rqadri's picture
Add files using upload-large-folder tool
d19a1c7 verified
Raw History Blame Contribute Delete
180 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9152902698282911,
"eval_steps": 500,
"global_step": 11194,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008176614881439084,
"grad_norm": 10.534998893737793,
"learning_rate": 1.1061797752808989e-06,
"loss": 2.2924,
"step": 10
},
{
"epoch": 0.001635322976287817,
"grad_norm": 6.318049430847168,
"learning_rate": 2.2241573033707866e-06,
"loss": 2.3794,
"step": 20
},
{
"epoch": 0.0024529844644317253,
"grad_norm": 2.999676465988159,
"learning_rate": 3.342134831460675e-06,
"loss": 1.9011,
"step": 30
},
{
"epoch": 0.003270645952575634,
"grad_norm": 1.3771625757217407,
"learning_rate": 4.4601123595505624e-06,
"loss": 1.5254,
"step": 40
},
{
"epoch": 0.004088307440719542,
"grad_norm": 0.8931247591972351,
"learning_rate": 5.5780898876404495e-06,
"loss": 1.226,
"step": 50
},
{
"epoch": 0.004905968928863451,
"grad_norm": 1.1269806623458862,
"learning_rate": 6.6960674157303374e-06,
"loss": 1.0716,
"step": 60
},
{
"epoch": 0.005723630417007359,
"grad_norm": 1.387283205986023,
"learning_rate": 7.814044943820226e-06,
"loss": 1.011,
"step": 70
},
{
"epoch": 0.006541291905151268,
"grad_norm": 0.6601500511169434,
"learning_rate": 8.932022471910112e-06,
"loss": 1.0061,
"step": 80
},
{
"epoch": 0.007358953393295176,
"grad_norm": 0.6297673583030701,
"learning_rate": 1.005e-05,
"loss": 0.9876,
"step": 90
},
{
"epoch": 0.008176614881439084,
"grad_norm": 0.5978390574455261,
"learning_rate": 1.1167977528089888e-05,
"loss": 0.93,
"step": 100
},
{
"epoch": 0.008994276369582993,
"grad_norm": 0.6085056662559509,
"learning_rate": 1.2285955056179776e-05,
"loss": 0.911,
"step": 110
},
{
"epoch": 0.009811937857726901,
"grad_norm": 0.630516767501831,
"learning_rate": 1.3403932584269664e-05,
"loss": 0.9181,
"step": 120
},
{
"epoch": 0.01062959934587081,
"grad_norm": 0.659934937953949,
"learning_rate": 1.4521910112359552e-05,
"loss": 0.8963,
"step": 130
},
{
"epoch": 0.011447260834014717,
"grad_norm": 0.5893965363502502,
"learning_rate": 1.563988764044944e-05,
"loss": 0.9024,
"step": 140
},
{
"epoch": 0.012264922322158627,
"grad_norm": 0.5959360003471375,
"learning_rate": 1.6757865168539328e-05,
"loss": 0.926,
"step": 150
},
{
"epoch": 0.013082583810302535,
"grad_norm": 0.5704044699668884,
"learning_rate": 1.7875842696629214e-05,
"loss": 0.8832,
"step": 160
},
{
"epoch": 0.013900245298446443,
"grad_norm": 0.6047025322914124,
"learning_rate": 1.89938202247191e-05,
"loss": 0.9009,
"step": 170
},
{
"epoch": 0.014717906786590351,
"grad_norm": 0.5899404883384705,
"learning_rate": 2e-05,
"loss": 0.8876,
"step": 180
},
{
"epoch": 0.01553556827473426,
"grad_norm": 0.5657159090042114,
"learning_rate": 2e-05,
"loss": 0.8758,
"step": 190
},
{
"epoch": 0.016353229762878167,
"grad_norm": 0.5136110186576843,
"learning_rate": 2e-05,
"loss": 0.892,
"step": 200
},
{
"epoch": 0.017170891251022075,
"grad_norm": 0.5496647357940674,
"learning_rate": 2e-05,
"loss": 0.9069,
"step": 210
},
{
"epoch": 0.017988552739165987,
"grad_norm": 0.5497331023216248,
"learning_rate": 2e-05,
"loss": 0.863,
"step": 220
},
{
"epoch": 0.018806214227309895,
"grad_norm": 0.5021716952323914,
"learning_rate": 2e-05,
"loss": 0.8952,
"step": 230
},
{
"epoch": 0.019623875715453803,
"grad_norm": 0.5393562912940979,
"learning_rate": 2e-05,
"loss": 0.8577,
"step": 240
},
{
"epoch": 0.02044153720359771,
"grad_norm": 0.5259848833084106,
"learning_rate": 2e-05,
"loss": 0.8904,
"step": 250
},
{
"epoch": 0.02125919869174162,
"grad_norm": 0.562362790107727,
"learning_rate": 2e-05,
"loss": 0.922,
"step": 260
},
{
"epoch": 0.022076860179885527,
"grad_norm": 0.514157235622406,
"learning_rate": 2e-05,
"loss": 0.8547,
"step": 270
},
{
"epoch": 0.022894521668029435,
"grad_norm": 0.5613844394683838,
"learning_rate": 2e-05,
"loss": 0.8867,
"step": 280
},
{
"epoch": 0.023712183156173343,
"grad_norm": 0.5137860774993896,
"learning_rate": 2e-05,
"loss": 0.864,
"step": 290
},
{
"epoch": 0.024529844644317254,
"grad_norm": 0.5275009274482727,
"learning_rate": 2e-05,
"loss": 0.927,
"step": 300
},
{
"epoch": 0.025347506132461162,
"grad_norm": 0.5223109722137451,
"learning_rate": 2e-05,
"loss": 0.8691,
"step": 310
},
{
"epoch": 0.02616516762060507,
"grad_norm": 0.5527063012123108,
"learning_rate": 2e-05,
"loss": 0.8648,
"step": 320
},
{
"epoch": 0.02698282910874898,
"grad_norm": 0.49740317463874817,
"learning_rate": 2e-05,
"loss": 0.8486,
"step": 330
},
{
"epoch": 0.027800490596892886,
"grad_norm": 0.49503129720687866,
"learning_rate": 2e-05,
"loss": 0.8395,
"step": 340
},
{
"epoch": 0.028618152085036794,
"grad_norm": 0.5311452150344849,
"learning_rate": 2e-05,
"loss": 0.8358,
"step": 350
},
{
"epoch": 0.029435813573180702,
"grad_norm": 0.5130277872085571,
"learning_rate": 2e-05,
"loss": 0.8554,
"step": 360
},
{
"epoch": 0.03025347506132461,
"grad_norm": 0.46086397767066956,
"learning_rate": 2e-05,
"loss": 0.8545,
"step": 370
},
{
"epoch": 0.03107113654946852,
"grad_norm": 0.49420565366744995,
"learning_rate": 2e-05,
"loss": 0.8401,
"step": 380
},
{
"epoch": 0.03188879803761243,
"grad_norm": 0.4804873764514923,
"learning_rate": 2e-05,
"loss": 0.8533,
"step": 390
},
{
"epoch": 0.032706459525756335,
"grad_norm": 0.4856545329093933,
"learning_rate": 2e-05,
"loss": 0.8172,
"step": 400
},
{
"epoch": 0.033524121013900246,
"grad_norm": 0.5700660347938538,
"learning_rate": 2e-05,
"loss": 0.87,
"step": 410
},
{
"epoch": 0.03434178250204415,
"grad_norm": 0.48742008209228516,
"learning_rate": 2e-05,
"loss": 0.8653,
"step": 420
},
{
"epoch": 0.03515944399018806,
"grad_norm": 0.5448684096336365,
"learning_rate": 2e-05,
"loss": 0.8574,
"step": 430
},
{
"epoch": 0.035977105478331974,
"grad_norm": 0.9145104289054871,
"learning_rate": 2e-05,
"loss": 0.8529,
"step": 440
},
{
"epoch": 0.03679476696647588,
"grad_norm": 0.6051157712936401,
"learning_rate": 2e-05,
"loss": 0.8492,
"step": 450
},
{
"epoch": 0.03761242845461979,
"grad_norm": 0.5123623013496399,
"learning_rate": 2e-05,
"loss": 0.8689,
"step": 460
},
{
"epoch": 0.038430089942763694,
"grad_norm": 0.49921512603759766,
"learning_rate": 2e-05,
"loss": 0.8126,
"step": 470
},
{
"epoch": 0.039247751430907606,
"grad_norm": 0.5155230760574341,
"learning_rate": 2e-05,
"loss": 0.7817,
"step": 480
},
{
"epoch": 0.04006541291905151,
"grad_norm": 0.46784713864326477,
"learning_rate": 2e-05,
"loss": 0.8324,
"step": 490
},
{
"epoch": 0.04088307440719542,
"grad_norm": 0.4902730882167816,
"learning_rate": 2e-05,
"loss": 0.8196,
"step": 500
},
{
"epoch": 0.041700735895339326,
"grad_norm": 0.47778692841529846,
"learning_rate": 2e-05,
"loss": 0.8537,
"step": 510
},
{
"epoch": 0.04251839738348324,
"grad_norm": 0.46416184306144714,
"learning_rate": 2e-05,
"loss": 0.815,
"step": 520
},
{
"epoch": 0.04333605887162715,
"grad_norm": 0.4957575798034668,
"learning_rate": 2e-05,
"loss": 0.8317,
"step": 530
},
{
"epoch": 0.044153720359771054,
"grad_norm": 0.4955309331417084,
"learning_rate": 2e-05,
"loss": 0.8053,
"step": 540
},
{
"epoch": 0.044971381847914965,
"grad_norm": 0.47801926732063293,
"learning_rate": 2e-05,
"loss": 0.8318,
"step": 550
},
{
"epoch": 0.04578904333605887,
"grad_norm": 0.4526337683200836,
"learning_rate": 2e-05,
"loss": 0.8297,
"step": 560
},
{
"epoch": 0.04660670482420278,
"grad_norm": 0.5001497864723206,
"learning_rate": 2e-05,
"loss": 0.7966,
"step": 570
},
{
"epoch": 0.047424366312346686,
"grad_norm": 0.4969994127750397,
"learning_rate": 2e-05,
"loss": 0.849,
"step": 580
},
{
"epoch": 0.0482420278004906,
"grad_norm": 0.4873654246330261,
"learning_rate": 2e-05,
"loss": 0.8098,
"step": 590
},
{
"epoch": 0.04905968928863451,
"grad_norm": 0.46147921681404114,
"learning_rate": 2e-05,
"loss": 0.8462,
"step": 600
},
{
"epoch": 0.04987735077677841,
"grad_norm": 0.48419851064682007,
"learning_rate": 2e-05,
"loss": 0.7878,
"step": 610
},
{
"epoch": 0.050695012264922325,
"grad_norm": 0.48362842202186584,
"learning_rate": 2e-05,
"loss": 0.8309,
"step": 620
},
{
"epoch": 0.05151267375306623,
"grad_norm": 0.4815925657749176,
"learning_rate": 2e-05,
"loss": 0.8409,
"step": 630
},
{
"epoch": 0.05233033524121014,
"grad_norm": 0.5199002027511597,
"learning_rate": 2e-05,
"loss": 0.8685,
"step": 640
},
{
"epoch": 0.053147996729354045,
"grad_norm": 0.4724765419960022,
"learning_rate": 2e-05,
"loss": 0.8279,
"step": 650
},
{
"epoch": 0.05396565821749796,
"grad_norm": 0.5170514583587646,
"learning_rate": 2e-05,
"loss": 0.8361,
"step": 660
},
{
"epoch": 0.05478331970564186,
"grad_norm": 0.5067883133888245,
"learning_rate": 2e-05,
"loss": 0.8425,
"step": 670
},
{
"epoch": 0.05560098119378577,
"grad_norm": 0.43147748708724976,
"learning_rate": 2e-05,
"loss": 0.8048,
"step": 680
},
{
"epoch": 0.056418642681929684,
"grad_norm": 0.508711576461792,
"learning_rate": 2e-05,
"loss": 0.848,
"step": 690
},
{
"epoch": 0.05723630417007359,
"grad_norm": 0.4623899459838867,
"learning_rate": 2e-05,
"loss": 0.826,
"step": 700
},
{
"epoch": 0.0580539656582175,
"grad_norm": 0.48188719153404236,
"learning_rate": 2e-05,
"loss": 0.8371,
"step": 710
},
{
"epoch": 0.058871627146361405,
"grad_norm": 0.4509360194206238,
"learning_rate": 2e-05,
"loss": 0.8107,
"step": 720
},
{
"epoch": 0.059689288634505316,
"grad_norm": 0.4348622262477875,
"learning_rate": 2e-05,
"loss": 0.7997,
"step": 730
},
{
"epoch": 0.06050695012264922,
"grad_norm": 0.4793204963207245,
"learning_rate": 2e-05,
"loss": 0.8357,
"step": 740
},
{
"epoch": 0.06132461161079313,
"grad_norm": 0.47185131907463074,
"learning_rate": 2e-05,
"loss": 0.814,
"step": 750
},
{
"epoch": 0.06214227309893704,
"grad_norm": 0.5104201436042786,
"learning_rate": 2e-05,
"loss": 0.8447,
"step": 760
},
{
"epoch": 0.06295993458708095,
"grad_norm": 0.45201629400253296,
"learning_rate": 2e-05,
"loss": 0.8265,
"step": 770
},
{
"epoch": 0.06377759607522486,
"grad_norm": 1.1489161252975464,
"learning_rate": 2e-05,
"loss": 0.8434,
"step": 780
},
{
"epoch": 0.06459525756336877,
"grad_norm": 0.5133305788040161,
"learning_rate": 2e-05,
"loss": 0.821,
"step": 790
},
{
"epoch": 0.06541291905151267,
"grad_norm": 0.45519399642944336,
"learning_rate": 2e-05,
"loss": 0.8432,
"step": 800
},
{
"epoch": 0.06623058053965658,
"grad_norm": 1.8952999114990234,
"learning_rate": 2e-05,
"loss": 0.7693,
"step": 810
},
{
"epoch": 0.06704824202780049,
"grad_norm": 0.4610855281352997,
"learning_rate": 2e-05,
"loss": 0.8228,
"step": 820
},
{
"epoch": 0.0678659035159444,
"grad_norm": 0.49510180950164795,
"learning_rate": 2e-05,
"loss": 0.8122,
"step": 830
},
{
"epoch": 0.0686835650040883,
"grad_norm": 0.4776274561882019,
"learning_rate": 2e-05,
"loss": 0.7767,
"step": 840
},
{
"epoch": 0.06950122649223221,
"grad_norm": 0.47551512718200684,
"learning_rate": 2e-05,
"loss": 0.8284,
"step": 850
},
{
"epoch": 0.07031888798037612,
"grad_norm": 0.49425825476646423,
"learning_rate": 2e-05,
"loss": 0.802,
"step": 860
},
{
"epoch": 0.07113654946852004,
"grad_norm": 0.46732547879219055,
"learning_rate": 2e-05,
"loss": 0.822,
"step": 870
},
{
"epoch": 0.07195421095666395,
"grad_norm": 0.44407016038894653,
"learning_rate": 2e-05,
"loss": 0.8251,
"step": 880
},
{
"epoch": 0.07277187244480784,
"grad_norm": 0.45862287282943726,
"learning_rate": 2e-05,
"loss": 0.8219,
"step": 890
},
{
"epoch": 0.07358953393295176,
"grad_norm": 0.48010408878326416,
"learning_rate": 2e-05,
"loss": 0.815,
"step": 900
},
{
"epoch": 0.07440719542109567,
"grad_norm": 0.4639987349510193,
"learning_rate": 2e-05,
"loss": 0.811,
"step": 910
},
{
"epoch": 0.07522485690923958,
"grad_norm": 0.4830052852630615,
"learning_rate": 2e-05,
"loss": 0.8209,
"step": 920
},
{
"epoch": 0.07604251839738348,
"grad_norm": 0.44569143652915955,
"learning_rate": 2e-05,
"loss": 0.7969,
"step": 930
},
{
"epoch": 0.07686017988552739,
"grad_norm": 0.4402197301387787,
"learning_rate": 2e-05,
"loss": 0.8048,
"step": 940
},
{
"epoch": 0.0776778413736713,
"grad_norm": 0.4393838346004486,
"learning_rate": 2e-05,
"loss": 0.7999,
"step": 950
},
{
"epoch": 0.07849550286181521,
"grad_norm": 0.45926329493522644,
"learning_rate": 2e-05,
"loss": 0.7993,
"step": 960
},
{
"epoch": 0.07931316434995912,
"grad_norm": 0.45618489384651184,
"learning_rate": 2e-05,
"loss": 0.7941,
"step": 970
},
{
"epoch": 0.08013082583810302,
"grad_norm": 0.4323713779449463,
"learning_rate": 2e-05,
"loss": 0.8128,
"step": 980
},
{
"epoch": 0.08094848732624693,
"grad_norm": 0.4577762484550476,
"learning_rate": 2e-05,
"loss": 0.8158,
"step": 990
},
{
"epoch": 0.08176614881439084,
"grad_norm": 0.44366759061813354,
"learning_rate": 2e-05,
"loss": 0.801,
"step": 1000
},
{
"epoch": 0.08258381030253475,
"grad_norm": 0.43710198998451233,
"learning_rate": 2e-05,
"loss": 0.8178,
"step": 1010
},
{
"epoch": 0.08340147179067865,
"grad_norm": 0.5216193795204163,
"learning_rate": 2e-05,
"loss": 0.8356,
"step": 1020
},
{
"epoch": 0.08421913327882256,
"grad_norm": 0.6718442440032959,
"learning_rate": 2e-05,
"loss": 0.8186,
"step": 1030
},
{
"epoch": 0.08503679476696648,
"grad_norm": 0.45956555008888245,
"learning_rate": 2e-05,
"loss": 0.7599,
"step": 1040
},
{
"epoch": 0.08585445625511039,
"grad_norm": 0.4390004575252533,
"learning_rate": 2e-05,
"loss": 0.7939,
"step": 1050
},
{
"epoch": 0.0866721177432543,
"grad_norm": 0.458306223154068,
"learning_rate": 2e-05,
"loss": 0.8154,
"step": 1060
},
{
"epoch": 0.0874897792313982,
"grad_norm": 0.46446549892425537,
"learning_rate": 2e-05,
"loss": 0.8013,
"step": 1070
},
{
"epoch": 0.08830744071954211,
"grad_norm": 0.46321365237236023,
"learning_rate": 2e-05,
"loss": 0.7987,
"step": 1080
},
{
"epoch": 0.08912510220768602,
"grad_norm": 0.4468840956687927,
"learning_rate": 2e-05,
"loss": 0.8049,
"step": 1090
},
{
"epoch": 0.08994276369582993,
"grad_norm": 0.4241533577442169,
"learning_rate": 2e-05,
"loss": 0.7683,
"step": 1100
},
{
"epoch": 0.09076042518397383,
"grad_norm": 0.4573608934879303,
"learning_rate": 2e-05,
"loss": 0.7942,
"step": 1110
},
{
"epoch": 0.09157808667211774,
"grad_norm": 0.4434383809566498,
"learning_rate": 2e-05,
"loss": 0.8005,
"step": 1120
},
{
"epoch": 0.09239574816026165,
"grad_norm": 0.5078239440917969,
"learning_rate": 2e-05,
"loss": 0.8192,
"step": 1130
},
{
"epoch": 0.09321340964840556,
"grad_norm": 0.4617445766925812,
"learning_rate": 2e-05,
"loss": 0.804,
"step": 1140
},
{
"epoch": 0.09403107113654947,
"grad_norm": 0.43501752614974976,
"learning_rate": 2e-05,
"loss": 0.8172,
"step": 1150
},
{
"epoch": 0.09484873262469337,
"grad_norm": 0.4166872501373291,
"learning_rate": 2e-05,
"loss": 0.8197,
"step": 1160
},
{
"epoch": 0.09566639411283728,
"grad_norm": 0.655982494354248,
"learning_rate": 2e-05,
"loss": 0.789,
"step": 1170
},
{
"epoch": 0.0964840556009812,
"grad_norm": 0.41751641035079956,
"learning_rate": 2e-05,
"loss": 0.7995,
"step": 1180
},
{
"epoch": 0.0973017170891251,
"grad_norm": 0.470846027135849,
"learning_rate": 2e-05,
"loss": 0.8322,
"step": 1190
},
{
"epoch": 0.09811937857726902,
"grad_norm": 0.4337841272354126,
"learning_rate": 2e-05,
"loss": 0.814,
"step": 1200
},
{
"epoch": 0.09893704006541292,
"grad_norm": 0.47191646695137024,
"learning_rate": 2e-05,
"loss": 0.8039,
"step": 1210
},
{
"epoch": 0.09975470155355683,
"grad_norm": 0.43299242854118347,
"learning_rate": 2e-05,
"loss": 0.8096,
"step": 1220
},
{
"epoch": 0.10057236304170074,
"grad_norm": 0.4392620623111725,
"learning_rate": 2e-05,
"loss": 0.811,
"step": 1230
},
{
"epoch": 0.10139002452984465,
"grad_norm": 0.4958798885345459,
"learning_rate": 2e-05,
"loss": 0.821,
"step": 1240
},
{
"epoch": 0.10220768601798855,
"grad_norm": 0.449624240398407,
"learning_rate": 2e-05,
"loss": 0.8024,
"step": 1250
},
{
"epoch": 0.10302534750613246,
"grad_norm": 0.5577616095542908,
"learning_rate": 2e-05,
"loss": 0.7876,
"step": 1260
},
{
"epoch": 0.10384300899427637,
"grad_norm": 0.4583219289779663,
"learning_rate": 2e-05,
"loss": 0.7821,
"step": 1270
},
{
"epoch": 0.10466067048242028,
"grad_norm": 0.4646966755390167,
"learning_rate": 2e-05,
"loss": 0.791,
"step": 1280
},
{
"epoch": 0.1054783319705642,
"grad_norm": 0.4247964918613434,
"learning_rate": 2e-05,
"loss": 0.78,
"step": 1290
},
{
"epoch": 0.10629599345870809,
"grad_norm": 0.46946001052856445,
"learning_rate": 2e-05,
"loss": 0.7948,
"step": 1300
},
{
"epoch": 0.107113654946852,
"grad_norm": 0.43030065298080444,
"learning_rate": 2e-05,
"loss": 0.8059,
"step": 1310
},
{
"epoch": 0.10793131643499591,
"grad_norm": 0.4443809986114502,
"learning_rate": 2e-05,
"loss": 0.7681,
"step": 1320
},
{
"epoch": 0.10874897792313983,
"grad_norm": 0.4212033748626709,
"learning_rate": 2e-05,
"loss": 0.7689,
"step": 1330
},
{
"epoch": 0.10956663941128372,
"grad_norm": 0.486328125,
"learning_rate": 2e-05,
"loss": 0.7612,
"step": 1340
},
{
"epoch": 0.11038430089942763,
"grad_norm": 0.43208643794059753,
"learning_rate": 2e-05,
"loss": 0.8189,
"step": 1350
},
{
"epoch": 0.11120196238757155,
"grad_norm": 0.4304050803184509,
"learning_rate": 2e-05,
"loss": 0.7886,
"step": 1360
},
{
"epoch": 0.11201962387571546,
"grad_norm": 0.4358374774456024,
"learning_rate": 2e-05,
"loss": 0.7838,
"step": 1370
},
{
"epoch": 0.11283728536385937,
"grad_norm": 0.4316851794719696,
"learning_rate": 2e-05,
"loss": 0.7945,
"step": 1380
},
{
"epoch": 0.11365494685200327,
"grad_norm": 0.43443015217781067,
"learning_rate": 2e-05,
"loss": 0.8029,
"step": 1390
},
{
"epoch": 0.11447260834014718,
"grad_norm": 0.4291702210903168,
"learning_rate": 2e-05,
"loss": 0.8028,
"step": 1400
},
{
"epoch": 0.11529026982829109,
"grad_norm": 0.4630155861377716,
"learning_rate": 2e-05,
"loss": 0.7724,
"step": 1410
},
{
"epoch": 0.116107931316435,
"grad_norm": 0.4348618686199188,
"learning_rate": 2e-05,
"loss": 0.7816,
"step": 1420
},
{
"epoch": 0.1169255928045789,
"grad_norm": 0.4198819100856781,
"learning_rate": 2e-05,
"loss": 0.7851,
"step": 1430
},
{
"epoch": 0.11774325429272281,
"grad_norm": 0.4539889991283417,
"learning_rate": 2e-05,
"loss": 0.7691,
"step": 1440
},
{
"epoch": 0.11856091578086672,
"grad_norm": 0.4621836245059967,
"learning_rate": 2e-05,
"loss": 0.7858,
"step": 1450
},
{
"epoch": 0.11937857726901063,
"grad_norm": 0.44036760926246643,
"learning_rate": 2e-05,
"loss": 0.7865,
"step": 1460
},
{
"epoch": 0.12019623875715454,
"grad_norm": 0.4228440225124359,
"learning_rate": 2e-05,
"loss": 0.7579,
"step": 1470
},
{
"epoch": 0.12101390024529844,
"grad_norm": 0.435478150844574,
"learning_rate": 2e-05,
"loss": 0.7823,
"step": 1480
},
{
"epoch": 0.12183156173344235,
"grad_norm": 0.4398757815361023,
"learning_rate": 2e-05,
"loss": 0.7727,
"step": 1490
},
{
"epoch": 0.12264922322158626,
"grad_norm": 0.42032548785209656,
"learning_rate": 2e-05,
"loss": 0.7772,
"step": 1500
},
{
"epoch": 0.12346688470973018,
"grad_norm": 1.2638468742370605,
"learning_rate": 2e-05,
"loss": 0.8029,
"step": 1510
},
{
"epoch": 0.12428454619787407,
"grad_norm": 0.4492338001728058,
"learning_rate": 2e-05,
"loss": 0.7976,
"step": 1520
},
{
"epoch": 0.12510220768601799,
"grad_norm": 0.4194067120552063,
"learning_rate": 2e-05,
"loss": 0.7729,
"step": 1530
},
{
"epoch": 0.1259198691741619,
"grad_norm": 0.4248548150062561,
"learning_rate": 2e-05,
"loss": 0.7832,
"step": 1540
},
{
"epoch": 0.1267375306623058,
"grad_norm": 0.44248121976852417,
"learning_rate": 2e-05,
"loss": 0.7993,
"step": 1550
},
{
"epoch": 0.12755519215044972,
"grad_norm": 0.4308151304721832,
"learning_rate": 2e-05,
"loss": 0.8292,
"step": 1560
},
{
"epoch": 0.12837285363859363,
"grad_norm": 0.46226322650909424,
"learning_rate": 2e-05,
"loss": 0.7998,
"step": 1570
},
{
"epoch": 0.12919051512673754,
"grad_norm": 0.6823664307594299,
"learning_rate": 2e-05,
"loss": 0.7797,
"step": 1580
},
{
"epoch": 0.13000817661488143,
"grad_norm": 0.49192625284194946,
"learning_rate": 2e-05,
"loss": 0.792,
"step": 1590
},
{
"epoch": 0.13082583810302534,
"grad_norm": 0.46564188599586487,
"learning_rate": 2e-05,
"loss": 0.7812,
"step": 1600
},
{
"epoch": 0.13164349959116925,
"grad_norm": 0.4175106883049011,
"learning_rate": 2e-05,
"loss": 0.8088,
"step": 1610
},
{
"epoch": 0.13246116107931316,
"grad_norm": 0.43853816390037537,
"learning_rate": 2e-05,
"loss": 0.7881,
"step": 1620
},
{
"epoch": 0.13327882256745707,
"grad_norm": 0.48302924633026123,
"learning_rate": 2e-05,
"loss": 0.7913,
"step": 1630
},
{
"epoch": 0.13409648405560098,
"grad_norm": 0.6604704856872559,
"learning_rate": 2e-05,
"loss": 0.7709,
"step": 1640
},
{
"epoch": 0.1349141455437449,
"grad_norm": 0.523360550403595,
"learning_rate": 2e-05,
"loss": 0.7948,
"step": 1650
},
{
"epoch": 0.1357318070318888,
"grad_norm": 0.41866135597229004,
"learning_rate": 2e-05,
"loss": 0.7374,
"step": 1660
},
{
"epoch": 0.13654946852003272,
"grad_norm": 0.43435412645339966,
"learning_rate": 2e-05,
"loss": 0.7516,
"step": 1670
},
{
"epoch": 0.1373671300081766,
"grad_norm": 0.4378770887851715,
"learning_rate": 2e-05,
"loss": 0.7793,
"step": 1680
},
{
"epoch": 0.1381847914963205,
"grad_norm": 0.4642718434333801,
"learning_rate": 2e-05,
"loss": 0.7891,
"step": 1690
},
{
"epoch": 0.13900245298446443,
"grad_norm": 0.4553477168083191,
"learning_rate": 2e-05,
"loss": 0.7693,
"step": 1700
},
{
"epoch": 0.13982011447260834,
"grad_norm": 0.44605308771133423,
"learning_rate": 2e-05,
"loss": 0.7894,
"step": 1710
},
{
"epoch": 0.14063777596075225,
"grad_norm": 0.4435379207134247,
"learning_rate": 2e-05,
"loss": 0.7834,
"step": 1720
},
{
"epoch": 0.14145543744889616,
"grad_norm": 0.43862536549568176,
"learning_rate": 2e-05,
"loss": 0.7783,
"step": 1730
},
{
"epoch": 0.14227309893704007,
"grad_norm": 0.3968248963356018,
"learning_rate": 2e-05,
"loss": 0.7753,
"step": 1740
},
{
"epoch": 0.14309076042518398,
"grad_norm": 0.43859657645225525,
"learning_rate": 2e-05,
"loss": 0.8167,
"step": 1750
},
{
"epoch": 0.1439084219133279,
"grad_norm": 0.4324052035808563,
"learning_rate": 2e-05,
"loss": 0.7865,
"step": 1760
},
{
"epoch": 0.14472608340147178,
"grad_norm": 0.43473488092422485,
"learning_rate": 2e-05,
"loss": 0.8224,
"step": 1770
},
{
"epoch": 0.1455437448896157,
"grad_norm": 0.42597419023513794,
"learning_rate": 2e-05,
"loss": 0.7985,
"step": 1780
},
{
"epoch": 0.1463614063777596,
"grad_norm": 0.48900550603866577,
"learning_rate": 2e-05,
"loss": 0.7562,
"step": 1790
},
{
"epoch": 0.1471790678659035,
"grad_norm": 0.4394257068634033,
"learning_rate": 2e-05,
"loss": 0.7721,
"step": 1800
},
{
"epoch": 0.14799672935404742,
"grad_norm": 0.43295830488204956,
"learning_rate": 2e-05,
"loss": 0.7727,
"step": 1810
},
{
"epoch": 0.14881439084219134,
"grad_norm": 0.4200812578201294,
"learning_rate": 2e-05,
"loss": 0.7866,
"step": 1820
},
{
"epoch": 0.14963205233033525,
"grad_norm": 0.4190243184566498,
"learning_rate": 2e-05,
"loss": 0.7765,
"step": 1830
},
{
"epoch": 0.15044971381847916,
"grad_norm": 0.45326995849609375,
"learning_rate": 2e-05,
"loss": 0.7871,
"step": 1840
},
{
"epoch": 0.15126737530662307,
"grad_norm": 0.4055960476398468,
"learning_rate": 2e-05,
"loss": 0.7632,
"step": 1850
},
{
"epoch": 0.15208503679476695,
"grad_norm": 0.48894283175468445,
"learning_rate": 2e-05,
"loss": 0.7756,
"step": 1860
},
{
"epoch": 0.15290269828291086,
"grad_norm": 0.4656607508659363,
"learning_rate": 2e-05,
"loss": 0.7726,
"step": 1870
},
{
"epoch": 0.15372035977105478,
"grad_norm": 0.4367874562740326,
"learning_rate": 2e-05,
"loss": 0.7772,
"step": 1880
},
{
"epoch": 0.1545380212591987,
"grad_norm": 0.4750969707965851,
"learning_rate": 2e-05,
"loss": 0.7712,
"step": 1890
},
{
"epoch": 0.1553556827473426,
"grad_norm": 0.4409693479537964,
"learning_rate": 2e-05,
"loss": 0.7904,
"step": 1900
},
{
"epoch": 0.1561733442354865,
"grad_norm": 0.4396913945674896,
"learning_rate": 2e-05,
"loss": 0.7597,
"step": 1910
},
{
"epoch": 0.15699100572363042,
"grad_norm": 0.41550886631011963,
"learning_rate": 2e-05,
"loss": 0.7812,
"step": 1920
},
{
"epoch": 0.15780866721177433,
"grad_norm": 0.42665189504623413,
"learning_rate": 2e-05,
"loss": 0.7871,
"step": 1930
},
{
"epoch": 0.15862632869991825,
"grad_norm": 0.4286573827266693,
"learning_rate": 2e-05,
"loss": 0.7736,
"step": 1940
},
{
"epoch": 0.15944399018806213,
"grad_norm": 0.42506206035614014,
"learning_rate": 2e-05,
"loss": 0.7784,
"step": 1950
},
{
"epoch": 0.16026165167620604,
"grad_norm": 0.46702292561531067,
"learning_rate": 2e-05,
"loss": 0.7703,
"step": 1960
},
{
"epoch": 0.16107931316434995,
"grad_norm": 0.43847423791885376,
"learning_rate": 2e-05,
"loss": 0.7656,
"step": 1970
},
{
"epoch": 0.16189697465249386,
"grad_norm": 0.4711706042289734,
"learning_rate": 2e-05,
"loss": 0.7657,
"step": 1980
},
{
"epoch": 0.16271463614063778,
"grad_norm": 0.5090168118476868,
"learning_rate": 2e-05,
"loss": 0.7592,
"step": 1990
},
{
"epoch": 0.1635322976287817,
"grad_norm": 0.46786943078041077,
"learning_rate": 2e-05,
"loss": 0.7921,
"step": 2000
},
{
"epoch": 0.1643499591169256,
"grad_norm": 0.4125913977622986,
"learning_rate": 2e-05,
"loss": 0.7831,
"step": 2010
},
{
"epoch": 0.1651676206050695,
"grad_norm": 0.4442942440509796,
"learning_rate": 2e-05,
"loss": 0.7536,
"step": 2020
},
{
"epoch": 0.16598528209321342,
"grad_norm": 0.4498803913593292,
"learning_rate": 2e-05,
"loss": 0.7825,
"step": 2030
},
{
"epoch": 0.1668029435813573,
"grad_norm": 0.4364321529865265,
"learning_rate": 2e-05,
"loss": 0.7899,
"step": 2040
},
{
"epoch": 0.16762060506950122,
"grad_norm": 0.4316538870334625,
"learning_rate": 2e-05,
"loss": 0.7791,
"step": 2050
},
{
"epoch": 0.16843826655764513,
"grad_norm": 0.45543980598449707,
"learning_rate": 2e-05,
"loss": 0.7882,
"step": 2060
},
{
"epoch": 0.16925592804578904,
"grad_norm": 0.4247235953807831,
"learning_rate": 2e-05,
"loss": 0.7986,
"step": 2070
},
{
"epoch": 0.17007358953393295,
"grad_norm": 0.4491972029209137,
"learning_rate": 2e-05,
"loss": 0.8298,
"step": 2080
},
{
"epoch": 0.17089125102207686,
"grad_norm": 0.4386272132396698,
"learning_rate": 2e-05,
"loss": 0.7572,
"step": 2090
},
{
"epoch": 0.17170891251022077,
"grad_norm": 0.4444292485713959,
"learning_rate": 2e-05,
"loss": 0.7801,
"step": 2100
},
{
"epoch": 0.17252657399836469,
"grad_norm": 0.4270576238632202,
"learning_rate": 2e-05,
"loss": 0.7656,
"step": 2110
},
{
"epoch": 0.1733442354865086,
"grad_norm": 0.436068594455719,
"learning_rate": 2e-05,
"loss": 0.7444,
"step": 2120
},
{
"epoch": 0.17416189697465248,
"grad_norm": 0.42339572310447693,
"learning_rate": 2e-05,
"loss": 0.7805,
"step": 2130
},
{
"epoch": 0.1749795584627964,
"grad_norm": 0.4199734628200531,
"learning_rate": 2e-05,
"loss": 0.7519,
"step": 2140
},
{
"epoch": 0.1757972199509403,
"grad_norm": 0.4218316376209259,
"learning_rate": 2e-05,
"loss": 0.7825,
"step": 2150
},
{
"epoch": 0.17661488143908421,
"grad_norm": 0.417795330286026,
"learning_rate": 2e-05,
"loss": 0.7871,
"step": 2160
},
{
"epoch": 0.17743254292722813,
"grad_norm": 0.4626673460006714,
"learning_rate": 2e-05,
"loss": 0.7771,
"step": 2170
},
{
"epoch": 0.17825020441537204,
"grad_norm": 0.468065470457077,
"learning_rate": 2e-05,
"loss": 0.7915,
"step": 2180
},
{
"epoch": 0.17906786590351595,
"grad_norm": 0.4166199564933777,
"learning_rate": 2e-05,
"loss": 0.7726,
"step": 2190
},
{
"epoch": 0.17988552739165986,
"grad_norm": 0.42800837755203247,
"learning_rate": 2e-05,
"loss": 0.7725,
"step": 2200
},
{
"epoch": 0.18070318887980377,
"grad_norm": 0.4120378792285919,
"learning_rate": 2e-05,
"loss": 0.743,
"step": 2210
},
{
"epoch": 0.18152085036794766,
"grad_norm": 0.5455228686332703,
"learning_rate": 2e-05,
"loss": 0.7318,
"step": 2220
},
{
"epoch": 0.18233851185609157,
"grad_norm": 0.4296571612358093,
"learning_rate": 2e-05,
"loss": 0.7928,
"step": 2230
},
{
"epoch": 0.18315617334423548,
"grad_norm": 0.4547067880630493,
"learning_rate": 2e-05,
"loss": 0.7747,
"step": 2240
},
{
"epoch": 0.1839738348323794,
"grad_norm": 0.4168989956378937,
"learning_rate": 2e-05,
"loss": 0.7923,
"step": 2250
},
{
"epoch": 0.1847914963205233,
"grad_norm": 0.44426074624061584,
"learning_rate": 2e-05,
"loss": 0.7748,
"step": 2260
},
{
"epoch": 0.1856091578086672,
"grad_norm": 0.4370101988315582,
"learning_rate": 2e-05,
"loss": 0.7758,
"step": 2270
},
{
"epoch": 0.18642681929681112,
"grad_norm": 0.49895063042640686,
"learning_rate": 2e-05,
"loss": 0.7619,
"step": 2280
},
{
"epoch": 0.18724448078495504,
"grad_norm": 0.40498825907707214,
"learning_rate": 2e-05,
"loss": 0.7709,
"step": 2290
},
{
"epoch": 0.18806214227309895,
"grad_norm": 0.4341062903404236,
"learning_rate": 2e-05,
"loss": 0.7862,
"step": 2300
},
{
"epoch": 0.18887980376124286,
"grad_norm": 0.46437665820121765,
"learning_rate": 2e-05,
"loss": 0.8152,
"step": 2310
},
{
"epoch": 0.18969746524938674,
"grad_norm": 0.43534669280052185,
"learning_rate": 2e-05,
"loss": 0.764,
"step": 2320
},
{
"epoch": 0.19051512673753065,
"grad_norm": 0.44681885838508606,
"learning_rate": 2e-05,
"loss": 0.7873,
"step": 2330
},
{
"epoch": 0.19133278822567457,
"grad_norm": 0.45268961787223816,
"learning_rate": 2e-05,
"loss": 0.7995,
"step": 2340
},
{
"epoch": 0.19215044971381848,
"grad_norm": 0.41534167528152466,
"learning_rate": 2e-05,
"loss": 0.7722,
"step": 2350
},
{
"epoch": 0.1929681112019624,
"grad_norm": 0.44464072585105896,
"learning_rate": 2e-05,
"loss": 0.7842,
"step": 2360
},
{
"epoch": 0.1937857726901063,
"grad_norm": 0.4211314618587494,
"learning_rate": 2e-05,
"loss": 0.7756,
"step": 2370
},
{
"epoch": 0.1946034341782502,
"grad_norm": 0.44980642199516296,
"learning_rate": 2e-05,
"loss": 0.7635,
"step": 2380
},
{
"epoch": 0.19542109566639412,
"grad_norm": 0.43298453092575073,
"learning_rate": 2e-05,
"loss": 0.7998,
"step": 2390
},
{
"epoch": 0.19623875715453803,
"grad_norm": 0.4239964485168457,
"learning_rate": 2e-05,
"loss": 0.7589,
"step": 2400
},
{
"epoch": 0.19705641864268192,
"grad_norm": 0.4066570997238159,
"learning_rate": 2e-05,
"loss": 0.779,
"step": 2410
},
{
"epoch": 0.19787408013082583,
"grad_norm": 0.47745639085769653,
"learning_rate": 2e-05,
"loss": 0.7877,
"step": 2420
},
{
"epoch": 0.19869174161896974,
"grad_norm": 0.4101332426071167,
"learning_rate": 2e-05,
"loss": 0.8011,
"step": 2430
},
{
"epoch": 0.19950940310711365,
"grad_norm": 0.4742473065853119,
"learning_rate": 2e-05,
"loss": 0.7922,
"step": 2440
},
{
"epoch": 0.20032706459525756,
"grad_norm": 0.4261021316051483,
"learning_rate": 2e-05,
"loss": 0.751,
"step": 2450
},
{
"epoch": 0.20114472608340148,
"grad_norm": 0.5145176649093628,
"learning_rate": 2e-05,
"loss": 0.7285,
"step": 2460
},
{
"epoch": 0.2019623875715454,
"grad_norm": 0.40921565890312195,
"learning_rate": 2e-05,
"loss": 0.7871,
"step": 2470
},
{
"epoch": 0.2027800490596893,
"grad_norm": 0.44564947485923767,
"learning_rate": 2e-05,
"loss": 0.8052,
"step": 2480
},
{
"epoch": 0.2035977105478332,
"grad_norm": 0.42048290371894836,
"learning_rate": 2e-05,
"loss": 0.8011,
"step": 2490
},
{
"epoch": 0.2044153720359771,
"grad_norm": 0.408529669046402,
"learning_rate": 2e-05,
"loss": 0.7733,
"step": 2500
},
{
"epoch": 0.205233033524121,
"grad_norm": 0.42771342396736145,
"learning_rate": 2e-05,
"loss": 0.7809,
"step": 2510
},
{
"epoch": 0.20605069501226492,
"grad_norm": 0.39116600155830383,
"learning_rate": 2e-05,
"loss": 0.7621,
"step": 2520
},
{
"epoch": 0.20686835650040883,
"grad_norm": 0.43292510509490967,
"learning_rate": 2e-05,
"loss": 0.7797,
"step": 2530
},
{
"epoch": 0.20768601798855274,
"grad_norm": 0.40812504291534424,
"learning_rate": 2e-05,
"loss": 0.7545,
"step": 2540
},
{
"epoch": 0.20850367947669665,
"grad_norm": 0.416536420583725,
"learning_rate": 2e-05,
"loss": 0.7432,
"step": 2550
},
{
"epoch": 0.20932134096484056,
"grad_norm": 0.3991974890232086,
"learning_rate": 2e-05,
"loss": 0.784,
"step": 2560
},
{
"epoch": 0.21013900245298447,
"grad_norm": 0.4071870744228363,
"learning_rate": 2e-05,
"loss": 0.7697,
"step": 2570
},
{
"epoch": 0.2109566639411284,
"grad_norm": 0.407435804605484,
"learning_rate": 2e-05,
"loss": 0.7512,
"step": 2580
},
{
"epoch": 0.21177432542927227,
"grad_norm": 0.3893747925758362,
"learning_rate": 2e-05,
"loss": 0.7493,
"step": 2590
},
{
"epoch": 0.21259198691741618,
"grad_norm": 0.4174593985080719,
"learning_rate": 2e-05,
"loss": 0.7752,
"step": 2600
},
{
"epoch": 0.2134096484055601,
"grad_norm": 0.4031468629837036,
"learning_rate": 2e-05,
"loss": 0.7418,
"step": 2610
},
{
"epoch": 0.214227309893704,
"grad_norm": 0.44015318155288696,
"learning_rate": 2e-05,
"loss": 0.7472,
"step": 2620
},
{
"epoch": 0.21504497138184792,
"grad_norm": 0.4368755519390106,
"learning_rate": 2e-05,
"loss": 0.7864,
"step": 2630
},
{
"epoch": 0.21586263286999183,
"grad_norm": 0.425601065158844,
"learning_rate": 2e-05,
"loss": 0.8093,
"step": 2640
},
{
"epoch": 0.21668029435813574,
"grad_norm": 0.43123167753219604,
"learning_rate": 2e-05,
"loss": 0.763,
"step": 2650
},
{
"epoch": 0.21749795584627965,
"grad_norm": 0.41291922330856323,
"learning_rate": 2e-05,
"loss": 0.7669,
"step": 2660
},
{
"epoch": 0.21831561733442356,
"grad_norm": 0.4302712380886078,
"learning_rate": 2e-05,
"loss": 0.7857,
"step": 2670
},
{
"epoch": 0.21913327882256745,
"grad_norm": 0.41941574215888977,
"learning_rate": 2e-05,
"loss": 0.7444,
"step": 2680
},
{
"epoch": 0.21995094031071136,
"grad_norm": 0.40381744503974915,
"learning_rate": 2e-05,
"loss": 0.7802,
"step": 2690
},
{
"epoch": 0.22076860179885527,
"grad_norm": 0.42162802815437317,
"learning_rate": 2e-05,
"loss": 0.7807,
"step": 2700
},
{
"epoch": 0.22158626328699918,
"grad_norm": 0.9290792346000671,
"learning_rate": 2e-05,
"loss": 0.8272,
"step": 2710
},
{
"epoch": 0.2224039247751431,
"grad_norm": 0.3997645080089569,
"learning_rate": 2e-05,
"loss": 0.7506,
"step": 2720
},
{
"epoch": 0.223221586263287,
"grad_norm": 0.4363479018211365,
"learning_rate": 2e-05,
"loss": 0.7878,
"step": 2730
},
{
"epoch": 0.22403924775143091,
"grad_norm": 0.4270990490913391,
"learning_rate": 2e-05,
"loss": 0.774,
"step": 2740
},
{
"epoch": 0.22485690923957483,
"grad_norm": 0.4099840521812439,
"learning_rate": 2e-05,
"loss": 0.7726,
"step": 2750
},
{
"epoch": 0.22567457072771874,
"grad_norm": 0.4205203652381897,
"learning_rate": 2e-05,
"loss": 0.7432,
"step": 2760
},
{
"epoch": 0.22649223221586262,
"grad_norm": 0.41790467500686646,
"learning_rate": 2e-05,
"loss": 0.7717,
"step": 2770
},
{
"epoch": 0.22730989370400653,
"grad_norm": 0.4241865277290344,
"learning_rate": 2e-05,
"loss": 0.7461,
"step": 2780
},
{
"epoch": 0.22812755519215044,
"grad_norm": 0.3862425684928894,
"learning_rate": 2e-05,
"loss": 0.7811,
"step": 2790
},
{
"epoch": 0.22894521668029436,
"grad_norm": 0.39432331919670105,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 2800
},
{
"epoch": 0.22976287816843827,
"grad_norm": 0.4487655460834503,
"learning_rate": 2e-05,
"loss": 0.7939,
"step": 2810
},
{
"epoch": 0.23058053965658218,
"grad_norm": 0.41441550850868225,
"learning_rate": 2e-05,
"loss": 0.773,
"step": 2820
},
{
"epoch": 0.2313982011447261,
"grad_norm": 0.41408640146255493,
"learning_rate": 2e-05,
"loss": 0.746,
"step": 2830
},
{
"epoch": 0.23221586263287,
"grad_norm": 0.4297657310962677,
"learning_rate": 2e-05,
"loss": 0.7557,
"step": 2840
},
{
"epoch": 0.2330335241210139,
"grad_norm": 0.4253140985965729,
"learning_rate": 2e-05,
"loss": 0.7718,
"step": 2850
},
{
"epoch": 0.2338511856091578,
"grad_norm": 0.4623691141605377,
"learning_rate": 2e-05,
"loss": 0.7645,
"step": 2860
},
{
"epoch": 0.2346688470973017,
"grad_norm": 0.425809383392334,
"learning_rate": 2e-05,
"loss": 0.7639,
"step": 2870
},
{
"epoch": 0.23548650858544562,
"grad_norm": 0.4268561601638794,
"learning_rate": 2e-05,
"loss": 0.7501,
"step": 2880
},
{
"epoch": 0.23630417007358953,
"grad_norm": 0.42480748891830444,
"learning_rate": 2e-05,
"loss": 0.737,
"step": 2890
},
{
"epoch": 0.23712183156173344,
"grad_norm": 0.3940679430961609,
"learning_rate": 2e-05,
"loss": 0.7478,
"step": 2900
},
{
"epoch": 0.23793949304987735,
"grad_norm": 0.3852456212043762,
"learning_rate": 2e-05,
"loss": 0.7685,
"step": 2910
},
{
"epoch": 0.23875715453802127,
"grad_norm": 0.4131034016609192,
"learning_rate": 2e-05,
"loss": 0.7467,
"step": 2920
},
{
"epoch": 0.23957481602616518,
"grad_norm": 0.44036728143692017,
"learning_rate": 2e-05,
"loss": 0.7303,
"step": 2930
},
{
"epoch": 0.2403924775143091,
"grad_norm": 0.43323394656181335,
"learning_rate": 2e-05,
"loss": 0.7649,
"step": 2940
},
{
"epoch": 0.24121013900245297,
"grad_norm": 0.44328805804252625,
"learning_rate": 2e-05,
"loss": 0.7589,
"step": 2950
},
{
"epoch": 0.24202780049059688,
"grad_norm": 0.41394710540771484,
"learning_rate": 2e-05,
"loss": 0.7534,
"step": 2960
},
{
"epoch": 0.2428454619787408,
"grad_norm": 0.41553813219070435,
"learning_rate": 2e-05,
"loss": 0.7706,
"step": 2970
},
{
"epoch": 0.2436631234668847,
"grad_norm": 0.39981281757354736,
"learning_rate": 2e-05,
"loss": 0.7355,
"step": 2980
},
{
"epoch": 0.24448078495502862,
"grad_norm": 0.42764589190483093,
"learning_rate": 2e-05,
"loss": 0.7612,
"step": 2990
},
{
"epoch": 0.24529844644317253,
"grad_norm": 0.46958279609680176,
"learning_rate": 2e-05,
"loss": 0.7685,
"step": 3000
},
{
"epoch": 0.24611610793131644,
"grad_norm": 0.43474486470222473,
"learning_rate": 2e-05,
"loss": 0.7476,
"step": 3010
},
{
"epoch": 0.24693376941946035,
"grad_norm": 0.4077485501766205,
"learning_rate": 2e-05,
"loss": 0.7395,
"step": 3020
},
{
"epoch": 0.24775143090760426,
"grad_norm": 0.4127010107040405,
"learning_rate": 2e-05,
"loss": 0.7449,
"step": 3030
},
{
"epoch": 0.24856909239574815,
"grad_norm": 0.38957399129867554,
"learning_rate": 2e-05,
"loss": 0.75,
"step": 3040
},
{
"epoch": 0.24938675388389206,
"grad_norm": 0.4030892848968506,
"learning_rate": 2e-05,
"loss": 0.7161,
"step": 3050
},
{
"epoch": 0.25020441537203597,
"grad_norm": 0.41952016949653625,
"learning_rate": 2e-05,
"loss": 0.7711,
"step": 3060
},
{
"epoch": 0.2510220768601799,
"grad_norm": 0.4226265251636505,
"learning_rate": 2e-05,
"loss": 0.7614,
"step": 3070
},
{
"epoch": 0.2518397383483238,
"grad_norm": 0.43005725741386414,
"learning_rate": 2e-05,
"loss": 0.7844,
"step": 3080
},
{
"epoch": 0.2526573998364677,
"grad_norm": 0.4619760811328888,
"learning_rate": 2e-05,
"loss": 0.7607,
"step": 3090
},
{
"epoch": 0.2534750613246116,
"grad_norm": 0.43441349267959595,
"learning_rate": 2e-05,
"loss": 0.7649,
"step": 3100
},
{
"epoch": 0.2542927228127555,
"grad_norm": 0.3946240246295929,
"learning_rate": 2e-05,
"loss": 0.7373,
"step": 3110
},
{
"epoch": 0.25511038430089944,
"grad_norm": 0.42219096422195435,
"learning_rate": 2e-05,
"loss": 0.7143,
"step": 3120
},
{
"epoch": 0.2559280457890433,
"grad_norm": 0.4020823836326599,
"learning_rate": 2e-05,
"loss": 0.7531,
"step": 3130
},
{
"epoch": 0.25674570727718726,
"grad_norm": 0.4099595546722412,
"learning_rate": 2e-05,
"loss": 0.7517,
"step": 3140
},
{
"epoch": 0.25756336876533115,
"grad_norm": 0.41835877299308777,
"learning_rate": 2e-05,
"loss": 0.7505,
"step": 3150
},
{
"epoch": 0.2583810302534751,
"grad_norm": 0.39703795313835144,
"learning_rate": 2e-05,
"loss": 0.7823,
"step": 3160
},
{
"epoch": 0.25919869174161897,
"grad_norm": 0.43623700737953186,
"learning_rate": 2e-05,
"loss": 0.7477,
"step": 3170
},
{
"epoch": 0.26001635322976285,
"grad_norm": 0.41160234808921814,
"learning_rate": 2e-05,
"loss": 0.7454,
"step": 3180
},
{
"epoch": 0.2608340147179068,
"grad_norm": 0.41497576236724854,
"learning_rate": 2e-05,
"loss": 0.7475,
"step": 3190
},
{
"epoch": 0.2616516762060507,
"grad_norm": 0.41480571031570435,
"learning_rate": 2e-05,
"loss": 0.775,
"step": 3200
},
{
"epoch": 0.2624693376941946,
"grad_norm": 0.42725762724876404,
"learning_rate": 2e-05,
"loss": 0.7756,
"step": 3210
},
{
"epoch": 0.2632869991823385,
"grad_norm": 0.4155767560005188,
"learning_rate": 2e-05,
"loss": 0.7756,
"step": 3220
},
{
"epoch": 0.26410466067048244,
"grad_norm": 0.3951760232448578,
"learning_rate": 2e-05,
"loss": 0.7702,
"step": 3230
},
{
"epoch": 0.2649223221586263,
"grad_norm": 0.40908899903297424,
"learning_rate": 2e-05,
"loss": 0.7464,
"step": 3240
},
{
"epoch": 0.26573998364677026,
"grad_norm": 0.4152882397174835,
"learning_rate": 2e-05,
"loss": 0.7756,
"step": 3250
},
{
"epoch": 0.26655764513491415,
"grad_norm": 0.41987019777297974,
"learning_rate": 2e-05,
"loss": 0.7585,
"step": 3260
},
{
"epoch": 0.26737530662305803,
"grad_norm": 0.3909783363342285,
"learning_rate": 2e-05,
"loss": 0.7568,
"step": 3270
},
{
"epoch": 0.26819296811120197,
"grad_norm": 0.417426198720932,
"learning_rate": 2e-05,
"loss": 0.7924,
"step": 3280
},
{
"epoch": 0.26901062959934585,
"grad_norm": 0.4050072729587555,
"learning_rate": 2e-05,
"loss": 0.7607,
"step": 3290
},
{
"epoch": 0.2698282910874898,
"grad_norm": 0.4422713816165924,
"learning_rate": 2e-05,
"loss": 0.7751,
"step": 3300
},
{
"epoch": 0.2706459525756337,
"grad_norm": 0.42104074358940125,
"learning_rate": 2e-05,
"loss": 0.7616,
"step": 3310
},
{
"epoch": 0.2714636140637776,
"grad_norm": 0.40387916564941406,
"learning_rate": 2e-05,
"loss": 0.7396,
"step": 3320
},
{
"epoch": 0.2722812755519215,
"grad_norm": 0.40367233753204346,
"learning_rate": 2e-05,
"loss": 0.7799,
"step": 3330
},
{
"epoch": 0.27309893704006544,
"grad_norm": 0.46548694372177124,
"learning_rate": 2e-05,
"loss": 0.7537,
"step": 3340
},
{
"epoch": 0.2739165985282093,
"grad_norm": 0.451147198677063,
"learning_rate": 2e-05,
"loss": 0.7961,
"step": 3350
},
{
"epoch": 0.2747342600163532,
"grad_norm": 0.44955241680145264,
"learning_rate": 2e-05,
"loss": 0.7614,
"step": 3360
},
{
"epoch": 0.27555192150449714,
"grad_norm": 0.4327465295791626,
"learning_rate": 2e-05,
"loss": 0.7484,
"step": 3370
},
{
"epoch": 0.276369582992641,
"grad_norm": 0.3961194157600403,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 3380
},
{
"epoch": 0.27718724448078497,
"grad_norm": 0.4116878807544708,
"learning_rate": 2e-05,
"loss": 0.7187,
"step": 3390
},
{
"epoch": 0.27800490596892885,
"grad_norm": 0.40448230504989624,
"learning_rate": 2e-05,
"loss": 0.7718,
"step": 3400
},
{
"epoch": 0.2788225674570728,
"grad_norm": 0.4147052764892578,
"learning_rate": 2e-05,
"loss": 0.7433,
"step": 3410
},
{
"epoch": 0.2796402289452167,
"grad_norm": 0.4361591935157776,
"learning_rate": 2e-05,
"loss": 0.7556,
"step": 3420
},
{
"epoch": 0.2804578904333606,
"grad_norm": 0.44823896884918213,
"learning_rate": 2e-05,
"loss": 0.7495,
"step": 3430
},
{
"epoch": 0.2812755519215045,
"grad_norm": 0.3785291016101837,
"learning_rate": 2e-05,
"loss": 0.74,
"step": 3440
},
{
"epoch": 0.2820932134096484,
"grad_norm": 0.38844063878059387,
"learning_rate": 2e-05,
"loss": 0.7416,
"step": 3450
},
{
"epoch": 0.2829108748977923,
"grad_norm": 0.43286728858947754,
"learning_rate": 2e-05,
"loss": 0.7566,
"step": 3460
},
{
"epoch": 0.2837285363859362,
"grad_norm": 0.4022294878959656,
"learning_rate": 2e-05,
"loss": 0.7529,
"step": 3470
},
{
"epoch": 0.28454619787408014,
"grad_norm": 0.4239967167377472,
"learning_rate": 2e-05,
"loss": 0.7432,
"step": 3480
},
{
"epoch": 0.285363859362224,
"grad_norm": 0.40576833486557007,
"learning_rate": 2e-05,
"loss": 0.7611,
"step": 3490
},
{
"epoch": 0.28618152085036797,
"grad_norm": 0.4220907986164093,
"learning_rate": 2e-05,
"loss": 0.7545,
"step": 3500
},
{
"epoch": 0.28699918233851185,
"grad_norm": 0.3993762731552124,
"learning_rate": 2e-05,
"loss": 0.7597,
"step": 3510
},
{
"epoch": 0.2878168438266558,
"grad_norm": 0.4114720821380615,
"learning_rate": 2e-05,
"loss": 0.736,
"step": 3520
},
{
"epoch": 0.28863450531479967,
"grad_norm": 0.43589070439338684,
"learning_rate": 2e-05,
"loss": 0.7691,
"step": 3530
},
{
"epoch": 0.28945216680294356,
"grad_norm": 0.4150513708591461,
"learning_rate": 2e-05,
"loss": 0.7698,
"step": 3540
},
{
"epoch": 0.2902698282910875,
"grad_norm": 0.39021164178848267,
"learning_rate": 2e-05,
"loss": 0.7489,
"step": 3550
},
{
"epoch": 0.2910874897792314,
"grad_norm": 0.5082252621650696,
"learning_rate": 2e-05,
"loss": 0.7374,
"step": 3560
},
{
"epoch": 0.2919051512673753,
"grad_norm": 0.43987151980400085,
"learning_rate": 2e-05,
"loss": 0.7452,
"step": 3570
},
{
"epoch": 0.2927228127555192,
"grad_norm": 0.4348721504211426,
"learning_rate": 2e-05,
"loss": 0.7691,
"step": 3580
},
{
"epoch": 0.29354047424366314,
"grad_norm": 0.3936716616153717,
"learning_rate": 2e-05,
"loss": 0.7405,
"step": 3590
},
{
"epoch": 0.294358135731807,
"grad_norm": 0.38866978883743286,
"learning_rate": 2e-05,
"loss": 0.7649,
"step": 3600
},
{
"epoch": 0.29517579721995096,
"grad_norm": 0.4239301085472107,
"learning_rate": 2e-05,
"loss": 0.7557,
"step": 3610
},
{
"epoch": 0.29599345870809485,
"grad_norm": 0.4139637351036072,
"learning_rate": 2e-05,
"loss": 0.7682,
"step": 3620
},
{
"epoch": 0.29681112019623873,
"grad_norm": 0.4086167812347412,
"learning_rate": 2e-05,
"loss": 0.7413,
"step": 3630
},
{
"epoch": 0.29762878168438267,
"grad_norm": 0.3776358962059021,
"learning_rate": 2e-05,
"loss": 0.7483,
"step": 3640
},
{
"epoch": 0.29844644317252655,
"grad_norm": 0.39760732650756836,
"learning_rate": 2e-05,
"loss": 0.746,
"step": 3650
},
{
"epoch": 0.2992641046606705,
"grad_norm": 0.37500932812690735,
"learning_rate": 2e-05,
"loss": 0.7874,
"step": 3660
},
{
"epoch": 0.3000817661488144,
"grad_norm": 0.43193116784095764,
"learning_rate": 2e-05,
"loss": 0.7651,
"step": 3670
},
{
"epoch": 0.3008994276369583,
"grad_norm": 0.41836270689964294,
"learning_rate": 2e-05,
"loss": 0.7507,
"step": 3680
},
{
"epoch": 0.3017170891251022,
"grad_norm": 0.3975755572319031,
"learning_rate": 2e-05,
"loss": 0.7365,
"step": 3690
},
{
"epoch": 0.30253475061324614,
"grad_norm": 0.4004629850387573,
"learning_rate": 2e-05,
"loss": 0.7406,
"step": 3700
},
{
"epoch": 0.30335241210139,
"grad_norm": 0.41910433769226074,
"learning_rate": 2e-05,
"loss": 0.8037,
"step": 3710
},
{
"epoch": 0.3041700735895339,
"grad_norm": 0.4036235511302948,
"learning_rate": 2e-05,
"loss": 0.7465,
"step": 3720
},
{
"epoch": 0.30498773507767785,
"grad_norm": 0.41243213415145874,
"learning_rate": 2e-05,
"loss": 0.7469,
"step": 3730
},
{
"epoch": 0.30580539656582173,
"grad_norm": 0.45436981320381165,
"learning_rate": 2e-05,
"loss": 0.7575,
"step": 3740
},
{
"epoch": 0.30662305805396567,
"grad_norm": 0.3992025554180145,
"learning_rate": 2e-05,
"loss": 0.7409,
"step": 3750
},
{
"epoch": 0.30744071954210955,
"grad_norm": 0.3761412501335144,
"learning_rate": 2e-05,
"loss": 0.7385,
"step": 3760
},
{
"epoch": 0.3082583810302535,
"grad_norm": 0.450571745634079,
"learning_rate": 2e-05,
"loss": 0.7623,
"step": 3770
},
{
"epoch": 0.3090760425183974,
"grad_norm": 0.43576136231422424,
"learning_rate": 2e-05,
"loss": 0.7465,
"step": 3780
},
{
"epoch": 0.3098937040065413,
"grad_norm": 0.38754603266716003,
"learning_rate": 2e-05,
"loss": 0.7577,
"step": 3790
},
{
"epoch": 0.3107113654946852,
"grad_norm": 0.41827720403671265,
"learning_rate": 2e-05,
"loss": 0.7554,
"step": 3800
},
{
"epoch": 0.3115290269828291,
"grad_norm": 0.4002350866794586,
"learning_rate": 2e-05,
"loss": 0.7514,
"step": 3810
},
{
"epoch": 0.312346688470973,
"grad_norm": 0.3866146504878998,
"learning_rate": 2e-05,
"loss": 0.7646,
"step": 3820
},
{
"epoch": 0.3131643499591169,
"grad_norm": 0.4048006236553192,
"learning_rate": 2e-05,
"loss": 0.7727,
"step": 3830
},
{
"epoch": 0.31398201144726084,
"grad_norm": 0.4159805476665497,
"learning_rate": 2e-05,
"loss": 0.7536,
"step": 3840
},
{
"epoch": 0.31479967293540473,
"grad_norm": 0.41745713353157043,
"learning_rate": 2e-05,
"loss": 0.7335,
"step": 3850
},
{
"epoch": 0.31561733442354867,
"grad_norm": 0.37902170419692993,
"learning_rate": 2e-05,
"loss": 0.7598,
"step": 3860
},
{
"epoch": 0.31643499591169255,
"grad_norm": 0.40632471442222595,
"learning_rate": 2e-05,
"loss": 0.7628,
"step": 3870
},
{
"epoch": 0.3172526573998365,
"grad_norm": 0.42005592584609985,
"learning_rate": 2e-05,
"loss": 0.753,
"step": 3880
},
{
"epoch": 0.3180703188879804,
"grad_norm": 0.4179593324661255,
"learning_rate": 2e-05,
"loss": 0.7511,
"step": 3890
},
{
"epoch": 0.31888798037612426,
"grad_norm": 0.4019879996776581,
"learning_rate": 2e-05,
"loss": 0.7428,
"step": 3900
},
{
"epoch": 0.3197056418642682,
"grad_norm": 0.3838173747062683,
"learning_rate": 2e-05,
"loss": 0.7218,
"step": 3910
},
{
"epoch": 0.3205233033524121,
"grad_norm": 0.3869258761405945,
"learning_rate": 2e-05,
"loss": 0.739,
"step": 3920
},
{
"epoch": 0.321340964840556,
"grad_norm": 0.3965376317501068,
"learning_rate": 2e-05,
"loss": 0.7557,
"step": 3930
},
{
"epoch": 0.3221586263286999,
"grad_norm": 0.3937940001487732,
"learning_rate": 2e-05,
"loss": 0.7434,
"step": 3940
},
{
"epoch": 0.32297628781684384,
"grad_norm": 0.4084051847457886,
"learning_rate": 2e-05,
"loss": 0.7375,
"step": 3950
},
{
"epoch": 0.3237939493049877,
"grad_norm": 0.38900405168533325,
"learning_rate": 2e-05,
"loss": 0.7534,
"step": 3960
},
{
"epoch": 0.32461161079313167,
"grad_norm": 0.397067129611969,
"learning_rate": 2e-05,
"loss": 0.7767,
"step": 3970
},
{
"epoch": 0.32542927228127555,
"grad_norm": 0.4044935405254364,
"learning_rate": 2e-05,
"loss": 0.7681,
"step": 3980
},
{
"epoch": 0.32624693376941943,
"grad_norm": 0.4075620770454407,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 3990
},
{
"epoch": 0.3270645952575634,
"grad_norm": 0.37622272968292236,
"learning_rate": 2e-05,
"loss": 0.7515,
"step": 4000
},
{
"epoch": 0.32788225674570726,
"grad_norm": 0.39748191833496094,
"learning_rate": 2e-05,
"loss": 0.7801,
"step": 4010
},
{
"epoch": 0.3286999182338512,
"grad_norm": 0.39921826124191284,
"learning_rate": 2e-05,
"loss": 0.7668,
"step": 4020
},
{
"epoch": 0.3295175797219951,
"grad_norm": 0.3993796706199646,
"learning_rate": 2e-05,
"loss": 0.7588,
"step": 4030
},
{
"epoch": 0.330335241210139,
"grad_norm": 0.40211594104766846,
"learning_rate": 2e-05,
"loss": 0.7293,
"step": 4040
},
{
"epoch": 0.3311529026982829,
"grad_norm": 0.3940093517303467,
"learning_rate": 2e-05,
"loss": 0.7089,
"step": 4050
},
{
"epoch": 0.33197056418642684,
"grad_norm": 0.41875386238098145,
"learning_rate": 2e-05,
"loss": 0.7188,
"step": 4060
},
{
"epoch": 0.3327882256745707,
"grad_norm": 0.4149841368198395,
"learning_rate": 2e-05,
"loss": 0.7494,
"step": 4070
},
{
"epoch": 0.3336058871627146,
"grad_norm": 0.40230244398117065,
"learning_rate": 2e-05,
"loss": 0.7384,
"step": 4080
},
{
"epoch": 0.33442354865085855,
"grad_norm": 0.5382981896400452,
"learning_rate": 2e-05,
"loss": 0.7529,
"step": 4090
},
{
"epoch": 0.33524121013900243,
"grad_norm": 0.3823828101158142,
"learning_rate": 2e-05,
"loss": 0.7814,
"step": 4100
},
{
"epoch": 0.33605887162714637,
"grad_norm": 0.4252546429634094,
"learning_rate": 2e-05,
"loss": 0.7531,
"step": 4110
},
{
"epoch": 0.33687653311529026,
"grad_norm": 0.40745359659194946,
"learning_rate": 2e-05,
"loss": 0.7765,
"step": 4120
},
{
"epoch": 0.3376941946034342,
"grad_norm": 0.3941667675971985,
"learning_rate": 2e-05,
"loss": 0.7566,
"step": 4130
},
{
"epoch": 0.3385118560915781,
"grad_norm": 0.40058091282844543,
"learning_rate": 2e-05,
"loss": 0.7448,
"step": 4140
},
{
"epoch": 0.339329517579722,
"grad_norm": 0.41285571455955505,
"learning_rate": 2e-05,
"loss": 0.7231,
"step": 4150
},
{
"epoch": 0.3401471790678659,
"grad_norm": 0.408279150724411,
"learning_rate": 2e-05,
"loss": 0.7582,
"step": 4160
},
{
"epoch": 0.3409648405560098,
"grad_norm": 0.3870287835597992,
"learning_rate": 2e-05,
"loss": 0.7542,
"step": 4170
},
{
"epoch": 0.3417825020441537,
"grad_norm": 0.4457457363605499,
"learning_rate": 2e-05,
"loss": 0.7583,
"step": 4180
},
{
"epoch": 0.3426001635322976,
"grad_norm": 0.38335496187210083,
"learning_rate": 2e-05,
"loss": 0.7482,
"step": 4190
},
{
"epoch": 0.34341782502044155,
"grad_norm": 0.432076632976532,
"learning_rate": 2e-05,
"loss": 0.7669,
"step": 4200
},
{
"epoch": 0.34423548650858543,
"grad_norm": 0.4170612096786499,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 4210
},
{
"epoch": 0.34505314799672937,
"grad_norm": 0.39313265681266785,
"learning_rate": 2e-05,
"loss": 0.7247,
"step": 4220
},
{
"epoch": 0.34587080948487325,
"grad_norm": 0.40817657113075256,
"learning_rate": 2e-05,
"loss": 0.7545,
"step": 4230
},
{
"epoch": 0.3466884709730172,
"grad_norm": 0.392434298992157,
"learning_rate": 2e-05,
"loss": 0.7402,
"step": 4240
},
{
"epoch": 0.3475061324611611,
"grad_norm": 0.42975276708602905,
"learning_rate": 2e-05,
"loss": 0.7636,
"step": 4250
},
{
"epoch": 0.34832379394930496,
"grad_norm": 0.3929069936275482,
"learning_rate": 2e-05,
"loss": 0.732,
"step": 4260
},
{
"epoch": 0.3491414554374489,
"grad_norm": 0.389665424823761,
"learning_rate": 2e-05,
"loss": 0.7468,
"step": 4270
},
{
"epoch": 0.3499591169255928,
"grad_norm": 0.38716766238212585,
"learning_rate": 2e-05,
"loss": 0.7405,
"step": 4280
},
{
"epoch": 0.3507767784137367,
"grad_norm": 0.3947966992855072,
"learning_rate": 2e-05,
"loss": 0.7251,
"step": 4290
},
{
"epoch": 0.3515944399018806,
"grad_norm": 0.37235280871391296,
"learning_rate": 2e-05,
"loss": 0.729,
"step": 4300
},
{
"epoch": 0.35241210139002455,
"grad_norm": 0.4839659035205841,
"learning_rate": 2e-05,
"loss": 0.7582,
"step": 4310
},
{
"epoch": 0.35322976287816843,
"grad_norm": 0.41242969036102295,
"learning_rate": 2e-05,
"loss": 0.7599,
"step": 4320
},
{
"epoch": 0.35404742436631237,
"grad_norm": 0.42157986760139465,
"learning_rate": 2e-05,
"loss": 0.7902,
"step": 4330
},
{
"epoch": 0.35486508585445625,
"grad_norm": 0.39237549901008606,
"learning_rate": 2e-05,
"loss": 0.7306,
"step": 4340
},
{
"epoch": 0.35568274734260014,
"grad_norm": 0.4073483943939209,
"learning_rate": 2e-05,
"loss": 0.7423,
"step": 4350
},
{
"epoch": 0.3565004088307441,
"grad_norm": 0.37807080149650574,
"learning_rate": 2e-05,
"loss": 0.7605,
"step": 4360
},
{
"epoch": 0.35731807031888796,
"grad_norm": 0.4259825348854065,
"learning_rate": 2e-05,
"loss": 0.7368,
"step": 4370
},
{
"epoch": 0.3581357318070319,
"grad_norm": 0.39345216751098633,
"learning_rate": 2e-05,
"loss": 0.7365,
"step": 4380
},
{
"epoch": 0.3589533932951758,
"grad_norm": 0.4019021689891815,
"learning_rate": 2e-05,
"loss": 0.7286,
"step": 4390
},
{
"epoch": 0.3597710547833197,
"grad_norm": 0.3999948799610138,
"learning_rate": 2e-05,
"loss": 0.7706,
"step": 4400
},
{
"epoch": 0.3605887162714636,
"grad_norm": 0.4193735420703888,
"learning_rate": 2e-05,
"loss": 0.7567,
"step": 4410
},
{
"epoch": 0.36140637775960754,
"grad_norm": 0.40287792682647705,
"learning_rate": 2e-05,
"loss": 0.7506,
"step": 4420
},
{
"epoch": 0.36222403924775143,
"grad_norm": 0.4220403730869293,
"learning_rate": 2e-05,
"loss": 0.7585,
"step": 4430
},
{
"epoch": 0.3630417007358953,
"grad_norm": 0.3977868854999542,
"learning_rate": 2e-05,
"loss": 0.7393,
"step": 4440
},
{
"epoch": 0.36385936222403925,
"grad_norm": 0.5432819128036499,
"learning_rate": 2e-05,
"loss": 0.7,
"step": 4450
},
{
"epoch": 0.36467702371218313,
"grad_norm": 0.4258385896682739,
"learning_rate": 2e-05,
"loss": 0.7682,
"step": 4460
},
{
"epoch": 0.3654946852003271,
"grad_norm": 0.3945724666118622,
"learning_rate": 2e-05,
"loss": 0.75,
"step": 4470
},
{
"epoch": 0.36631234668847096,
"grad_norm": 0.41011112928390503,
"learning_rate": 2e-05,
"loss": 0.7485,
"step": 4480
},
{
"epoch": 0.3671300081766149,
"grad_norm": 0.3658716082572937,
"learning_rate": 2e-05,
"loss": 0.719,
"step": 4490
},
{
"epoch": 0.3679476696647588,
"grad_norm": 0.38978058099746704,
"learning_rate": 2e-05,
"loss": 0.7624,
"step": 4500
},
{
"epoch": 0.3687653311529027,
"grad_norm": 0.41422396898269653,
"learning_rate": 2e-05,
"loss": 0.7292,
"step": 4510
},
{
"epoch": 0.3695829926410466,
"grad_norm": 0.6165766716003418,
"learning_rate": 2e-05,
"loss": 0.7435,
"step": 4520
},
{
"epoch": 0.3704006541291905,
"grad_norm": 0.39596688747406006,
"learning_rate": 2e-05,
"loss": 0.7539,
"step": 4530
},
{
"epoch": 0.3712183156173344,
"grad_norm": 0.4293264150619507,
"learning_rate": 2e-05,
"loss": 0.7695,
"step": 4540
},
{
"epoch": 0.3720359771054783,
"grad_norm": 0.5381836295127869,
"learning_rate": 2e-05,
"loss": 0.7286,
"step": 4550
},
{
"epoch": 0.37285363859362225,
"grad_norm": 0.42194926738739014,
"learning_rate": 2e-05,
"loss": 0.7436,
"step": 4560
},
{
"epoch": 0.37367130008176613,
"grad_norm": 0.3940383791923523,
"learning_rate": 2e-05,
"loss": 0.7405,
"step": 4570
},
{
"epoch": 0.3744889615699101,
"grad_norm": 0.3936174511909485,
"learning_rate": 2e-05,
"loss": 0.7148,
"step": 4580
},
{
"epoch": 0.37530662305805396,
"grad_norm": 0.4172099232673645,
"learning_rate": 2e-05,
"loss": 0.7578,
"step": 4590
},
{
"epoch": 0.3761242845461979,
"grad_norm": 0.3652750849723816,
"learning_rate": 2e-05,
"loss": 0.7628,
"step": 4600
},
{
"epoch": 0.3769419460343418,
"grad_norm": 0.40293166041374207,
"learning_rate": 2e-05,
"loss": 0.7226,
"step": 4610
},
{
"epoch": 0.3777596075224857,
"grad_norm": 0.4645305871963501,
"learning_rate": 2e-05,
"loss": 0.7235,
"step": 4620
},
{
"epoch": 0.3785772690106296,
"grad_norm": 0.47156858444213867,
"learning_rate": 2e-05,
"loss": 0.7149,
"step": 4630
},
{
"epoch": 0.3793949304987735,
"grad_norm": 0.3718200922012329,
"learning_rate": 2e-05,
"loss": 0.7276,
"step": 4640
},
{
"epoch": 0.3802125919869174,
"grad_norm": 0.38712579011917114,
"learning_rate": 2e-05,
"loss": 0.7327,
"step": 4650
},
{
"epoch": 0.3810302534750613,
"grad_norm": 0.38667309284210205,
"learning_rate": 2e-05,
"loss": 0.7565,
"step": 4660
},
{
"epoch": 0.38184791496320525,
"grad_norm": 0.3759092092514038,
"learning_rate": 2e-05,
"loss": 0.764,
"step": 4670
},
{
"epoch": 0.38266557645134913,
"grad_norm": 0.39515963196754456,
"learning_rate": 2e-05,
"loss": 0.7443,
"step": 4680
},
{
"epoch": 0.38348323793949307,
"grad_norm": 0.41865164041519165,
"learning_rate": 2e-05,
"loss": 0.7549,
"step": 4690
},
{
"epoch": 0.38430089942763695,
"grad_norm": 0.4113163352012634,
"learning_rate": 2e-05,
"loss": 0.7343,
"step": 4700
},
{
"epoch": 0.3851185609157809,
"grad_norm": 0.39225131273269653,
"learning_rate": 2e-05,
"loss": 0.7493,
"step": 4710
},
{
"epoch": 0.3859362224039248,
"grad_norm": 0.39372774958610535,
"learning_rate": 2e-05,
"loss": 0.7385,
"step": 4720
},
{
"epoch": 0.38675388389206866,
"grad_norm": 0.3898671567440033,
"learning_rate": 2e-05,
"loss": 0.7735,
"step": 4730
},
{
"epoch": 0.3875715453802126,
"grad_norm": 0.415476530790329,
"learning_rate": 2e-05,
"loss": 0.7246,
"step": 4740
},
{
"epoch": 0.3883892068683565,
"grad_norm": 0.39973869919776917,
"learning_rate": 2e-05,
"loss": 0.7139,
"step": 4750
},
{
"epoch": 0.3892068683565004,
"grad_norm": 0.3770519495010376,
"learning_rate": 2e-05,
"loss": 0.7483,
"step": 4760
},
{
"epoch": 0.3900245298446443,
"grad_norm": 0.4224013388156891,
"learning_rate": 2e-05,
"loss": 0.7415,
"step": 4770
},
{
"epoch": 0.39084219133278825,
"grad_norm": 0.43009740114212036,
"learning_rate": 2e-05,
"loss": 0.7763,
"step": 4780
},
{
"epoch": 0.39165985282093213,
"grad_norm": 0.3691262900829315,
"learning_rate": 2e-05,
"loss": 0.704,
"step": 4790
},
{
"epoch": 0.39247751430907607,
"grad_norm": 0.42145824432373047,
"learning_rate": 2e-05,
"loss": 0.7226,
"step": 4800
},
{
"epoch": 0.39329517579721995,
"grad_norm": 0.40462708473205566,
"learning_rate": 2e-05,
"loss": 0.7571,
"step": 4810
},
{
"epoch": 0.39411283728536384,
"grad_norm": 0.3951694369316101,
"learning_rate": 2e-05,
"loss": 0.7414,
"step": 4820
},
{
"epoch": 0.3949304987735078,
"grad_norm": 0.3983229696750641,
"learning_rate": 2e-05,
"loss": 0.7525,
"step": 4830
},
{
"epoch": 0.39574816026165166,
"grad_norm": 0.4490920901298523,
"learning_rate": 2e-05,
"loss": 0.7582,
"step": 4840
},
{
"epoch": 0.3965658217497956,
"grad_norm": 0.39146313071250916,
"learning_rate": 2e-05,
"loss": 0.7707,
"step": 4850
},
{
"epoch": 0.3973834832379395,
"grad_norm": 0.4129073917865753,
"learning_rate": 2e-05,
"loss": 0.7246,
"step": 4860
},
{
"epoch": 0.3982011447260834,
"grad_norm": 0.41237565875053406,
"learning_rate": 2e-05,
"loss": 0.7214,
"step": 4870
},
{
"epoch": 0.3990188062142273,
"grad_norm": 0.4975462257862091,
"learning_rate": 2e-05,
"loss": 0.7427,
"step": 4880
},
{
"epoch": 0.39983646770237125,
"grad_norm": 0.444686621427536,
"learning_rate": 2e-05,
"loss": 0.7753,
"step": 4890
},
{
"epoch": 0.40065412919051513,
"grad_norm": 0.40740424394607544,
"learning_rate": 2e-05,
"loss": 0.7414,
"step": 4900
},
{
"epoch": 0.401471790678659,
"grad_norm": 0.4233196973800659,
"learning_rate": 2e-05,
"loss": 0.7496,
"step": 4910
},
{
"epoch": 0.40228945216680295,
"grad_norm": 0.39530307054519653,
"learning_rate": 2e-05,
"loss": 0.769,
"step": 4920
},
{
"epoch": 0.40310711365494684,
"grad_norm": 0.37782007455825806,
"learning_rate": 2e-05,
"loss": 0.7062,
"step": 4930
},
{
"epoch": 0.4039247751430908,
"grad_norm": 0.3919887840747833,
"learning_rate": 2e-05,
"loss": 0.7376,
"step": 4940
},
{
"epoch": 0.40474243663123466,
"grad_norm": 0.4067898690700531,
"learning_rate": 2e-05,
"loss": 0.7607,
"step": 4950
},
{
"epoch": 0.4055600981193786,
"grad_norm": 0.39617830514907837,
"learning_rate": 2e-05,
"loss": 0.7676,
"step": 4960
},
{
"epoch": 0.4063777596075225,
"grad_norm": 0.427827388048172,
"learning_rate": 2e-05,
"loss": 0.7547,
"step": 4970
},
{
"epoch": 0.4071954210956664,
"grad_norm": 0.39542731642723083,
"learning_rate": 2e-05,
"loss": 0.7444,
"step": 4980
},
{
"epoch": 0.4080130825838103,
"grad_norm": 0.386767715215683,
"learning_rate": 2e-05,
"loss": 0.7404,
"step": 4990
},
{
"epoch": 0.4088307440719542,
"grad_norm": 0.38438013195991516,
"learning_rate": 2e-05,
"loss": 0.7154,
"step": 5000
},
{
"epoch": 0.40964840556009813,
"grad_norm": 0.41563791036605835,
"learning_rate": 2e-05,
"loss": 0.7221,
"step": 5010
},
{
"epoch": 0.410466067048242,
"grad_norm": 0.3918059468269348,
"learning_rate": 2e-05,
"loss": 0.7136,
"step": 5020
},
{
"epoch": 0.41128372853638595,
"grad_norm": 0.37461262941360474,
"learning_rate": 2e-05,
"loss": 0.7057,
"step": 5030
},
{
"epoch": 0.41210139002452983,
"grad_norm": 0.4226076006889343,
"learning_rate": 2e-05,
"loss": 0.7531,
"step": 5040
},
{
"epoch": 0.4129190515126738,
"grad_norm": 0.3859594464302063,
"learning_rate": 2e-05,
"loss": 0.746,
"step": 5050
},
{
"epoch": 0.41373671300081766,
"grad_norm": 0.4011143445968628,
"learning_rate": 2e-05,
"loss": 0.7418,
"step": 5060
},
{
"epoch": 0.4145543744889616,
"grad_norm": 0.3988434076309204,
"learning_rate": 2e-05,
"loss": 0.7307,
"step": 5070
},
{
"epoch": 0.4153720359771055,
"grad_norm": 0.4151720106601715,
"learning_rate": 2e-05,
"loss": 0.6879,
"step": 5080
},
{
"epoch": 0.41618969746524936,
"grad_norm": 0.37193483114242554,
"learning_rate": 2e-05,
"loss": 0.76,
"step": 5090
},
{
"epoch": 0.4170073589533933,
"grad_norm": 0.4051276743412018,
"learning_rate": 2e-05,
"loss": 0.7402,
"step": 5100
},
{
"epoch": 0.4178250204415372,
"grad_norm": 0.38233277201652527,
"learning_rate": 2e-05,
"loss": 0.7154,
"step": 5110
},
{
"epoch": 0.4186426819296811,
"grad_norm": 0.6284881830215454,
"learning_rate": 2e-05,
"loss": 0.7324,
"step": 5120
},
{
"epoch": 0.419460343417825,
"grad_norm": 0.4044628441333771,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 5130
},
{
"epoch": 0.42027800490596895,
"grad_norm": 0.4031095802783966,
"learning_rate": 2e-05,
"loss": 0.7286,
"step": 5140
},
{
"epoch": 0.42109566639411283,
"grad_norm": 0.3735601603984833,
"learning_rate": 2e-05,
"loss": 0.7119,
"step": 5150
},
{
"epoch": 0.4219133278822568,
"grad_norm": 0.40492406487464905,
"learning_rate": 2e-05,
"loss": 0.74,
"step": 5160
},
{
"epoch": 0.42273098937040066,
"grad_norm": 0.39481648802757263,
"learning_rate": 2e-05,
"loss": 0.7391,
"step": 5170
},
{
"epoch": 0.42354865085854454,
"grad_norm": 0.5594878792762756,
"learning_rate": 2e-05,
"loss": 0.7757,
"step": 5180
},
{
"epoch": 0.4243663123466885,
"grad_norm": 0.4194161891937256,
"learning_rate": 2e-05,
"loss": 0.7708,
"step": 5190
},
{
"epoch": 0.42518397383483236,
"grad_norm": 0.37851884961128235,
"learning_rate": 2e-05,
"loss": 0.723,
"step": 5200
},
{
"epoch": 0.4260016353229763,
"grad_norm": 0.3945302367210388,
"learning_rate": 2e-05,
"loss": 0.7741,
"step": 5210
},
{
"epoch": 0.4268192968111202,
"grad_norm": 0.39186879992485046,
"learning_rate": 2e-05,
"loss": 0.7644,
"step": 5220
},
{
"epoch": 0.4276369582992641,
"grad_norm": 0.38638511300086975,
"learning_rate": 2e-05,
"loss": 0.7289,
"step": 5230
},
{
"epoch": 0.428454619787408,
"grad_norm": 0.374289333820343,
"learning_rate": 2e-05,
"loss": 0.727,
"step": 5240
},
{
"epoch": 0.42927228127555195,
"grad_norm": 0.3913343548774719,
"learning_rate": 2e-05,
"loss": 0.7107,
"step": 5250
},
{
"epoch": 0.43008994276369583,
"grad_norm": 0.3803686499595642,
"learning_rate": 2e-05,
"loss": 0.7596,
"step": 5260
},
{
"epoch": 0.4309076042518397,
"grad_norm": 0.3877560794353485,
"learning_rate": 2e-05,
"loss": 0.7221,
"step": 5270
},
{
"epoch": 0.43172526573998365,
"grad_norm": 0.39545392990112305,
"learning_rate": 2e-05,
"loss": 0.7344,
"step": 5280
},
{
"epoch": 0.43254292722812754,
"grad_norm": 0.3907477557659149,
"learning_rate": 2e-05,
"loss": 0.6987,
"step": 5290
},
{
"epoch": 0.4333605887162715,
"grad_norm": 0.4057869017124176,
"learning_rate": 2e-05,
"loss": 0.7338,
"step": 5300
},
{
"epoch": 0.43417825020441536,
"grad_norm": 0.39841634035110474,
"learning_rate": 2e-05,
"loss": 0.7711,
"step": 5310
},
{
"epoch": 0.4349959116925593,
"grad_norm": 0.38393041491508484,
"learning_rate": 2e-05,
"loss": 0.7349,
"step": 5320
},
{
"epoch": 0.4358135731807032,
"grad_norm": 0.3844257593154907,
"learning_rate": 2e-05,
"loss": 0.7234,
"step": 5330
},
{
"epoch": 0.4366312346688471,
"grad_norm": 0.40129122138023376,
"learning_rate": 2e-05,
"loss": 0.7341,
"step": 5340
},
{
"epoch": 0.437448896156991,
"grad_norm": 0.4176855683326721,
"learning_rate": 2e-05,
"loss": 0.7457,
"step": 5350
},
{
"epoch": 0.4382665576451349,
"grad_norm": 0.36444923281669617,
"learning_rate": 2e-05,
"loss": 0.7053,
"step": 5360
},
{
"epoch": 0.43908421913327883,
"grad_norm": 0.3858892321586609,
"learning_rate": 2e-05,
"loss": 0.719,
"step": 5370
},
{
"epoch": 0.4399018806214227,
"grad_norm": 0.45291462540626526,
"learning_rate": 2e-05,
"loss": 0.7638,
"step": 5380
},
{
"epoch": 0.44071954210956665,
"grad_norm": 0.41140055656433105,
"learning_rate": 2e-05,
"loss": 0.7652,
"step": 5390
},
{
"epoch": 0.44153720359771054,
"grad_norm": 0.4094192683696747,
"learning_rate": 2e-05,
"loss": 0.7516,
"step": 5400
},
{
"epoch": 0.4423548650858545,
"grad_norm": 0.3956548273563385,
"learning_rate": 2e-05,
"loss": 0.7184,
"step": 5410
},
{
"epoch": 0.44317252657399836,
"grad_norm": 0.3922652006149292,
"learning_rate": 2e-05,
"loss": 0.7147,
"step": 5420
},
{
"epoch": 0.4439901880621423,
"grad_norm": 0.38257864117622375,
"learning_rate": 2e-05,
"loss": 0.7619,
"step": 5430
},
{
"epoch": 0.4448078495502862,
"grad_norm": 0.37260591983795166,
"learning_rate": 2e-05,
"loss": 0.7281,
"step": 5440
},
{
"epoch": 0.44562551103843007,
"grad_norm": 0.4042538106441498,
"learning_rate": 2e-05,
"loss": 0.7078,
"step": 5450
},
{
"epoch": 0.446443172526574,
"grad_norm": 0.3874247968196869,
"learning_rate": 2e-05,
"loss": 0.7351,
"step": 5460
},
{
"epoch": 0.4472608340147179,
"grad_norm": 0.41795167326927185,
"learning_rate": 2e-05,
"loss": 0.7236,
"step": 5470
},
{
"epoch": 0.44807849550286183,
"grad_norm": 0.3942680060863495,
"learning_rate": 2e-05,
"loss": 0.7461,
"step": 5480
},
{
"epoch": 0.4488961569910057,
"grad_norm": 0.3911779820919037,
"learning_rate": 2e-05,
"loss": 0.724,
"step": 5490
},
{
"epoch": 0.44971381847914965,
"grad_norm": 0.39125683903694153,
"learning_rate": 2e-05,
"loss": 0.7299,
"step": 5500
},
{
"epoch": 0.45053147996729354,
"grad_norm": 0.3991476595401764,
"learning_rate": 2e-05,
"loss": 0.7181,
"step": 5510
},
{
"epoch": 0.4513491414554375,
"grad_norm": 0.3854283094406128,
"learning_rate": 2e-05,
"loss": 0.6959,
"step": 5520
},
{
"epoch": 0.45216680294358136,
"grad_norm": 0.4002152681350708,
"learning_rate": 2e-05,
"loss": 0.7655,
"step": 5530
},
{
"epoch": 0.45298446443172524,
"grad_norm": 0.39789167046546936,
"learning_rate": 2e-05,
"loss": 0.727,
"step": 5540
},
{
"epoch": 0.4538021259198692,
"grad_norm": 0.39868685603141785,
"learning_rate": 2e-05,
"loss": 0.7068,
"step": 5550
},
{
"epoch": 0.45461978740801307,
"grad_norm": 0.3992416560649872,
"learning_rate": 2e-05,
"loss": 0.712,
"step": 5560
},
{
"epoch": 0.455437448896157,
"grad_norm": 0.3877620995044708,
"learning_rate": 2e-05,
"loss": 0.7429,
"step": 5570
},
{
"epoch": 0.4562551103843009,
"grad_norm": 0.4050482511520386,
"learning_rate": 2e-05,
"loss": 0.7203,
"step": 5580
},
{
"epoch": 0.4570727718724448,
"grad_norm": 0.3726508617401123,
"learning_rate": 2e-05,
"loss": 0.7545,
"step": 5590
},
{
"epoch": 0.4578904333605887,
"grad_norm": 0.41111037135124207,
"learning_rate": 2e-05,
"loss": 0.7638,
"step": 5600
},
{
"epoch": 0.45870809484873265,
"grad_norm": 0.3929140865802765,
"learning_rate": 2e-05,
"loss": 0.7349,
"step": 5610
},
{
"epoch": 0.45952575633687653,
"grad_norm": 0.4351554811000824,
"learning_rate": 2e-05,
"loss": 0.7575,
"step": 5620
},
{
"epoch": 0.4603434178250204,
"grad_norm": 0.38412046432495117,
"learning_rate": 2e-05,
"loss": 0.722,
"step": 5630
},
{
"epoch": 0.46116107931316436,
"grad_norm": 0.38781240582466125,
"learning_rate": 2e-05,
"loss": 0.744,
"step": 5640
},
{
"epoch": 0.46197874080130824,
"grad_norm": 0.40069451928138733,
"learning_rate": 2e-05,
"loss": 0.7662,
"step": 5650
},
{
"epoch": 0.4627964022894522,
"grad_norm": 0.439037561416626,
"learning_rate": 2e-05,
"loss": 0.7209,
"step": 5660
},
{
"epoch": 0.46361406377759606,
"grad_norm": 0.38045382499694824,
"learning_rate": 2e-05,
"loss": 0.7455,
"step": 5670
},
{
"epoch": 0.46443172526574,
"grad_norm": 0.431621253490448,
"learning_rate": 2e-05,
"loss": 0.729,
"step": 5680
},
{
"epoch": 0.4652493867538839,
"grad_norm": 0.4068882167339325,
"learning_rate": 2e-05,
"loss": 0.7174,
"step": 5690
},
{
"epoch": 0.4660670482420278,
"grad_norm": 0.4053855538368225,
"learning_rate": 2e-05,
"loss": 0.7341,
"step": 5700
},
{
"epoch": 0.4668847097301717,
"grad_norm": 0.4215748608112335,
"learning_rate": 2e-05,
"loss": 0.7326,
"step": 5710
},
{
"epoch": 0.4677023712183156,
"grad_norm": 0.4851236641407013,
"learning_rate": 2e-05,
"loss": 0.7468,
"step": 5720
},
{
"epoch": 0.46852003270645953,
"grad_norm": 0.4110400974750519,
"learning_rate": 2e-05,
"loss": 0.7418,
"step": 5730
},
{
"epoch": 0.4693376941946034,
"grad_norm": 0.3751293122768402,
"learning_rate": 2e-05,
"loss": 0.7046,
"step": 5740
},
{
"epoch": 0.47015535568274736,
"grad_norm": 0.4214579463005066,
"learning_rate": 2e-05,
"loss": 0.7435,
"step": 5750
},
{
"epoch": 0.47097301717089124,
"grad_norm": 0.3882138431072235,
"learning_rate": 2e-05,
"loss": 0.7236,
"step": 5760
},
{
"epoch": 0.4717906786590352,
"grad_norm": 0.4066796600818634,
"learning_rate": 2e-05,
"loss": 0.7126,
"step": 5770
},
{
"epoch": 0.47260834014717906,
"grad_norm": 0.43531790375709534,
"learning_rate": 2e-05,
"loss": 0.7382,
"step": 5780
},
{
"epoch": 0.473426001635323,
"grad_norm": 0.38278236985206604,
"learning_rate": 2e-05,
"loss": 0.7294,
"step": 5790
},
{
"epoch": 0.4742436631234669,
"grad_norm": 0.37796634435653687,
"learning_rate": 2e-05,
"loss": 0.7396,
"step": 5800
},
{
"epoch": 0.47506132461161077,
"grad_norm": 0.4083399474620819,
"learning_rate": 2e-05,
"loss": 0.7139,
"step": 5810
},
{
"epoch": 0.4758789860997547,
"grad_norm": 0.36090272665023804,
"learning_rate": 2e-05,
"loss": 0.7388,
"step": 5820
},
{
"epoch": 0.4766966475878986,
"grad_norm": 0.37881138920783997,
"learning_rate": 2e-05,
"loss": 0.7342,
"step": 5830
},
{
"epoch": 0.47751430907604253,
"grad_norm": 0.3956334590911865,
"learning_rate": 2e-05,
"loss": 0.7316,
"step": 5840
},
{
"epoch": 0.4783319705641864,
"grad_norm": 1.154186487197876,
"learning_rate": 2e-05,
"loss": 0.7002,
"step": 5850
},
{
"epoch": 0.47914963205233035,
"grad_norm": 0.3899221420288086,
"learning_rate": 2e-05,
"loss": 0.7263,
"step": 5860
},
{
"epoch": 0.47996729354047424,
"grad_norm": 0.39512503147125244,
"learning_rate": 2e-05,
"loss": 0.7516,
"step": 5870
},
{
"epoch": 0.4807849550286182,
"grad_norm": 0.41432517766952515,
"learning_rate": 2e-05,
"loss": 0.7366,
"step": 5880
},
{
"epoch": 0.48160261651676206,
"grad_norm": 0.5273881554603577,
"learning_rate": 2e-05,
"loss": 0.7388,
"step": 5890
},
{
"epoch": 0.48242027800490594,
"grad_norm": 0.39111030101776123,
"learning_rate": 2e-05,
"loss": 0.7161,
"step": 5900
},
{
"epoch": 0.4832379394930499,
"grad_norm": 0.3799116611480713,
"learning_rate": 2e-05,
"loss": 0.7266,
"step": 5910
},
{
"epoch": 0.48405560098119377,
"grad_norm": 0.430471271276474,
"learning_rate": 2e-05,
"loss": 0.7356,
"step": 5920
},
{
"epoch": 0.4848732624693377,
"grad_norm": 0.3964878022670746,
"learning_rate": 2e-05,
"loss": 0.7223,
"step": 5930
},
{
"epoch": 0.4856909239574816,
"grad_norm": 0.36826828122138977,
"learning_rate": 2e-05,
"loss": 0.7023,
"step": 5940
},
{
"epoch": 0.48650858544562553,
"grad_norm": 0.37714770436286926,
"learning_rate": 2e-05,
"loss": 0.7582,
"step": 5950
},
{
"epoch": 0.4873262469337694,
"grad_norm": 0.38691821694374084,
"learning_rate": 2e-05,
"loss": 0.7337,
"step": 5960
},
{
"epoch": 0.48814390842191335,
"grad_norm": 0.3870263993740082,
"learning_rate": 2e-05,
"loss": 0.7216,
"step": 5970
},
{
"epoch": 0.48896156991005724,
"grad_norm": 0.38973450660705566,
"learning_rate": 2e-05,
"loss": 0.7364,
"step": 5980
},
{
"epoch": 0.4897792313982011,
"grad_norm": 0.38969680666923523,
"learning_rate": 2e-05,
"loss": 0.7216,
"step": 5990
},
{
"epoch": 0.49059689288634506,
"grad_norm": 0.39326009154319763,
"learning_rate": 2e-05,
"loss": 0.711,
"step": 6000
},
{
"epoch": 0.49141455437448894,
"grad_norm": 0.521453320980072,
"learning_rate": 2e-05,
"loss": 0.7482,
"step": 6010
},
{
"epoch": 0.4922322158626329,
"grad_norm": 0.3774580657482147,
"learning_rate": 2e-05,
"loss": 0.7301,
"step": 6020
},
{
"epoch": 0.49304987735077677,
"grad_norm": 0.40430912375450134,
"learning_rate": 2e-05,
"loss": 0.7346,
"step": 6030
},
{
"epoch": 0.4938675388389207,
"grad_norm": 0.4193746745586395,
"learning_rate": 2e-05,
"loss": 0.7105,
"step": 6040
},
{
"epoch": 0.4946852003270646,
"grad_norm": 0.42411455512046814,
"learning_rate": 2e-05,
"loss": 0.724,
"step": 6050
},
{
"epoch": 0.49550286181520853,
"grad_norm": 0.3788541555404663,
"learning_rate": 2e-05,
"loss": 0.7292,
"step": 6060
},
{
"epoch": 0.4963205233033524,
"grad_norm": 0.5293607711791992,
"learning_rate": 2e-05,
"loss": 0.7143,
"step": 6070
},
{
"epoch": 0.4971381847914963,
"grad_norm": 0.4039698839187622,
"learning_rate": 2e-05,
"loss": 0.7351,
"step": 6080
},
{
"epoch": 0.49795584627964024,
"grad_norm": 0.38631147146224976,
"learning_rate": 2e-05,
"loss": 0.7435,
"step": 6090
},
{
"epoch": 0.4987735077677841,
"grad_norm": 0.38360661268234253,
"learning_rate": 2e-05,
"loss": 0.6995,
"step": 6100
},
{
"epoch": 0.49959116925592806,
"grad_norm": 0.3665395677089691,
"learning_rate": 2e-05,
"loss": 0.7173,
"step": 6110
},
{
"epoch": 0.5004088307440719,
"grad_norm": 0.4371582865715027,
"learning_rate": 2e-05,
"loss": 0.7576,
"step": 6120
},
{
"epoch": 0.5012264922322158,
"grad_norm": 0.3966323435306549,
"learning_rate": 2e-05,
"loss": 0.7153,
"step": 6130
},
{
"epoch": 0.5020441537203598,
"grad_norm": 0.42034411430358887,
"learning_rate": 2e-05,
"loss": 0.7226,
"step": 6140
},
{
"epoch": 0.5028618152085037,
"grad_norm": 0.40780723094940186,
"learning_rate": 2e-05,
"loss": 0.6716,
"step": 6150
},
{
"epoch": 0.5036794766966476,
"grad_norm": 0.40354254841804504,
"learning_rate": 2e-05,
"loss": 0.7159,
"step": 6160
},
{
"epoch": 0.5044971381847915,
"grad_norm": 0.4011727571487427,
"learning_rate": 2e-05,
"loss": 0.7635,
"step": 6170
},
{
"epoch": 0.5053147996729354,
"grad_norm": 0.36710065603256226,
"learning_rate": 2e-05,
"loss": 0.7446,
"step": 6180
},
{
"epoch": 0.5061324611610794,
"grad_norm": 0.4147448241710663,
"learning_rate": 2e-05,
"loss": 0.7237,
"step": 6190
},
{
"epoch": 0.5069501226492232,
"grad_norm": 0.41342294216156006,
"learning_rate": 2e-05,
"loss": 0.7307,
"step": 6200
},
{
"epoch": 0.5077677841373671,
"grad_norm": 0.45012757182121277,
"learning_rate": 2e-05,
"loss": 0.7755,
"step": 6210
},
{
"epoch": 0.508585445625511,
"grad_norm": 0.39716026186943054,
"learning_rate": 2e-05,
"loss": 0.7233,
"step": 6220
},
{
"epoch": 0.509403107113655,
"grad_norm": 0.38225623965263367,
"learning_rate": 2e-05,
"loss": 0.7113,
"step": 6230
},
{
"epoch": 0.5102207686017989,
"grad_norm": 0.4065392315387726,
"learning_rate": 2e-05,
"loss": 0.7033,
"step": 6240
},
{
"epoch": 0.5110384300899428,
"grad_norm": 0.4059799611568451,
"learning_rate": 2e-05,
"loss": 0.7104,
"step": 6250
},
{
"epoch": 0.5118560915780866,
"grad_norm": 0.4158405661582947,
"learning_rate": 2e-05,
"loss": 0.7569,
"step": 6260
},
{
"epoch": 0.5126737530662305,
"grad_norm": 0.4016571640968323,
"learning_rate": 2e-05,
"loss": 0.7127,
"step": 6270
},
{
"epoch": 0.5134914145543745,
"grad_norm": 0.39826682209968567,
"learning_rate": 2e-05,
"loss": 0.7368,
"step": 6280
},
{
"epoch": 0.5143090760425184,
"grad_norm": 0.40669143199920654,
"learning_rate": 2e-05,
"loss": 0.7252,
"step": 6290
},
{
"epoch": 0.5151267375306623,
"grad_norm": 0.39983949065208435,
"learning_rate": 2e-05,
"loss": 0.7089,
"step": 6300
},
{
"epoch": 0.5159443990188062,
"grad_norm": 0.4313853681087494,
"learning_rate": 2e-05,
"loss": 0.7521,
"step": 6310
},
{
"epoch": 0.5167620605069502,
"grad_norm": 0.3945882320404053,
"learning_rate": 2e-05,
"loss": 0.7083,
"step": 6320
},
{
"epoch": 0.517579721995094,
"grad_norm": 0.45251569151878357,
"learning_rate": 2e-05,
"loss": 0.7492,
"step": 6330
},
{
"epoch": 0.5183973834832379,
"grad_norm": 0.3848367929458618,
"learning_rate": 2e-05,
"loss": 0.7297,
"step": 6340
},
{
"epoch": 0.5192150449713818,
"grad_norm": 0.3864355981349945,
"learning_rate": 2e-05,
"loss": 0.7234,
"step": 6350
},
{
"epoch": 0.5200327064595257,
"grad_norm": 0.38432541489601135,
"learning_rate": 2e-05,
"loss": 0.7471,
"step": 6360
},
{
"epoch": 0.5208503679476697,
"grad_norm": 0.3781149387359619,
"learning_rate": 2e-05,
"loss": 0.7093,
"step": 6370
},
{
"epoch": 0.5216680294358136,
"grad_norm": 0.39279818534851074,
"learning_rate": 2e-05,
"loss": 0.7472,
"step": 6380
},
{
"epoch": 0.5224856909239575,
"grad_norm": 0.3886862099170685,
"learning_rate": 2e-05,
"loss": 0.7127,
"step": 6390
},
{
"epoch": 0.5233033524121014,
"grad_norm": 0.4260784089565277,
"learning_rate": 2e-05,
"loss": 0.7453,
"step": 6400
},
{
"epoch": 0.5241210139002453,
"grad_norm": 0.4057619869709015,
"learning_rate": 2e-05,
"loss": 0.724,
"step": 6410
},
{
"epoch": 0.5249386753883892,
"grad_norm": 0.42176517844200134,
"learning_rate": 2e-05,
"loss": 0.7349,
"step": 6420
},
{
"epoch": 0.5257563368765331,
"grad_norm": 0.43129077553749084,
"learning_rate": 2e-05,
"loss": 0.7238,
"step": 6430
},
{
"epoch": 0.526573998364677,
"grad_norm": 0.4219561219215393,
"learning_rate": 2e-05,
"loss": 0.7545,
"step": 6440
},
{
"epoch": 0.5273916598528209,
"grad_norm": 0.3911988437175751,
"learning_rate": 2e-05,
"loss": 0.7236,
"step": 6450
},
{
"epoch": 0.5282093213409649,
"grad_norm": 0.37093886733055115,
"learning_rate": 2e-05,
"loss": 0.7711,
"step": 6460
},
{
"epoch": 0.5290269828291088,
"grad_norm": 0.38755398988723755,
"learning_rate": 2e-05,
"loss": 0.7353,
"step": 6470
},
{
"epoch": 0.5298446443172526,
"grad_norm": 0.37228506803512573,
"learning_rate": 2e-05,
"loss": 0.7285,
"step": 6480
},
{
"epoch": 0.5306623058053965,
"grad_norm": 0.37210145592689514,
"learning_rate": 2e-05,
"loss": 0.7249,
"step": 6490
},
{
"epoch": 0.5314799672935405,
"grad_norm": 0.3976168930530548,
"learning_rate": 2e-05,
"loss": 0.7211,
"step": 6500
},
{
"epoch": 0.5322976287816844,
"grad_norm": 0.3910818099975586,
"learning_rate": 2e-05,
"loss": 0.7478,
"step": 6510
},
{
"epoch": 0.5331152902698283,
"grad_norm": 0.3939887285232544,
"learning_rate": 2e-05,
"loss": 0.7762,
"step": 6520
},
{
"epoch": 0.5339329517579722,
"grad_norm": 0.4082985818386078,
"learning_rate": 2e-05,
"loss": 0.7361,
"step": 6530
},
{
"epoch": 0.5347506132461161,
"grad_norm": 0.3842323124408722,
"learning_rate": 2e-05,
"loss": 0.6993,
"step": 6540
},
{
"epoch": 0.53556827473426,
"grad_norm": 0.3859511911869049,
"learning_rate": 2e-05,
"loss": 0.7331,
"step": 6550
},
{
"epoch": 0.5363859362224039,
"grad_norm": 0.38415318727493286,
"learning_rate": 2e-05,
"loss": 0.7408,
"step": 6560
},
{
"epoch": 0.5372035977105478,
"grad_norm": 0.4089614748954773,
"learning_rate": 2e-05,
"loss": 0.7303,
"step": 6570
},
{
"epoch": 0.5380212591986917,
"grad_norm": 0.4016498923301697,
"learning_rate": 2e-05,
"loss": 0.7213,
"step": 6580
},
{
"epoch": 0.5388389206868357,
"grad_norm": 0.4609025716781616,
"learning_rate": 2e-05,
"loss": 0.7351,
"step": 6590
},
{
"epoch": 0.5396565821749796,
"grad_norm": 0.3984414041042328,
"learning_rate": 2e-05,
"loss": 0.7555,
"step": 6600
},
{
"epoch": 0.5404742436631235,
"grad_norm": 0.41029107570648193,
"learning_rate": 2e-05,
"loss": 0.7302,
"step": 6610
},
{
"epoch": 0.5412919051512673,
"grad_norm": 0.37710443139076233,
"learning_rate": 2e-05,
"loss": 0.7525,
"step": 6620
},
{
"epoch": 0.5421095666394112,
"grad_norm": 0.4033324122428894,
"learning_rate": 2e-05,
"loss": 0.7266,
"step": 6630
},
{
"epoch": 0.5429272281275552,
"grad_norm": 0.4084174931049347,
"learning_rate": 2e-05,
"loss": 0.7144,
"step": 6640
},
{
"epoch": 0.5437448896156991,
"grad_norm": 0.4083803594112396,
"learning_rate": 2e-05,
"loss": 0.7137,
"step": 6650
},
{
"epoch": 0.544562551103843,
"grad_norm": 0.4336593449115753,
"learning_rate": 2e-05,
"loss": 0.7562,
"step": 6660
},
{
"epoch": 0.5453802125919869,
"grad_norm": 0.4053659439086914,
"learning_rate": 2e-05,
"loss": 0.7192,
"step": 6670
},
{
"epoch": 0.5461978740801309,
"grad_norm": 0.4452054798603058,
"learning_rate": 2e-05,
"loss": 0.7237,
"step": 6680
},
{
"epoch": 0.5470155355682748,
"grad_norm": 0.3796253204345703,
"learning_rate": 2e-05,
"loss": 0.707,
"step": 6690
},
{
"epoch": 0.5478331970564186,
"grad_norm": 0.4349365532398224,
"learning_rate": 2e-05,
"loss": 0.7487,
"step": 6700
},
{
"epoch": 0.5486508585445625,
"grad_norm": 0.39009132981300354,
"learning_rate": 2e-05,
"loss": 0.7321,
"step": 6710
},
{
"epoch": 0.5494685200327064,
"grad_norm": 0.4136855900287628,
"learning_rate": 2e-05,
"loss": 0.7003,
"step": 6720
},
{
"epoch": 0.5502861815208504,
"grad_norm": 0.414394736289978,
"learning_rate": 2e-05,
"loss": 0.7354,
"step": 6730
},
{
"epoch": 0.5511038430089943,
"grad_norm": 0.4299166798591614,
"learning_rate": 2e-05,
"loss": 0.7387,
"step": 6740
},
{
"epoch": 0.5519215044971382,
"grad_norm": 0.3743124008178711,
"learning_rate": 2e-05,
"loss": 0.704,
"step": 6750
},
{
"epoch": 0.552739165985282,
"grad_norm": 0.45423370599746704,
"learning_rate": 2e-05,
"loss": 0.7217,
"step": 6760
},
{
"epoch": 0.553556827473426,
"grad_norm": 0.39603668451309204,
"learning_rate": 2e-05,
"loss": 0.7217,
"step": 6770
},
{
"epoch": 0.5543744889615699,
"grad_norm": 0.40863049030303955,
"learning_rate": 2e-05,
"loss": 0.7277,
"step": 6780
},
{
"epoch": 0.5551921504497138,
"grad_norm": 0.36929062008857727,
"learning_rate": 2e-05,
"loss": 0.7405,
"step": 6790
},
{
"epoch": 0.5560098119378577,
"grad_norm": 0.3728216290473938,
"learning_rate": 2e-05,
"loss": 0.7215,
"step": 6800
},
{
"epoch": 0.5568274734260016,
"grad_norm": 0.3949262797832489,
"learning_rate": 2e-05,
"loss": 0.7046,
"step": 6810
},
{
"epoch": 0.5576451349141456,
"grad_norm": 0.3965948224067688,
"learning_rate": 2e-05,
"loss": 0.7532,
"step": 6820
},
{
"epoch": 0.5584627964022895,
"grad_norm": 0.383021742105484,
"learning_rate": 2e-05,
"loss": 0.7225,
"step": 6830
},
{
"epoch": 0.5592804578904333,
"grad_norm": 0.415387362241745,
"learning_rate": 2e-05,
"loss": 0.7394,
"step": 6840
},
{
"epoch": 0.5600981193785772,
"grad_norm": 0.38204288482666016,
"learning_rate": 2e-05,
"loss": 0.731,
"step": 6850
},
{
"epoch": 0.5609157808667212,
"grad_norm": 0.3826631009578705,
"learning_rate": 2e-05,
"loss": 0.7214,
"step": 6860
},
{
"epoch": 0.5617334423548651,
"grad_norm": 0.3958834707736969,
"learning_rate": 2e-05,
"loss": 0.7331,
"step": 6870
},
{
"epoch": 0.562551103843009,
"grad_norm": 0.4243982434272766,
"learning_rate": 2e-05,
"loss": 0.7169,
"step": 6880
},
{
"epoch": 0.5633687653311529,
"grad_norm": 0.38992610573768616,
"learning_rate": 2e-05,
"loss": 0.7043,
"step": 6890
},
{
"epoch": 0.5641864268192968,
"grad_norm": 0.3799871802330017,
"learning_rate": 2e-05,
"loss": 0.7158,
"step": 6900
},
{
"epoch": 0.5650040883074408,
"grad_norm": 0.43410569429397583,
"learning_rate": 2e-05,
"loss": 0.7677,
"step": 6910
},
{
"epoch": 0.5658217497955846,
"grad_norm": 0.3734298646450043,
"learning_rate": 2e-05,
"loss": 0.7228,
"step": 6920
},
{
"epoch": 0.5666394112837285,
"grad_norm": 0.38182392716407776,
"learning_rate": 2e-05,
"loss": 0.7141,
"step": 6930
},
{
"epoch": 0.5674570727718724,
"grad_norm": 0.43535903096199036,
"learning_rate": 2e-05,
"loss": 0.7236,
"step": 6940
},
{
"epoch": 0.5682747342600164,
"grad_norm": 0.37895089387893677,
"learning_rate": 2e-05,
"loss": 0.7249,
"step": 6950
},
{
"epoch": 0.5690923957481603,
"grad_norm": 0.3477756977081299,
"learning_rate": 2e-05,
"loss": 0.7222,
"step": 6960
},
{
"epoch": 0.5699100572363042,
"grad_norm": 0.49796661734580994,
"learning_rate": 2e-05,
"loss": 0.7269,
"step": 6970
},
{
"epoch": 0.570727718724448,
"grad_norm": 0.40733978152275085,
"learning_rate": 2e-05,
"loss": 0.7195,
"step": 6980
},
{
"epoch": 0.5715453802125919,
"grad_norm": 0.3971668481826782,
"learning_rate": 2e-05,
"loss": 0.7181,
"step": 6990
},
{
"epoch": 0.5723630417007359,
"grad_norm": 0.3817703425884247,
"learning_rate": 2e-05,
"loss": 0.7242,
"step": 7000
},
{
"epoch": 0.5731807031888798,
"grad_norm": 0.38494381308555603,
"learning_rate": 2e-05,
"loss": 0.7033,
"step": 7010
},
{
"epoch": 0.5739983646770237,
"grad_norm": 0.38035619258880615,
"learning_rate": 2e-05,
"loss": 0.7169,
"step": 7020
},
{
"epoch": 0.5748160261651676,
"grad_norm": 0.37850189208984375,
"learning_rate": 2e-05,
"loss": 0.742,
"step": 7030
},
{
"epoch": 0.5756336876533116,
"grad_norm": 0.4150765836238861,
"learning_rate": 2e-05,
"loss": 0.7215,
"step": 7040
},
{
"epoch": 0.5764513491414555,
"grad_norm": 0.36908048391342163,
"learning_rate": 2e-05,
"loss": 0.69,
"step": 7050
},
{
"epoch": 0.5772690106295993,
"grad_norm": 0.38219964504241943,
"learning_rate": 2e-05,
"loss": 0.7277,
"step": 7060
},
{
"epoch": 0.5780866721177432,
"grad_norm": 0.4135884642601013,
"learning_rate": 2e-05,
"loss": 0.6966,
"step": 7070
},
{
"epoch": 0.5789043336058871,
"grad_norm": 0.39122429490089417,
"learning_rate": 2e-05,
"loss": 0.7055,
"step": 7080
},
{
"epoch": 0.5797219950940311,
"grad_norm": 0.37881603837013245,
"learning_rate": 2e-05,
"loss": 0.74,
"step": 7090
},
{
"epoch": 0.580539656582175,
"grad_norm": 0.39096370339393616,
"learning_rate": 2e-05,
"loss": 0.7688,
"step": 7100
},
{
"epoch": 0.5813573180703189,
"grad_norm": 0.38468700647354126,
"learning_rate": 2e-05,
"loss": 0.7111,
"step": 7110
},
{
"epoch": 0.5821749795584628,
"grad_norm": 0.3719991445541382,
"learning_rate": 2e-05,
"loss": 0.7318,
"step": 7120
},
{
"epoch": 0.5829926410466068,
"grad_norm": 0.4110388159751892,
"learning_rate": 2e-05,
"loss": 0.7359,
"step": 7130
},
{
"epoch": 0.5838103025347506,
"grad_norm": 0.3928191661834717,
"learning_rate": 2e-05,
"loss": 0.7099,
"step": 7140
},
{
"epoch": 0.5846279640228945,
"grad_norm": 0.38766661286354065,
"learning_rate": 2e-05,
"loss": 0.7294,
"step": 7150
},
{
"epoch": 0.5854456255110384,
"grad_norm": 0.3949083685874939,
"learning_rate": 2e-05,
"loss": 0.7635,
"step": 7160
},
{
"epoch": 0.5862632869991823,
"grad_norm": 0.3682543635368347,
"learning_rate": 2e-05,
"loss": 0.7311,
"step": 7170
},
{
"epoch": 0.5870809484873263,
"grad_norm": 0.398599773645401,
"learning_rate": 2e-05,
"loss": 0.7179,
"step": 7180
},
{
"epoch": 0.5878986099754702,
"grad_norm": 0.41542503237724304,
"learning_rate": 2e-05,
"loss": 0.7329,
"step": 7190
},
{
"epoch": 0.588716271463614,
"grad_norm": 0.38999563455581665,
"learning_rate": 2e-05,
"loss": 0.7355,
"step": 7200
},
{
"epoch": 0.5895339329517579,
"grad_norm": 0.38349151611328125,
"learning_rate": 2e-05,
"loss": 0.7009,
"step": 7210
},
{
"epoch": 0.5903515944399019,
"grad_norm": 0.40047788619995117,
"learning_rate": 2e-05,
"loss": 0.7455,
"step": 7220
},
{
"epoch": 0.5911692559280458,
"grad_norm": 0.3965801000595093,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 7230
},
{
"epoch": 0.5919869174161897,
"grad_norm": 0.39024582505226135,
"learning_rate": 2e-05,
"loss": 0.7368,
"step": 7240
},
{
"epoch": 0.5928045789043336,
"grad_norm": 0.3906495273113251,
"learning_rate": 2e-05,
"loss": 0.6758,
"step": 7250
},
{
"epoch": 0.5936222403924775,
"grad_norm": 0.36905229091644287,
"learning_rate": 2e-05,
"loss": 0.7214,
"step": 7260
},
{
"epoch": 0.5944399018806215,
"grad_norm": 0.3941437602043152,
"learning_rate": 2e-05,
"loss": 0.7058,
"step": 7270
},
{
"epoch": 0.5952575633687653,
"grad_norm": 0.3895969092845917,
"learning_rate": 2e-05,
"loss": 0.7229,
"step": 7280
},
{
"epoch": 0.5960752248569092,
"grad_norm": 0.3826324939727783,
"learning_rate": 2e-05,
"loss": 0.7187,
"step": 7290
},
{
"epoch": 0.5968928863450531,
"grad_norm": 0.38967013359069824,
"learning_rate": 2e-05,
"loss": 0.7314,
"step": 7300
},
{
"epoch": 0.5977105478331971,
"grad_norm": 0.3706248998641968,
"learning_rate": 2e-05,
"loss": 0.703,
"step": 7310
},
{
"epoch": 0.598528209321341,
"grad_norm": 0.3898773193359375,
"learning_rate": 2e-05,
"loss": 0.7429,
"step": 7320
},
{
"epoch": 0.5993458708094849,
"grad_norm": 0.40563520789146423,
"learning_rate": 2e-05,
"loss": 0.7523,
"step": 7330
},
{
"epoch": 0.6001635322976288,
"grad_norm": 0.3647954761981964,
"learning_rate": 2e-05,
"loss": 0.7207,
"step": 7340
},
{
"epoch": 0.6009811937857726,
"grad_norm": 0.4183512032032013,
"learning_rate": 2e-05,
"loss": 0.7416,
"step": 7350
},
{
"epoch": 0.6017988552739166,
"grad_norm": 0.40611550211906433,
"learning_rate": 2e-05,
"loss": 0.7041,
"step": 7360
},
{
"epoch": 0.6026165167620605,
"grad_norm": 0.4383697509765625,
"learning_rate": 2e-05,
"loss": 0.7414,
"step": 7370
},
{
"epoch": 0.6034341782502044,
"grad_norm": 0.3718542158603668,
"learning_rate": 2e-05,
"loss": 0.7117,
"step": 7380
},
{
"epoch": 0.6042518397383483,
"grad_norm": 0.3631412386894226,
"learning_rate": 2e-05,
"loss": 0.6972,
"step": 7390
},
{
"epoch": 0.6050695012264923,
"grad_norm": 0.3969319760799408,
"learning_rate": 2e-05,
"loss": 0.7399,
"step": 7400
},
{
"epoch": 0.6058871627146362,
"grad_norm": 0.42158883810043335,
"learning_rate": 2e-05,
"loss": 0.718,
"step": 7410
},
{
"epoch": 0.60670482420278,
"grad_norm": 0.4003793001174927,
"learning_rate": 2e-05,
"loss": 0.7081,
"step": 7420
},
{
"epoch": 0.6075224856909239,
"grad_norm": 0.36187267303466797,
"learning_rate": 2e-05,
"loss": 0.7004,
"step": 7430
},
{
"epoch": 0.6083401471790678,
"grad_norm": 0.37817618250846863,
"learning_rate": 2e-05,
"loss": 0.694,
"step": 7440
},
{
"epoch": 0.6091578086672118,
"grad_norm": 0.38521474599838257,
"learning_rate": 2e-05,
"loss": 0.7173,
"step": 7450
},
{
"epoch": 0.6099754701553557,
"grad_norm": 0.3693437874317169,
"learning_rate": 2e-05,
"loss": 0.737,
"step": 7460
},
{
"epoch": 0.6107931316434996,
"grad_norm": 0.40069666504859924,
"learning_rate": 2e-05,
"loss": 0.7471,
"step": 7470
},
{
"epoch": 0.6116107931316435,
"grad_norm": 0.3843598961830139,
"learning_rate": 2e-05,
"loss": 0.6904,
"step": 7480
},
{
"epoch": 0.6124284546197875,
"grad_norm": 0.3898952603340149,
"learning_rate": 2e-05,
"loss": 0.7124,
"step": 7490
},
{
"epoch": 0.6132461161079313,
"grad_norm": 0.36968499422073364,
"learning_rate": 2e-05,
"loss": 0.7134,
"step": 7500
},
{
"epoch": 0.6140637775960752,
"grad_norm": 0.46105071902275085,
"learning_rate": 2e-05,
"loss": 0.7513,
"step": 7510
},
{
"epoch": 0.6148814390842191,
"grad_norm": 0.39693546295166016,
"learning_rate": 2e-05,
"loss": 0.6899,
"step": 7520
},
{
"epoch": 0.615699100572363,
"grad_norm": 0.3977808654308319,
"learning_rate": 2e-05,
"loss": 0.716,
"step": 7530
},
{
"epoch": 0.616516762060507,
"grad_norm": 0.38357222080230713,
"learning_rate": 2e-05,
"loss": 0.7262,
"step": 7540
},
{
"epoch": 0.6173344235486509,
"grad_norm": 0.41151443123817444,
"learning_rate": 2e-05,
"loss": 0.7181,
"step": 7550
},
{
"epoch": 0.6181520850367948,
"grad_norm": 0.4048464894294739,
"learning_rate": 2e-05,
"loss": 0.7273,
"step": 7560
},
{
"epoch": 0.6189697465249386,
"grad_norm": 0.4024839997291565,
"learning_rate": 2e-05,
"loss": 0.7171,
"step": 7570
},
{
"epoch": 0.6197874080130826,
"grad_norm": 0.38937562704086304,
"learning_rate": 2e-05,
"loss": 0.7329,
"step": 7580
},
{
"epoch": 0.6206050695012265,
"grad_norm": 0.3667527735233307,
"learning_rate": 2e-05,
"loss": 0.7134,
"step": 7590
},
{
"epoch": 0.6214227309893704,
"grad_norm": 0.37961307168006897,
"learning_rate": 2e-05,
"loss": 0.7071,
"step": 7600
},
{
"epoch": 0.6222403924775143,
"grad_norm": 0.38057851791381836,
"learning_rate": 2e-05,
"loss": 0.7324,
"step": 7610
},
{
"epoch": 0.6230580539656582,
"grad_norm": 0.3772267699241638,
"learning_rate": 2e-05,
"loss": 0.6937,
"step": 7620
},
{
"epoch": 0.6238757154538022,
"grad_norm": 0.3708053529262543,
"learning_rate": 2e-05,
"loss": 0.6973,
"step": 7630
},
{
"epoch": 0.624693376941946,
"grad_norm": 0.3787739872932434,
"learning_rate": 2e-05,
"loss": 0.7102,
"step": 7640
},
{
"epoch": 0.6255110384300899,
"grad_norm": 0.3880579471588135,
"learning_rate": 2e-05,
"loss": 0.6806,
"step": 7650
},
{
"epoch": 0.6263286999182338,
"grad_norm": 0.3797479569911957,
"learning_rate": 2e-05,
"loss": 0.7325,
"step": 7660
},
{
"epoch": 0.6271463614063778,
"grad_norm": 0.3906855881214142,
"learning_rate": 2e-05,
"loss": 0.6996,
"step": 7670
},
{
"epoch": 0.6279640228945217,
"grad_norm": 0.3645192086696625,
"learning_rate": 2e-05,
"loss": 0.7065,
"step": 7680
},
{
"epoch": 0.6287816843826656,
"grad_norm": 0.4076537489891052,
"learning_rate": 2e-05,
"loss": 0.7263,
"step": 7690
},
{
"epoch": 0.6295993458708095,
"grad_norm": 0.39263322949409485,
"learning_rate": 2e-05,
"loss": 0.7284,
"step": 7700
},
{
"epoch": 0.6304170073589533,
"grad_norm": 0.398645281791687,
"learning_rate": 2e-05,
"loss": 0.7009,
"step": 7710
},
{
"epoch": 0.6312346688470973,
"grad_norm": 0.3786969482898712,
"learning_rate": 2e-05,
"loss": 0.7112,
"step": 7720
},
{
"epoch": 0.6320523303352412,
"grad_norm": 0.417774498462677,
"learning_rate": 2e-05,
"loss": 0.7307,
"step": 7730
},
{
"epoch": 0.6328699918233851,
"grad_norm": 0.37405505776405334,
"learning_rate": 2e-05,
"loss": 0.7227,
"step": 7740
},
{
"epoch": 0.633687653311529,
"grad_norm": 0.3912171423435211,
"learning_rate": 2e-05,
"loss": 0.7159,
"step": 7750
},
{
"epoch": 0.634505314799673,
"grad_norm": 0.37108883261680603,
"learning_rate": 2e-05,
"loss": 0.7131,
"step": 7760
},
{
"epoch": 0.6353229762878169,
"grad_norm": 0.38099753856658936,
"learning_rate": 2e-05,
"loss": 0.7161,
"step": 7770
},
{
"epoch": 0.6361406377759607,
"grad_norm": 0.3800545930862427,
"learning_rate": 2e-05,
"loss": 0.7364,
"step": 7780
},
{
"epoch": 0.6369582992641046,
"grad_norm": 0.37537217140197754,
"learning_rate": 2e-05,
"loss": 0.682,
"step": 7790
},
{
"epoch": 0.6377759607522485,
"grad_norm": 0.3771691918373108,
"learning_rate": 2e-05,
"loss": 0.7254,
"step": 7800
},
{
"epoch": 0.6385936222403925,
"grad_norm": 0.3979094624519348,
"learning_rate": 2e-05,
"loss": 0.7218,
"step": 7810
},
{
"epoch": 0.6394112837285364,
"grad_norm": 0.3696105182170868,
"learning_rate": 2e-05,
"loss": 0.7069,
"step": 7820
},
{
"epoch": 0.6402289452166803,
"grad_norm": 0.40755414962768555,
"learning_rate": 2e-05,
"loss": 0.7455,
"step": 7830
},
{
"epoch": 0.6410466067048242,
"grad_norm": 0.3854145109653473,
"learning_rate": 2e-05,
"loss": 0.7108,
"step": 7840
},
{
"epoch": 0.6418642681929682,
"grad_norm": 0.37715867161750793,
"learning_rate": 2e-05,
"loss": 0.739,
"step": 7850
},
{
"epoch": 0.642681929681112,
"grad_norm": 0.438903272151947,
"learning_rate": 2e-05,
"loss": 0.7358,
"step": 7860
},
{
"epoch": 0.6434995911692559,
"grad_norm": 0.39477598667144775,
"learning_rate": 2e-05,
"loss": 0.7381,
"step": 7870
},
{
"epoch": 0.6443172526573998,
"grad_norm": 0.40717998147010803,
"learning_rate": 2e-05,
"loss": 0.692,
"step": 7880
},
{
"epoch": 0.6451349141455437,
"grad_norm": 0.38728293776512146,
"learning_rate": 2e-05,
"loss": 0.679,
"step": 7890
},
{
"epoch": 0.6459525756336877,
"grad_norm": 0.40416544675827026,
"learning_rate": 2e-05,
"loss": 0.7495,
"step": 7900
},
{
"epoch": 0.6467702371218316,
"grad_norm": 0.40616118907928467,
"learning_rate": 2e-05,
"loss": 0.7133,
"step": 7910
},
{
"epoch": 0.6475878986099755,
"grad_norm": 0.39391154050827026,
"learning_rate": 2e-05,
"loss": 0.6979,
"step": 7920
},
{
"epoch": 0.6484055600981193,
"grad_norm": 0.44841650128364563,
"learning_rate": 2e-05,
"loss": 0.7292,
"step": 7930
},
{
"epoch": 0.6492232215862633,
"grad_norm": 0.39034605026245117,
"learning_rate": 2e-05,
"loss": 0.6994,
"step": 7940
},
{
"epoch": 0.6500408830744072,
"grad_norm": 0.42051780223846436,
"learning_rate": 2e-05,
"loss": 0.7296,
"step": 7950
},
{
"epoch": 0.6508585445625511,
"grad_norm": 0.4429325759410858,
"learning_rate": 2e-05,
"loss": 0.7301,
"step": 7960
},
{
"epoch": 0.651676206050695,
"grad_norm": 0.3801387846469879,
"learning_rate": 2e-05,
"loss": 0.7016,
"step": 7970
},
{
"epoch": 0.6524938675388389,
"grad_norm": 0.36322730779647827,
"learning_rate": 2e-05,
"loss": 0.7137,
"step": 7980
},
{
"epoch": 0.6533115290269829,
"grad_norm": 0.37573617696762085,
"learning_rate": 2e-05,
"loss": 0.6941,
"step": 7990
},
{
"epoch": 0.6541291905151267,
"grad_norm": 0.37574875354766846,
"learning_rate": 2e-05,
"loss": 0.7156,
"step": 8000
},
{
"epoch": 0.6549468520032706,
"grad_norm": 0.36992698907852173,
"learning_rate": 2e-05,
"loss": 0.6922,
"step": 8010
},
{
"epoch": 0.6557645134914145,
"grad_norm": 0.3979203402996063,
"learning_rate": 2e-05,
"loss": 0.7184,
"step": 8020
},
{
"epoch": 0.6565821749795585,
"grad_norm": 0.3656754791736603,
"learning_rate": 2e-05,
"loss": 0.7081,
"step": 8030
},
{
"epoch": 0.6573998364677024,
"grad_norm": 0.42211803793907166,
"learning_rate": 2e-05,
"loss": 0.7315,
"step": 8040
},
{
"epoch": 0.6582174979558463,
"grad_norm": 0.3930283188819885,
"learning_rate": 2e-05,
"loss": 0.7317,
"step": 8050
},
{
"epoch": 0.6590351594439902,
"grad_norm": 0.39533066749572754,
"learning_rate": 2e-05,
"loss": 0.7258,
"step": 8060
},
{
"epoch": 0.659852820932134,
"grad_norm": 0.39132604002952576,
"learning_rate": 2e-05,
"loss": 0.7002,
"step": 8070
},
{
"epoch": 0.660670482420278,
"grad_norm": 0.39889493584632874,
"learning_rate": 2e-05,
"loss": 0.7242,
"step": 8080
},
{
"epoch": 0.6614881439084219,
"grad_norm": 0.4047960340976715,
"learning_rate": 2e-05,
"loss": 0.7314,
"step": 8090
},
{
"epoch": 0.6623058053965658,
"grad_norm": 0.40466776490211487,
"learning_rate": 2e-05,
"loss": 0.7339,
"step": 8100
},
{
"epoch": 0.6631234668847097,
"grad_norm": 0.368439644575119,
"learning_rate": 2e-05,
"loss": 0.6953,
"step": 8110
},
{
"epoch": 0.6639411283728537,
"grad_norm": 0.3857625722885132,
"learning_rate": 2e-05,
"loss": 0.7205,
"step": 8120
},
{
"epoch": 0.6647587898609976,
"grad_norm": 0.3679870367050171,
"learning_rate": 2e-05,
"loss": 0.7028,
"step": 8130
},
{
"epoch": 0.6655764513491415,
"grad_norm": 0.3780772387981415,
"learning_rate": 2e-05,
"loss": 0.732,
"step": 8140
},
{
"epoch": 0.6663941128372853,
"grad_norm": 0.37176239490509033,
"learning_rate": 2e-05,
"loss": 0.7174,
"step": 8150
},
{
"epoch": 0.6672117743254292,
"grad_norm": 0.3872131407260895,
"learning_rate": 2e-05,
"loss": 0.7008,
"step": 8160
},
{
"epoch": 0.6680294358135732,
"grad_norm": 0.4394111931324005,
"learning_rate": 2e-05,
"loss": 0.7282,
"step": 8170
},
{
"epoch": 0.6688470973017171,
"grad_norm": 0.39667439460754395,
"learning_rate": 2e-05,
"loss": 0.6996,
"step": 8180
},
{
"epoch": 0.669664758789861,
"grad_norm": 0.36127331852912903,
"learning_rate": 2e-05,
"loss": 0.73,
"step": 8190
},
{
"epoch": 0.6704824202780049,
"grad_norm": 0.390909343957901,
"learning_rate": 2e-05,
"loss": 0.7095,
"step": 8200
},
{
"epoch": 0.6713000817661489,
"grad_norm": 0.4127674400806427,
"learning_rate": 2e-05,
"loss": 0.7135,
"step": 8210
},
{
"epoch": 0.6721177432542927,
"grad_norm": 0.3576696515083313,
"learning_rate": 2e-05,
"loss": 0.7204,
"step": 8220
},
{
"epoch": 0.6729354047424366,
"grad_norm": 0.3902510702610016,
"learning_rate": 2e-05,
"loss": 0.7203,
"step": 8230
},
{
"epoch": 0.6737530662305805,
"grad_norm": 0.38031139969825745,
"learning_rate": 2e-05,
"loss": 0.7187,
"step": 8240
},
{
"epoch": 0.6745707277187244,
"grad_norm": 0.38591304421424866,
"learning_rate": 2e-05,
"loss": 0.6947,
"step": 8250
},
{
"epoch": 0.6753883892068684,
"grad_norm": 0.41441458463668823,
"learning_rate": 2e-05,
"loss": 0.7076,
"step": 8260
},
{
"epoch": 0.6762060506950123,
"grad_norm": 0.37392786145210266,
"learning_rate": 2e-05,
"loss": 0.6782,
"step": 8270
},
{
"epoch": 0.6770237121831562,
"grad_norm": 0.3913317024707794,
"learning_rate": 2e-05,
"loss": 0.7572,
"step": 8280
},
{
"epoch": 0.6778413736713,
"grad_norm": 0.36678120493888855,
"learning_rate": 2e-05,
"loss": 0.6992,
"step": 8290
},
{
"epoch": 0.678659035159444,
"grad_norm": 0.3854345679283142,
"learning_rate": 2e-05,
"loss": 0.6969,
"step": 8300
},
{
"epoch": 0.6794766966475879,
"grad_norm": 0.38369277119636536,
"learning_rate": 2e-05,
"loss": 0.7325,
"step": 8310
},
{
"epoch": 0.6802943581357318,
"grad_norm": 0.3791201710700989,
"learning_rate": 2e-05,
"loss": 0.7142,
"step": 8320
},
{
"epoch": 0.6811120196238757,
"grad_norm": 0.40714752674102783,
"learning_rate": 2e-05,
"loss": 0.712,
"step": 8330
},
{
"epoch": 0.6819296811120196,
"grad_norm": 0.3877699375152588,
"learning_rate": 2e-05,
"loss": 0.6943,
"step": 8340
},
{
"epoch": 0.6827473426001636,
"grad_norm": 0.3655410706996918,
"learning_rate": 2e-05,
"loss": 0.7067,
"step": 8350
},
{
"epoch": 0.6835650040883074,
"grad_norm": 0.3765968084335327,
"learning_rate": 2e-05,
"loss": 0.6957,
"step": 8360
},
{
"epoch": 0.6843826655764513,
"grad_norm": 0.41230982542037964,
"learning_rate": 2e-05,
"loss": 0.7171,
"step": 8370
},
{
"epoch": 0.6852003270645952,
"grad_norm": 0.4245662987232208,
"learning_rate": 2e-05,
"loss": 0.7176,
"step": 8380
},
{
"epoch": 0.6860179885527392,
"grad_norm": 0.39811086654663086,
"learning_rate": 2e-05,
"loss": 0.7191,
"step": 8390
},
{
"epoch": 0.6868356500408831,
"grad_norm": 0.4246889650821686,
"learning_rate": 2e-05,
"loss": 0.7268,
"step": 8400
},
{
"epoch": 0.687653311529027,
"grad_norm": 0.3825872540473938,
"learning_rate": 2e-05,
"loss": 0.6999,
"step": 8410
},
{
"epoch": 0.6884709730171709,
"grad_norm": 0.3795316517353058,
"learning_rate": 2e-05,
"loss": 0.7081,
"step": 8420
},
{
"epoch": 0.6892886345053147,
"grad_norm": 0.375752717256546,
"learning_rate": 2e-05,
"loss": 0.7175,
"step": 8430
},
{
"epoch": 0.6901062959934587,
"grad_norm": 0.44567587971687317,
"learning_rate": 2e-05,
"loss": 0.7021,
"step": 8440
},
{
"epoch": 0.6909239574816026,
"grad_norm": 0.3814459443092346,
"learning_rate": 2e-05,
"loss": 0.7309,
"step": 8450
},
{
"epoch": 0.6917416189697465,
"grad_norm": 0.4140760004520416,
"learning_rate": 2e-05,
"loss": 0.7201,
"step": 8460
},
{
"epoch": 0.6925592804578904,
"grad_norm": 0.38124480843544006,
"learning_rate": 2e-05,
"loss": 0.7103,
"step": 8470
},
{
"epoch": 0.6933769419460344,
"grad_norm": 0.38113194704055786,
"learning_rate": 2e-05,
"loss": 0.7274,
"step": 8480
},
{
"epoch": 0.6941946034341783,
"grad_norm": 0.4052790105342865,
"learning_rate": 2e-05,
"loss": 0.7303,
"step": 8490
},
{
"epoch": 0.6950122649223222,
"grad_norm": 0.37977078557014465,
"learning_rate": 2e-05,
"loss": 0.7384,
"step": 8500
},
{
"epoch": 0.695829926410466,
"grad_norm": 0.372354656457901,
"learning_rate": 2e-05,
"loss": 0.7156,
"step": 8510
},
{
"epoch": 0.6966475878986099,
"grad_norm": 0.35826221108436584,
"learning_rate": 2e-05,
"loss": 0.7015,
"step": 8520
},
{
"epoch": 0.6974652493867539,
"grad_norm": 0.3903435468673706,
"learning_rate": 2e-05,
"loss": 0.7104,
"step": 8530
},
{
"epoch": 0.6982829108748978,
"grad_norm": 0.37950003147125244,
"learning_rate": 2e-05,
"loss": 0.677,
"step": 8540
},
{
"epoch": 0.6991005723630417,
"grad_norm": 0.37612470984458923,
"learning_rate": 2e-05,
"loss": 0.7138,
"step": 8550
},
{
"epoch": 0.6999182338511856,
"grad_norm": 0.40326419472694397,
"learning_rate": 2e-05,
"loss": 0.7125,
"step": 8560
},
{
"epoch": 0.7007358953393296,
"grad_norm": 0.41126900911331177,
"learning_rate": 2e-05,
"loss": 0.7542,
"step": 8570
},
{
"epoch": 0.7015535568274734,
"grad_norm": 0.3865297734737396,
"learning_rate": 2e-05,
"loss": 0.6896,
"step": 8580
},
{
"epoch": 0.7023712183156173,
"grad_norm": 0.3726907968521118,
"learning_rate": 2e-05,
"loss": 0.7147,
"step": 8590
},
{
"epoch": 0.7031888798037612,
"grad_norm": 0.3943450152873993,
"learning_rate": 2e-05,
"loss": 0.714,
"step": 8600
},
{
"epoch": 0.7040065412919051,
"grad_norm": 0.3850589990615845,
"learning_rate": 2e-05,
"loss": 0.7225,
"step": 8610
},
{
"epoch": 0.7048242027800491,
"grad_norm": 0.384355366230011,
"learning_rate": 2e-05,
"loss": 0.6945,
"step": 8620
},
{
"epoch": 0.705641864268193,
"grad_norm": 0.36848247051239014,
"learning_rate": 2e-05,
"loss": 0.6983,
"step": 8630
},
{
"epoch": 0.7064595257563369,
"grad_norm": 0.42313793301582336,
"learning_rate": 2e-05,
"loss": 0.7028,
"step": 8640
},
{
"epoch": 0.7072771872444807,
"grad_norm": 0.39240145683288574,
"learning_rate": 2e-05,
"loss": 0.7014,
"step": 8650
},
{
"epoch": 0.7080948487326247,
"grad_norm": 0.3745633661746979,
"learning_rate": 2e-05,
"loss": 0.7014,
"step": 8660
},
{
"epoch": 0.7089125102207686,
"grad_norm": 0.38909491896629333,
"learning_rate": 2e-05,
"loss": 0.7008,
"step": 8670
},
{
"epoch": 0.7097301717089125,
"grad_norm": 0.4325014352798462,
"learning_rate": 2e-05,
"loss": 0.7064,
"step": 8680
},
{
"epoch": 0.7105478331970564,
"grad_norm": 0.3801126182079315,
"learning_rate": 2e-05,
"loss": 0.6973,
"step": 8690
},
{
"epoch": 0.7113654946852003,
"grad_norm": 0.38816311955451965,
"learning_rate": 2e-05,
"loss": 0.7228,
"step": 8700
},
{
"epoch": 0.7121831561733443,
"grad_norm": 0.36710384488105774,
"learning_rate": 2e-05,
"loss": 0.6889,
"step": 8710
},
{
"epoch": 0.7130008176614882,
"grad_norm": 0.40296927094459534,
"learning_rate": 2e-05,
"loss": 0.7053,
"step": 8720
},
{
"epoch": 0.713818479149632,
"grad_norm": 0.38554781675338745,
"learning_rate": 2e-05,
"loss": 0.6962,
"step": 8730
},
{
"epoch": 0.7146361406377759,
"grad_norm": 0.3844847083091736,
"learning_rate": 2e-05,
"loss": 0.7044,
"step": 8740
},
{
"epoch": 0.7154538021259199,
"grad_norm": 0.42103567719459534,
"learning_rate": 2e-05,
"loss": 0.7026,
"step": 8750
},
{
"epoch": 0.7162714636140638,
"grad_norm": 0.3701580762863159,
"learning_rate": 2e-05,
"loss": 0.7313,
"step": 8760
},
{
"epoch": 0.7170891251022077,
"grad_norm": 0.38155385851860046,
"learning_rate": 2e-05,
"loss": 0.7101,
"step": 8770
},
{
"epoch": 0.7179067865903516,
"grad_norm": 0.3899162709712982,
"learning_rate": 2e-05,
"loss": 0.7255,
"step": 8780
},
{
"epoch": 0.7187244480784954,
"grad_norm": 0.4215267598628998,
"learning_rate": 2e-05,
"loss": 0.7178,
"step": 8790
},
{
"epoch": 0.7195421095666394,
"grad_norm": 0.36979684233665466,
"learning_rate": 2e-05,
"loss": 0.6829,
"step": 8800
},
{
"epoch": 0.7203597710547833,
"grad_norm": 0.3744196891784668,
"learning_rate": 2e-05,
"loss": 0.7007,
"step": 8810
},
{
"epoch": 0.7211774325429272,
"grad_norm": 0.5159342288970947,
"learning_rate": 2e-05,
"loss": 0.6958,
"step": 8820
},
{
"epoch": 0.7219950940310711,
"grad_norm": 0.3716892898082733,
"learning_rate": 2e-05,
"loss": 0.7245,
"step": 8830
},
{
"epoch": 0.7228127555192151,
"grad_norm": 0.37018489837646484,
"learning_rate": 2e-05,
"loss": 0.7285,
"step": 8840
},
{
"epoch": 0.723630417007359,
"grad_norm": 0.37027207016944885,
"learning_rate": 2e-05,
"loss": 0.6894,
"step": 8850
},
{
"epoch": 0.7244480784955029,
"grad_norm": 0.3884619176387787,
"learning_rate": 2e-05,
"loss": 0.7104,
"step": 8860
},
{
"epoch": 0.7252657399836467,
"grad_norm": 0.41086649894714355,
"learning_rate": 2e-05,
"loss": 0.7062,
"step": 8870
},
{
"epoch": 0.7260834014717906,
"grad_norm": 0.40272092819213867,
"learning_rate": 2e-05,
"loss": 0.7287,
"step": 8880
},
{
"epoch": 0.7269010629599346,
"grad_norm": 0.3882623314857483,
"learning_rate": 2e-05,
"loss": 0.7307,
"step": 8890
},
{
"epoch": 0.7277187244480785,
"grad_norm": 0.39466193318367004,
"learning_rate": 2e-05,
"loss": 0.7076,
"step": 8900
},
{
"epoch": 0.7285363859362224,
"grad_norm": 0.3991258144378662,
"learning_rate": 2e-05,
"loss": 0.7473,
"step": 8910
},
{
"epoch": 0.7293540474243663,
"grad_norm": 0.3717644214630127,
"learning_rate": 2e-05,
"loss": 0.7139,
"step": 8920
},
{
"epoch": 0.7301717089125103,
"grad_norm": 0.40505483746528625,
"learning_rate": 2e-05,
"loss": 0.7297,
"step": 8930
},
{
"epoch": 0.7309893704006541,
"grad_norm": 0.40505021810531616,
"learning_rate": 2e-05,
"loss": 0.7063,
"step": 8940
},
{
"epoch": 0.731807031888798,
"grad_norm": 0.39017975330352783,
"learning_rate": 2e-05,
"loss": 0.7148,
"step": 8950
},
{
"epoch": 0.7326246933769419,
"grad_norm": 0.6285465359687805,
"learning_rate": 2e-05,
"loss": 0.7141,
"step": 8960
},
{
"epoch": 0.7334423548650858,
"grad_norm": 0.3755154311656952,
"learning_rate": 2e-05,
"loss": 0.7148,
"step": 8970
},
{
"epoch": 0.7342600163532298,
"grad_norm": 0.34206828474998474,
"learning_rate": 2e-05,
"loss": 0.7272,
"step": 8980
},
{
"epoch": 0.7350776778413737,
"grad_norm": 0.4102880656719208,
"learning_rate": 2e-05,
"loss": 0.7019,
"step": 8990
},
{
"epoch": 0.7358953393295176,
"grad_norm": 0.38105693459510803,
"learning_rate": 2e-05,
"loss": 0.7308,
"step": 9000
},
{
"epoch": 0.7367130008176614,
"grad_norm": 0.37958770990371704,
"learning_rate": 2e-05,
"loss": 0.7291,
"step": 9010
},
{
"epoch": 0.7375306623058054,
"grad_norm": 0.394216388463974,
"learning_rate": 2e-05,
"loss": 0.6685,
"step": 9020
},
{
"epoch": 0.7383483237939493,
"grad_norm": 0.3768483102321625,
"learning_rate": 2e-05,
"loss": 0.7334,
"step": 9030
},
{
"epoch": 0.7391659852820932,
"grad_norm": 0.40919482707977295,
"learning_rate": 2e-05,
"loss": 0.723,
"step": 9040
},
{
"epoch": 0.7399836467702371,
"grad_norm": 0.3880603611469269,
"learning_rate": 2e-05,
"loss": 0.6984,
"step": 9050
},
{
"epoch": 0.740801308258381,
"grad_norm": 0.4032613933086395,
"learning_rate": 2e-05,
"loss": 0.7244,
"step": 9060
},
{
"epoch": 0.741618969746525,
"grad_norm": 0.4074728786945343,
"learning_rate": 2e-05,
"loss": 0.6826,
"step": 9070
},
{
"epoch": 0.7424366312346689,
"grad_norm": 0.35620808601379395,
"learning_rate": 2e-05,
"loss": 0.6694,
"step": 9080
},
{
"epoch": 0.7432542927228127,
"grad_norm": 0.3847275972366333,
"learning_rate": 2e-05,
"loss": 0.7575,
"step": 9090
},
{
"epoch": 0.7440719542109566,
"grad_norm": 0.39583247900009155,
"learning_rate": 2e-05,
"loss": 0.707,
"step": 9100
},
{
"epoch": 0.7448896156991006,
"grad_norm": 0.4044800400733948,
"learning_rate": 2e-05,
"loss": 0.735,
"step": 9110
},
{
"epoch": 0.7457072771872445,
"grad_norm": 0.36118680238723755,
"learning_rate": 2e-05,
"loss": 0.7156,
"step": 9120
},
{
"epoch": 0.7465249386753884,
"grad_norm": 0.39010629057884216,
"learning_rate": 2e-05,
"loss": 0.6926,
"step": 9130
},
{
"epoch": 0.7473426001635323,
"grad_norm": 0.37552088499069214,
"learning_rate": 2e-05,
"loss": 0.7091,
"step": 9140
},
{
"epoch": 0.7481602616516762,
"grad_norm": 0.3881646692752838,
"learning_rate": 2e-05,
"loss": 0.7117,
"step": 9150
},
{
"epoch": 0.7489779231398201,
"grad_norm": 0.40854597091674805,
"learning_rate": 2e-05,
"loss": 0.7084,
"step": 9160
},
{
"epoch": 0.749795584627964,
"grad_norm": 0.37144872546195984,
"learning_rate": 2e-05,
"loss": 0.6877,
"step": 9170
},
{
"epoch": 0.7506132461161079,
"grad_norm": 0.40291646122932434,
"learning_rate": 2e-05,
"loss": 0.7186,
"step": 9180
},
{
"epoch": 0.7514309076042518,
"grad_norm": 0.3799823224544525,
"learning_rate": 2e-05,
"loss": 0.6979,
"step": 9190
},
{
"epoch": 0.7522485690923958,
"grad_norm": 0.38580408692359924,
"learning_rate": 2e-05,
"loss": 0.7194,
"step": 9200
},
{
"epoch": 0.7530662305805397,
"grad_norm": 0.36788588762283325,
"learning_rate": 2e-05,
"loss": 0.6798,
"step": 9210
},
{
"epoch": 0.7538838920686836,
"grad_norm": 0.37825724482536316,
"learning_rate": 2e-05,
"loss": 0.73,
"step": 9220
},
{
"epoch": 0.7547015535568274,
"grad_norm": 0.40458667278289795,
"learning_rate": 2e-05,
"loss": 0.716,
"step": 9230
},
{
"epoch": 0.7555192150449714,
"grad_norm": 0.3898790776729584,
"learning_rate": 2e-05,
"loss": 0.7215,
"step": 9240
},
{
"epoch": 0.7563368765331153,
"grad_norm": 0.3553779721260071,
"learning_rate": 2e-05,
"loss": 0.6919,
"step": 9250
},
{
"epoch": 0.7571545380212592,
"grad_norm": 0.37732797861099243,
"learning_rate": 2e-05,
"loss": 0.7224,
"step": 9260
},
{
"epoch": 0.7579721995094031,
"grad_norm": 0.38592857122421265,
"learning_rate": 2e-05,
"loss": 0.7093,
"step": 9270
},
{
"epoch": 0.758789860997547,
"grad_norm": 0.36830833554267883,
"learning_rate": 2e-05,
"loss": 0.6876,
"step": 9280
},
{
"epoch": 0.759607522485691,
"grad_norm": 0.3792353570461273,
"learning_rate": 2e-05,
"loss": 0.6901,
"step": 9290
},
{
"epoch": 0.7604251839738349,
"grad_norm": 0.3909231126308441,
"learning_rate": 2e-05,
"loss": 0.7094,
"step": 9300
},
{
"epoch": 0.7612428454619787,
"grad_norm": 0.39044252038002014,
"learning_rate": 2e-05,
"loss": 0.7101,
"step": 9310
},
{
"epoch": 0.7620605069501226,
"grad_norm": 0.3698933720588684,
"learning_rate": 2e-05,
"loss": 0.7131,
"step": 9320
},
{
"epoch": 0.7628781684382666,
"grad_norm": 0.39745262265205383,
"learning_rate": 2e-05,
"loss": 0.6894,
"step": 9330
},
{
"epoch": 0.7636958299264105,
"grad_norm": 0.3926052451133728,
"learning_rate": 2e-05,
"loss": 0.6951,
"step": 9340
},
{
"epoch": 0.7645134914145544,
"grad_norm": 0.39207562804222107,
"learning_rate": 2e-05,
"loss": 0.7436,
"step": 9350
},
{
"epoch": 0.7653311529026983,
"grad_norm": 0.3587401211261749,
"learning_rate": 2e-05,
"loss": 0.7178,
"step": 9360
},
{
"epoch": 0.7661488143908421,
"grad_norm": 0.43905872106552124,
"learning_rate": 2e-05,
"loss": 0.7253,
"step": 9370
},
{
"epoch": 0.7669664758789861,
"grad_norm": 0.353103905916214,
"learning_rate": 2e-05,
"loss": 0.6959,
"step": 9380
},
{
"epoch": 0.76778413736713,
"grad_norm": 0.3872166574001312,
"learning_rate": 2e-05,
"loss": 0.695,
"step": 9390
},
{
"epoch": 0.7686017988552739,
"grad_norm": 0.43390828371047974,
"learning_rate": 2e-05,
"loss": 0.7135,
"step": 9400
},
{
"epoch": 0.7694194603434178,
"grad_norm": 0.40270760655403137,
"learning_rate": 2e-05,
"loss": 0.7046,
"step": 9410
},
{
"epoch": 0.7702371218315618,
"grad_norm": 0.402137815952301,
"learning_rate": 2e-05,
"loss": 0.7018,
"step": 9420
},
{
"epoch": 0.7710547833197057,
"grad_norm": 0.3770756423473358,
"learning_rate": 2e-05,
"loss": 0.7417,
"step": 9430
},
{
"epoch": 0.7718724448078496,
"grad_norm": 0.38187167048454285,
"learning_rate": 2e-05,
"loss": 0.712,
"step": 9440
},
{
"epoch": 0.7726901062959934,
"grad_norm": 0.35748592019081116,
"learning_rate": 2e-05,
"loss": 0.6664,
"step": 9450
},
{
"epoch": 0.7735077677841373,
"grad_norm": 0.3726420998573303,
"learning_rate": 2e-05,
"loss": 0.7177,
"step": 9460
},
{
"epoch": 0.7743254292722813,
"grad_norm": 0.42980751395225525,
"learning_rate": 2e-05,
"loss": 0.6995,
"step": 9470
},
{
"epoch": 0.7751430907604252,
"grad_norm": 0.40147361159324646,
"learning_rate": 2e-05,
"loss": 0.7397,
"step": 9480
},
{
"epoch": 0.7759607522485691,
"grad_norm": 0.40417081117630005,
"learning_rate": 2e-05,
"loss": 0.6912,
"step": 9490
},
{
"epoch": 0.776778413736713,
"grad_norm": 0.4044056534767151,
"learning_rate": 2e-05,
"loss": 0.693,
"step": 9500
},
{
"epoch": 0.777596075224857,
"grad_norm": 0.410415917634964,
"learning_rate": 2e-05,
"loss": 0.7188,
"step": 9510
},
{
"epoch": 0.7784137367130008,
"grad_norm": 0.41199836134910583,
"learning_rate": 2e-05,
"loss": 0.71,
"step": 9520
},
{
"epoch": 0.7792313982011447,
"grad_norm": 0.40095558762550354,
"learning_rate": 2e-05,
"loss": 0.7032,
"step": 9530
},
{
"epoch": 0.7800490596892886,
"grad_norm": 0.3814401924610138,
"learning_rate": 2e-05,
"loss": 0.707,
"step": 9540
},
{
"epoch": 0.7808667211774325,
"grad_norm": 0.36641961336135864,
"learning_rate": 2e-05,
"loss": 0.6941,
"step": 9550
},
{
"epoch": 0.7816843826655765,
"grad_norm": 0.39270785450935364,
"learning_rate": 2e-05,
"loss": 0.6949,
"step": 9560
},
{
"epoch": 0.7825020441537204,
"grad_norm": 0.3597510755062103,
"learning_rate": 2e-05,
"loss": 0.71,
"step": 9570
},
{
"epoch": 0.7833197056418643,
"grad_norm": 0.4060615003108978,
"learning_rate": 2e-05,
"loss": 0.7032,
"step": 9580
},
{
"epoch": 0.7841373671300081,
"grad_norm": 0.39678332209587097,
"learning_rate": 2e-05,
"loss": 0.7338,
"step": 9590
},
{
"epoch": 0.7849550286181521,
"grad_norm": 0.4351567029953003,
"learning_rate": 2e-05,
"loss": 0.6881,
"step": 9600
},
{
"epoch": 0.785772690106296,
"grad_norm": 0.38145536184310913,
"learning_rate": 2e-05,
"loss": 0.7058,
"step": 9610
},
{
"epoch": 0.7865903515944399,
"grad_norm": 0.3586829602718353,
"learning_rate": 2e-05,
"loss": 0.6878,
"step": 9620
},
{
"epoch": 0.7874080130825838,
"grad_norm": 0.38323840498924255,
"learning_rate": 2e-05,
"loss": 0.7163,
"step": 9630
},
{
"epoch": 0.7882256745707277,
"grad_norm": 0.36338692903518677,
"learning_rate": 2e-05,
"loss": 0.7255,
"step": 9640
},
{
"epoch": 0.7890433360588717,
"grad_norm": 0.37059319019317627,
"learning_rate": 2e-05,
"loss": 0.7033,
"step": 9650
},
{
"epoch": 0.7898609975470156,
"grad_norm": 0.39110058546066284,
"learning_rate": 2e-05,
"loss": 0.7236,
"step": 9660
},
{
"epoch": 0.7906786590351594,
"grad_norm": 0.38414162397384644,
"learning_rate": 2e-05,
"loss": 0.7257,
"step": 9670
},
{
"epoch": 0.7914963205233033,
"grad_norm": 0.3816734254360199,
"learning_rate": 2e-05,
"loss": 0.701,
"step": 9680
},
{
"epoch": 0.7923139820114473,
"grad_norm": 0.38287994265556335,
"learning_rate": 2e-05,
"loss": 0.6947,
"step": 9690
},
{
"epoch": 0.7931316434995912,
"grad_norm": 0.38201695680618286,
"learning_rate": 2e-05,
"loss": 0.6863,
"step": 9700
},
{
"epoch": 0.7939493049877351,
"grad_norm": 0.3740134537220001,
"learning_rate": 2e-05,
"loss": 0.7443,
"step": 9710
},
{
"epoch": 0.794766966475879,
"grad_norm": 0.44060218334198,
"learning_rate": 2e-05,
"loss": 0.7125,
"step": 9720
},
{
"epoch": 0.7955846279640229,
"grad_norm": 0.39556771516799927,
"learning_rate": 2e-05,
"loss": 0.6752,
"step": 9730
},
{
"epoch": 0.7964022894521668,
"grad_norm": 0.3801727890968323,
"learning_rate": 2e-05,
"loss": 0.6797,
"step": 9740
},
{
"epoch": 0.7972199509403107,
"grad_norm": 0.37364211678504944,
"learning_rate": 2e-05,
"loss": 0.6963,
"step": 9750
},
{
"epoch": 0.7980376124284546,
"grad_norm": 0.38798683881759644,
"learning_rate": 2e-05,
"loss": 0.7025,
"step": 9760
},
{
"epoch": 0.7988552739165985,
"grad_norm": 0.3733680546283722,
"learning_rate": 2e-05,
"loss": 0.6967,
"step": 9770
},
{
"epoch": 0.7996729354047425,
"grad_norm": 0.3645612299442291,
"learning_rate": 2e-05,
"loss": 0.6958,
"step": 9780
},
{
"epoch": 0.8004905968928864,
"grad_norm": 0.3862253427505493,
"learning_rate": 2e-05,
"loss": 0.7429,
"step": 9790
},
{
"epoch": 0.8013082583810303,
"grad_norm": 0.37967002391815186,
"learning_rate": 2e-05,
"loss": 0.7055,
"step": 9800
},
{
"epoch": 0.8021259198691741,
"grad_norm": 0.4187982380390167,
"learning_rate": 2e-05,
"loss": 0.7291,
"step": 9810
},
{
"epoch": 0.802943581357318,
"grad_norm": 0.3889312148094177,
"learning_rate": 2e-05,
"loss": 0.7159,
"step": 9820
},
{
"epoch": 0.803761242845462,
"grad_norm": 0.39745205640792847,
"learning_rate": 2e-05,
"loss": 0.7131,
"step": 9830
},
{
"epoch": 0.8045789043336059,
"grad_norm": 0.39822646975517273,
"learning_rate": 2e-05,
"loss": 0.7121,
"step": 9840
},
{
"epoch": 0.8053965658217498,
"grad_norm": 0.3866014778614044,
"learning_rate": 2e-05,
"loss": 0.7214,
"step": 9850
},
{
"epoch": 0.8062142273098937,
"grad_norm": 0.4019232988357544,
"learning_rate": 2e-05,
"loss": 0.6924,
"step": 9860
},
{
"epoch": 0.8070318887980377,
"grad_norm": 0.4176042973995209,
"learning_rate": 2e-05,
"loss": 0.7101,
"step": 9870
},
{
"epoch": 0.8078495502861816,
"grad_norm": 0.38012629747390747,
"learning_rate": 2e-05,
"loss": 0.7263,
"step": 9880
},
{
"epoch": 0.8086672117743254,
"grad_norm": 0.4058390259742737,
"learning_rate": 2e-05,
"loss": 0.6919,
"step": 9890
},
{
"epoch": 0.8094848732624693,
"grad_norm": 0.39357295632362366,
"learning_rate": 2e-05,
"loss": 0.7107,
"step": 9900
},
{
"epoch": 0.8103025347506132,
"grad_norm": 0.409150093793869,
"learning_rate": 2e-05,
"loss": 0.7033,
"step": 9910
},
{
"epoch": 0.8111201962387572,
"grad_norm": 0.3608975410461426,
"learning_rate": 2e-05,
"loss": 0.7127,
"step": 9920
},
{
"epoch": 0.8119378577269011,
"grad_norm": 0.3726581037044525,
"learning_rate": 2e-05,
"loss": 0.694,
"step": 9930
},
{
"epoch": 0.812755519215045,
"grad_norm": 0.35783761739730835,
"learning_rate": 2e-05,
"loss": 0.6902,
"step": 9940
},
{
"epoch": 0.8135731807031888,
"grad_norm": 0.3849797546863556,
"learning_rate": 2e-05,
"loss": 0.695,
"step": 9950
},
{
"epoch": 0.8143908421913328,
"grad_norm": 0.38283759355545044,
"learning_rate": 2e-05,
"loss": 0.6909,
"step": 9960
},
{
"epoch": 0.8152085036794767,
"grad_norm": 0.4008461534976959,
"learning_rate": 2e-05,
"loss": 0.701,
"step": 9970
},
{
"epoch": 0.8160261651676206,
"grad_norm": 0.390794038772583,
"learning_rate": 2e-05,
"loss": 0.6685,
"step": 9980
},
{
"epoch": 0.8168438266557645,
"grad_norm": 0.3822691738605499,
"learning_rate": 2e-05,
"loss": 0.6855,
"step": 9990
},
{
"epoch": 0.8176614881439084,
"grad_norm": 0.38428863883018494,
"learning_rate": 2e-05,
"loss": 0.6821,
"step": 10000
},
{
"epoch": 0.8184791496320524,
"grad_norm": 0.3738388419151306,
"learning_rate": 2e-05,
"loss": 0.6843,
"step": 10010
},
{
"epoch": 0.8192968111201963,
"grad_norm": 0.38718411326408386,
"learning_rate": 2e-05,
"loss": 0.6992,
"step": 10020
},
{
"epoch": 0.8201144726083401,
"grad_norm": 0.402376264333725,
"learning_rate": 2e-05,
"loss": 0.6918,
"step": 10030
},
{
"epoch": 0.820932134096484,
"grad_norm": 0.3827599585056305,
"learning_rate": 2e-05,
"loss": 0.6776,
"step": 10040
},
{
"epoch": 0.821749795584628,
"grad_norm": 0.367160826921463,
"learning_rate": 2e-05,
"loss": 0.7229,
"step": 10050
},
{
"epoch": 0.8225674570727719,
"grad_norm": 0.36382490396499634,
"learning_rate": 2e-05,
"loss": 0.7211,
"step": 10060
},
{
"epoch": 0.8233851185609158,
"grad_norm": 0.3898535370826721,
"learning_rate": 2e-05,
"loss": 0.6903,
"step": 10070
},
{
"epoch": 0.8242027800490597,
"grad_norm": 0.37079155445098877,
"learning_rate": 2e-05,
"loss": 0.7168,
"step": 10080
},
{
"epoch": 0.8250204415372036,
"grad_norm": 0.4050709307193756,
"learning_rate": 2e-05,
"loss": 0.6709,
"step": 10090
},
{
"epoch": 0.8258381030253475,
"grad_norm": 0.8351162672042847,
"learning_rate": 2e-05,
"loss": 0.71,
"step": 10100
},
{
"epoch": 0.8266557645134914,
"grad_norm": 0.39785236120224,
"learning_rate": 2e-05,
"loss": 0.7188,
"step": 10110
},
{
"epoch": 0.8274734260016353,
"grad_norm": 0.39478611946105957,
"learning_rate": 2e-05,
"loss": 0.7265,
"step": 10120
},
{
"epoch": 0.8282910874897792,
"grad_norm": 0.3996069133281708,
"learning_rate": 2e-05,
"loss": 0.7226,
"step": 10130
},
{
"epoch": 0.8291087489779232,
"grad_norm": 0.4063313901424408,
"learning_rate": 2e-05,
"loss": 0.6986,
"step": 10140
},
{
"epoch": 0.8299264104660671,
"grad_norm": 0.3814488351345062,
"learning_rate": 2e-05,
"loss": 0.6847,
"step": 10150
},
{
"epoch": 0.830744071954211,
"grad_norm": 0.37787821888923645,
"learning_rate": 2e-05,
"loss": 0.6938,
"step": 10160
},
{
"epoch": 0.8315617334423548,
"grad_norm": 0.37715527415275574,
"learning_rate": 2e-05,
"loss": 0.6964,
"step": 10170
},
{
"epoch": 0.8323793949304987,
"grad_norm": 0.3893508315086365,
"learning_rate": 1.999957358086795e-05,
"loss": 0.7163,
"step": 10180
},
{
"epoch": 0.8331970564186427,
"grad_norm": 0.37114375829696655,
"learning_rate": 1.9991993813020483e-05,
"loss": 0.7134,
"step": 10190
},
{
"epoch": 0.8340147179067866,
"grad_norm": 0.3585938513278961,
"learning_rate": 1.9974946373028628e-05,
"loss": 0.6978,
"step": 10200
},
{
"epoch": 0.8348323793949305,
"grad_norm": 0.3655489683151245,
"learning_rate": 1.994844749501764e-05,
"loss": 0.7,
"step": 10210
},
{
"epoch": 0.8356500408830744,
"grad_norm": 0.3851862847805023,
"learning_rate": 1.9912522413632988e-05,
"loss": 0.7054,
"step": 10220
},
{
"epoch": 0.8364677023712184,
"grad_norm": 0.3986681401729584,
"learning_rate": 1.9867205340009656e-05,
"loss": 0.6839,
"step": 10230
},
{
"epoch": 0.8372853638593623,
"grad_norm": 0.3750159740447998,
"learning_rate": 1.9812539429193164e-05,
"loss": 0.7189,
"step": 10240
},
{
"epoch": 0.8381030253475061,
"grad_norm": 0.3610939383506775,
"learning_rate": 1.9748576739043423e-05,
"loss": 0.6904,
"step": 10250
},
{
"epoch": 0.83892068683565,
"grad_norm": 0.3807199001312256,
"learning_rate": 1.967537818066047e-05,
"loss": 0.7132,
"step": 10260
},
{
"epoch": 0.8397383483237939,
"grad_norm": 0.40728816390037537,
"learning_rate": 1.959301346037941e-05,
"loss": 0.6742,
"step": 10270
},
{
"epoch": 0.8405560098119379,
"grad_norm": 0.4602898955345154,
"learning_rate": 1.9501561013389646e-05,
"loss": 0.7415,
"step": 10280
},
{
"epoch": 0.8413736713000818,
"grad_norm": 0.3729492723941803,
"learning_rate": 1.9401107929041795e-05,
"loss": 0.7198,
"step": 10290
},
{
"epoch": 0.8421913327882257,
"grad_norm": 0.37476179003715515,
"learning_rate": 1.9291749867913225e-05,
"loss": 0.7124,
"step": 10300
},
{
"epoch": 0.8430089942763695,
"grad_norm": 0.380525141954422,
"learning_rate": 1.91735909707114e-05,
"loss": 0.7281,
"step": 10310
},
{
"epoch": 0.8438266557645135,
"grad_norm": 0.4078218638896942,
"learning_rate": 1.9046743759101545e-05,
"loss": 0.7055,
"step": 10320
},
{
"epoch": 0.8446443172526574,
"grad_norm": 0.3895138204097748,
"learning_rate": 1.8911329028553324e-05,
"loss": 0.7132,
"step": 10330
},
{
"epoch": 0.8454619787408013,
"grad_norm": 0.3766283094882965,
"learning_rate": 1.876747573330835e-05,
"loss": 0.7003,
"step": 10340
},
{
"epoch": 0.8462796402289452,
"grad_norm": 0.3981696665287018,
"learning_rate": 1.8615320863578244e-05,
"loss": 0.6988,
"step": 10350
},
{
"epoch": 0.8470973017170891,
"grad_norm": 0.39509448409080505,
"learning_rate": 1.8455009315090036e-05,
"loss": 0.6943,
"step": 10360
},
{
"epoch": 0.8479149632052331,
"grad_norm": 0.36399543285369873,
"learning_rate": 1.8286693751103294e-05,
"loss": 0.725,
"step": 10370
},
{
"epoch": 0.848732624693377,
"grad_norm": 0.4151357114315033,
"learning_rate": 1.811053445703024e-05,
"loss": 0.6965,
"step": 10380
},
{
"epoch": 0.8495502861815208,
"grad_norm": 0.4020456373691559,
"learning_rate": 1.7926699187797375e-05,
"loss": 0.7248,
"step": 10390
},
{
"epoch": 0.8503679476696647,
"grad_norm": 0.40458643436431885,
"learning_rate": 1.7735363008093995e-05,
"loss": 0.6812,
"step": 10400
},
{
"epoch": 0.8511856091578087,
"grad_norm": 0.3973253667354584,
"learning_rate": 1.7536708125659577e-05,
"loss": 0.6942,
"step": 10410
},
{
"epoch": 0.8520032706459526,
"grad_norm": 0.3637685775756836,
"learning_rate": 1.733092371776905e-05,
"loss": 0.6936,
"step": 10420
},
{
"epoch": 0.8528209321340965,
"grad_norm": 0.37673094868659973,
"learning_rate": 1.7118205751080915e-05,
"loss": 0.6878,
"step": 10430
},
{
"epoch": 0.8536385936222404,
"grad_norm": 0.37734436988830566,
"learning_rate": 1.6898756795019935e-05,
"loss": 0.6947,
"step": 10440
},
{
"epoch": 0.8544562551103843,
"grad_norm": 0.38562697172164917,
"learning_rate": 1.667278582887211e-05,
"loss": 0.6976,
"step": 10450
},
{
"epoch": 0.8552739165985282,
"grad_norm": 0.38020020723342896,
"learning_rate": 1.6440508042775545e-05,
"loss": 0.7004,
"step": 10460
},
{
"epoch": 0.8560915780866721,
"grad_norm": 0.3792610168457031,
"learning_rate": 1.6202144632796796e-05,
"loss": 0.6735,
"step": 10470
},
{
"epoch": 0.856909239574816,
"grad_norm": 0.38012149930000305,
"learning_rate": 1.595792259028783e-05,
"loss": 0.7122,
"step": 10480
},
{
"epoch": 0.8577269010629599,
"grad_norm": 0.3814740478992462,
"learning_rate": 1.570807448572419e-05,
"loss": 0.6642,
"step": 10490
},
{
"epoch": 0.8585445625511039,
"grad_norm": 0.36654943227767944,
"learning_rate": 1.5452838247230197e-05,
"loss": 0.6833,
"step": 10500
},
{
"epoch": 0.8593622240392478,
"grad_norm": 0.37783777713775635,
"learning_rate": 1.5192456934002106e-05,
"loss": 0.7283,
"step": 10510
},
{
"epoch": 0.8601798855273917,
"grad_norm": 0.43871939182281494,
"learning_rate": 1.492717850484502e-05,
"loss": 0.6762,
"step": 10520
},
{
"epoch": 0.8609975470155355,
"grad_norm": 0.39402177929878235,
"learning_rate": 1.4657255582043904e-05,
"loss": 0.691,
"step": 10530
},
{
"epoch": 0.8618152085036794,
"grad_norm": 0.37783652544021606,
"learning_rate": 1.4382945210793663e-05,
"loss": 0.7066,
"step": 10540
},
{
"epoch": 0.8626328699918234,
"grad_norm": 0.3981021046638489,
"learning_rate": 1.4104508614417267e-05,
"loss": 0.6829,
"step": 10550
},
{
"epoch": 0.8634505314799673,
"grad_norm": 0.378268837928772,
"learning_rate": 1.382221094560512e-05,
"loss": 0.7275,
"step": 10560
},
{
"epoch": 0.8642681929681112,
"grad_norm": 0.3925245404243469,
"learning_rate": 1.3536321033912506e-05,
"loss": 0.6944,
"step": 10570
},
{
"epoch": 0.8650858544562551,
"grad_norm": 0.38472065329551697,
"learning_rate": 1.3247111129755609e-05,
"loss": 0.6846,
"step": 10580
},
{
"epoch": 0.8659035159443991,
"grad_norm": 0.38513079285621643,
"learning_rate": 1.2954856645149868e-05,
"loss": 0.6844,
"step": 10590
},
{
"epoch": 0.866721177432543,
"grad_norm": 0.4379119575023651,
"learning_rate": 1.2659835891437553e-05,
"loss": 0.7115,
"step": 10600
},
{
"epoch": 0.8675388389206868,
"grad_norm": 0.3622902035713196,
"learning_rate": 1.236232981425436e-05,
"loss": 0.6735,
"step": 10610
},
{
"epoch": 0.8683565004088307,
"grad_norm": 0.3786431849002838,
"learning_rate": 1.2062621725987402e-05,
"loss": 0.7138,
"step": 10620
},
{
"epoch": 0.8691741618969746,
"grad_norm": 0.39203810691833496,
"learning_rate": 1.1760997035979318e-05,
"loss": 0.6634,
"step": 10630
},
{
"epoch": 0.8699918233851186,
"grad_norm": 0.37007516622543335,
"learning_rate": 1.1457742978735514e-05,
"loss": 0.7008,
"step": 10640
},
{
"epoch": 0.8708094848732625,
"grad_norm": 0.4103783071041107,
"learning_rate": 1.1153148340393298e-05,
"loss": 0.6888,
"step": 10650
},
{
"epoch": 0.8716271463614064,
"grad_norm": 0.37784343957901,
"learning_rate": 1.0847503183713424e-05,
"loss": 0.6873,
"step": 10660
},
{
"epoch": 0.8724448078495503,
"grad_norm": 0.37784522771835327,
"learning_rate": 1.0541098571855896e-05,
"loss": 0.6891,
"step": 10670
},
{
"epoch": 0.8732624693376942,
"grad_norm": 0.41790395975112915,
"learning_rate": 1.0234226291203123e-05,
"loss": 0.6888,
"step": 10680
},
{
"epoch": 0.8740801308258381,
"grad_norm": 0.38745781779289246,
"learning_rate": 9.927178573494333e-06,
"loss": 0.69,
"step": 10690
},
{
"epoch": 0.874897792313982,
"grad_norm": 0.398221492767334,
"learning_rate": 9.620247817535902e-06,
"loss": 0.7129,
"step": 10700
},
{
"epoch": 0.8757154538021259,
"grad_norm": 0.394243448972702,
"learning_rate": 9.313726310752581e-06,
"loss": 0.7139,
"step": 10710
},
{
"epoch": 0.8765331152902698,
"grad_norm": 0.34937936067581177,
"learning_rate": 9.00790595084477e-06,
"loss": 0.6657,
"step": 10720
},
{
"epoch": 0.8773507767784138,
"grad_norm": 0.3711390495300293,
"learning_rate": 8.703077967816944e-06,
"loss": 0.6728,
"step": 10730
},
{
"epoch": 0.8781684382665577,
"grad_norm": 0.3892441689968109,
"learning_rate": 8.39953264664192e-06,
"loss": 0.716,
"step": 10740
},
{
"epoch": 0.8789860997547015,
"grad_norm": 0.3458738327026367,
"learning_rate": 8.097559050825058e-06,
"loss": 0.6848,
"step": 10750
},
{
"epoch": 0.8798037612428454,
"grad_norm": 0.36444976925849915,
"learning_rate": 7.79744474713166e-06,
"loss": 0.6923,
"step": 10760
},
{
"epoch": 0.8806214227309894,
"grad_norm": 0.3681751787662506,
"learning_rate": 7.499475531739696e-06,
"loss": 0.6572,
"step": 10770
},
{
"epoch": 0.8814390842191333,
"grad_norm": 0.3733675181865692,
"learning_rate": 7.203935158078662e-06,
"loss": 0.6765,
"step": 10780
},
{
"epoch": 0.8822567457072772,
"grad_norm": 0.41421762108802795,
"learning_rate": 6.911105066613709e-06,
"loss": 0.6718,
"step": 10790
},
{
"epoch": 0.8830744071954211,
"grad_norm": 0.3620806932449341,
"learning_rate": 6.621264116832394e-06,
"loss": 0.6816,
"step": 10800
},
{
"epoch": 0.883892068683565,
"grad_norm": 0.37829500436782837,
"learning_rate": 6.334688321689274e-06,
"loss": 0.6804,
"step": 10810
},
{
"epoch": 0.884709730171709,
"grad_norm": 0.32967308163642883,
"learning_rate": 6.051650584761237e-06,
"loss": 0.6502,
"step": 10820
},
{
"epoch": 0.8855273916598528,
"grad_norm": 0.35177892446517944,
"learning_rate": 5.772420440363841e-06,
"loss": 0.6457,
"step": 10830
},
{
"epoch": 0.8863450531479967,
"grad_norm": 0.3710830807685852,
"learning_rate": 5.497263796876207e-06,
"loss": 0.6656,
"step": 10840
},
{
"epoch": 0.8871627146361406,
"grad_norm": 0.4050544798374176,
"learning_rate": 5.226442683518813e-06,
"loss": 0.6785,
"step": 10850
},
{
"epoch": 0.8879803761242846,
"grad_norm": 0.4040084779262543,
"learning_rate": 4.960215000825432e-06,
"loss": 0.7288,
"step": 10860
},
{
"epoch": 0.8887980376124285,
"grad_norm": 0.4372066557407379,
"learning_rate": 4.698834275046711e-06,
"loss": 0.6837,
"step": 10870
},
{
"epoch": 0.8896156991005724,
"grad_norm": 0.3768942058086395,
"learning_rate": 4.442549416719423e-06,
"loss": 0.6782,
"step": 10880
},
{
"epoch": 0.8904333605887162,
"grad_norm": 0.3646766245365143,
"learning_rate": 4.1916044836311565e-06,
"loss": 0.6755,
"step": 10890
},
{
"epoch": 0.8912510220768601,
"grad_norm": 0.3826245963573456,
"learning_rate": 3.94623844840626e-06,
"loss": 0.6898,
"step": 10900
},
{
"epoch": 0.8920686835650041,
"grad_norm": 0.39837127923965454,
"learning_rate": 3.7066849709343467e-06,
"loss": 0.7064,
"step": 10910
},
{
"epoch": 0.892886345053148,
"grad_norm": 0.362407386302948,
"learning_rate": 3.473172175858059e-06,
"loss": 0.6442,
"step": 10920
},
{
"epoch": 0.8937040065412919,
"grad_norm": 0.3900721073150635,
"learning_rate": 3.2459224353320046e-06,
"loss": 0.7004,
"step": 10930
},
{
"epoch": 0.8945216680294358,
"grad_norm": 0.36550894379615784,
"learning_rate": 3.025152157259735e-06,
"loss": 0.6889,
"step": 10940
},
{
"epoch": 0.8953393295175798,
"grad_norm": 0.37971267104148865,
"learning_rate": 2.811071579210425e-06,
"loss": 0.6817,
"step": 10950
},
{
"epoch": 0.8961569910057237,
"grad_norm": 0.37165409326553345,
"learning_rate": 2.603884568211513e-06,
"loss": 0.7084,
"step": 10960
},
{
"epoch": 0.8969746524938675,
"grad_norm": 0.37599119544029236,
"learning_rate": 2.4037884266079434e-06,
"loss": 0.6631,
"step": 10970
},
{
"epoch": 0.8977923139820114,
"grad_norm": 0.3620017170906067,
"learning_rate": 2.2109737041728955e-06,
"loss": 0.6825,
"step": 10980
},
{
"epoch": 0.8986099754701553,
"grad_norm": 0.38728195428848267,
"learning_rate": 2.0256240166489454e-06,
"loss": 0.6773,
"step": 10990
},
{
"epoch": 0.8994276369582993,
"grad_norm": 0.37711861729621887,
"learning_rate": 1.847915870892413e-06,
"loss": 0.6647,
"step": 11000
},
{
"epoch": 0.9002452984464432,
"grad_norm": 0.36904317140579224,
"learning_rate": 1.6780184967874715e-06,
"loss": 0.6763,
"step": 11010
},
{
"epoch": 0.9010629599345871,
"grad_norm": 0.34700000286102295,
"learning_rate": 1.5160936860900294e-06,
"loss": 0.6896,
"step": 11020
},
{
"epoch": 0.901880621422731,
"grad_norm": 0.3383204936981201,
"learning_rate": 1.3622956383548762e-06,
"loss": 0.6581,
"step": 11030
},
{
"epoch": 0.902698282910875,
"grad_norm": 0.3764679431915283,
"learning_rate": 1.216770814092834e-06,
"loss": 0.6922,
"step": 11040
},
{
"epoch": 0.9035159443990188,
"grad_norm": 0.3605528175830841,
"learning_rate": 1.0796577952977013e-06,
"loss": 0.6685,
"step": 11050
},
{
"epoch": 0.9043336058871627,
"grad_norm": 0.3455122411251068,
"learning_rate": 9.510871534758705e-07,
"loss": 0.6644,
"step": 11060
},
{
"epoch": 0.9051512673753066,
"grad_norm": 0.35580572485923767,
"learning_rate": 8.311813253042334e-07,
"loss": 0.6522,
"step": 11070
},
{
"epoch": 0.9059689288634505,
"grad_norm": 0.3604564666748047,
"learning_rate": 7.200544960348308e-07,
"loss": 0.6648,
"step": 11080
},
{
"epoch": 0.9067865903515945,
"grad_norm": 0.3734298348426819,
"learning_rate": 6.178124907572545e-07,
"loss": 0.69,
"step": 11090
},
{
"epoch": 0.9076042518397384,
"grad_norm": 0.35744336247444153,
"learning_rate": 5.245526736223528e-07,
"loss": 0.6706,
"step": 11100
},
{
"epoch": 0.9084219133278822,
"grad_norm": 0.373531699180603,
"learning_rate": 4.403638551232205e-07,
"loss": 0.6839,
"step": 11110
},
{
"epoch": 0.9092395748160261,
"grad_norm": 0.368693083524704,
"learning_rate": 3.6532620752175616e-07,
"loss": 0.6764,
"step": 11120
},
{
"epoch": 0.9100572363041701,
"grad_norm": 0.37689825892448425,
"learning_rate": 2.9951118850133247e-07,
"loss": 0.6707,
"step": 11130
},
{
"epoch": 0.910874897792314,
"grad_norm": 0.35353395342826843,
"learning_rate": 2.429814731182774e-07,
"loss": 0.6643,
"step": 11140
},
{
"epoch": 0.9116925592804579,
"grad_norm": 0.36653703451156616,
"learning_rate": 1.957908941169813e-07,
"loss": 0.642,
"step": 11150
},
{
"epoch": 0.9125102207686018,
"grad_norm": 0.35942766070365906,
"learning_rate": 1.5798439066545254e-07,
"loss": 0.6801,
"step": 11160
},
{
"epoch": 0.9133278822567457,
"grad_norm": 0.3710387647151947,
"learning_rate": 1.2959796556014471e-07,
"loss": 0.6623,
"step": 11170
},
{
"epoch": 0.9141455437448897,
"grad_norm": 0.36466386914253235,
"learning_rate": 1.1065865094081884e-07,
"loss": 0.6644,
"step": 11180
},
{
"epoch": 0.9149632052330335,
"grad_norm": 0.3551817238330841,
"learning_rate": 1.011844825480734e-07,
"loss": 0.6699,
"step": 11190
},
{
"epoch": 0.9152902698282911,
"step": 11194,
"total_flos": 4.547514943786923e+19,
"train_loss": 0.003823117611769035,
"train_runtime": 609.8376,
"train_samples_per_second": 220.268,
"train_steps_per_second": 18.356
}
],
"logging_steps": 10,
"max_steps": 11194,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 159,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.547514943786923e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}