "rewrite6_full_passes": 2.0,
      "step": 118560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4053632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8285e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.415245056152344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0121,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16902666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.330048230130824,
      "rewrite6_current_pass_fraction": 0.14259702829545676,
      "rewrite6_full_passes": 2.0,
      "step": 118570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4078208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.829e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.415245056152344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9937,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16952,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.95939580231382,
      "rewrite6_current_pass_fraction": 0.14326400654320268,
      "rewrite6_full_passes": 2.0,
      "step": 118580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4102784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8295e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.415245056152344,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0337,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.16981333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.77395419741855,
      "rewrite6_current_pass_fraction": 0.1438823002473175,
      "rewrite6_full_passes": 2.0,
      "step": 118590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.412736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.83e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0193,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17021333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.47695279920925,
      "rewrite6_current_pass_fraction": 0.14452980467761095,
      "rewrite6_full_passes": 2.0,
      "step": 118600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4151936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8305e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0502,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17057333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.11956791376016,
      "rewrite6_current_pass_fraction": 0.14511888765554712,
      "rewrite6_full_passes": 2.0,
      "step": 118610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4176512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.831e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0179,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17078666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.44830512334259,
      "rewrite6_current_pass_fraction": 0.14569336527039395,
      "rewrite6_full_passes": 2.0,
      "step": 118620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4201088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8315e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.014,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17098666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.368712040756797,
      "rewrite6_current_pass_fraction": 0.14620942143288349,
      "rewrite6_full_passes": 2.0,
      "step": 118630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4225664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.832e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0404,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17117333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.913607005029263,
      "rewrite6_current_pass_fraction": 0.14678389904773032,
      "rewrite6_full_passes": 2.0,
      "step": 118640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.425024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8325e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0184,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17153333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.458531832368465,
      "rewrite6_current_pass_fraction": 0.14740219275184516,
      "rewrite6_full_passes": 2.0,
      "step": 118650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4274816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.833e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0133,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17193333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.354458931498517,
      "rewrite6_current_pass_fraction": 0.14802048645595997,
      "rewrite6_full_passes": 2.0,
      "step": 118660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4299392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8335e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0194,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17232,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.479000596877352,
      "rewrite6_current_pass_fraction": 0.14860470097953302,
      "rewrite6_full_passes": 2.0,
      "step": 118670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4323968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.834e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0241,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17254666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.57547844502625,
      "rewrite6_current_pass_fraction": 0.14916944168565363,
      "rewrite6_full_passes": 2.0,
      "step": 118680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4348544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8345e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0201,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17284,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.49334091582123,
      "rewrite6_current_pass_fraction": 0.1496514186676014,
      "rewrite6_full_passes": 2.0,
      "step": 118690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.437312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.835e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0264,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17310666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.622856509337947,
      "rewrite6_current_pass_fraction": 0.15018694864754337,
      "rewrite6_full_passes": 2.0,
      "step": 118700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4397696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8355e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17336,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.1506981363556698,
      "rewrite6_full_passes": 2.0,
      "step": 118710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4422272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.836e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17368,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.821789281331817,
      "rewrite6_current_pass_fraction": 0.15127748242487976,
      "rewrite6_full_passes": 2.0,
      "step": 118720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4446848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8365e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0168,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17396,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.425824354396642,
      "rewrite6_current_pass_fraction": 0.15178867013300618,
      "rewrite6_full_passes": 2.0,
      "step": 118730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4471424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.837e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0343,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17425333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.78642230999673,
      "rewrite6_current_pass_fraction": 0.15228038402368016,
      "rewrite6_full_passes": 2.0,
      "step": 118740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8375e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0403,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17449333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.911515748893304,
      "rewrite6_current_pass_fraction": 0.15278183482308036,
      "rewrite6_full_passes": 2.0,
      "step": 118750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4520576,
      "curriculum_late_phase": 1.0,
      "epoch": 1.838e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9951,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17486666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.987358525776234,
      "rewrite6_current_pass_fraction": 0.15337578625537965,
      "rewrite6_full_passes": 2.0,
      "step": 118760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4545152,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8385e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0214,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17522666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.51999958339129,
      "rewrite6_current_pass_fraction": 0.15387723705477985,
      "rewrite6_full_passes": 2.0,
      "step": 118770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4569728,
      "curriculum_late_phase": 1.0,
      "epoch": 1.839e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0366,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1756,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.83428610357238,
      "rewrite6_current_pass_fraction": 0.15451987303071021,
      "rewrite6_full_passes": 2.0,
      "step": 118780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4594304,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8395e-05,
      "grad_norm": 0.5078125,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0124,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17586666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.33614815954352,
      "rewrite6_current_pass_fraction": 0.1551284298260988,
      "rewrite6_full_passes": 2.0,
      "step": 118790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.461888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.84e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0378,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17614666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.859302253584733,
      "rewrite6_current_pass_fraction": 0.1557321181671243,
      "rewrite6_full_passes": 2.0,
      "step": 118800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4643456,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8405e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0296,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1766,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.688955351911602,
      "rewrite6_current_pass_fraction": 0.15636988568869156,
      "rewrite6_full_passes": 2.0,
      "step": 118810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4668032,
      "curriculum_late_phase": 1.0,
      "epoch": 1.841e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0439,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17690666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.986932874965774,
      "rewrite6_current_pass_fraction": 0.15686646803372864,
      "rewrite6_full_passes": 2.0,
      "step": 118820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4692608,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8415e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 16.41616439819336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0022,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17726666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.129773747082783,
      "rewrite6_current_pass_fraction": 0.1575626570076532,
      "rewrite6_full_passes": 2.0,
      "step": 118830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4717184,
      "curriculum_late_phase": 1.0,
      "epoch": 1.842e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.4176025390625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9997,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17756,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.07951216586949,
      "rewrite6_current_pass_fraction": 0.15823450370976222,
      "rewrite6_full_passes": 2.0,
      "step": 118840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.474176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8425e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.417827606201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0339,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17781333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.77810940376482,
      "rewrite6_current_pass_fraction": 0.15876029678097797,
      "rewrite6_full_passes": 2.0,
      "step": 118850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4766336,
      "curriculum_late_phase": 1.0,
      "epoch": 1.843e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 16.417827606201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0179,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17829333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.44830512334259,
      "rewrite6_current_pass_fraction": 0.15931043212400928,
      "rewrite6_full_passes": 2.0,
      "step": 118860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4790912,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8435e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 16.417827606201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0252,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17874666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.59812392404582,
      "rewrite6_current_pass_fraction": 0.1597631983797784,
      "rewrite6_full_passes": 2.0,
      "step": 118870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4815488,
      "curriculum_late_phase": 1.0,
      "epoch": 1.844e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.417827606201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0476,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17906666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.064728359497874,
      "rewrite6_current_pass_fraction": 0.16034254444898832,
      "rewrite6_full_passes": 2.0,
      "step": 118880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4840064,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8445e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.417827606201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0436,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17937333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.980637739420835,
      "rewrite6_current_pass_fraction": 0.16089754824638275,
      "rewrite6_full_passes": 2.0,
      "step": 118890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.486464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.845e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 16.418411254882812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0286,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17961333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.668276737590073,
      "rewrite6_current_pass_fraction": 0.16145742049814024,
      "rewrite6_full_passes": 2.0,
      "step": 118900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4889216,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8455e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.418411254882812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0211,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.17984,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.513844506823926,
      "rewrite6_current_pass_fraction": 0.16202216120426088,
      "rewrite6_full_passes": 2.0,
      "step": 118910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4913792,
      "curriculum_late_phase": 1.0,
      "epoch": 1.846e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.418411254882812,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0319,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1802,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.736594713485793,
      "rewrite6_current_pass_fraction": 0.16263071799964948,
      "rewrite6_full_passes": 2.0,
      "step": 118920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4938368,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8465e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.934398651123047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0145,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18050666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.378898943290586,
      "rewrite6_current_pass_fraction": 0.1632587486124905,
      "rewrite6_full_passes": 2.0,
      "step": 118930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.4962944,
      "curriculum_late_phase": 1.0,
      "epoch": 1.847e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.66576385498047,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0472,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18074666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.05630415310768,
      "rewrite6_current_pass_fraction": 0.16390625304278397,
      "rewrite6_full_passes": 2.0,
      "step": 118940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.498752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8475e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 17.147502899169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0342,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18112,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.784343771694374,
      "rewrite6_current_pass_fraction": 0.16444178302272594,
      "rewrite6_full_passes": 2.0,
      "step": 118950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5012096,
      "curriculum_late_phase": 1.0,
      "epoch": 1.848e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 17.147502899169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0218,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18152,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.528209225043515,
      "rewrite6_current_pass_fraction": 0.16505033981811454,
      "rewrite6_full_passes": 2.0,
      "step": 118960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5036672,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8485e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 17.147502899169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0457,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18189333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.024743373380456,
      "rewrite6_current_pass_fraction": 0.1656248174329614,
      "rewrite6_full_passes": 2.0,
      "step": 118970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5061248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.849e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 17.147502899169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0178,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1822,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.446260395068368,
      "rewrite6_current_pass_fraction": 0.16626258495452864,
      "rewrite6_full_passes": 2.0,
      "step": 118980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5085824,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8495e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 17.147502899169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0118,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18245333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.323950130422475,
      "rewrite6_current_pass_fraction": 0.16682732566064926,
      "rewrite6_full_passes": 2.0,
      "step": 118990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.51104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.85e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 17.147502899169922,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18276,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.511793224939044,
      "rewrite6_current_pass_fraction": 0.16748456699966893,
      "rewrite6_full_passes": 2.0,
      "step": 119000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.85e-05,
      "eval_loss": 3.081606388092041,
      "eval_perplexity": 21.79338292333707,
      "eval_runtime": 0.0495,
      "eval_samples_per_second": 646.101,
      "eval_steps_per_second": 20.191,
      "step": 119000
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5134976,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8505e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 17.15176773071289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0467,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18305333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.045778632630523,
      "rewrite6_current_pass_fraction": 0.1680200969796109,
      "rewrite6_full_passes": 2.0,
      "step": 119010,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5159552,
      "curriculum_late_phase": 1.0,
      "epoch": 1.851e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 17.15176773071289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18329333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.041569897791508,
      "rewrite6_current_pass_fraction": 0.16860917995754707,
      "rewrite6_full_passes": 2.0,
      "step": 119020,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5184128,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8515e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 17.15176773071289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0542,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18358666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.20421536745936,
      "rewrite6_current_pass_fraction": 0.1691203676656735,
      "rewrite6_full_passes": 2.0,
      "step": 119030,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5208704,
      "curriculum_late_phase": 1.0,
      "epoch": 1.852e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 17.15176773071289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.017,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18385333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.42990992781124,
      "rewrite6_current_pass_fraction": 0.1696169500107106,
      "rewrite6_full_passes": 2.0,
      "step": 119040,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.523328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8525e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 17.02906036376953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0295,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18414666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.686886559817744,
      "rewrite6_current_pass_fraction": 0.1702839282584565,
      "rewrite6_full_passes": 2.0,
      "step": 119050,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5257856,
      "curriculum_late_phase": 1.0,
      "epoch": 1.853e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 17.02906036376953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0291,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18450666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.678613459924104,
      "rewrite6_current_pass_fraction": 0.17087787969075577,
      "rewrite6_full_passes": 2.0,
      "step": 119060,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5282432,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8535e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 17.02906036376953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0362,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18474666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.825954055651632,
      "rewrite6_current_pass_fraction": 0.1715010418492337,
      "rewrite6_full_passes": 2.0,
      "step": 119070,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5307008,
      "curriculum_late_phase": 1.0,
      "epoch": 1.854e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 17.02906036376953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.033,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18513333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.759417517911764,
      "rewrite6_current_pass_fraction": 0.1720463087379019,
      "rewrite6_full_passes": 2.0,
      "step": 119080,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5331584,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8545e-05,
      "grad_norm": 0.5234375,
      "host_rss_gib": 17.02906036376953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0366,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18548,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.83428610357238,
      "rewrite6_current_pass_fraction": 0.17263539171583805,
      "rewrite6_full_passes": 2.0,
      "step": 119090,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.535616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.855e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0327,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18570666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.753190626736775,
      "rewrite6_current_pass_fraction": 0.17320986933068488,
      "rewrite6_full_passes": 2.0,
      "step": 119100,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5380736,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8555e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0265,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.186,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.624918898106596,
      "rewrite6_current_pass_fraction": 0.17368697785826956,
      "rewrite6_full_passes": 2.0,
      "step": 119110,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5405312,
      "curriculum_late_phase": 1.0,
      "epoch": 1.856e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0398,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18625333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.901062604522732,
      "rewrite6_current_pass_fraction": 0.17423711320130086,
      "rewrite6_full_passes": 2.0,
      "step": 119120,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5429888,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8565e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0545,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1866,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.210577586354717,
      "rewrite6_current_pass_fraction": 0.17495277599267783,
      "rewrite6_full_passes": 2.0,
      "step": 119130,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5454464,
      "curriculum_late_phase": 1.0,
      "epoch": 1.857e-05,
      "grad_norm": 0.59765625,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0114,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18690666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.31582217606955,
      "rewrite6_current_pass_fraction": 0.175541858970614,
      "rewrite6_full_passes": 2.0,
      "step": 119140,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.547904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8575e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0519,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18721333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.155501714289947,
      "rewrite6_current_pass_fraction": 0.1760287044069249,
      "rewrite6_full_passes": 2.0,
      "step": 119150,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5503616,
      "curriculum_late_phase": 1.0,
      "epoch": 1.858e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0208,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18749333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.507691276502573,
      "rewrite6_current_pass_fraction": 0.1765788397499562,
      "rewrite6_full_passes": 2.0,
      "step": 119160,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5528192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8585e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0121,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18793333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.330048230130824,
      "rewrite6_current_pass_fraction": 0.17705594827754084,
      "rewrite6_full_passes": 2.0,
      "step": 119170,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5552768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.859e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0464,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1882,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.039465846006077,
      "rewrite6_current_pass_fraction": 0.1776352943467508,
      "rewrite6_full_passes": 2.0,
      "step": 119180,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5577344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8595e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0423,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1886,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.953380631318563,
      "rewrite6_current_pass_fraction": 0.1782438511421394,
      "rewrite6_full_passes": 2.0,
      "step": 119190,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.560192,
      "curriculum_late_phase": 1.0,
      "epoch": 1.86e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 17.029155731201172,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0299,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18889333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.69516296961328,
      "rewrite6_current_pass_fraction": 0.17886701330061733,
      "rewrite6_full_passes": 2.0,
      "step": 119200,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5626496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8605e-05,
      "grad_norm": 0.455078125,
      "host_rss_gib": 15.29825210571289,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.03,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18926666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.6972325893895,
      "rewrite6_current_pass_fraction": 0.1794707016416428,
      "rewrite6_full_passes": 2.0,
      "step": 119210,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5651072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.861e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.299972534179688,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0122,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18968,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.332081336607462,
      "rewrite6_current_pass_fraction": 0.1800743899826683,
      "rewrite6_full_passes": 2.0,
      "step": 119220,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5675648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8615e-05,
      "grad_norm": 0.462890625,
      "host_rss_gib": 15.301929473876953,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0126,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.18990666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.34021579592551,
      "rewrite6_current_pass_fraction": 0.18065373605187826,
      "rewrite6_full_passes": 2.0,
      "step": 119230,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5700224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.862e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.301979064941406,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0257,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19010666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.60842556120252,
      "rewrite6_current_pass_fraction": 0.18122334521236197,
      "rewrite6_full_passes": 2.0,
      "step": 119240,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.57248,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8625e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.304279327392578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0272,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19049333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.639361395619666,
      "rewrite6_current_pass_fraction": 0.18185137582520303,
      "rewrite6_full_passes": 2.0,
      "step": 119250,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5749376,
      "curriculum_late_phase": 1.0,
      "epoch": 1.863e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.304279327392578,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.018,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1908,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.45035005609985,
      "rewrite6_current_pass_fraction": 0.18235282662460323,
      "rewrite6_full_passes": 2.0,
      "step": 119260,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5773952,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8635e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.30434799194336,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0262,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19106666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.618732350465706,
      "rewrite6_current_pass_fraction": 0.18291756733072384,
      "rewrite6_full_passes": 2.0,
      "step": 119270,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5798528,
      "curriculum_late_phase": 1.0,
      "epoch": 1.864e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.304828643798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.042,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19138666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.947095559937008,
      "rewrite6_current_pass_fraction": 0.18342875503885026,
      "rewrite6_full_passes": 2.0,
      "step": 119280,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5823104,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8645e-05,
      "grad_norm": 0.494140625,
      "host_rss_gib": 15.304828643798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0121,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19172,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.330048230130824,
      "rewrite6_current_pass_fraction": 0.18412007555841173,
      "rewrite6_full_passes": 2.0,
      "step": 119290,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.584768,
      "curriculum_late_phase": 1.0,
      "epoch": 1.865e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 15.304828643798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0092,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19208,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.271176495537865,
      "rewrite6_current_pass_fraction": 0.18468968471889544,
      "rewrite6_full_passes": 2.0,
      "step": 119300,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5872256,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8655e-05,
      "grad_norm": 0.451171875,
      "host_rss_gib": 15.304828643798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0243,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19241333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.579593952252267,
      "rewrite6_current_pass_fraction": 0.1853810052384569,
      "rewrite6_full_passes": 2.0,
      "step": 119310,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5896832,
      "curriculum_late_phase": 1.0,
      "epoch": 1.866e-05,
      "grad_norm": 0.447265625,
      "host_rss_gib": 15.304828643798828,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0326,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19273333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.75111541143659,
      "rewrite6_current_pass_fraction": 0.18596521976202995,
      "rewrite6_full_passes": 2.0,
      "step": 119320,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5921408,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8665e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.304832458496094,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0435,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19308,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.97853978054658,
      "rewrite6_current_pass_fraction": 0.1865348289225137,
      "rewrite6_full_passes": 2.0,
      "step": 119330,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5945984,
      "curriculum_late_phase": 1.0,
      "epoch": 1.867e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0358,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19330666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.817625339883573,
      "rewrite6_current_pass_fraction": 0.18710930653736052,
      "rewrite6_full_passes": 2.0,
      "step": 119340,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.597056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8675e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0381,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19356,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.86556098302329,
      "rewrite6_current_pass_fraction": 0.18761562579112384,
      "rewrite6_full_passes": 2.0,
      "step": 119350,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.5995136,
      "curriculum_late_phase": 1.0,
      "epoch": 1.868e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0117,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19389333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.32191783702579,
      "rewrite6_current_pass_fraction": 0.18817549804288133,
      "rewrite6_full_passes": 2.0,
      "step": 119360,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6019712,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8685e-05,
      "grad_norm": 0.4453125,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0419,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19422666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.945000955113006,
      "rewrite6_current_pass_fraction": 0.18883273938190104,
      "rewrite6_full_passes": 2.0,
      "step": 119370,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6044288,
      "curriculum_late_phase": 1.0,
      "epoch": 1.869e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0181,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19449333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.452395193360626,
      "rewrite6_current_pass_fraction": 0.18938774317929544,
      "rewrite6_full_passes": 2.0,
      "step": 119380,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6068864,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8695e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0165,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19485333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.419697526160515,
      "rewrite6_current_pass_fraction": 0.19000603688341025,
      "rewrite6_full_passes": 2.0,
      "step": 119390,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.609344,
      "curriculum_late_phase": 1.0,
      "epoch": 1.87e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0213,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19516,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.517947686029533,
      "rewrite6_current_pass_fraction": 0.19059998831570954,
      "rewrite6_full_passes": 2.0,
      "step": 119400,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6118016,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8705e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0195,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1954,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.481048599335452,
      "rewrite6_current_pass_fraction": 0.19115012365874082,
      "rewrite6_full_passes": 2.0,
      "step": 119410,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6142592,
      "curriculum_late_phase": 1.0,
      "epoch": 1.871e-05,
      "grad_norm": 0.484375,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0754,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19561333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.65854359649817,
      "rewrite6_current_pass_fraction": 0.1916369690950517,
      "rewrite6_full_passes": 2.0,
      "step": 119420,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6167168,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8715e-05,
      "grad_norm": 0.56640625,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0389,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.196,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.882260110570105,
      "rewrite6_current_pass_fraction": 0.19216276216626746,
      "rewrite6_full_passes": 2.0,
      "step": 119430,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6191744,
      "curriculum_late_phase": 1.0,
      "epoch": 1.872e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9922,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19626666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.929479151707675,
      "rewrite6_current_pass_fraction": 0.19278105587038227,
      "rewrite6_full_passes": 2.0,
      "step": 119440,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.621632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8725e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0568,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19664,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.259418059817307,
      "rewrite6_current_pass_fraction": 0.19342856030067573,
      "rewrite6_full_passes": 2.0,
      "step": 119450,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6240896,
      "curriculum_late_phase": 1.0,
      "epoch": 1.873e-05,
      "grad_norm": 0.478515625,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0237,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19698666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.567249899467065,
      "rewrite6_current_pass_fraction": 0.1939056688282604,
      "rewrite6_full_passes": 2.0,
      "step": 119460,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6265472,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8735e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0456,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19729333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.022641004163326,
      "rewrite6_current_pass_fraction": 0.19442172499074994,
      "rewrite6_full_passes": 2.0,
      "step": 119470,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6290048,
      "curriculum_late_phase": 1.0,
      "epoch": 1.874e-05,
      "grad_norm": 0.421875,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0284,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19764,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.664143495580532,
      "rewrite6_current_pass_fraction": 0.19504001869486476,
      "rewrite6_full_passes": 2.0,
      "step": 119480,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6314624,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8745e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0112,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19792,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.311759417923692,
      "rewrite6_current_pass_fraction": 0.19564857549025336,
      "rewrite6_full_passes": 2.0,
      "step": 119490,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.63392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.875e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 15.3052978515625,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0449,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19822666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.007930304805882,
      "rewrite6_current_pass_fraction": 0.19623765846818952,
      "rewrite6_full_passes": 2.0,
      "step": 119500,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.875e-05,
      "eval_loss": 3.081559181213379,
      "eval_perplexity": 21.79235415003655,
      "eval_runtime": 0.0497,
      "eval_samples_per_second": 644.092,
      "eval_steps_per_second": 20.128,
      "step": 119500
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6363776,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8755e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0047,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19861333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.180161139447517,
      "rewrite6_current_pass_fraction": 0.19680239917431014,
      "rewrite6_full_passes": 2.0,
      "step": 119510,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6388352,
      "curriculum_late_phase": 1.0,
      "epoch": 1.876e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0229,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19890666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.55080267931274,
      "rewrite6_current_pass_fraction": 0.19734766606297832,
      "rewrite6_full_passes": 2.0,
      "step": 119520,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6412928,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8765e-05,
      "grad_norm": 0.4765625,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0195,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19921333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.481048599335452,
      "rewrite6_current_pass_fraction": 0.1979367490409145,
      "rewrite6_full_passes": 2.0,
      "step": 119530,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6437504,
      "curriculum_late_phase": 1.0,
      "epoch": 1.877e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0051,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.1996,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.188234818531463,
      "rewrite6_current_pass_fraction": 0.19848201592958267,
      "rewrite6_full_passes": 2.0,
      "step": 119540,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.646208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8775e-05,
      "grad_norm": 0.51953125,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.19984,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.1990467566357033,
      "rewrite6_full_passes": 2.0,
      "step": 119550,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6486656,
      "curriculum_late_phase": 1.0,
      "epoch": 1.878e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0171,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20016,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.431953020956982,
      "rewrite6_current_pass_fraction": 0.1995968919787346,
      "rewrite6_full_passes": 2.0,
      "step": 119560,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6511232,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8785e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20042666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.81346223114035,
      "rewrite6_current_pass_fraction": 0.20016650113921833,
      "rewrite6_full_passes": 2.0,
      "step": 119570,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6535808,
      "curriculum_late_phase": 1.0,
      "epoch": 1.879e-05,
      "grad_norm": 0.51171875,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0367,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20064,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.836369636357645,
      "rewrite6_current_pass_fraction": 0.20068742575607096,
      "rewrite6_full_passes": 2.0,
      "step": 119580,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6560384,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8795e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0198,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20105333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.487193835654615,
      "rewrite6_current_pass_fraction": 0.20130085100582268,
      "rewrite6_full_passes": 2.0,
      "step": 119590,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.658496,
      "curriculum_late_phase": 1.0,
      "epoch": 1.88e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0266,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20148,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.626981493124443,
      "rewrite6_current_pass_fraction": 0.2018655917119433,
      "rewrite6_full_passes": 2.0,
      "step": 119600,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6609536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8805e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 15.386096954345703,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9981,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20170666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.047410634477544,
      "rewrite6_current_pass_fraction": 0.2024205955093377,
      "rewrite6_full_passes": 2.0,
      "step": 119610,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6634112,
      "curriculum_late_phase": 1.0,
      "epoch": 1.881e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 15.386493682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0248,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20213333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.589886322106427,
      "rewrite6_current_pass_fraction": 0.20301454694163698,
      "rewrite6_full_passes": 2.0,
      "step": 119620,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6658688,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8815e-05,
      "grad_norm": 0.50390625,
      "host_rss_gib": 15.386493682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0111,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20256,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.30972834353731,
      "rewrite6_current_pass_fraction": 0.20357441919339447,
      "rewrite6_full_passes": 2.0,
      "step": 119630,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6683264,
      "curriculum_late_phase": 1.0,
      "epoch": 1.882e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.386493682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.036,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20288,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.821789281331817,
      "rewrite6_current_pass_fraction": 0.20428034507604526,
      "rewrite6_full_passes": 2.0,
      "step": 119640,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.670784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8825e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 15.386493682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0347,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20328,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.79473854205626,
      "rewrite6_current_pass_fraction": 0.2049570602325174,
      "rewrite6_full_passes": 2.0,
      "step": 119650,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6732416,
      "curriculum_late_phase": 1.0,
      "epoch": 1.883e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 15.386493682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.016,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20354666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.409490229434265,
      "rewrite6_current_pass_fraction": 0.20556074857354287,
      "rewrite6_full_passes": 2.0,
      "step": 119660,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6756992,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8835e-05,
      "grad_norm": 0.470703125,
      "host_rss_gib": 15.386493682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20388,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.20610114700784796,
      "rewrite6_full_passes": 2.0,
      "step": 119670,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6781568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.884e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 15.386493682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0275,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20417333333333335,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.645554132902497,
      "rewrite6_current_pass_fraction": 0.20664154544215302,
      "rewrite6_full_passes": 2.0,
      "step": 119680,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6806144,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8845e-05,
      "grad_norm": 0.482421875,
      "host_rss_gib": 15.386493682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0432,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20450666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.972247162552314,
      "rewrite6_current_pass_fraction": 0.20724523378317852,
      "rewrite6_full_passes": 2.0,
      "step": 119690,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.683072,
      "curriculum_late_phase": 1.0,
      "epoch": 1.885e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 15.386493682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0483,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20481333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.079478831412377,
      "rewrite6_current_pass_fraction": 0.2077953691262098,
      "rewrite6_full_passes": 2.0,
      "step": 119700,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6855296,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8855e-05,
      "grad_norm": 0.57421875,
      "host_rss_gib": 15.386493682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0278,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20525333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.651748728285217,
      "rewrite6_current_pass_fraction": 0.2084574789195926,
      "rewrite6_full_passes": 2.0,
      "step": 119710,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6879872,
      "curriculum_late_phase": 1.0,
      "epoch": 1.886e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.573993682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0538,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20548,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.19573537742345,
      "rewrite6_current_pass_fraction": 0.20904656189752877,
      "rewrite6_full_passes": 2.0,
      "step": 119720,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6904448,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8865e-05,
      "grad_norm": 0.439453125,
      "host_rss_gib": 14.573993682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0162,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20573333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.413572535697167,
      "rewrite6_current_pass_fraction": 0.20962103951237562,
      "rewrite6_full_passes": 2.0,
      "step": 119730,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6929024,
      "curriculum_late_phase": 1.0,
      "epoch": 1.887e-05,
      "grad_norm": 0.57421875,
      "host_rss_gib": 14.573993682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0228,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20593333333333333,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.5487477017954,
      "rewrite6_current_pass_fraction": 0.21013709567486513,
      "rewrite6_full_passes": 2.0,
      "step": 119740,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.69536,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8875e-05,
      "grad_norm": 0.443359375,
      "host_rss_gib": 14.573993682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0275,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20634666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.645554132902497,
      "rewrite6_current_pass_fraction": 0.21085762692060525,
      "rewrite6_full_passes": 2.0,
      "step": 119750,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.6978176,
      "curriculum_late_phase": 1.0,
      "epoch": 1.888e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 14.573993682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.009,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20661333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.267122665635263,
      "rewrite6_current_pass_fraction": 0.2113882884461841,
      "rewrite6_full_passes": 2.0,
      "step": 119760,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7002752,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8885e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 14.573993682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0169,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20696,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.42786703896461,
      "rewrite6_current_pass_fraction": 0.21200171369593582,
      "rewrite6_full_passes": 2.0,
      "step": 119770,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7027328,
      "curriculum_late_phase": 1.0,
      "epoch": 1.889e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 14.573993682861328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0138,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20726666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.36463870569573,
      "rewrite6_current_pass_fraction": 0.21263461276313997,
      "rewrite6_full_passes": 2.0,
      "step": 119780,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7051904,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8895e-05,
      "grad_norm": 0.45703125,
      "host_rss_gib": 15.124946594238281,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.029,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20754666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.67654570196773,
      "rewrite6_current_pass_fraction": 0.21322369574107614,
      "rewrite6_full_passes": 2.0,
      "step": 119790,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.707648,
      "curriculum_late_phase": 1.0,
      "epoch": 1.89e-05,
      "grad_norm": 0.486328125,
      "host_rss_gib": 16.110870361328125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0356,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20788,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.81346223114035,
      "rewrite6_current_pass_fraction": 0.2138273840821016,
      "rewrite6_full_passes": 2.0,
      "step": 119800,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7101056,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8905e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0253,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20817333333333332,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.600183839432283,
      "rewrite6_current_pass_fraction": 0.2143580456076805,
      "rewrite6_full_passes": 2.0,
      "step": 119810,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7125632,
      "curriculum_late_phase": 1.0,
      "epoch": 1.891e-05,
      "grad_norm": 0.54296875,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0267,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2084,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.629044294412097,
      "rewrite6_current_pass_fraction": 0.21487897022453312,
      "rewrite6_full_passes": 2.0,
      "step": 119820,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7150208,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8915e-05,
      "grad_norm": 0.44921875,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0283,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20870666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.662077184548252,
      "rewrite6_current_pass_fraction": 0.21543884247629064,
      "rewrite6_full_passes": 2.0,
      "step": 119830,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7174784,
      "curriculum_late_phase": 1.0,
      "epoch": 1.892e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0294,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20898666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.684817974592743,
      "rewrite6_current_pass_fraction": 0.21611555763276274,
      "rewrite6_full_passes": 2.0,
      "step": 119840,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.719936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8925e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.021,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20922666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.511793224939044,
      "rewrite6_current_pass_fraction": 0.2166949037019727,
      "rewrite6_full_passes": 2.0,
      "step": 119850,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7223936,
      "curriculum_late_phase": 1.0,
      "epoch": 1.893e-05,
      "grad_norm": 0.49609375,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0221,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20946666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.53436861167283,
      "rewrite6_current_pass_fraction": 0.21733753967790306,
      "rewrite6_full_passes": 2.0,
      "step": 119860,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7248512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8935e-05,
      "grad_norm": 0.48828125,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0573,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20968,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.270050426717436,
      "rewrite6_current_pass_fraction": 0.21798991256255965,
      "rewrite6_full_passes": 2.0,
      "step": 119870,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7273088,
      "curriculum_late_phase": 1.0,
      "epoch": 1.894e-05,
      "grad_norm": 0.466796875,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0183,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.20996,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.456486081474473,
      "rewrite6_current_pass_fraction": 0.21839886272906078,
      "rewrite6_full_passes": 2.0,
      "step": 119880,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7297664,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8945e-05,
      "grad_norm": 0.46484375,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0185,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21036,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.460577787847765,
      "rewrite6_current_pass_fraction": 0.21895386652645518,
      "rewrite6_full_passes": 2.0,
      "step": 119890,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.732224,
      "curriculum_late_phase": 1.0,
      "epoch": 1.895e-05,
      "grad_norm": 0.48046875,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0154,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21074666666666667,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.397248208270213,
      "rewrite6_current_pass_fraction": 0.2195672917762069,
      "rewrite6_full_passes": 2.0,
      "step": 119900,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7346816,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8955e-05,
      "grad_norm": 0.490234375,
      "host_rss_gib": 16.335968017578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0302,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21105333333333334,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.701372449879624,
      "rewrite6_current_pass_fraction": 0.22018558548032172,
      "rewrite6_full_passes": 2.0,
      "step": 119910,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7371392,
      "curriculum_late_phase": 1.0,
      "epoch": 1.896e-05,
      "grad_norm": 0.5,
      "host_rss_gib": 16.339874267578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0284,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21134666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.664143495580532,
      "rewrite6_current_pass_fraction": 0.22062861482736462,
      "rewrite6_full_passes": 2.0,
      "step": 119920,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7395968,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8965e-05,
      "grad_norm": 0.44140625,
      "host_rss_gib": 16.339874267578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0425,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.2118,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.957571726540376,
      "rewrite6_current_pass_fraction": 0.22122256625966388,
      "rewrite6_full_passes": 2.0,
      "step": 119930,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7420544,
      "curriculum_late_phase": 1.0,
      "epoch": 1.897e-05,
      "grad_norm": 0.47265625,
      "host_rss_gib": 16.339874267578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0484,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21206666666666665,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.08158688469643,
      "rewrite6_current_pass_fraction": 0.22181651769196314,
      "rewrite6_full_passes": 2.0,
      "step": 119940,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.744512,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8975e-05,
      "grad_norm": 0.4609375,
      "host_rss_gib": 16.339874267578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0217,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21226666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.52615650675864,
      "rewrite6_current_pass_fraction": 0.22247375903098285,
      "rewrite6_full_passes": 2.0,
      "step": 119950,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7469696,
      "curriculum_late_phase": 1.0,
      "epoch": 1.898e-05,
      "grad_norm": 0.458984375,
      "host_rss_gib": 16.339874267578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0333,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21268,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.765646277434353,
      "rewrite6_current_pass_fraction": 0.22313100037000252,
      "rewrite6_full_passes": 2.0,
      "step": 119960,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7494272,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8985e-05,
      "grad_norm": 0.4921875,
      "host_rss_gib": 16.339874267578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0149,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.213,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.387052133397212,
      "rewrite6_current_pass_fraction": 0.22361784580631341,
      "rewrite6_full_passes": 2.0,
      "step": 119970,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7518848,
      "curriculum_late_phase": 1.0,
      "epoch": 1.899e-05,
      "grad_norm": 0.46875,
      "host_rss_gib": 16.339874267578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.9934,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21326666666666666,
      "no_prompt_full_passes": 1.0,
      "perplexity": 19.953408881656124,
      "rewrite6_current_pass_fraction": 0.2242312710560651,
      "rewrite6_full_passes": 2.0,
      "step": 119980,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7543424,
      "curriculum_late_phase": 1.0,
      "epoch": 1.8995e-05,
      "grad_norm": 0.474609375,
      "host_rss_gib": 16.339874267578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0465,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21352,
      "no_prompt_full_passes": 1.0,
      "perplexity": 21.041569897791508,
      "rewrite6_current_pass_fraction": 0.22484956476017995,
      "rewrite6_full_passes": 2.0,
      "step": 119990,
      "tokens_per_update": 245760.0
    },
    {
      "batch_size": 240.0,
      "cuda_allocated_gib": 0.22137832641601562,
      "cuda_peak_allocated_gib": 83.48212242126465,
      "cuda_peak_reserved_gib": 83.958984375,
      "cuda_reserved_gib": 83.958984375,
      "cumulative_tokens_billion": 23.7568,
      "curriculum_late_phase": 1.0,
      "epoch": 1.9e-05,
      "grad_norm": 0.453125,
      "host_rss_gib": 16.339874267578125,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.0273,
      "mix_fineweb": 0.8,
      "mix_no_prompt": 0.01,
      "mix_rewrite6": 0.05,
      "mix_ultra_multi": 0.09,
      "mix_ultra_qa": 0.05,
      "no_prompt_current_pass_fraction": 0.21378666666666668,
      "no_prompt_full_passes": 1.0,
      "perplexity": 20.641425434959473,
      "rewrite6_current_pass_fraction": 0.22542404237502678,
      "rewrite6_full_passes": 2.0,
      "step": 120000,
      "tokens_per_update": 245760.0
    },
    {
      "epoch": 1.9e-05,
      "eval_loss": 3.0809221267700195,
      "eval_perplexity": 21.778475655141797,
      "eval_runtime": 0.0489,
      "eval_samples_per_second": 653.926,
      "eval_steps_per_second": 20.435,
      "step": 120000
    }
  ],
  "logging_steps": 10,
  "max_steps": 2000000000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.2595637174272e+18,
  "train_batch_size": 240,
  "trial_name": null,
  "trial_params": null
}