stay4s-1b-base / trainer_state.json
miesdevries's picture
Upload folder using huggingface_hub
637a2ae verified
Raw History Blame Contribute Delete
134 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8398685605702707,
"eval_steps": 500,
"global_step": 20000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0010498357007128385,
"grad_norm": 8.5625,
"learning_rate": 1.4999999999999999e-05,
"loss": 11.6589,
"step": 25
},
{
"epoch": 0.002099671401425677,
"grad_norm": 6.125,
"learning_rate": 2.9999999999999997e-05,
"loss": 8.8573,
"step": 50
},
{
"epoch": 0.0031495071021385153,
"grad_norm": 11.125,
"learning_rate": 4.4999999999999996e-05,
"loss": 6.794,
"step": 75
},
{
"epoch": 0.004199342802851354,
"grad_norm": 6.625,
"learning_rate": 5.9999999999999995e-05,
"loss": 4.9229,
"step": 100
},
{
"epoch": 0.005249178503564192,
"grad_norm": 19.375,
"learning_rate": 7.5e-05,
"loss": 3.6685,
"step": 125
},
{
"epoch": 0.0062990142042770305,
"grad_norm": 2.75,
"learning_rate": 8.999999999999999e-05,
"loss": 3.3577,
"step": 150
},
{
"epoch": 0.007348849904989869,
"grad_norm": 2.875,
"learning_rate": 0.00010499999999999999,
"loss": 2.9583,
"step": 175
},
{
"epoch": 0.008398685605702708,
"grad_norm": 1.3984375,
"learning_rate": 0.00011999999999999999,
"loss": 3.0188,
"step": 200
},
{
"epoch": 0.009448521306415545,
"grad_norm": 1.828125,
"learning_rate": 0.000135,
"loss": 2.8262,
"step": 225
},
{
"epoch": 0.010498357007128385,
"grad_norm": 2.140625,
"learning_rate": 0.00015,
"loss": 2.6889,
"step": 250
},
{
"epoch": 0.011548192707841224,
"grad_norm": 0.94921875,
"learning_rate": 0.000165,
"loss": 2.7896,
"step": 275
},
{
"epoch": 0.012598028408554061,
"grad_norm": 1.125,
"learning_rate": 0.00017999999999999998,
"loss": 2.3196,
"step": 300
},
{
"epoch": 0.0136478641092669,
"grad_norm": 0.7890625,
"learning_rate": 0.000195,
"loss": 2.3639,
"step": 325
},
{
"epoch": 0.014697699809979738,
"grad_norm": 1.28125,
"learning_rate": 0.00020999999999999998,
"loss": 2.402,
"step": 350
},
{
"epoch": 0.015747535510692575,
"grad_norm": 0.8671875,
"learning_rate": 0.000225,
"loss": 2.7208,
"step": 375
},
{
"epoch": 0.016797371211405416,
"grad_norm": 1.515625,
"learning_rate": 0.00023999999999999998,
"loss": 1.8824,
"step": 400
},
{
"epoch": 0.017847206912118253,
"grad_norm": 1.0859375,
"learning_rate": 0.00025499999999999996,
"loss": 2.4477,
"step": 425
},
{
"epoch": 0.01889704261283109,
"grad_norm": 0.9921875,
"learning_rate": 0.00027,
"loss": 2.2581,
"step": 450
},
{
"epoch": 0.01994687831354393,
"grad_norm": 1.921875,
"learning_rate": 0.000285,
"loss": 2.3087,
"step": 475
},
{
"epoch": 0.02099671401425677,
"grad_norm": 0.8984375,
"learning_rate": 0.0003,
"loss": 2.4734,
"step": 500
},
{
"epoch": 0.022046549714969606,
"grad_norm": 0.7734375,
"learning_rate": 0.00029999878333443556,
"loss": 2.2547,
"step": 525
},
{
"epoch": 0.023096385415682447,
"grad_norm": 0.65234375,
"learning_rate": 0.00029999513335747914,
"loss": 2.2643,
"step": 550
},
{
"epoch": 0.024146221116395285,
"grad_norm": 0.52734375,
"learning_rate": 0.0002999890501283416,
"loss": 2.2354,
"step": 575
},
{
"epoch": 0.025196056817108122,
"grad_norm": 0.66796875,
"learning_rate": 0.0002999805337457062,
"loss": 2.3773,
"step": 600
},
{
"epoch": 0.02624589251782096,
"grad_norm": 0.4765625,
"learning_rate": 0.0002999695843477275,
"loss": 2.1924,
"step": 625
},
{
"epoch": 0.0272957282185338,
"grad_norm": 0.54296875,
"learning_rate": 0.0002999562021120289,
"loss": 2.2009,
"step": 650
},
{
"epoch": 0.028345563919246638,
"grad_norm": 0.71875,
"learning_rate": 0.0002999403872556999,
"loss": 2.3172,
"step": 675
},
{
"epoch": 0.029395399619959475,
"grad_norm": 0.8125,
"learning_rate": 0.0002999221400352922,
"loss": 2.2046,
"step": 700
},
{
"epoch": 0.030445235320672316,
"grad_norm": 0.515625,
"learning_rate": 0.0002999014607468162,
"loss": 1.9667,
"step": 725
},
{
"epoch": 0.03149507102138515,
"grad_norm": 0.78125,
"learning_rate": 0.0002998783497257354,
"loss": 2.061,
"step": 750
},
{
"epoch": 0.03254490672209799,
"grad_norm": 0.99609375,
"learning_rate": 0.0002998528073469618,
"loss": 2.1627,
"step": 775
},
{
"epoch": 0.03359474242281083,
"grad_norm": 0.640625,
"learning_rate": 0.000299824834024849,
"loss": 1.9714,
"step": 800
},
{
"epoch": 0.034644578123523666,
"grad_norm": 1.125,
"learning_rate": 0.000299794430213186,
"loss": 2.076,
"step": 825
},
{
"epoch": 0.03569441382423651,
"grad_norm": 0.7421875,
"learning_rate": 0.00029976159640519,
"loss": 2.1965,
"step": 850
},
{
"epoch": 0.03674424952494935,
"grad_norm": 0.81640625,
"learning_rate": 0.0002997263331334976,
"loss": 1.9331,
"step": 875
},
{
"epoch": 0.03779408522566218,
"grad_norm": 0.74609375,
"learning_rate": 0.000299688640970157,
"loss": 2.156,
"step": 900
},
{
"epoch": 0.03884392092637502,
"grad_norm": 0.7578125,
"learning_rate": 0.0002996485205266183,
"loss": 2.0264,
"step": 925
},
{
"epoch": 0.03989375662708786,
"grad_norm": 0.6171875,
"learning_rate": 0.00029960597245372365,
"loss": 1.8579,
"step": 950
},
{
"epoch": 0.0409435923278007,
"grad_norm": 0.79296875,
"learning_rate": 0.0002995609974416967,
"loss": 2.0668,
"step": 975
},
{
"epoch": 0.04199342802851354,
"grad_norm": 0.7265625,
"learning_rate": 0.0002995135962201315,
"loss": 1.9539,
"step": 1000
},
{
"epoch": 0.04304326372922638,
"grad_norm": 0.80859375,
"learning_rate": 0.0002994637695579804,
"loss": 2.1986,
"step": 1025
},
{
"epoch": 0.04409309942993921,
"grad_norm": 0.58984375,
"learning_rate": 0.0002994115182635419,
"loss": 1.7772,
"step": 1050
},
{
"epoch": 0.045142935130652054,
"grad_norm": 0.7265625,
"learning_rate": 0.00029935684318444736,
"loss": 2.0095,
"step": 1075
},
{
"epoch": 0.046192770831364895,
"grad_norm": 0.63671875,
"learning_rate": 0.00029929974520764727,
"loss": 1.7553,
"step": 1100
},
{
"epoch": 0.04724260653207773,
"grad_norm": 0.74609375,
"learning_rate": 0.0002992402252593968,
"loss": 2.1066,
"step": 1125
},
{
"epoch": 0.04829244223279057,
"grad_norm": 0.5390625,
"learning_rate": 0.00029917828430524096,
"loss": 1.8091,
"step": 1150
},
{
"epoch": 0.0493422779335034,
"grad_norm": 0.43359375,
"learning_rate": 0.0002991139233499988,
"loss": 1.5769,
"step": 1175
},
{
"epoch": 0.050392113634216244,
"grad_norm": 0.404296875,
"learning_rate": 0.00029904714343774696,
"loss": 1.6034,
"step": 1200
},
{
"epoch": 0.051441949334929085,
"grad_norm": 0.74609375,
"learning_rate": 0.0002989779456518031,
"loss": 1.9894,
"step": 1225
},
{
"epoch": 0.05249178503564192,
"grad_norm": 0.75390625,
"learning_rate": 0.00029890633111470807,
"loss": 2.1371,
"step": 1250
},
{
"epoch": 0.05354162073635476,
"grad_norm": 0.67578125,
"learning_rate": 0.0002988323009882078,
"loss": 1.9529,
"step": 1275
},
{
"epoch": 0.0545914564370676,
"grad_norm": 0.69140625,
"learning_rate": 0.0002987558564732343,
"loss": 1.882,
"step": 1300
},
{
"epoch": 0.055641292137780435,
"grad_norm": 0.5703125,
"learning_rate": 0.00029867699880988636,
"loss": 1.949,
"step": 1325
},
{
"epoch": 0.056691127838493276,
"grad_norm": 0.7421875,
"learning_rate": 0.00029859572927740937,
"loss": 1.8379,
"step": 1350
},
{
"epoch": 0.05774096353920612,
"grad_norm": 0.5546875,
"learning_rate": 0.0002985120491941746,
"loss": 1.6345,
"step": 1375
},
{
"epoch": 0.05879079923991895,
"grad_norm": 0.640625,
"learning_rate": 0.0002984259599176576,
"loss": 1.8074,
"step": 1400
},
{
"epoch": 0.05984063494063179,
"grad_norm": 0.6484375,
"learning_rate": 0.0002983374628444167,
"loss": 1.6384,
"step": 1425
},
{
"epoch": 0.06089047064134463,
"grad_norm": 0.75390625,
"learning_rate": 0.00029824655941006954,
"loss": 1.613,
"step": 1450
},
{
"epoch": 0.061940306342057466,
"grad_norm": 0.341796875,
"learning_rate": 0.00029815325108927063,
"loss": 1.7496,
"step": 1475
},
{
"epoch": 0.0629901420427703,
"grad_norm": 0.67578125,
"learning_rate": 0.00029805753939568687,
"loss": 1.5794,
"step": 1500
},
{
"epoch": 0.06403997774348315,
"grad_norm": 0.2890625,
"learning_rate": 0.0002979594258819733,
"loss": 1.746,
"step": 1525
},
{
"epoch": 0.06508981344419598,
"grad_norm": 0.5234375,
"learning_rate": 0.0002978589121397476,
"loss": 1.7576,
"step": 1550
},
{
"epoch": 0.06613964914490882,
"grad_norm": 0.69140625,
"learning_rate": 0.0002977559997995646,
"loss": 1.8052,
"step": 1575
},
{
"epoch": 0.06718948484562166,
"grad_norm": 0.60546875,
"learning_rate": 0.00029765069053088963,
"loss": 1.7931,
"step": 1600
},
{
"epoch": 0.0682393205463345,
"grad_norm": 0.86328125,
"learning_rate": 0.00029754298604207154,
"loss": 1.843,
"step": 1625
},
{
"epoch": 0.06928915624704733,
"grad_norm": 0.43359375,
"learning_rate": 0.00029743288808031483,
"loss": 1.8037,
"step": 1650
},
{
"epoch": 0.07033899194776018,
"grad_norm": 0.6875,
"learning_rate": 0.00029732039843165145,
"loss": 1.8496,
"step": 1675
},
{
"epoch": 0.07138882764847301,
"grad_norm": 0.478515625,
"learning_rate": 0.000297205518920912,
"loss": 1.693,
"step": 1700
},
{
"epoch": 0.07243866334918585,
"grad_norm": 0.765625,
"learning_rate": 0.0002970882514116956,
"loss": 1.7673,
"step": 1725
},
{
"epoch": 0.0734884990498987,
"grad_norm": 0.58203125,
"learning_rate": 0.0002969685978063401,
"loss": 1.787,
"step": 1750
},
{
"epoch": 0.07453833475061153,
"grad_norm": 0.64453125,
"learning_rate": 0.00029684656004589125,
"loss": 1.7582,
"step": 1775
},
{
"epoch": 0.07558817045132436,
"grad_norm": 0.5625,
"learning_rate": 0.0002967221401100708,
"loss": 1.6549,
"step": 1800
},
{
"epoch": 0.07663800615203721,
"grad_norm": 0.68359375,
"learning_rate": 0.00029659534001724486,
"loss": 1.6057,
"step": 1825
},
{
"epoch": 0.07768784185275004,
"grad_norm": 0.73046875,
"learning_rate": 0.0002964661618243908,
"loss": 1.6964,
"step": 1850
},
{
"epoch": 0.07873767755346288,
"grad_norm": 0.474609375,
"learning_rate": 0.00029633460762706405,
"loss": 1.7172,
"step": 1875
},
{
"epoch": 0.07978751325417573,
"grad_norm": 0.6796875,
"learning_rate": 0.0002962006795593642,
"loss": 1.758,
"step": 1900
},
{
"epoch": 0.08083734895488856,
"grad_norm": 0.5390625,
"learning_rate": 0.0002960643797939,
"loss": 1.7098,
"step": 1925
},
{
"epoch": 0.0818871846556014,
"grad_norm": 0.5859375,
"learning_rate": 0.0002959257105417546,
"loss": 1.5477,
"step": 1950
},
{
"epoch": 0.08293702035631424,
"grad_norm": 0.75,
"learning_rate": 0.00029578467405244954,
"loss": 1.6237,
"step": 1975
},
{
"epoch": 0.08398685605702708,
"grad_norm": 0.66796875,
"learning_rate": 0.00029564127261390776,
"loss": 1.6889,
"step": 2000
},
{
"epoch": 0.08503669175773991,
"grad_norm": 0.70703125,
"learning_rate": 0.00029549550855241736,
"loss": 1.6056,
"step": 2025
},
{
"epoch": 0.08608652745845276,
"grad_norm": 0.71875,
"learning_rate": 0.00029534738423259307,
"loss": 1.8107,
"step": 2050
},
{
"epoch": 0.08713636315916559,
"grad_norm": 0.984375,
"learning_rate": 0.0002951969020573384,
"loss": 1.7317,
"step": 2075
},
{
"epoch": 0.08818619885987843,
"grad_norm": 0.77734375,
"learning_rate": 0.0002950440644678064,
"loss": 1.8898,
"step": 2100
},
{
"epoch": 0.08923603456059127,
"grad_norm": 0.1845703125,
"learning_rate": 0.0002948888739433602,
"loss": 1.5502,
"step": 2125
},
{
"epoch": 0.09028587026130411,
"grad_norm": 0.58203125,
"learning_rate": 0.0002947313330015327,
"loss": 1.5836,
"step": 2150
},
{
"epoch": 0.09133570596201694,
"grad_norm": 0.42578125,
"learning_rate": 0.0002945714441979857,
"loss": 1.7206,
"step": 2175
},
{
"epoch": 0.09238554166272979,
"grad_norm": 0.4453125,
"learning_rate": 0.00029440921012646856,
"loss": 1.4853,
"step": 2200
},
{
"epoch": 0.09343537736344262,
"grad_norm": 0.70703125,
"learning_rate": 0.0002942446334187761,
"loss": 1.6863,
"step": 2225
},
{
"epoch": 0.09448521306415546,
"grad_norm": 0.515625,
"learning_rate": 0.0002940777167447058,
"loss": 1.3892,
"step": 2250
},
{
"epoch": 0.0955350487648683,
"grad_norm": 0.66796875,
"learning_rate": 0.00029390846281201463,
"loss": 1.7194,
"step": 2275
},
{
"epoch": 0.09658488446558114,
"grad_norm": 0.53125,
"learning_rate": 0.0002937368743663749,
"loss": 1.1978,
"step": 2300
},
{
"epoch": 0.09763472016629397,
"grad_norm": 0.7421875,
"learning_rate": 0.0002935629541913301,
"loss": 1.5214,
"step": 2325
},
{
"epoch": 0.0986845558670068,
"grad_norm": 0.7265625,
"learning_rate": 0.0002933867051082493,
"loss": 1.4986,
"step": 2350
},
{
"epoch": 0.09973439156771965,
"grad_norm": 0.7734375,
"learning_rate": 0.0002932081299762818,
"loss": 1.739,
"step": 2375
},
{
"epoch": 0.10078422726843249,
"grad_norm": 0.81640625,
"learning_rate": 0.0002930272316923104,
"loss": 1.6137,
"step": 2400
},
{
"epoch": 0.10183406296914532,
"grad_norm": 0.71875,
"learning_rate": 0.00029284401319090464,
"loss": 1.4206,
"step": 2425
},
{
"epoch": 0.10288389866985817,
"grad_norm": 0.7109375,
"learning_rate": 0.00029265847744427303,
"loss": 1.6706,
"step": 2450
},
{
"epoch": 0.103933734370571,
"grad_norm": 0.82421875,
"learning_rate": 0.0002924706274622149,
"loss": 1.7199,
"step": 2475
},
{
"epoch": 0.10498357007128384,
"grad_norm": 0.796875,
"learning_rate": 0.00029228046629207175,
"loss": 1.6273,
"step": 2500
},
{
"epoch": 0.10603340577199669,
"grad_norm": 0.65234375,
"learning_rate": 0.0002920879970186776,
"loss": 1.7317,
"step": 2525
},
{
"epoch": 0.10708324147270952,
"grad_norm": 0.61328125,
"learning_rate": 0.00029189322276430873,
"loss": 1.5458,
"step": 2550
},
{
"epoch": 0.10813307717342235,
"grad_norm": 0.6015625,
"learning_rate": 0.0002916961466886337,
"loss": 1.556,
"step": 2575
},
{
"epoch": 0.1091829128741352,
"grad_norm": 0.58984375,
"learning_rate": 0.0002914967719886614,
"loss": 1.5998,
"step": 2600
},
{
"epoch": 0.11023274857484804,
"grad_norm": 0.6953125,
"learning_rate": 0.00029129510189868974,
"loss": 1.5598,
"step": 2625
},
{
"epoch": 0.11128258427556087,
"grad_norm": 0.5703125,
"learning_rate": 0.0002910911396902526,
"loss": 1.4931,
"step": 2650
},
{
"epoch": 0.11233241997627372,
"grad_norm": 0.59375,
"learning_rate": 0.0002908848886720673,
"loss": 1.5169,
"step": 2675
},
{
"epoch": 0.11338225567698655,
"grad_norm": 0.65625,
"learning_rate": 0.00029067635218998065,
"loss": 1.5185,
"step": 2700
},
{
"epoch": 0.11443209137769939,
"grad_norm": 0.68359375,
"learning_rate": 0.0002904655336269148,
"loss": 1.7869,
"step": 2725
},
{
"epoch": 0.11548192707841223,
"grad_norm": 0.61328125,
"learning_rate": 0.00029025243640281223,
"loss": 1.6423,
"step": 2750
},
{
"epoch": 0.11653176277912507,
"grad_norm": 0.546875,
"learning_rate": 0.00029003706397458023,
"loss": 1.5646,
"step": 2775
},
{
"epoch": 0.1175815984798379,
"grad_norm": 0.71484375,
"learning_rate": 0.00028981941983603516,
"loss": 1.6321,
"step": 2800
},
{
"epoch": 0.11863143418055075,
"grad_norm": 0.66015625,
"learning_rate": 0.00028959950751784527,
"loss": 1.7333,
"step": 2825
},
{
"epoch": 0.11968126988126358,
"grad_norm": 0.6953125,
"learning_rate": 0.00028937733058747396,
"loss": 1.6051,
"step": 2850
},
{
"epoch": 0.12073110558197642,
"grad_norm": 0.69921875,
"learning_rate": 0.0002891528926491214,
"loss": 1.614,
"step": 2875
},
{
"epoch": 0.12178094128268926,
"grad_norm": 0.306640625,
"learning_rate": 0.0002889261973436664,
"loss": 1.4731,
"step": 2900
},
{
"epoch": 0.1228307769834021,
"grad_norm": 0.6015625,
"learning_rate": 0.0002886972483486074,
"loss": 1.2093,
"step": 2925
},
{
"epoch": 0.12388061268411493,
"grad_norm": 0.6796875,
"learning_rate": 0.0002884660493780023,
"loss": 1.5286,
"step": 2950
},
{
"epoch": 0.12493044838482778,
"grad_norm": 0.6875,
"learning_rate": 0.00028823260418240894,
"loss": 1.5924,
"step": 2975
},
{
"epoch": 0.1259802840855406,
"grad_norm": 0.65234375,
"learning_rate": 0.0002879969165488236,
"loss": 1.4546,
"step": 3000
},
{
"epoch": 0.12703011978625345,
"grad_norm": 0.58203125,
"learning_rate": 0.0002877589903006201,
"loss": 1.4099,
"step": 3025
},
{
"epoch": 0.1280799554869663,
"grad_norm": 0.5,
"learning_rate": 0.00028751882929748735,
"loss": 1.5862,
"step": 3050
},
{
"epoch": 0.12912979118767912,
"grad_norm": 0.474609375,
"learning_rate": 0.000287276437435367,
"loss": 1.497,
"step": 3075
},
{
"epoch": 0.13017962688839196,
"grad_norm": 0.69921875,
"learning_rate": 0.0002870318186463901,
"loss": 1.6673,
"step": 3100
},
{
"epoch": 0.1312294625891048,
"grad_norm": 0.482421875,
"learning_rate": 0.0002867849768988135,
"loss": 1.5323,
"step": 3125
},
{
"epoch": 0.13227929828981763,
"grad_norm": 0.5390625,
"learning_rate": 0.00028653591619695514,
"loss": 1.3955,
"step": 3150
},
{
"epoch": 0.13332913399053048,
"grad_norm": 0.50390625,
"learning_rate": 0.0002862846405811295,
"loss": 1.4374,
"step": 3175
},
{
"epoch": 0.13437896969124333,
"grad_norm": 0.81640625,
"learning_rate": 0.0002860311541275818,
"loss": 1.4831,
"step": 3200
},
{
"epoch": 0.13542880539195615,
"grad_norm": 0.765625,
"learning_rate": 0.0002857754609484218,
"loss": 1.6497,
"step": 3225
},
{
"epoch": 0.136478641092669,
"grad_norm": 0.55078125,
"learning_rate": 0.0002855175651915573,
"loss": 1.4062,
"step": 3250
},
{
"epoch": 0.13752847679338184,
"grad_norm": 0.5390625,
"learning_rate": 0.0002852574710406269,
"loss": 1.5309,
"step": 3275
},
{
"epoch": 0.13857831249409466,
"grad_norm": 0.375,
"learning_rate": 0.0002849951827149318,
"loss": 1.7072,
"step": 3300
},
{
"epoch": 0.1396281481948075,
"grad_norm": 0.640625,
"learning_rate": 0.0002847307044693677,
"loss": 1.513,
"step": 3325
},
{
"epoch": 0.14067798389552036,
"grad_norm": 0.5859375,
"learning_rate": 0.0002844640405943555,
"loss": 1.5214,
"step": 3350
},
{
"epoch": 0.14172781959623318,
"grad_norm": 0.390625,
"learning_rate": 0.0002841951954157721,
"loss": 1.446,
"step": 3375
},
{
"epoch": 0.14277765529694603,
"grad_norm": 0.671875,
"learning_rate": 0.0002839241732948796,
"loss": 1.496,
"step": 3400
},
{
"epoch": 0.14382749099765887,
"grad_norm": 0.71484375,
"learning_rate": 0.00028365097862825513,
"loss": 1.4588,
"step": 3425
},
{
"epoch": 0.1448773266983717,
"grad_norm": 0.72265625,
"learning_rate": 0.0002833756158477194,
"loss": 1.4188,
"step": 3450
},
{
"epoch": 0.14592716239908454,
"grad_norm": 0.546875,
"learning_rate": 0.00028309808942026434,
"loss": 1.4055,
"step": 3475
},
{
"epoch": 0.1469769980997974,
"grad_norm": 0.68359375,
"learning_rate": 0.0002828184038479814,
"loss": 1.3029,
"step": 3500
},
{
"epoch": 0.1480268338005102,
"grad_norm": 0.609375,
"learning_rate": 0.00028253656366798805,
"loss": 1.5277,
"step": 3525
},
{
"epoch": 0.14907666950122306,
"grad_norm": 0.6875,
"learning_rate": 0.0002822525734523541,
"loss": 1.5112,
"step": 3550
},
{
"epoch": 0.1501265052019359,
"grad_norm": 0.74609375,
"learning_rate": 0.00028196643780802766,
"loss": 1.4601,
"step": 3575
},
{
"epoch": 0.15117634090264873,
"grad_norm": 0.39453125,
"learning_rate": 0.0002816781613767607,
"loss": 1.4925,
"step": 3600
},
{
"epoch": 0.15222617660336157,
"grad_norm": 0.375,
"learning_rate": 0.0002813877488350332,
"loss": 1.5456,
"step": 3625
},
{
"epoch": 0.15327601230407442,
"grad_norm": 0.67578125,
"learning_rate": 0.0002810952048939777,
"loss": 1.7381,
"step": 3650
},
{
"epoch": 0.15432584800478724,
"grad_norm": 0.66796875,
"learning_rate": 0.0002808005342993027,
"loss": 1.5479,
"step": 3675
},
{
"epoch": 0.1553756837055001,
"grad_norm": 0.53125,
"learning_rate": 0.00028050374183121585,
"loss": 1.3238,
"step": 3700
},
{
"epoch": 0.15642551940621294,
"grad_norm": 0.6875,
"learning_rate": 0.0002802048323043461,
"loss": 1.6131,
"step": 3725
},
{
"epoch": 0.15747535510692576,
"grad_norm": 0.6328125,
"learning_rate": 0.0002799038105676658,
"loss": 1.5243,
"step": 3750
},
{
"epoch": 0.1585251908076386,
"grad_norm": 0.73828125,
"learning_rate": 0.000279600681504412,
"loss": 1.324,
"step": 3775
},
{
"epoch": 0.15957502650835145,
"grad_norm": 0.58984375,
"learning_rate": 0.00027929545003200736,
"loss": 1.4907,
"step": 3800
},
{
"epoch": 0.16062486220906427,
"grad_norm": 0.61328125,
"learning_rate": 0.00027898812110198015,
"loss": 1.3824,
"step": 3825
},
{
"epoch": 0.16167469790977712,
"grad_norm": 0.251953125,
"learning_rate": 0.00027867869969988395,
"loss": 1.1719,
"step": 3850
},
{
"epoch": 0.16272453361048997,
"grad_norm": 0.69921875,
"learning_rate": 0.0002783671908452171,
"loss": 1.693,
"step": 3875
},
{
"epoch": 0.1637743693112028,
"grad_norm": 0.65625,
"learning_rate": 0.0002780535995913408,
"loss": 1.5577,
"step": 3900
},
{
"epoch": 0.16482420501191564,
"grad_norm": 0.72265625,
"learning_rate": 0.0002777379310253975,
"loss": 1.5117,
"step": 3925
},
{
"epoch": 0.16587404071262848,
"grad_norm": 0.59765625,
"learning_rate": 0.00027742019026822816,
"loss": 1.3249,
"step": 3950
},
{
"epoch": 0.1669238764133413,
"grad_norm": 0.6171875,
"learning_rate": 0.00027710038247428923,
"loss": 1.44,
"step": 3975
},
{
"epoch": 0.16797371211405415,
"grad_norm": 0.66796875,
"learning_rate": 0.0002767785128315692,
"loss": 1.3328,
"step": 4000
},
{
"epoch": 0.169023547814767,
"grad_norm": 0.62109375,
"learning_rate": 0.00027645458656150415,
"loss": 1.2369,
"step": 4025
},
{
"epoch": 0.17007338351547982,
"grad_norm": 0.6953125,
"learning_rate": 0.0002761286089188933,
"loss": 1.3771,
"step": 4050
},
{
"epoch": 0.17112321921619267,
"grad_norm": 0.68359375,
"learning_rate": 0.0002758005851918136,
"loss": 1.3358,
"step": 4075
},
{
"epoch": 0.17217305491690552,
"grad_norm": 0.5703125,
"learning_rate": 0.00027547052070153396,
"loss": 1.4317,
"step": 4100
},
{
"epoch": 0.17322289061761834,
"grad_norm": 0.6875,
"learning_rate": 0.00027513842080242916,
"loss": 1.4766,
"step": 4125
},
{
"epoch": 0.17427272631833118,
"grad_norm": 0.69140625,
"learning_rate": 0.0002748042908818926,
"loss": 1.4279,
"step": 4150
},
{
"epoch": 0.17532256201904403,
"grad_norm": 0.640625,
"learning_rate": 0.00027446813636024925,
"loss": 1.4856,
"step": 4175
},
{
"epoch": 0.17637239771975685,
"grad_norm": 0.85546875,
"learning_rate": 0.0002741299626906675,
"loss": 1.5237,
"step": 4200
},
{
"epoch": 0.1774222334204697,
"grad_norm": 0.5546875,
"learning_rate": 0.00027378977535907065,
"loss": 1.4809,
"step": 4225
},
{
"epoch": 0.17847206912118255,
"grad_norm": 0.6171875,
"learning_rate": 0.00027344757988404845,
"loss": 1.5505,
"step": 4250
},
{
"epoch": 0.17952190482189537,
"grad_norm": 0.6171875,
"learning_rate": 0.00027310338181676675,
"loss": 1.4133,
"step": 4275
},
{
"epoch": 0.18057174052260822,
"grad_norm": 0.4453125,
"learning_rate": 0.000272757186740878,
"loss": 1.4246,
"step": 4300
},
{
"epoch": 0.18162157622332106,
"grad_norm": 0.67578125,
"learning_rate": 0.00027240900027243066,
"loss": 1.4885,
"step": 4325
},
{
"epoch": 0.18267141192403388,
"grad_norm": 0.25,
"learning_rate": 0.0002720588280597779,
"loss": 1.4127,
"step": 4350
},
{
"epoch": 0.18372124762474673,
"grad_norm": 0.5625,
"learning_rate": 0.00027170667578348595,
"loss": 1.4847,
"step": 4375
},
{
"epoch": 0.18477108332545958,
"grad_norm": 0.7734375,
"learning_rate": 0.0002713525491562421,
"loss": 1.4588,
"step": 4400
},
{
"epoch": 0.1858209190261724,
"grad_norm": 0.60546875,
"learning_rate": 0.00027099645392276203,
"loss": 1.5837,
"step": 4425
},
{
"epoch": 0.18687075472688525,
"grad_norm": 0.70703125,
"learning_rate": 0.0002706383958596965,
"loss": 1.2415,
"step": 4450
},
{
"epoch": 0.1879205904275981,
"grad_norm": 0.345703125,
"learning_rate": 0.0002702783807755377,
"loss": 1.4206,
"step": 4475
},
{
"epoch": 0.18897042612831091,
"grad_norm": 0.8515625,
"learning_rate": 0.00026991641451052517,
"loss": 1.5103,
"step": 4500
},
{
"epoch": 0.19002026182902376,
"grad_norm": 0.55859375,
"learning_rate": 0.0002695525029365506,
"loss": 1.2414,
"step": 4525
},
{
"epoch": 0.1910700975297366,
"grad_norm": 1.1328125,
"learning_rate": 0.00026918665195706313,
"loss": 1.5864,
"step": 4550
},
{
"epoch": 0.19211993323044943,
"grad_norm": 0.703125,
"learning_rate": 0.00026881886750697325,
"loss": 1.4257,
"step": 4575
},
{
"epoch": 0.19316976893116228,
"grad_norm": 0.62109375,
"learning_rate": 0.0002684491555525567,
"loss": 1.39,
"step": 4600
},
{
"epoch": 0.1942196046318751,
"grad_norm": 0.70703125,
"learning_rate": 0.0002680775220913575,
"loss": 1.2638,
"step": 4625
},
{
"epoch": 0.19526944033258795,
"grad_norm": 0.68359375,
"learning_rate": 0.00026770397315209075,
"loss": 1.4858,
"step": 4650
},
{
"epoch": 0.1963192760333008,
"grad_norm": 0.46484375,
"learning_rate": 0.0002673285147945449,
"loss": 1.4532,
"step": 4675
},
{
"epoch": 0.1973691117340136,
"grad_norm": 0.4765625,
"learning_rate": 0.00026695115310948334,
"loss": 1.3984,
"step": 4700
},
{
"epoch": 0.19841894743472646,
"grad_norm": 0.60546875,
"learning_rate": 0.0002665718942185456,
"loss": 1.2777,
"step": 4725
},
{
"epoch": 0.1994687831354393,
"grad_norm": 0.796875,
"learning_rate": 0.00026619074427414814,
"loss": 1.4835,
"step": 4750
},
{
"epoch": 0.20051861883615213,
"grad_norm": 0.51171875,
"learning_rate": 0.0002658077094593845,
"loss": 1.2569,
"step": 4775
},
{
"epoch": 0.20156845453686498,
"grad_norm": 0.52734375,
"learning_rate": 0.0002654227959879249,
"loss": 1.4715,
"step": 4800
},
{
"epoch": 0.20261829023757782,
"grad_norm": 0.59765625,
"learning_rate": 0.0002650360101039155,
"loss": 1.2005,
"step": 4825
},
{
"epoch": 0.20366812593829065,
"grad_norm": 0.6328125,
"learning_rate": 0.0002646473580818772,
"loss": 1.2418,
"step": 4850
},
{
"epoch": 0.2047179616390035,
"grad_norm": 0.63671875,
"learning_rate": 0.00026425684622660386,
"loss": 1.3937,
"step": 4875
},
{
"epoch": 0.20576779733971634,
"grad_norm": 0.6328125,
"learning_rate": 0.0002638644808730597,
"loss": 1.4241,
"step": 4900
},
{
"epoch": 0.20681763304042916,
"grad_norm": 0.53515625,
"learning_rate": 0.000263470268386277,
"loss": 1.3458,
"step": 4925
},
{
"epoch": 0.207867468741142,
"grad_norm": 0.625,
"learning_rate": 0.0002630742151612525,
"loss": 1.4296,
"step": 4950
},
{
"epoch": 0.20891730444185486,
"grad_norm": 0.404296875,
"learning_rate": 0.0002626763276228438,
"loss": 1.3122,
"step": 4975
},
{
"epoch": 0.20996714014256768,
"grad_norm": 0.7265625,
"learning_rate": 0.00026227661222566516,
"loss": 1.4778,
"step": 5000
},
{
"epoch": 0.21101697584328052,
"grad_norm": 0.66796875,
"learning_rate": 0.00026187507545398264,
"loss": 1.5412,
"step": 5025
},
{
"epoch": 0.21206681154399337,
"grad_norm": 0.71875,
"learning_rate": 0.00026147172382160914,
"loss": 1.2923,
"step": 5050
},
{
"epoch": 0.2131166472447062,
"grad_norm": 0.515625,
"learning_rate": 0.0002610665638717985,
"loss": 1.4284,
"step": 5075
},
{
"epoch": 0.21416648294541904,
"grad_norm": 0.6875,
"learning_rate": 0.00026065960217713953,
"loss": 1.4662,
"step": 5100
},
{
"epoch": 0.2152163186461319,
"grad_norm": 0.45703125,
"learning_rate": 0.0002602508453394493,
"loss": 1.2495,
"step": 5125
},
{
"epoch": 0.2162661543468447,
"grad_norm": 0.5703125,
"learning_rate": 0.00025984029998966604,
"loss": 1.4569,
"step": 5150
},
{
"epoch": 0.21731599004755756,
"grad_norm": 0.64453125,
"learning_rate": 0.0002594279727877416,
"loss": 1.5058,
"step": 5175
},
{
"epoch": 0.2183658257482704,
"grad_norm": 0.59375,
"learning_rate": 0.00025901387042253347,
"loss": 1.4171,
"step": 5200
},
{
"epoch": 0.21941566144898322,
"grad_norm": 0.58203125,
"learning_rate": 0.00025859799961169616,
"loss": 1.3593,
"step": 5225
},
{
"epoch": 0.22046549714969607,
"grad_norm": 0.62890625,
"learning_rate": 0.0002581803671015722,
"loss": 1.2547,
"step": 5250
},
{
"epoch": 0.22151533285040892,
"grad_norm": 0.32421875,
"learning_rate": 0.0002577609796670828,
"loss": 1.1967,
"step": 5275
},
{
"epoch": 0.22256516855112174,
"grad_norm": 0.3515625,
"learning_rate": 0.00025733984411161815,
"loss": 1.2644,
"step": 5300
},
{
"epoch": 0.2236150042518346,
"grad_norm": 0.64453125,
"learning_rate": 0.00025691696726692643,
"loss": 1.4234,
"step": 5325
},
{
"epoch": 0.22466483995254743,
"grad_norm": 0.54296875,
"learning_rate": 0.0002564923559930037,
"loss": 1.3127,
"step": 5350
},
{
"epoch": 0.22571467565326025,
"grad_norm": 0.7265625,
"learning_rate": 0.00025606601717798207,
"loss": 1.2659,
"step": 5375
},
{
"epoch": 0.2267645113539731,
"grad_norm": 0.53125,
"learning_rate": 0.0002556379577380184,
"loss": 1.293,
"step": 5400
},
{
"epoch": 0.22781434705468595,
"grad_norm": 0.6171875,
"learning_rate": 0.0002552081846171816,
"loss": 1.1604,
"step": 5425
},
{
"epoch": 0.22886418275539877,
"grad_norm": 0.390625,
"learning_rate": 0.00025477670478734054,
"loss": 1.1308,
"step": 5450
},
{
"epoch": 0.22991401845611162,
"grad_norm": 0.7109375,
"learning_rate": 0.0002543435252480505,
"loss": 1.2044,
"step": 5475
},
{
"epoch": 0.23096385415682447,
"grad_norm": 0.453125,
"learning_rate": 0.0002539086530264399,
"loss": 1.3691,
"step": 5500
},
{
"epoch": 0.2320136898575373,
"grad_norm": 0.734375,
"learning_rate": 0.0002534720951770961,
"loss": 1.3728,
"step": 5525
},
{
"epoch": 0.23306352555825013,
"grad_norm": 0.318359375,
"learning_rate": 0.00025303385878195126,
"loss": 1.4762,
"step": 5550
},
{
"epoch": 0.23411336125896298,
"grad_norm": 0.67578125,
"learning_rate": 0.00025259395095016704,
"loss": 1.377,
"step": 5575
},
{
"epoch": 0.2351631969596758,
"grad_norm": 0.609375,
"learning_rate": 0.00025215237881801955,
"loss": 1.1406,
"step": 5600
},
{
"epoch": 0.23621303266038865,
"grad_norm": 0.64453125,
"learning_rate": 0.0002517091495487836,
"loss": 1.3844,
"step": 5625
},
{
"epoch": 0.2372628683611015,
"grad_norm": 0.5546875,
"learning_rate": 0.0002512642703326164,
"loss": 1.2723,
"step": 5650
},
{
"epoch": 0.23831270406181432,
"grad_norm": 0.65625,
"learning_rate": 0.00025081774838644084,
"loss": 1.522,
"step": 5675
},
{
"epoch": 0.23936253976252717,
"grad_norm": 0.6875,
"learning_rate": 0.0002503695909538287,
"loss": 1.3442,
"step": 5700
},
{
"epoch": 0.24041237546324,
"grad_norm": 0.443359375,
"learning_rate": 0.0002499198053048828,
"loss": 1.3833,
"step": 5725
},
{
"epoch": 0.24146221116395283,
"grad_norm": 0.58203125,
"learning_rate": 0.00024946839873611924,
"loss": 1.3046,
"step": 5750
},
{
"epoch": 0.24251204686466568,
"grad_norm": 0.6796875,
"learning_rate": 0.00024901537857034917,
"loss": 1.3707,
"step": 5775
},
{
"epoch": 0.24356188256537853,
"grad_norm": 0.71875,
"learning_rate": 0.00024856075215655963,
"loss": 1.2845,
"step": 5800
},
{
"epoch": 0.24461171826609135,
"grad_norm": 0.546875,
"learning_rate": 0.00024810452686979467,
"loss": 1.2731,
"step": 5825
},
{
"epoch": 0.2456615539668042,
"grad_norm": 0.48828125,
"learning_rate": 0.0002476467101110356,
"loss": 1.3633,
"step": 5850
},
{
"epoch": 0.24671138966751704,
"grad_norm": 0.59375,
"learning_rate": 0.0002471873093070808,
"loss": 1.3021,
"step": 5875
},
{
"epoch": 0.24776122536822986,
"grad_norm": 0.515625,
"learning_rate": 0.0002467263319104256,
"loss": 1.5275,
"step": 5900
},
{
"epoch": 0.2488110610689427,
"grad_norm": 0.5390625,
"learning_rate": 0.00024626378539914083,
"loss": 1.1582,
"step": 5925
},
{
"epoch": 0.24986089676965556,
"grad_norm": 0.5703125,
"learning_rate": 0.00024579967727675215,
"loss": 1.22,
"step": 5950
},
{
"epoch": 0.2509107324703684,
"grad_norm": 0.7109375,
"learning_rate": 0.0002453340150721178,
"loss": 1.3823,
"step": 5975
},
{
"epoch": 0.2519605681710812,
"grad_norm": 0.640625,
"learning_rate": 0.0002448668063393066,
"loss": 1.4315,
"step": 6000
},
{
"epoch": 0.2530104038717941,
"grad_norm": 0.5703125,
"learning_rate": 0.0002443980586574756,
"loss": 1.4359,
"step": 6025
},
{
"epoch": 0.2540602395725069,
"grad_norm": 1.15625,
"learning_rate": 0.00024392777963074704,
"loss": 1.3333,
"step": 6050
},
{
"epoch": 0.2551100752732197,
"grad_norm": 0.63671875,
"learning_rate": 0.00024345597688808486,
"loss": 1.281,
"step": 6075
},
{
"epoch": 0.2561599109739326,
"grad_norm": 0.60546875,
"learning_rate": 0.00024298265808317103,
"loss": 1.3109,
"step": 6100
},
{
"epoch": 0.2572097466746454,
"grad_norm": 0.640625,
"learning_rate": 0.00024250783089428148,
"loss": 1.405,
"step": 6125
},
{
"epoch": 0.25825958237535823,
"grad_norm": 0.78125,
"learning_rate": 0.0002420315030241613,
"loss": 1.2182,
"step": 6150
},
{
"epoch": 0.2593094180760711,
"grad_norm": 0.51171875,
"learning_rate": 0.0002415536821999002,
"loss": 1.3447,
"step": 6175
},
{
"epoch": 0.2603592537767839,
"grad_norm": 0.443359375,
"learning_rate": 0.00024107437617280667,
"loss": 1.2388,
"step": 6200
},
{
"epoch": 0.26140908947749675,
"grad_norm": 0.5546875,
"learning_rate": 0.00024059359271828254,
"loss": 1.4317,
"step": 6225
},
{
"epoch": 0.2624589251782096,
"grad_norm": 0.66796875,
"learning_rate": 0.0002401113396356968,
"loss": 1.2541,
"step": 6250
},
{
"epoch": 0.26350876087892244,
"grad_norm": 0.62890625,
"learning_rate": 0.00023962762474825908,
"loss": 1.2702,
"step": 6275
},
{
"epoch": 0.26455859657963526,
"grad_norm": 0.76171875,
"learning_rate": 0.0002391424559028926,
"loss": 1.2235,
"step": 6300
},
{
"epoch": 0.26560843228034814,
"grad_norm": 0.546875,
"learning_rate": 0.00023865584097010708,
"loss": 1.3224,
"step": 6325
},
{
"epoch": 0.26665826798106096,
"grad_norm": 0.640625,
"learning_rate": 0.00023816778784387094,
"loss": 1.0074,
"step": 6350
},
{
"epoch": 0.2677081036817738,
"grad_norm": 0.59375,
"learning_rate": 0.00023767830444148332,
"loss": 1.2888,
"step": 6375
},
{
"epoch": 0.26875793938248665,
"grad_norm": 0.84765625,
"learning_rate": 0.00023718739870344548,
"loss": 1.2824,
"step": 6400
},
{
"epoch": 0.2698077750831995,
"grad_norm": 0.71875,
"learning_rate": 0.00023669507859333224,
"loss": 1.399,
"step": 6425
},
{
"epoch": 0.2708576107839123,
"grad_norm": 0.69140625,
"learning_rate": 0.00023620135209766253,
"loss": 1.2657,
"step": 6450
},
{
"epoch": 0.27190744648462517,
"grad_norm": 0.6484375,
"learning_rate": 0.0002357062272257701,
"loss": 1.3002,
"step": 6475
},
{
"epoch": 0.272957282185338,
"grad_norm": 0.65234375,
"learning_rate": 0.00023520971200967334,
"loss": 1.2187,
"step": 6500
},
{
"epoch": 0.2740071178860508,
"grad_norm": 0.46484375,
"learning_rate": 0.00023471181450394513,
"loss": 1.2968,
"step": 6525
},
{
"epoch": 0.2750569535867637,
"grad_norm": 0.61328125,
"learning_rate": 0.00023421254278558213,
"loss": 1.0866,
"step": 6550
},
{
"epoch": 0.2761067892874765,
"grad_norm": 0.69140625,
"learning_rate": 0.00023371190495387372,
"loss": 1.3556,
"step": 6575
},
{
"epoch": 0.2771566249881893,
"grad_norm": 0.65234375,
"learning_rate": 0.0002332099091302708,
"loss": 1.3677,
"step": 6600
},
{
"epoch": 0.2782064606889022,
"grad_norm": 1.0625,
"learning_rate": 0.00023270656345825375,
"loss": 1.2858,
"step": 6625
},
{
"epoch": 0.279256296389615,
"grad_norm": 0.5625,
"learning_rate": 0.0002322018761032005,
"loss": 1.3507,
"step": 6650
},
{
"epoch": 0.28030613209032784,
"grad_norm": 0.60546875,
"learning_rate": 0.00023169585525225405,
"loss": 1.2169,
"step": 6675
},
{
"epoch": 0.2813559677910407,
"grad_norm": 0.62890625,
"learning_rate": 0.0002311885091141897,
"loss": 1.3232,
"step": 6700
},
{
"epoch": 0.28240580349175354,
"grad_norm": 0.55859375,
"learning_rate": 0.00023067984591928177,
"loss": 1.3938,
"step": 6725
},
{
"epoch": 0.28345563919246636,
"grad_norm": 0.42578125,
"learning_rate": 0.00023016987391917014,
"loss": 1.1407,
"step": 6750
},
{
"epoch": 0.28450547489317923,
"grad_norm": 0.63671875,
"learning_rate": 0.00022965860138672646,
"loss": 1.2632,
"step": 6775
},
{
"epoch": 0.28555531059389205,
"grad_norm": 0.55078125,
"learning_rate": 0.00022914603661591988,
"loss": 1.2075,
"step": 6800
},
{
"epoch": 0.2866051462946049,
"grad_norm": 0.5859375,
"learning_rate": 0.00022863218792168244,
"loss": 1.3991,
"step": 6825
},
{
"epoch": 0.28765498199531775,
"grad_norm": 0.60546875,
"learning_rate": 0.00022811706363977434,
"loss": 1.4537,
"step": 6850
},
{
"epoch": 0.28870481769603057,
"grad_norm": 0.625,
"learning_rate": 0.0002276006721266485,
"loss": 1.3088,
"step": 6875
},
{
"epoch": 0.2897546533967434,
"grad_norm": 0.50390625,
"learning_rate": 0.0002270830217593153,
"loss": 1.3302,
"step": 6900
},
{
"epoch": 0.29080448909745626,
"grad_norm": 0.64453125,
"learning_rate": 0.00022656412093520636,
"loss": 1.2292,
"step": 6925
},
{
"epoch": 0.2918543247981691,
"grad_norm": 0.76171875,
"learning_rate": 0.00022604397807203855,
"loss": 1.3177,
"step": 6950
},
{
"epoch": 0.2929041604988819,
"grad_norm": 0.66015625,
"learning_rate": 0.00022552260160767736,
"loss": 1.3096,
"step": 6975
},
{
"epoch": 0.2939539961995948,
"grad_norm": 0.486328125,
"learning_rate": 0.000225,
"loss": 1.2031,
"step": 7000
},
{
"epoch": 0.2950038319003076,
"grad_norm": 0.64453125,
"learning_rate": 0.0002244761817267582,
"loss": 1.2405,
"step": 7025
},
{
"epoch": 0.2960536676010204,
"grad_norm": 0.640625,
"learning_rate": 0.0002239511552854406,
"loss": 1.1314,
"step": 7050
},
{
"epoch": 0.2971035033017333,
"grad_norm": 0.62890625,
"learning_rate": 0.0002234249291931352,
"loss": 1.2731,
"step": 7075
},
{
"epoch": 0.2981533390024461,
"grad_norm": 0.498046875,
"learning_rate": 0.00022289751198639087,
"loss": 1.2711,
"step": 7100
},
{
"epoch": 0.29920317470315894,
"grad_norm": 0.61328125,
"learning_rate": 0.000222368912221079,
"loss": 1.515,
"step": 7125
},
{
"epoch": 0.3002530104038718,
"grad_norm": 0.2890625,
"learning_rate": 0.0002218391384722546,
"loss": 1.1751,
"step": 7150
},
{
"epoch": 0.30130284610458463,
"grad_norm": 0.6015625,
"learning_rate": 0.0002213081993340174,
"loss": 1.253,
"step": 7175
},
{
"epoch": 0.30235268180529745,
"grad_norm": 0.6171875,
"learning_rate": 0.0002207761034193723,
"loss": 1.4119,
"step": 7200
},
{
"epoch": 0.3034025175060103,
"grad_norm": 0.64453125,
"learning_rate": 0.00022024285936008968,
"loss": 1.2355,
"step": 7225
},
{
"epoch": 0.30445235320672315,
"grad_norm": 0.482421875,
"learning_rate": 0.00021970847580656525,
"loss": 1.1818,
"step": 7250
},
{
"epoch": 0.30550218890743597,
"grad_norm": 0.7265625,
"learning_rate": 0.00021917296142768002,
"loss": 1.1232,
"step": 7275
},
{
"epoch": 0.30655202460814884,
"grad_norm": 0.6484375,
"learning_rate": 0.0002186363249106593,
"loss": 1.2814,
"step": 7300
},
{
"epoch": 0.30760186030886166,
"grad_norm": 0.58984375,
"learning_rate": 0.00021809857496093199,
"loss": 1.4748,
"step": 7325
},
{
"epoch": 0.3086516960095745,
"grad_norm": 0.57421875,
"learning_rate": 0.0002175597203019895,
"loss": 1.1877,
"step": 7350
},
{
"epoch": 0.30970153171028736,
"grad_norm": 0.6796875,
"learning_rate": 0.00021701976967524386,
"loss": 1.2221,
"step": 7375
},
{
"epoch": 0.3107513674110002,
"grad_norm": 0.6796875,
"learning_rate": 0.0002164787318398862,
"loss": 1.4531,
"step": 7400
},
{
"epoch": 0.311801203111713,
"grad_norm": 0.67578125,
"learning_rate": 0.00021593661557274454,
"loss": 1.2176,
"step": 7425
},
{
"epoch": 0.3128510388124259,
"grad_norm": 0.625,
"learning_rate": 0.00021539342966814148,
"loss": 1.2691,
"step": 7450
},
{
"epoch": 0.3139008745131387,
"grad_norm": 0.67578125,
"learning_rate": 0.00021484918293775157,
"loss": 1.3566,
"step": 7475
},
{
"epoch": 0.3149507102138515,
"grad_norm": 0.373046875,
"learning_rate": 0.0002143038842104581,
"loss": 1.1841,
"step": 7500
},
{
"epoch": 0.3160005459145644,
"grad_norm": 0.50390625,
"learning_rate": 0.00021375754233221028,
"loss": 1.343,
"step": 7525
},
{
"epoch": 0.3170503816152772,
"grad_norm": 0.37109375,
"learning_rate": 0.00021321016616587937,
"loss": 1.2309,
"step": 7550
},
{
"epoch": 0.31810021731599003,
"grad_norm": 0.62109375,
"learning_rate": 0.00021266176459111514,
"loss": 1.3821,
"step": 7575
},
{
"epoch": 0.3191500530167029,
"grad_norm": 0.51171875,
"learning_rate": 0.00021211234650420177,
"loss": 1.1745,
"step": 7600
},
{
"epoch": 0.3201998887174157,
"grad_norm": 0.64453125,
"learning_rate": 0.0002115619208179135,
"loss": 1.2192,
"step": 7625
},
{
"epoch": 0.32124972441812855,
"grad_norm": 0.640625,
"learning_rate": 0.00021101049646137003,
"loss": 1.1591,
"step": 7650
},
{
"epoch": 0.3222995601188414,
"grad_norm": 0.5859375,
"learning_rate": 0.00021045808237989168,
"loss": 1.1559,
"step": 7675
},
{
"epoch": 0.32334939581955424,
"grad_norm": 0.474609375,
"learning_rate": 0.0002099046875348543,
"loss": 1.1429,
"step": 7700
},
{
"epoch": 0.32439923152026706,
"grad_norm": 0.53515625,
"learning_rate": 0.00020935032090354397,
"loss": 0.9901,
"step": 7725
},
{
"epoch": 0.32544906722097994,
"grad_norm": 0.640625,
"learning_rate": 0.00020879499147901124,
"loss": 1.1671,
"step": 7750
},
{
"epoch": 0.32649890292169276,
"grad_norm": 0.1708984375,
"learning_rate": 0.00020823870826992523,
"loss": 1.0083,
"step": 7775
},
{
"epoch": 0.3275487386224056,
"grad_norm": 0.54296875,
"learning_rate": 0.00020768148030042764,
"loss": 1.0704,
"step": 7800
},
{
"epoch": 0.32859857432311845,
"grad_norm": 0.5546875,
"learning_rate": 0.00020712331660998614,
"loss": 1.0862,
"step": 7825
},
{
"epoch": 0.3296484100238313,
"grad_norm": 0.48828125,
"learning_rate": 0.00020656422625324805,
"loss": 1.1361,
"step": 7850
},
{
"epoch": 0.3306982457245441,
"grad_norm": 0.380859375,
"learning_rate": 0.00020600421829989314,
"loss": 1.176,
"step": 7875
},
{
"epoch": 0.33174808142525697,
"grad_norm": 0.69921875,
"learning_rate": 0.00020544330183448664,
"loss": 1.1419,
"step": 7900
},
{
"epoch": 0.3327979171259698,
"grad_norm": 0.625,
"learning_rate": 0.00020488148595633183,
"loss": 1.1829,
"step": 7925
},
{
"epoch": 0.3338477528266826,
"grad_norm": 0.466796875,
"learning_rate": 0.00020431877977932248,
"loss": 1.2213,
"step": 7950
},
{
"epoch": 0.3348975885273955,
"grad_norm": 0.5859375,
"learning_rate": 0.000203755192431795,
"loss": 1.2982,
"step": 7975
},
{
"epoch": 0.3359474242281083,
"grad_norm": 0.49609375,
"learning_rate": 0.0002031907330563803,
"loss": 0.9316,
"step": 8000
},
{
"epoch": 0.3369972599288211,
"grad_norm": 0.7109375,
"learning_rate": 0.00020262541080985554,
"loss": 1.2339,
"step": 8025
},
{
"epoch": 0.338047095629534,
"grad_norm": 0.5078125,
"learning_rate": 0.0002020592348629954,
"loss": 1.1747,
"step": 8050
},
{
"epoch": 0.3390969313302468,
"grad_norm": 0.578125,
"learning_rate": 0.00020149221440042374,
"loss": 1.1988,
"step": 8075
},
{
"epoch": 0.34014676703095964,
"grad_norm": 0.625,
"learning_rate": 0.00020092435862046406,
"loss": 1.4721,
"step": 8100
},
{
"epoch": 0.3411966027316725,
"grad_norm": 0.6171875,
"learning_rate": 0.0002003556767349907,
"loss": 1.2069,
"step": 8125
},
{
"epoch": 0.34224643843238534,
"grad_norm": 0.53125,
"learning_rate": 0.00019978617796927926,
"loss": 1.2075,
"step": 8150
},
{
"epoch": 0.34329627413309816,
"grad_norm": 0.494140625,
"learning_rate": 0.00019921587156185687,
"loss": 1.1632,
"step": 8175
},
{
"epoch": 0.34434610983381103,
"grad_norm": 0.75390625,
"learning_rate": 0.0001986447667643524,
"loss": 1.2552,
"step": 8200
},
{
"epoch": 0.34539594553452385,
"grad_norm": 0.7890625,
"learning_rate": 0.00019807287284134644,
"loss": 1.2759,
"step": 8225
},
{
"epoch": 0.34644578123523667,
"grad_norm": 0.66796875,
"learning_rate": 0.0001975001990702209,
"loss": 1.2171,
"step": 8250
},
{
"epoch": 0.34749561693594955,
"grad_norm": 0.53125,
"learning_rate": 0.00019692675474100845,
"loss": 1.2413,
"step": 8275
},
{
"epoch": 0.34854545263666237,
"grad_norm": 0.5234375,
"learning_rate": 0.0001963525491562421,
"loss": 1.1921,
"step": 8300
},
{
"epoch": 0.3495952883373752,
"grad_norm": 0.50390625,
"learning_rate": 0.00019577759163080398,
"loss": 1.133,
"step": 8325
},
{
"epoch": 0.35064512403808806,
"grad_norm": 0.56640625,
"learning_rate": 0.00019520189149177438,
"loss": 1.1137,
"step": 8350
},
{
"epoch": 0.3516949597388009,
"grad_norm": 0.515625,
"learning_rate": 0.00019462545807828044,
"loss": 1.0703,
"step": 8375
},
{
"epoch": 0.3527447954395137,
"grad_norm": 0.453125,
"learning_rate": 0.0001940483007413446,
"loss": 1.0421,
"step": 8400
},
{
"epoch": 0.3537946311402266,
"grad_norm": 0.671875,
"learning_rate": 0.000193470428843733,
"loss": 1.2304,
"step": 8425
},
{
"epoch": 0.3548444668409394,
"grad_norm": 0.56640625,
"learning_rate": 0.00019289185175980336,
"loss": 1.0303,
"step": 8450
},
{
"epoch": 0.3558943025416522,
"grad_norm": 0.6328125,
"learning_rate": 0.0001923125788753534,
"loss": 1.3677,
"step": 8475
},
{
"epoch": 0.3569441382423651,
"grad_norm": 0.353515625,
"learning_rate": 0.0001917326195874679,
"loss": 1.0805,
"step": 8500
},
{
"epoch": 0.3579939739430779,
"grad_norm": 0.64453125,
"learning_rate": 0.00019115198330436686,
"loss": 1.1183,
"step": 8525
},
{
"epoch": 0.35904380964379073,
"grad_norm": 0.56640625,
"learning_rate": 0.00019057067944525257,
"loss": 1.0904,
"step": 8550
},
{
"epoch": 0.3600936453445036,
"grad_norm": 0.609375,
"learning_rate": 0.00018998871744015677,
"loss": 1.0845,
"step": 8575
},
{
"epoch": 0.36114348104521643,
"grad_norm": 0.765625,
"learning_rate": 0.00018940610672978802,
"loss": 1.0644,
"step": 8600
},
{
"epoch": 0.36219331674592925,
"grad_norm": 0.57421875,
"learning_rate": 0.0001888228567653781,
"loss": 1.2777,
"step": 8625
},
{
"epoch": 0.3632431524466421,
"grad_norm": 1.0078125,
"learning_rate": 0.00018823897700852898,
"loss": 1.0352,
"step": 8650
},
{
"epoch": 0.36429298814735495,
"grad_norm": 0.609375,
"learning_rate": 0.00018765447693105925,
"loss": 1.2105,
"step": 8675
},
{
"epoch": 0.36534282384806777,
"grad_norm": 0.44921875,
"learning_rate": 0.00018706936601485042,
"loss": 1.1375,
"step": 8700
},
{
"epoch": 0.36639265954878064,
"grad_norm": 0.56640625,
"learning_rate": 0.00018648365375169325,
"loss": 1.1686,
"step": 8725
},
{
"epoch": 0.36744249524949346,
"grad_norm": 0.4921875,
"learning_rate": 0.00018589734964313365,
"loss": 1.1878,
"step": 8750
},
{
"epoch": 0.3684923309502063,
"grad_norm": 0.640625,
"learning_rate": 0.0001853104632003185,
"loss": 1.1263,
"step": 8775
},
{
"epoch": 0.36954216665091916,
"grad_norm": 0.515625,
"learning_rate": 0.00018472300394384146,
"loss": 1.2784,
"step": 8800
},
{
"epoch": 0.370592002351632,
"grad_norm": 0.7890625,
"learning_rate": 0.00018413498140358846,
"loss": 1.2941,
"step": 8825
},
{
"epoch": 0.3716418380523448,
"grad_norm": 0.5625,
"learning_rate": 0.00018354640511858333,
"loss": 1.2202,
"step": 8850
},
{
"epoch": 0.3726916737530577,
"grad_norm": 0.65625,
"learning_rate": 0.00018295728463683258,
"loss": 1.1567,
"step": 8875
},
{
"epoch": 0.3737415094537705,
"grad_norm": 0.63671875,
"learning_rate": 0.00018236762951517093,
"loss": 1.3015,
"step": 8900
},
{
"epoch": 0.3747913451544833,
"grad_norm": 0.37890625,
"learning_rate": 0.00018177744931910614,
"loss": 1.0761,
"step": 8925
},
{
"epoch": 0.3758411808551962,
"grad_norm": 0.6171875,
"learning_rate": 0.00018118675362266385,
"loss": 1.2619,
"step": 8950
},
{
"epoch": 0.376891016555909,
"grad_norm": 0.58203125,
"learning_rate": 0.0001805955520082323,
"loss": 1.0349,
"step": 8975
},
{
"epoch": 0.37794085225662183,
"grad_norm": 0.482421875,
"learning_rate": 0.00018000385406640666,
"loss": 1.1587,
"step": 9000
},
{
"epoch": 0.3789906879573347,
"grad_norm": 0.435546875,
"learning_rate": 0.00017941166939583377,
"loss": 1.2204,
"step": 9025
},
{
"epoch": 0.3800405236580475,
"grad_norm": 0.279296875,
"learning_rate": 0.00017881900760305624,
"loss": 1.138,
"step": 9050
},
{
"epoch": 0.38109035935876034,
"grad_norm": 0.6015625,
"learning_rate": 0.0001782258783023567,
"loss": 1.2844,
"step": 9075
},
{
"epoch": 0.3821401950594732,
"grad_norm": 0.51953125,
"learning_rate": 0.0001776322911156017,
"loss": 1.0862,
"step": 9100
},
{
"epoch": 0.38319003076018604,
"grad_norm": 0.63671875,
"learning_rate": 0.00017703825567208587,
"loss": 1.2521,
"step": 9125
},
{
"epoch": 0.38423986646089886,
"grad_norm": 0.7109375,
"learning_rate": 0.00017644378160837536,
"loss": 1.1552,
"step": 9150
},
{
"epoch": 0.3852897021616117,
"grad_norm": 0.439453125,
"learning_rate": 0.00017584887856815178,
"loss": 1.1468,
"step": 9175
},
{
"epoch": 0.38633953786232456,
"grad_norm": 0.546875,
"learning_rate": 0.00017525355620205574,
"loss": 1.1263,
"step": 9200
},
{
"epoch": 0.3873893735630374,
"grad_norm": 0.0947265625,
"learning_rate": 0.00017465782416753027,
"loss": 1.1752,
"step": 9225
},
{
"epoch": 0.3884392092637502,
"grad_norm": 0.59375,
"learning_rate": 0.00017406169212866402,
"loss": 1.1728,
"step": 9250
},
{
"epoch": 0.38948904496446307,
"grad_norm": 0.59765625,
"learning_rate": 0.00017346516975603462,
"loss": 1.2319,
"step": 9275
},
{
"epoch": 0.3905388806651759,
"grad_norm": 0.5859375,
"learning_rate": 0.0001728682667265518,
"loss": 1.1663,
"step": 9300
},
{
"epoch": 0.3915887163658887,
"grad_norm": 0.625,
"learning_rate": 0.00017227099272330035,
"loss": 1.0206,
"step": 9325
},
{
"epoch": 0.3926385520666016,
"grad_norm": 0.58203125,
"learning_rate": 0.00017167335743538317,
"loss": 1.1324,
"step": 9350
},
{
"epoch": 0.3936883877673144,
"grad_norm": 0.8046875,
"learning_rate": 0.00017107537055776391,
"loss": 1.2566,
"step": 9375
},
{
"epoch": 0.3947382234680272,
"grad_norm": 0.5859375,
"learning_rate": 0.00017047704179110976,
"loss": 1.0822,
"step": 9400
},
{
"epoch": 0.3957880591687401,
"grad_norm": 0.625,
"learning_rate": 0.00016987838084163415,
"loss": 1.0014,
"step": 9425
},
{
"epoch": 0.3968378948694529,
"grad_norm": 0.5390625,
"learning_rate": 0.00016927939742093917,
"loss": 1.2111,
"step": 9450
},
{
"epoch": 0.39788773057016574,
"grad_norm": 0.55859375,
"learning_rate": 0.00016868010124585831,
"loss": 1.0861,
"step": 9475
},
{
"epoch": 0.3989375662708786,
"grad_norm": 0.404296875,
"learning_rate": 0.00016808050203829842,
"loss": 1.0733,
"step": 9500
},
{
"epoch": 0.39998740197159144,
"grad_norm": 0.5,
"learning_rate": 0.00016748060952508233,
"loss": 1.0351,
"step": 9525
},
{
"epoch": 0.40103723767230426,
"grad_norm": 0.68359375,
"learning_rate": 0.00016688043343779081,
"loss": 1.1101,
"step": 9550
},
{
"epoch": 0.40208707337301713,
"grad_norm": 0.61328125,
"learning_rate": 0.00016627998351260493,
"loss": 1.1334,
"step": 9575
},
{
"epoch": 0.40313690907372995,
"grad_norm": 0.671875,
"learning_rate": 0.000165679269490148,
"loss": 0.996,
"step": 9600
},
{
"epoch": 0.4041867447744428,
"grad_norm": 0.66015625,
"learning_rate": 0.00016507830111532756,
"loss": 1.0872,
"step": 9625
},
{
"epoch": 0.40523658047515565,
"grad_norm": 0.322265625,
"learning_rate": 0.00016447708813717725,
"loss": 1.0642,
"step": 9650
},
{
"epoch": 0.40628641617586847,
"grad_norm": 0.5234375,
"learning_rate": 0.00016387564030869884,
"loss": 1.0756,
"step": 9675
},
{
"epoch": 0.4073362518765813,
"grad_norm": 0.265625,
"learning_rate": 0.00016327396738670375,
"loss": 1.0704,
"step": 9700
},
{
"epoch": 0.40838608757729417,
"grad_norm": 0.5078125,
"learning_rate": 0.00016267207913165508,
"loss": 1.1415,
"step": 9725
},
{
"epoch": 0.409435923278007,
"grad_norm": 0.48828125,
"learning_rate": 0.0001620699853075089,
"loss": 1.0856,
"step": 9750
},
{
"epoch": 0.4104857589787198,
"grad_norm": 0.55859375,
"learning_rate": 0.00016146769568155624,
"loss": 1.2584,
"step": 9775
},
{
"epoch": 0.4115355946794327,
"grad_norm": 0.546875,
"learning_rate": 0.00016086522002426435,
"loss": 1.1531,
"step": 9800
},
{
"epoch": 0.4125854303801455,
"grad_norm": 0.78125,
"learning_rate": 0.00016026256810911841,
"loss": 1.0243,
"step": 9825
},
{
"epoch": 0.4136352660808583,
"grad_norm": 0.447265625,
"learning_rate": 0.00015965974971246288,
"loss": 1.1546,
"step": 9850
},
{
"epoch": 0.4146851017815712,
"grad_norm": 0.5546875,
"learning_rate": 0.0001590567746133429,
"loss": 1.0287,
"step": 9875
},
{
"epoch": 0.415734937482284,
"grad_norm": 0.64453125,
"learning_rate": 0.0001584536525933456,
"loss": 1.1281,
"step": 9900
},
{
"epoch": 0.41678477318299684,
"grad_norm": 0.11669921875,
"learning_rate": 0.0001578503934364416,
"loss": 1.0934,
"step": 9925
},
{
"epoch": 0.4178346088837097,
"grad_norm": 0.6484375,
"learning_rate": 0.00015724700692882604,
"loss": 1.2266,
"step": 9950
},
{
"epoch": 0.41888444458442253,
"grad_norm": 0.486328125,
"learning_rate": 0.0001566435028587602,
"loss": 1.1236,
"step": 9975
},
{
"epoch": 0.41993428028513535,
"grad_norm": 0.58984375,
"learning_rate": 0.00015603989101641228,
"loss": 1.0814,
"step": 10000
},
{
"epoch": 0.42098411598584823,
"grad_norm": 0.5078125,
"learning_rate": 0.00015543618119369883,
"loss": 1.1815,
"step": 10025
},
{
"epoch": 0.42203395168656105,
"grad_norm": 0.349609375,
"learning_rate": 0.00015483238318412596,
"loss": 1.1933,
"step": 10050
},
{
"epoch": 0.42308378738727387,
"grad_norm": 0.515625,
"learning_rate": 0.00015422850678263017,
"loss": 1.0942,
"step": 10075
},
{
"epoch": 0.42413362308798674,
"grad_norm": 0.47265625,
"learning_rate": 0.00015362456178541982,
"loss": 1.0045,
"step": 10100
},
{
"epoch": 0.42518345878869956,
"grad_norm": 0.4765625,
"learning_rate": 0.00015302055798981605,
"loss": 1.2429,
"step": 10125
},
{
"epoch": 0.4262332944894124,
"grad_norm": 0.3828125,
"learning_rate": 0.0001524165051940937,
"loss": 0.9375,
"step": 10150
},
{
"epoch": 0.42728313019012526,
"grad_norm": 0.62109375,
"learning_rate": 0.00015181241319732258,
"loss": 1.2556,
"step": 10175
},
{
"epoch": 0.4283329658908381,
"grad_norm": 0.51953125,
"learning_rate": 0.00015120829179920847,
"loss": 1.0118,
"step": 10200
},
{
"epoch": 0.4293828015915509,
"grad_norm": 0.625,
"learning_rate": 0.00015060415079993405,
"loss": 1.0527,
"step": 10225
},
{
"epoch": 0.4304326372922638,
"grad_norm": 0.49609375,
"learning_rate": 0.00015,
"loss": 0.9751,
"step": 10250
},
{
"epoch": 0.4314824729929766,
"grad_norm": 0.458984375,
"learning_rate": 0.00014939584920006593,
"loss": 1.1399,
"step": 10275
},
{
"epoch": 0.4325323086936894,
"grad_norm": 0.466796875,
"learning_rate": 0.00014879170820079153,
"loss": 1.0368,
"step": 10300
},
{
"epoch": 0.4335821443944023,
"grad_norm": 0.62890625,
"learning_rate": 0.0001481875868026774,
"loss": 1.2455,
"step": 10325
},
{
"epoch": 0.4346319800951151,
"grad_norm": 0.640625,
"learning_rate": 0.0001475834948059063,
"loss": 1.1758,
"step": 10350
},
{
"epoch": 0.43568181579582793,
"grad_norm": 0.60546875,
"learning_rate": 0.00014697944201018398,
"loss": 1.2452,
"step": 10375
},
{
"epoch": 0.4367316514965408,
"grad_norm": 0.5,
"learning_rate": 0.00014637543821458018,
"loss": 0.9339,
"step": 10400
},
{
"epoch": 0.4377814871972536,
"grad_norm": 0.54296875,
"learning_rate": 0.00014577149321736983,
"loss": 1.2017,
"step": 10425
},
{
"epoch": 0.43883132289796645,
"grad_norm": 0.423828125,
"learning_rate": 0.00014516761681587407,
"loss": 1.1525,
"step": 10450
},
{
"epoch": 0.4398811585986793,
"grad_norm": 0.625,
"learning_rate": 0.0001445638188063012,
"loss": 1.0734,
"step": 10475
},
{
"epoch": 0.44093099429939214,
"grad_norm": 0.357421875,
"learning_rate": 0.00014396010898358775,
"loss": 1.1867,
"step": 10500
},
{
"epoch": 0.44198083000010496,
"grad_norm": 0.63671875,
"learning_rate": 0.00014335649714123982,
"loss": 1.0687,
"step": 10525
},
{
"epoch": 0.44303066570081784,
"grad_norm": 0.1083984375,
"learning_rate": 0.00014275299307117394,
"loss": 0.9275,
"step": 10550
},
{
"epoch": 0.44408050140153066,
"grad_norm": 0.5234375,
"learning_rate": 0.0001421496065635584,
"loss": 1.0585,
"step": 10575
},
{
"epoch": 0.4451303371022435,
"grad_norm": 0.498046875,
"learning_rate": 0.00014154634740665435,
"loss": 1.2308,
"step": 10600
},
{
"epoch": 0.44618017280295635,
"grad_norm": 0.515625,
"learning_rate": 0.00014094322538665707,
"loss": 0.9778,
"step": 10625
},
{
"epoch": 0.4472300085036692,
"grad_norm": 0.609375,
"learning_rate": 0.00014034025028753706,
"loss": 1.1593,
"step": 10650
},
{
"epoch": 0.448279844204382,
"grad_norm": 0.67578125,
"learning_rate": 0.00013973743189088153,
"loss": 1.2109,
"step": 10675
},
{
"epoch": 0.44932967990509487,
"grad_norm": 0.4921875,
"learning_rate": 0.00013913477997573565,
"loss": 1.091,
"step": 10700
},
{
"epoch": 0.4503795156058077,
"grad_norm": 0.55078125,
"learning_rate": 0.00013853230431844376,
"loss": 1.1002,
"step": 10725
},
{
"epoch": 0.4514293513065205,
"grad_norm": 0.42578125,
"learning_rate": 0.0001379300146924911,
"loss": 1.2307,
"step": 10750
},
{
"epoch": 0.4524791870072334,
"grad_norm": 0.62109375,
"learning_rate": 0.00013732792086834495,
"loss": 1.2844,
"step": 10775
},
{
"epoch": 0.4535290227079462,
"grad_norm": 0.390625,
"learning_rate": 0.00013672603261329625,
"loss": 0.9445,
"step": 10800
},
{
"epoch": 0.454578858408659,
"grad_norm": 0.5703125,
"learning_rate": 0.00013612435969130113,
"loss": 1.1876,
"step": 10825
},
{
"epoch": 0.4556286941093719,
"grad_norm": 0.59765625,
"learning_rate": 0.00013552291186282273,
"loss": 1.1389,
"step": 10850
},
{
"epoch": 0.4566785298100847,
"grad_norm": 0.55859375,
"learning_rate": 0.00013492169888467244,
"loss": 0.97,
"step": 10875
},
{
"epoch": 0.45772836551079754,
"grad_norm": 0.56640625,
"learning_rate": 0.000134320730509852,
"loss": 1.0841,
"step": 10900
},
{
"epoch": 0.4587782012115104,
"grad_norm": 0.54296875,
"learning_rate": 0.00013372001648739507,
"loss": 1.0902,
"step": 10925
},
{
"epoch": 0.45982803691222324,
"grad_norm": 0.57421875,
"learning_rate": 0.0001331195665622092,
"loss": 0.9931,
"step": 10950
},
{
"epoch": 0.46087787261293606,
"grad_norm": 0.55078125,
"learning_rate": 0.00013251939047491767,
"loss": 1.0705,
"step": 10975
},
{
"epoch": 0.46192770831364893,
"grad_norm": 0.65234375,
"learning_rate": 0.00013191949796170156,
"loss": 1.0991,
"step": 11000
},
{
"epoch": 0.46297754401436175,
"grad_norm": 0.60546875,
"learning_rate": 0.0001313198987541417,
"loss": 1.211,
"step": 11025
},
{
"epoch": 0.4640273797150746,
"grad_norm": 0.5859375,
"learning_rate": 0.0001307206025790608,
"loss": 1.146,
"step": 11050
},
{
"epoch": 0.46507721541578745,
"grad_norm": 0.5625,
"learning_rate": 0.0001301216191583659,
"loss": 1.031,
"step": 11075
},
{
"epoch": 0.46612705111650027,
"grad_norm": 0.58984375,
"learning_rate": 0.00012952295820889027,
"loss": 1.188,
"step": 11100
},
{
"epoch": 0.4671768868172131,
"grad_norm": 0.63671875,
"learning_rate": 0.0001289246294422361,
"loss": 1.0754,
"step": 11125
},
{
"epoch": 0.46822672251792596,
"grad_norm": 0.4609375,
"learning_rate": 0.00012832664256461684,
"loss": 1.0702,
"step": 11150
},
{
"epoch": 0.4692765582186388,
"grad_norm": 0.6796875,
"learning_rate": 0.0001277290072766996,
"loss": 1.0649,
"step": 11175
},
{
"epoch": 0.4703263939193516,
"grad_norm": 0.5859375,
"learning_rate": 0.00012713173327344817,
"loss": 1.1837,
"step": 11200
},
{
"epoch": 0.4713762296200645,
"grad_norm": 0.5,
"learning_rate": 0.00012653483024396533,
"loss": 1.0212,
"step": 11225
},
{
"epoch": 0.4724260653207773,
"grad_norm": 0.53125,
"learning_rate": 0.00012593830787133593,
"loss": 1.2251,
"step": 11250
},
{
"epoch": 0.4734759010214901,
"grad_norm": 0.333984375,
"learning_rate": 0.0001253421758324697,
"loss": 1.0215,
"step": 11275
},
{
"epoch": 0.474525736722203,
"grad_norm": 0.1064453125,
"learning_rate": 0.0001247464437979442,
"loss": 0.9063,
"step": 11300
},
{
"epoch": 0.4755755724229158,
"grad_norm": 0.298828125,
"learning_rate": 0.00012415112143184822,
"loss": 0.8319,
"step": 11325
},
{
"epoch": 0.47662540812362864,
"grad_norm": 0.48046875,
"learning_rate": 0.00012355621839162464,
"loss": 0.9301,
"step": 11350
},
{
"epoch": 0.4776752438243415,
"grad_norm": 0.494140625,
"learning_rate": 0.00012296174432791413,
"loss": 0.9361,
"step": 11375
},
{
"epoch": 0.47872507952505433,
"grad_norm": 0.443359375,
"learning_rate": 0.00012236770888439827,
"loss": 1.1178,
"step": 11400
},
{
"epoch": 0.47977491522576715,
"grad_norm": 0.26953125,
"learning_rate": 0.00012177412169764329,
"loss": 1.0684,
"step": 11425
},
{
"epoch": 0.48082475092648,
"grad_norm": 0.625,
"learning_rate": 0.00012118099239694372,
"loss": 1.0906,
"step": 11450
},
{
"epoch": 0.48187458662719285,
"grad_norm": 0.76953125,
"learning_rate": 0.00012058833060416623,
"loss": 1.0151,
"step": 11475
},
{
"epoch": 0.48292442232790567,
"grad_norm": 0.51171875,
"learning_rate": 0.00011999614593359334,
"loss": 1.1089,
"step": 11500
},
{
"epoch": 0.48397425802861854,
"grad_norm": 0.2578125,
"learning_rate": 0.0001194044479917677,
"loss": 1.0821,
"step": 11525
},
{
"epoch": 0.48502409372933136,
"grad_norm": 0.546875,
"learning_rate": 0.00011881324637733611,
"loss": 1.1728,
"step": 11550
},
{
"epoch": 0.4860739294300442,
"grad_norm": 0.51953125,
"learning_rate": 0.00011822255068089384,
"loss": 1.1581,
"step": 11575
},
{
"epoch": 0.48712376513075706,
"grad_norm": 0.5390625,
"learning_rate": 0.00011763237048482908,
"loss": 0.9753,
"step": 11600
},
{
"epoch": 0.4881736008314699,
"grad_norm": 0.4921875,
"learning_rate": 0.00011704271536316746,
"loss": 0.985,
"step": 11625
},
{
"epoch": 0.4892234365321827,
"grad_norm": 0.443359375,
"learning_rate": 0.0001164535948814167,
"loss": 1.0448,
"step": 11650
},
{
"epoch": 0.4902732722328956,
"grad_norm": 0.71875,
"learning_rate": 0.0001158650185964115,
"loss": 1.1458,
"step": 11675
},
{
"epoch": 0.4913231079336084,
"grad_norm": 0.314453125,
"learning_rate": 0.00011527699605615856,
"loss": 1.0963,
"step": 11700
},
{
"epoch": 0.4923729436343212,
"grad_norm": 0.396484375,
"learning_rate": 0.00011468953679968153,
"loss": 1.1794,
"step": 11725
},
{
"epoch": 0.4934227793350341,
"grad_norm": 0.59375,
"learning_rate": 0.00011410265035686637,
"loss": 0.9767,
"step": 11750
},
{
"epoch": 0.4944726150357469,
"grad_norm": 0.337890625,
"learning_rate": 0.00011351634624830668,
"loss": 1.1573,
"step": 11775
},
{
"epoch": 0.49552245073645973,
"grad_norm": 0.6015625,
"learning_rate": 0.00011293063398514954,
"loss": 1.0924,
"step": 11800
},
{
"epoch": 0.4965722864371726,
"grad_norm": 0.54296875,
"learning_rate": 0.00011234552306894074,
"loss": 1.1425,
"step": 11825
},
{
"epoch": 0.4976221221378854,
"grad_norm": 0.48046875,
"learning_rate": 0.00011176102299147097,
"loss": 0.9907,
"step": 11850
},
{
"epoch": 0.49867195783859825,
"grad_norm": 0.59765625,
"learning_rate": 0.00011117714323462186,
"loss": 1.3243,
"step": 11875
},
{
"epoch": 0.4997217935393111,
"grad_norm": 0.5859375,
"learning_rate": 0.00011059389327021195,
"loss": 1.1395,
"step": 11900
},
{
"epoch": 0.5007716292400239,
"grad_norm": 0.61328125,
"learning_rate": 0.0001100112825598432,
"loss": 1.1725,
"step": 11925
},
{
"epoch": 0.5018214649407368,
"grad_norm": 0.59765625,
"learning_rate": 0.00010942932055474743,
"loss": 1.1107,
"step": 11950
},
{
"epoch": 0.5028713006414496,
"grad_norm": 0.43359375,
"learning_rate": 0.00010884801669563313,
"loss": 0.8768,
"step": 11975
},
{
"epoch": 0.5039211363421624,
"grad_norm": 0.51171875,
"learning_rate": 0.0001082673804125321,
"loss": 1.1317,
"step": 12000
},
{
"epoch": 0.5049709720428753,
"grad_norm": 0.5078125,
"learning_rate": 0.00010768742112464662,
"loss": 1.0004,
"step": 12025
},
{
"epoch": 0.5060208077435882,
"grad_norm": 0.4609375,
"learning_rate": 0.00010710814824019663,
"loss": 1.14,
"step": 12050
},
{
"epoch": 0.507070643444301,
"grad_norm": 0.443359375,
"learning_rate": 0.00010652957115626703,
"loss": 0.8955,
"step": 12075
},
{
"epoch": 0.5081204791450138,
"grad_norm": 0.63671875,
"learning_rate": 0.0001059516992586554,
"loss": 1.0887,
"step": 12100
},
{
"epoch": 0.5091703148457266,
"grad_norm": 0.58984375,
"learning_rate": 0.00010537454192171957,
"loss": 1.1575,
"step": 12125
},
{
"epoch": 0.5102201505464394,
"grad_norm": 0.41015625,
"learning_rate": 0.00010479810850822563,
"loss": 1.0744,
"step": 12150
},
{
"epoch": 0.5112699862471524,
"grad_norm": 0.390625,
"learning_rate": 0.00010422240836919604,
"loss": 1.2597,
"step": 12175
},
{
"epoch": 0.5123198219478652,
"grad_norm": 0.318359375,
"learning_rate": 0.0001036474508437579,
"loss": 1.1041,
"step": 12200
},
{
"epoch": 0.513369657648578,
"grad_norm": 0.478515625,
"learning_rate": 0.00010307324525899155,
"loss": 1.0814,
"step": 12225
},
{
"epoch": 0.5144194933492908,
"grad_norm": 0.447265625,
"learning_rate": 0.00010249980092977915,
"loss": 1.1349,
"step": 12250
},
{
"epoch": 0.5154693290500036,
"grad_norm": 0.478515625,
"learning_rate": 0.00010192712715865358,
"loss": 1.2416,
"step": 12275
},
{
"epoch": 0.5165191647507165,
"grad_norm": 0.62109375,
"learning_rate": 0.0001013552332356476,
"loss": 1.1884,
"step": 12300
},
{
"epoch": 0.5175690004514294,
"grad_norm": 0.41796875,
"learning_rate": 0.00010078412843814317,
"loss": 1.0958,
"step": 12325
},
{
"epoch": 0.5186188361521422,
"grad_norm": 0.490234375,
"learning_rate": 0.00010021382203072075,
"loss": 1.067,
"step": 12350
},
{
"epoch": 0.519668671852855,
"grad_norm": 0.60546875,
"learning_rate": 9.964432326500931e-05,
"loss": 1.0752,
"step": 12375
},
{
"epoch": 0.5207185075535679,
"grad_norm": 0.7578125,
"learning_rate": 9.907564137953592e-05,
"loss": 1.2035,
"step": 12400
},
{
"epoch": 0.5217683432542807,
"grad_norm": 0.6328125,
"learning_rate": 9.850778559957623e-05,
"loss": 1.171,
"step": 12425
},
{
"epoch": 0.5228181789549935,
"grad_norm": 0.267578125,
"learning_rate": 9.794076513700455e-05,
"loss": 0.8032,
"step": 12450
},
{
"epoch": 0.5238680146557064,
"grad_norm": 0.6171875,
"learning_rate": 9.737458919014442e-05,
"loss": 1.1434,
"step": 12475
},
{
"epoch": 0.5249178503564192,
"grad_norm": 0.380859375,
"learning_rate": 9.680926694361964e-05,
"loss": 1.1448,
"step": 12500
},
{
"epoch": 0.5259676860571321,
"grad_norm": 0.6640625,
"learning_rate": 9.624480756820496e-05,
"loss": 1.1123,
"step": 12525
},
{
"epoch": 0.5270175217578449,
"grad_norm": 0.5234375,
"learning_rate": 9.568122022067749e-05,
"loss": 0.975,
"step": 12550
},
{
"epoch": 0.5280673574585577,
"grad_norm": 0.6875,
"learning_rate": 9.511851404366815e-05,
"loss": 1.1167,
"step": 12575
},
{
"epoch": 0.5291171931592705,
"grad_norm": 0.51171875,
"learning_rate": 9.455669816551335e-05,
"loss": 1.201,
"step": 12600
},
{
"epoch": 0.5301670288599835,
"grad_norm": 0.625,
"learning_rate": 9.399578170010684e-05,
"loss": 1.2179,
"step": 12625
},
{
"epoch": 0.5312168645606963,
"grad_norm": 0.625,
"learning_rate": 9.343577374675193e-05,
"loss": 0.9526,
"step": 12650
},
{
"epoch": 0.5322667002614091,
"grad_norm": 0.427734375,
"learning_rate": 9.287668339001386e-05,
"loss": 0.9028,
"step": 12675
},
{
"epoch": 0.5333165359621219,
"grad_norm": 0.52734375,
"learning_rate": 9.231851969957238e-05,
"loss": 0.9427,
"step": 12700
},
{
"epoch": 0.5343663716628347,
"grad_norm": 0.57421875,
"learning_rate": 9.176129173007475e-05,
"loss": 1.0787,
"step": 12725
},
{
"epoch": 0.5354162073635476,
"grad_norm": 0.46484375,
"learning_rate": 9.120500852098876e-05,
"loss": 1.0251,
"step": 12750
},
{
"epoch": 0.5364660430642605,
"grad_norm": 0.58984375,
"learning_rate": 9.0649679096456e-05,
"loss": 0.9541,
"step": 12775
},
{
"epoch": 0.5375158787649733,
"grad_norm": 0.76953125,
"learning_rate": 9.009531246514569e-05,
"loss": 1.1522,
"step": 12800
},
{
"epoch": 0.5385657144656861,
"grad_norm": 0.341796875,
"learning_rate": 8.954191762010833e-05,
"loss": 1.0069,
"step": 12825
},
{
"epoch": 0.539615550166399,
"grad_norm": 0.11328125,
"learning_rate": 8.898950353862998e-05,
"loss": 1.133,
"step": 12850
},
{
"epoch": 0.5406653858671118,
"grad_norm": 0.451171875,
"learning_rate": 8.84380791820865e-05,
"loss": 1.1475,
"step": 12875
},
{
"epoch": 0.5417152215678246,
"grad_norm": 0.3984375,
"learning_rate": 8.788765349579823e-05,
"loss": 1.0686,
"step": 12900
},
{
"epoch": 0.5427650572685375,
"grad_norm": 0.546875,
"learning_rate": 8.733823540888489e-05,
"loss": 1.2948,
"step": 12925
},
{
"epoch": 0.5438148929692503,
"grad_norm": 1.0703125,
"learning_rate": 8.678983383412066e-05,
"loss": 1.0897,
"step": 12950
},
{
"epoch": 0.5448647286699632,
"grad_norm": 0.51171875,
"learning_rate": 8.624245766778971e-05,
"loss": 1.1607,
"step": 12975
},
{
"epoch": 0.545914564370676,
"grad_norm": 0.56640625,
"learning_rate": 8.569611578954184e-05,
"loss": 1.0567,
"step": 13000
},
{
"epoch": 0.5469644000713888,
"grad_norm": 0.58984375,
"learning_rate": 8.515081706224837e-05,
"loss": 1.1421,
"step": 13025
},
{
"epoch": 0.5480142357721016,
"grad_norm": 0.4921875,
"learning_rate": 8.460657033185846e-05,
"loss": 1.206,
"step": 13050
},
{
"epoch": 0.5490640714728146,
"grad_norm": 0.3671875,
"learning_rate": 8.406338442725541e-05,
"loss": 1.2295,
"step": 13075
},
{
"epoch": 0.5501139071735274,
"grad_norm": 0.59375,
"learning_rate": 8.352126816011381e-05,
"loss": 1.1154,
"step": 13100
},
{
"epoch": 0.5511637428742402,
"grad_norm": 0.58203125,
"learning_rate": 8.298023032475612e-05,
"loss": 1.0917,
"step": 13125
},
{
"epoch": 0.552213578574953,
"grad_norm": 0.59765625,
"learning_rate": 8.244027969801046e-05,
"loss": 0.9473,
"step": 13150
},
{
"epoch": 0.5532634142756658,
"grad_norm": 0.11669921875,
"learning_rate": 8.190142503906798e-05,
"loss": 0.9594,
"step": 13175
},
{
"epoch": 0.5543132499763787,
"grad_norm": 0.53125,
"learning_rate": 8.136367508934071e-05,
"loss": 1.1299,
"step": 13200
},
{
"epoch": 0.5553630856770916,
"grad_norm": 0.55078125,
"learning_rate": 8.082703857232e-05,
"loss": 1.0247,
"step": 13225
},
{
"epoch": 0.5564129213778044,
"grad_norm": 0.18359375,
"learning_rate": 8.02915241934347e-05,
"loss": 1.0369,
"step": 13250
},
{
"epoch": 0.5574627570785172,
"grad_norm": 0.439453125,
"learning_rate": 7.97571406399103e-05,
"loss": 1.0932,
"step": 13275
},
{
"epoch": 0.55851259277923,
"grad_norm": 0.63671875,
"learning_rate": 7.922389658062769e-05,
"loss": 1.1498,
"step": 13300
},
{
"epoch": 0.5595624284799429,
"grad_norm": 0.64453125,
"learning_rate": 7.869180066598258e-05,
"loss": 1.1687,
"step": 13325
},
{
"epoch": 0.5606122641806557,
"grad_norm": 0.484375,
"learning_rate": 7.816086152774542e-05,
"loss": 1.0859,
"step": 13350
},
{
"epoch": 0.5616620998813686,
"grad_norm": 0.54296875,
"learning_rate": 7.7631087778921e-05,
"loss": 1.0387,
"step": 13375
},
{
"epoch": 0.5627119355820814,
"grad_norm": 0.609375,
"learning_rate": 7.710248801360908e-05,
"loss": 1.2117,
"step": 13400
},
{
"epoch": 0.5637617712827943,
"grad_norm": 0.53125,
"learning_rate": 7.65750708068648e-05,
"loss": 1.2659,
"step": 13425
},
{
"epoch": 0.5648116069835071,
"grad_norm": 0.52734375,
"learning_rate": 7.604884471455938e-05,
"loss": 1.0915,
"step": 13450
},
{
"epoch": 0.5658614426842199,
"grad_norm": 0.58203125,
"learning_rate": 7.552381827324186e-05,
"loss": 1.2546,
"step": 13475
},
{
"epoch": 0.5669112783849327,
"grad_norm": 0.435546875,
"learning_rate": 7.500000000000002e-05,
"loss": 1.1008,
"step": 13500
},
{
"epoch": 0.5679611140856456,
"grad_norm": 0.6796875,
"learning_rate": 7.447739839232263e-05,
"loss": 1.1686,
"step": 13525
},
{
"epoch": 0.5690109497863585,
"grad_norm": 0.57421875,
"learning_rate": 7.395602192796148e-05,
"loss": 1.1541,
"step": 13550
},
{
"epoch": 0.5700607854870713,
"grad_norm": 0.5703125,
"learning_rate": 7.343587906479366e-05,
"loss": 1.0637,
"step": 13575
},
{
"epoch": 0.5711106211877841,
"grad_norm": 0.5859375,
"learning_rate": 7.291697824068471e-05,
"loss": 1.0259,
"step": 13600
},
{
"epoch": 0.5721604568884969,
"grad_norm": 0.5390625,
"learning_rate": 7.239932787335146e-05,
"loss": 1.1753,
"step": 13625
},
{
"epoch": 0.5732102925892097,
"grad_norm": 0.578125,
"learning_rate": 7.188293636022562e-05,
"loss": 1.1472,
"step": 13650
},
{
"epoch": 0.5742601282899227,
"grad_norm": 0.5390625,
"learning_rate": 7.136781207831751e-05,
"loss": 1.2726,
"step": 13675
},
{
"epoch": 0.5753099639906355,
"grad_norm": 0.55078125,
"learning_rate": 7.085396338408006e-05,
"loss": 1.0084,
"step": 13700
},
{
"epoch": 0.5763597996913483,
"grad_norm": 0.3671875,
"learning_rate": 7.03413986132735e-05,
"loss": 0.916,
"step": 13725
},
{
"epoch": 0.5774096353920611,
"grad_norm": 0.4765625,
"learning_rate": 6.983012608082982e-05,
"loss": 1.0645,
"step": 13750
},
{
"epoch": 0.578459471092774,
"grad_norm": 0.58984375,
"learning_rate": 6.93201540807182e-05,
"loss": 1.1948,
"step": 13775
},
{
"epoch": 0.5795093067934868,
"grad_norm": 0.42578125,
"learning_rate": 6.88114908858103e-05,
"loss": 1.0131,
"step": 13800
},
{
"epoch": 0.5805591424941997,
"grad_norm": 0.53515625,
"learning_rate": 6.830414474774594e-05,
"loss": 1.2336,
"step": 13825
},
{
"epoch": 0.5816089781949125,
"grad_norm": 0.58203125,
"learning_rate": 6.779812389679952e-05,
"loss": 1.207,
"step": 13850
},
{
"epoch": 0.5826588138956253,
"grad_norm": 0.6328125,
"learning_rate": 6.729343654174625e-05,
"loss": 1.0024,
"step": 13875
},
{
"epoch": 0.5837086495963382,
"grad_norm": 0.515625,
"learning_rate": 6.679009086972916e-05,
"loss": 1.0531,
"step": 13900
},
{
"epoch": 0.584758485297051,
"grad_norm": 0.5703125,
"learning_rate": 6.628809504612626e-05,
"loss": 0.9854,
"step": 13925
},
{
"epoch": 0.5858083209977638,
"grad_norm": 0.546875,
"learning_rate": 6.578745721441787e-05,
"loss": 1.2319,
"step": 13950
},
{
"epoch": 0.5868581566984767,
"grad_norm": 0.40625,
"learning_rate": 6.528818549605488e-05,
"loss": 0.9618,
"step": 13975
},
{
"epoch": 0.5879079923991896,
"grad_norm": 0.62109375,
"learning_rate": 6.479028799032664e-05,
"loss": 1.2905,
"step": 14000
},
{
"epoch": 0.5889578280999024,
"grad_norm": 0.5390625,
"learning_rate": 6.429377277422985e-05,
"loss": 1.2429,
"step": 14025
},
{
"epoch": 0.5900076638006152,
"grad_norm": 0.55859375,
"learning_rate": 6.379864790233747e-05,
"loss": 1.2236,
"step": 14050
},
{
"epoch": 0.591057499501328,
"grad_norm": 0.62890625,
"learning_rate": 6.330492140666776e-05,
"loss": 1.1221,
"step": 14075
},
{
"epoch": 0.5921073352020408,
"grad_norm": 0.388671875,
"learning_rate": 6.281260129655455e-05,
"loss": 0.9457,
"step": 14100
},
{
"epoch": 0.5931571709027538,
"grad_norm": 0.57421875,
"learning_rate": 6.23216955585167e-05,
"loss": 1.0886,
"step": 14125
},
{
"epoch": 0.5942070066034666,
"grad_norm": 0.62109375,
"learning_rate": 6.183221215612904e-05,
"loss": 1.1392,
"step": 14150
},
{
"epoch": 0.5952568423041794,
"grad_norm": 0.5390625,
"learning_rate": 6.134415902989295e-05,
"loss": 0.8879,
"step": 14175
},
{
"epoch": 0.5963066780048922,
"grad_norm": 0.68359375,
"learning_rate": 6.08575440971074e-05,
"loss": 1.1193,
"step": 14200
},
{
"epoch": 0.597356513705605,
"grad_norm": 0.6953125,
"learning_rate": 6.03723752517409e-05,
"loss": 1.0994,
"step": 14225
},
{
"epoch": 0.5984063494063179,
"grad_norm": 0.4609375,
"learning_rate": 5.9888660364303137e-05,
"loss": 1.173,
"step": 14250
},
{
"epoch": 0.5994561851070308,
"grad_norm": 0.298828125,
"learning_rate": 5.9406407281717405e-05,
"loss": 1.0254,
"step": 14275
},
{
"epoch": 0.6005060208077436,
"grad_norm": 0.546875,
"learning_rate": 5.892562382719332e-05,
"loss": 1.0322,
"step": 14300
},
{
"epoch": 0.6015558565084564,
"grad_norm": 0.57421875,
"learning_rate": 5.844631780009976e-05,
"loss": 1.0679,
"step": 14325
},
{
"epoch": 0.6026056922091693,
"grad_norm": 0.173828125,
"learning_rate": 5.7968496975838674e-05,
"loss": 1.2121,
"step": 14350
},
{
"epoch": 0.6036555279098821,
"grad_norm": 0.41015625,
"learning_rate": 5.7492169105718535e-05,
"loss": 1.1169,
"step": 14375
},
{
"epoch": 0.6047053636105949,
"grad_norm": 0.60546875,
"learning_rate": 5.701734191682894e-05,
"loss": 1.2163,
"step": 14400
},
{
"epoch": 0.6057551993113078,
"grad_norm": 0.482421875,
"learning_rate": 5.654402311191515e-05,
"loss": 1.0457,
"step": 14425
},
{
"epoch": 0.6068050350120207,
"grad_norm": 0.470703125,
"learning_rate": 5.607222036925294e-05,
"loss": 1.1437,
"step": 14450
},
{
"epoch": 0.6078548707127335,
"grad_norm": 0.58984375,
"learning_rate": 5.56019413425244e-05,
"loss": 1.0453,
"step": 14475
},
{
"epoch": 0.6089047064134463,
"grad_norm": 0.55078125,
"learning_rate": 5.513319366069342e-05,
"loss": 1.1764,
"step": 14500
},
{
"epoch": 0.6099545421141591,
"grad_norm": 0.65234375,
"learning_rate": 5.4665984927882206e-05,
"loss": 1.0844,
"step": 14525
},
{
"epoch": 0.6110043778148719,
"grad_norm": 0.404296875,
"learning_rate": 5.420032272324783e-05,
"loss": 0.98,
"step": 14550
},
{
"epoch": 0.6120542135155849,
"grad_norm": 0.458984375,
"learning_rate": 5.373621460085911e-05,
"loss": 1.211,
"step": 14575
},
{
"epoch": 0.6131040492162977,
"grad_norm": 0.6796875,
"learning_rate": 5.327366808957444e-05,
"loss": 1.0328,
"step": 14600
},
{
"epoch": 0.6141538849170105,
"grad_norm": 0.55859375,
"learning_rate": 5.2812690692919195e-05,
"loss": 1.088,
"step": 14625
},
{
"epoch": 0.6152037206177233,
"grad_norm": 0.59765625,
"learning_rate": 5.235328988896443e-05,
"loss": 0.8702,
"step": 14650
},
{
"epoch": 0.6162535563184361,
"grad_norm": 0.60546875,
"learning_rate": 5.1895473130205374e-05,
"loss": 1.135,
"step": 14675
},
{
"epoch": 0.617303392019149,
"grad_norm": 0.6015625,
"learning_rate": 5.1439247843440405e-05,
"loss": 1.2242,
"step": 14700
},
{
"epoch": 0.6183532277198618,
"grad_norm": 0.42578125,
"learning_rate": 5.098462142965087e-05,
"loss": 1.2759,
"step": 14725
},
{
"epoch": 0.6194030634205747,
"grad_norm": 0.482421875,
"learning_rate": 5.0531601263880747e-05,
"loss": 1.1016,
"step": 14750
},
{
"epoch": 0.6204528991212875,
"grad_norm": 0.546875,
"learning_rate": 5.008019469511719e-05,
"loss": 0.997,
"step": 14775
},
{
"epoch": 0.6215027348220004,
"grad_norm": 0.6015625,
"learning_rate": 4.963040904617131e-05,
"loss": 1.1841,
"step": 14800
},
{
"epoch": 0.6225525705227132,
"grad_norm": 0.625,
"learning_rate": 4.9182251613559074e-05,
"loss": 0.985,
"step": 14825
},
{
"epoch": 0.623602406223426,
"grad_norm": 0.5625,
"learning_rate": 4.873572966738359e-05,
"loss": 1.1212,
"step": 14850
},
{
"epoch": 0.6246522419241388,
"grad_norm": 0.404296875,
"learning_rate": 4.8290850451216354e-05,
"loss": 1.1319,
"step": 14875
},
{
"epoch": 0.6257020776248517,
"grad_norm": 0.609375,
"learning_rate": 4.7847621181980404e-05,
"loss": 1.1624,
"step": 14900
},
{
"epoch": 0.6267519133255646,
"grad_norm": 0.53515625,
"learning_rate": 4.7406049049832955e-05,
"loss": 1.1851,
"step": 14925
},
{
"epoch": 0.6278017490262774,
"grad_norm": 0.5390625,
"learning_rate": 4.69661412180487e-05,
"loss": 0.917,
"step": 14950
},
{
"epoch": 0.6288515847269902,
"grad_norm": 0.3671875,
"learning_rate": 4.652790482290385e-05,
"loss": 1.0829,
"step": 14975
},
{
"epoch": 0.629901420427703,
"grad_norm": 0.64453125,
"learning_rate": 4.609134697356009e-05,
"loss": 0.9791,
"step": 15000
},
{
"epoch": 0.6309512561284158,
"grad_norm": 0.5625,
"learning_rate": 4.5656474751949455e-05,
"loss": 1.0311,
"step": 15025
},
{
"epoch": 0.6320010918291288,
"grad_norm": 0.51171875,
"learning_rate": 4.5223295212659464e-05,
"loss": 1.179,
"step": 15050
},
{
"epoch": 0.6330509275298416,
"grad_norm": 0.50390625,
"learning_rate": 4.4791815382818365e-05,
"loss": 1.1793,
"step": 15075
},
{
"epoch": 0.6341007632305544,
"grad_norm": 0.71875,
"learning_rate": 4.436204226198162e-05,
"loss": 1.1232,
"step": 15100
},
{
"epoch": 0.6351505989312672,
"grad_norm": 0.55859375,
"learning_rate": 4.3933982822017876e-05,
"loss": 1.0392,
"step": 15125
},
{
"epoch": 0.6362004346319801,
"grad_norm": 0.53125,
"learning_rate": 4.350764400699626e-05,
"loss": 1.031,
"step": 15150
},
{
"epoch": 0.6372502703326929,
"grad_norm": 0.4140625,
"learning_rate": 4.3083032733073536e-05,
"loss": 1.0691,
"step": 15175
},
{
"epoch": 0.6383001060334058,
"grad_norm": 0.421875,
"learning_rate": 4.266015588838185e-05,
"loss": 0.9525,
"step": 15200
},
{
"epoch": 0.6393499417341186,
"grad_norm": 0.6875,
"learning_rate": 4.223902033291719e-05,
"loss": 1.0429,
"step": 15225
},
{
"epoch": 0.6403997774348315,
"grad_norm": 0.66796875,
"learning_rate": 4.181963289842783e-05,
"loss": 1.0999,
"step": 15250
},
{
"epoch": 0.6414496131355443,
"grad_norm": 0.37890625,
"learning_rate": 4.140200038830387e-05,
"loss": 1.0902,
"step": 15275
},
{
"epoch": 0.6424994488362571,
"grad_norm": 0.376953125,
"learning_rate": 4.0986129577466545e-05,
"loss": 0.858,
"step": 15300
},
{
"epoch": 0.6435492845369699,
"grad_norm": 0.2138671875,
"learning_rate": 4.0572027212258404e-05,
"loss": 1.1549,
"step": 15325
},
{
"epoch": 0.6445991202376828,
"grad_norm": 0.482421875,
"learning_rate": 4.015970001033401e-05,
"loss": 1.0091,
"step": 15350
},
{
"epoch": 0.6456489559383957,
"grad_norm": 0.5078125,
"learning_rate": 3.974915466055074e-05,
"loss": 1.0572,
"step": 15375
},
{
"epoch": 0.6466987916391085,
"grad_norm": 0.462890625,
"learning_rate": 3.934039782286048e-05,
"loss": 1.1396,
"step": 15400
},
{
"epoch": 0.6477486273398213,
"grad_norm": 0.72265625,
"learning_rate": 3.893343612820155e-05,
"loss": 1.1633,
"step": 15425
},
{
"epoch": 0.6487984630405341,
"grad_norm": 0.423828125,
"learning_rate": 3.852827617839084e-05,
"loss": 0.9902,
"step": 15450
},
{
"epoch": 0.6498482987412469,
"grad_norm": 0.59765625,
"learning_rate": 3.812492454601735e-05,
"loss": 1.0629,
"step": 15475
},
{
"epoch": 0.6508981344419599,
"grad_norm": 0.267578125,
"learning_rate": 3.7723387774334816e-05,
"loss": 1.1956,
"step": 15500
},
{
"epoch": 0.6519479701426727,
"grad_norm": 0.6171875,
"learning_rate": 3.732367237715614e-05,
"loss": 1.0655,
"step": 15525
},
{
"epoch": 0.6529978058433855,
"grad_norm": 0.69921875,
"learning_rate": 3.692578483874748e-05,
"loss": 1.0292,
"step": 15550
},
{
"epoch": 0.6540476415440983,
"grad_norm": 0.388671875,
"learning_rate": 3.652973161372295e-05,
"loss": 1.1644,
"step": 15575
},
{
"epoch": 0.6550974772448112,
"grad_norm": 0.412109375,
"learning_rate": 3.613551912694028e-05,
"loss": 1.0857,
"step": 15600
},
{
"epoch": 0.656147312945524,
"grad_norm": 0.384765625,
"learning_rate": 3.574315377339613e-05,
"loss": 1.1966,
"step": 15625
},
{
"epoch": 0.6571971486462369,
"grad_norm": 0.58984375,
"learning_rate": 3.535264191812274e-05,
"loss": 1.1279,
"step": 15650
},
{
"epoch": 0.6582469843469497,
"grad_norm": 0.404296875,
"learning_rate": 3.49639898960845e-05,
"loss": 1.035,
"step": 15675
},
{
"epoch": 0.6592968200476625,
"grad_norm": 0.6171875,
"learning_rate": 3.45772040120751e-05,
"loss": 1.031,
"step": 15700
},
{
"epoch": 0.6603466557483754,
"grad_norm": 0.59765625,
"learning_rate": 3.4192290540615506e-05,
"loss": 1.0585,
"step": 15725
},
{
"epoch": 0.6613964914490882,
"grad_norm": 0.439453125,
"learning_rate": 3.3809255725851825e-05,
"loss": 0.9671,
"step": 15750
},
{
"epoch": 0.662446327149801,
"grad_norm": 0.494140625,
"learning_rate": 3.342810578145436e-05,
"loss": 0.9874,
"step": 15775
},
{
"epoch": 0.6634961628505139,
"grad_norm": 0.59375,
"learning_rate": 3.304884689051666e-05,
"loss": 1.0242,
"step": 15800
},
{
"epoch": 0.6645459985512268,
"grad_norm": 0.392578125,
"learning_rate": 3.2671485205455064e-05,
"loss": 0.9905,
"step": 15825
},
{
"epoch": 0.6655958342519396,
"grad_norm": 0.51953125,
"learning_rate": 3.229602684790924e-05,
"loss": 1.1478,
"step": 15850
},
{
"epoch": 0.6666456699526524,
"grad_norm": 0.51171875,
"learning_rate": 3.192247790864249e-05,
"loss": 1.2598,
"step": 15875
},
{
"epoch": 0.6676955056533652,
"grad_norm": 0.458984375,
"learning_rate": 3.155084444744326e-05,
"loss": 1.1662,
"step": 15900
},
{
"epoch": 0.668745341354078,
"grad_norm": 0.68359375,
"learning_rate": 3.1181132493026734e-05,
"loss": 1.1146,
"step": 15925
},
{
"epoch": 0.669795177054791,
"grad_norm": 0.51171875,
"learning_rate": 3.0813348042936876e-05,
"loss": 1.0196,
"step": 15950
},
{
"epoch": 0.6708450127555038,
"grad_norm": 0.240234375,
"learning_rate": 3.044749706344944e-05,
"loss": 0.8744,
"step": 15975
},
{
"epoch": 0.6718948484562166,
"grad_norm": 0.63671875,
"learning_rate": 3.0083585489474855e-05,
"loss": 1.0892,
"step": 16000
},
{
"epoch": 0.6729446841569294,
"grad_norm": 0.55078125,
"learning_rate": 2.9721619224462236e-05,
"loss": 1.0506,
"step": 16025
},
{
"epoch": 0.6739945198576422,
"grad_norm": 0.376953125,
"learning_rate": 2.9361604140303467e-05,
"loss": 1.0366,
"step": 16050
},
{
"epoch": 0.6750443555583551,
"grad_norm": 0.578125,
"learning_rate": 2.9003546077237933e-05,
"loss": 1.2576,
"step": 16075
},
{
"epoch": 0.676094191259068,
"grad_norm": 0.408203125,
"learning_rate": 2.8647450843757897e-05,
"loss": 1.0275,
"step": 16100
},
{
"epoch": 0.6771440269597808,
"grad_norm": 0.56640625,
"learning_rate": 2.8293324216514033e-05,
"loss": 1.1833,
"step": 16125
},
{
"epoch": 0.6781938626604936,
"grad_norm": 0.26953125,
"learning_rate": 2.7941171940222052e-05,
"loss": 0.945,
"step": 16150
},
{
"epoch": 0.6792436983612065,
"grad_norm": 0.78125,
"learning_rate": 2.7590999727569285e-05,
"loss": 1.1001,
"step": 16175
},
{
"epoch": 0.6802935340619193,
"grad_norm": 0.671875,
"learning_rate": 2.7242813259121936e-05,
"loss": 1.1169,
"step": 16200
},
{
"epoch": 0.6813433697626321,
"grad_norm": 0.474609375,
"learning_rate": 2.689661818323326e-05,
"loss": 1.0022,
"step": 16225
},
{
"epoch": 0.682393205463345,
"grad_norm": 0.408203125,
"learning_rate": 2.6552420115951544e-05,
"loss": 0.9877,
"step": 16250
},
{
"epoch": 0.6834430411640579,
"grad_norm": 0.58984375,
"learning_rate": 2.6210224640929283e-05,
"loss": 1.1879,
"step": 16275
},
{
"epoch": 0.6844928768647707,
"grad_norm": 0.4140625,
"learning_rate": 2.5870037309332535e-05,
"loss": 1.0557,
"step": 16300
},
{
"epoch": 0.6855427125654835,
"grad_norm": 0.53125,
"learning_rate": 2.5531863639750738e-05,
"loss": 1.1362,
"step": 16325
},
{
"epoch": 0.6865925482661963,
"grad_norm": 0.55078125,
"learning_rate": 2.5195709118107372e-05,
"loss": 1.0474,
"step": 16350
},
{
"epoch": 0.6876423839669091,
"grad_norm": 0.49609375,
"learning_rate": 2.4861579197570803e-05,
"loss": 0.9323,
"step": 16375
},
{
"epoch": 0.6886922196676221,
"grad_norm": 0.6015625,
"learning_rate": 2.4529479298465986e-05,
"loss": 1.3069,
"step": 16400
},
{
"epoch": 0.6897420553683349,
"grad_norm": 0.5859375,
"learning_rate": 2.4199414808186406e-05,
"loss": 1.1598,
"step": 16425
},
{
"epoch": 0.6907918910690477,
"grad_norm": 0.490234375,
"learning_rate": 2.387139108110669e-05,
"loss": 1.0227,
"step": 16450
},
{
"epoch": 0.6918417267697605,
"grad_norm": 0.44140625,
"learning_rate": 2.3545413438495857e-05,
"loss": 1.1178,
"step": 16475
},
{
"epoch": 0.6928915624704733,
"grad_norm": 0.625,
"learning_rate": 2.3221487168430806e-05,
"loss": 1.0225,
"step": 16500
},
{
"epoch": 0.6939413981711862,
"grad_norm": 0.609375,
"learning_rate": 2.289961752571075e-05,
"loss": 1.0469,
"step": 16525
},
{
"epoch": 0.6949912338718991,
"grad_norm": 0.55859375,
"learning_rate": 2.257980973177187e-05,
"loss": 1.0234,
"step": 16550
},
{
"epoch": 0.6960410695726119,
"grad_norm": 0.6015625,
"learning_rate": 2.2262068974602503e-05,
"loss": 1.1441,
"step": 16575
},
{
"epoch": 0.6970909052733247,
"grad_norm": 0.62109375,
"learning_rate": 2.1946400408659198e-05,
"loss": 1.1231,
"step": 16600
},
{
"epoch": 0.6981407409740376,
"grad_norm": 0.5625,
"learning_rate": 2.1632809154782886e-05,
"loss": 1.0681,
"step": 16625
},
{
"epoch": 0.6991905766747504,
"grad_norm": 0.29296875,
"learning_rate": 2.1321300300116013e-05,
"loss": 0.9589,
"step": 16650
},
{
"epoch": 0.7002404123754632,
"grad_norm": 0.4609375,
"learning_rate": 2.1011878898019853e-05,
"loss": 0.9615,
"step": 16675
},
{
"epoch": 0.7012902480761761,
"grad_norm": 0.625,
"learning_rate": 2.0704549967992605e-05,
"loss": 1.2253,
"step": 16700
},
{
"epoch": 0.702340083776889,
"grad_norm": 0.640625,
"learning_rate": 2.039931849558798e-05,
"loss": 1.255,
"step": 16725
},
{
"epoch": 0.7033899194776018,
"grad_norm": 0.47265625,
"learning_rate": 2.009618943233419e-05,
"loss": 1.0251,
"step": 16750
},
{
"epoch": 0.7044397551783146,
"grad_norm": 0.421875,
"learning_rate": 1.9795167695653878e-05,
"loss": 1.1894,
"step": 16775
},
{
"epoch": 0.7054895908790274,
"grad_norm": 0.515625,
"learning_rate": 1.9496258168784122e-05,
"loss": 1.0545,
"step": 16800
},
{
"epoch": 0.7065394265797402,
"grad_norm": 0.6328125,
"learning_rate": 1.9199465700697247e-05,
"loss": 1.037,
"step": 16825
},
{
"epoch": 0.7075892622804532,
"grad_norm": 0.482421875,
"learning_rate": 1.8904795106022313e-05,
"loss": 1.068,
"step": 16850
},
{
"epoch": 0.708639097981166,
"grad_norm": 0.453125,
"learning_rate": 1.861225116496682e-05,
"loss": 1.1304,
"step": 16875
},
{
"epoch": 0.7096889336818788,
"grad_norm": 0.61328125,
"learning_rate": 1.8321838623239306e-05,
"loss": 1.0944,
"step": 16900
},
{
"epoch": 0.7107387693825916,
"grad_norm": 0.427734375,
"learning_rate": 1.803356219197234e-05,
"loss": 1.1014,
"step": 16925
},
{
"epoch": 0.7117886050833044,
"grad_norm": 1.0234375,
"learning_rate": 1.774742654764591e-05,
"loss": 1.1269,
"step": 16950
},
{
"epoch": 0.7128384407840173,
"grad_norm": 0.62890625,
"learning_rate": 1.7463436332011933e-05,
"loss": 0.9335,
"step": 16975
},
{
"epoch": 0.7138882764847302,
"grad_norm": 0.51171875,
"learning_rate": 1.718159615201853e-05,
"loss": 1.0583,
"step": 17000
},
{
"epoch": 0.714938112185443,
"grad_norm": 0.396484375,
"learning_rate": 1.6901910579735623e-05,
"loss": 1.1465,
"step": 17025
},
{
"epoch": 0.7159879478861558,
"grad_norm": 0.6328125,
"learning_rate": 1.6624384152280646e-05,
"loss": 1.2525,
"step": 17050
},
{
"epoch": 0.7170377835868686,
"grad_norm": 0.52734375,
"learning_rate": 1.634902137174483e-05,
"loss": 1.1994,
"step": 17075
},
{
"epoch": 0.7180876192875815,
"grad_norm": 0.84375,
"learning_rate": 1.6075826705120404e-05,
"loss": 1.1688,
"step": 17100
},
{
"epoch": 0.7191374549882943,
"grad_norm": 0.451171875,
"learning_rate": 1.580480458422791e-05,
"loss": 1.1071,
"step": 17125
},
{
"epoch": 0.7201872906890072,
"grad_norm": 0.84765625,
"learning_rate": 1.5535959405644434e-05,
"loss": 1.0429,
"step": 17150
},
{
"epoch": 0.72123712638972,
"grad_norm": 0.5546875,
"learning_rate": 1.5269295530632307e-05,
"loss": 0.9043,
"step": 17175
},
{
"epoch": 0.7222869620904329,
"grad_norm": 0.62109375,
"learning_rate": 1.5004817285068194e-05,
"loss": 0.9717,
"step": 17200
},
{
"epoch": 0.7233367977911457,
"grad_norm": 0.58203125,
"learning_rate": 1.4742528959373107e-05,
"loss": 1.1211,
"step": 17225
},
{
"epoch": 0.7243866334918585,
"grad_norm": 0.609375,
"learning_rate": 1.4482434808442684e-05,
"loss": 1.1302,
"step": 17250
},
{
"epoch": 0.7254364691925713,
"grad_norm": 0.7421875,
"learning_rate": 1.4224539051578198e-05,
"loss": 1.1155,
"step": 17275
},
{
"epoch": 0.7264863048932843,
"grad_norm": 0.578125,
"learning_rate": 1.3968845872418188e-05,
"loss": 1.0724,
"step": 17300
},
{
"epoch": 0.7275361405939971,
"grad_norm": 0.63671875,
"learning_rate": 1.371535941887042e-05,
"loss": 1.0348,
"step": 17325
},
{
"epoch": 0.7285859762947099,
"grad_norm": 0.546875,
"learning_rate": 1.3464083803044812e-05,
"loss": 1.1203,
"step": 17350
},
{
"epoch": 0.7296358119954227,
"grad_norm": 0.66796875,
"learning_rate": 1.3215023101186489e-05,
"loss": 0.9517,
"step": 17375
},
{
"epoch": 0.7306856476961355,
"grad_norm": 0.46484375,
"learning_rate": 1.2968181353609852e-05,
"loss": 1.0962,
"step": 17400
},
{
"epoch": 0.7317354833968484,
"grad_norm": 0.486328125,
"learning_rate": 1.2723562564632993e-05,
"loss": 1.0109,
"step": 17425
},
{
"epoch": 0.7327853190975613,
"grad_norm": 0.56640625,
"learning_rate": 1.2481170702512638e-05,
"loss": 1.0531,
"step": 17450
},
{
"epoch": 0.7338351547982741,
"grad_norm": 0.578125,
"learning_rate": 1.2241009699379905e-05,
"loss": 1.0217,
"step": 17475
},
{
"epoch": 0.7348849904989869,
"grad_norm": 0.482421875,
"learning_rate": 1.2003083451176366e-05,
"loss": 1.2119,
"step": 17500
},
{
"epoch": 0.7359348261996997,
"grad_norm": 0.490234375,
"learning_rate": 1.1767395817591053e-05,
"loss": 0.9211,
"step": 17525
},
{
"epoch": 0.7369846619004126,
"grad_norm": 0.333984375,
"learning_rate": 1.1533950621997684e-05,
"loss": 1.0552,
"step": 17550
},
{
"epoch": 0.7380344976011254,
"grad_norm": 0.490234375,
"learning_rate": 1.1302751651392594e-05,
"loss": 1.1606,
"step": 17575
},
{
"epoch": 0.7390843333018383,
"grad_norm": 0.57421875,
"learning_rate": 1.1073802656333546e-05,
"loss": 1.2232,
"step": 17600
},
{
"epoch": 0.7401341690025511,
"grad_norm": 0.63671875,
"learning_rate": 1.084710735087857e-05,
"loss": 1.0956,
"step": 17625
},
{
"epoch": 0.741184004703264,
"grad_norm": 0.51171875,
"learning_rate": 1.0622669412526019e-05,
"loss": 0.9505,
"step": 17650
},
{
"epoch": 0.7422338404039768,
"grad_norm": 0.6171875,
"learning_rate": 1.0400492482154694e-05,
"loss": 0.9808,
"step": 17675
},
{
"epoch": 0.7432836761046896,
"grad_norm": 0.6328125,
"learning_rate": 1.0180580163964847e-05,
"loss": 0.9399,
"step": 17700
},
{
"epoch": 0.7443335118054024,
"grad_norm": 0.427734375,
"learning_rate": 9.962936025419754e-06,
"loss": 0.9044,
"step": 17725
},
{
"epoch": 0.7453833475061153,
"grad_norm": 0.455078125,
"learning_rate": 9.747563597187791e-06,
"loss": 0.9535,
"step": 17750
},
{
"epoch": 0.7464331832068282,
"grad_norm": 0.6171875,
"learning_rate": 9.534466373085176e-06,
"loss": 1.2258,
"step": 17775
},
{
"epoch": 0.747483018907541,
"grad_norm": 0.625,
"learning_rate": 9.323647810019353e-06,
"loss": 1.1561,
"step": 17800
},
{
"epoch": 0.7485328546082538,
"grad_norm": 0.55859375,
"learning_rate": 9.115111327932717e-06,
"loss": 1.0656,
"step": 17825
},
{
"epoch": 0.7495826903089666,
"grad_norm": 0.6328125,
"learning_rate": 8.908860309747429e-06,
"loss": 1.1332,
"step": 17850
},
{
"epoch": 0.7506325260096794,
"grad_norm": 0.515625,
"learning_rate": 8.70489810131027e-06,
"loss": 1.1022,
"step": 17875
},
{
"epoch": 0.7516823617103924,
"grad_norm": 0.365234375,
"learning_rate": 8.503228011338525e-06,
"loss": 0.9377,
"step": 17900
},
{
"epoch": 0.7527321974111052,
"grad_norm": 0.59765625,
"learning_rate": 8.30385331136627e-06,
"loss": 1.1226,
"step": 17925
},
{
"epoch": 0.753782033111818,
"grad_norm": 0.54296875,
"learning_rate": 8.10677723569122e-06,
"loss": 1.0944,
"step": 17950
},
{
"epoch": 0.7548318688125308,
"grad_norm": 0.5546875,
"learning_rate": 7.912002981322396e-06,
"loss": 1.0833,
"step": 17975
},
{
"epoch": 0.7558817045132437,
"grad_norm": 0.455078125,
"learning_rate": 7.719533707928177e-06,
"loss": 1.042,
"step": 18000
},
{
"epoch": 0.7569315402139565,
"grad_norm": 0.447265625,
"learning_rate": 7.529372537785034e-06,
"loss": 0.9277,
"step": 18025
},
{
"epoch": 0.7579813759146694,
"grad_norm": 0.5078125,
"learning_rate": 7.34152255572697e-06,
"loss": 0.9699,
"step": 18050
},
{
"epoch": 0.7590312116153822,
"grad_norm": 0.5546875,
"learning_rate": 7.155986809095332e-06,
"loss": 1.0953,
"step": 18075
},
{
"epoch": 0.760081047316095,
"grad_norm": 0.6171875,
"learning_rate": 6.9727683076895635e-06,
"loss": 1.0535,
"step": 18100
},
{
"epoch": 0.7611308830168079,
"grad_norm": 0.5859375,
"learning_rate": 6.79187002371816e-06,
"loss": 1.1895,
"step": 18125
},
{
"epoch": 0.7621807187175207,
"grad_norm": 0.51953125,
"learning_rate": 6.613294891750642e-06,
"loss": 0.89,
"step": 18150
},
{
"epoch": 0.7632305544182335,
"grad_norm": 0.42578125,
"learning_rate": 6.437045808669894e-06,
"loss": 1.1992,
"step": 18175
},
{
"epoch": 0.7642803901189464,
"grad_norm": 0.43359375,
"learning_rate": 6.263125633625066e-06,
"loss": 1.161,
"step": 18200
},
{
"epoch": 0.7653302258196593,
"grad_norm": 0.46484375,
"learning_rate": 6.091537187985379e-06,
"loss": 1.234,
"step": 18225
},
{
"epoch": 0.7663800615203721,
"grad_norm": 0.64453125,
"learning_rate": 5.9222832552941635e-06,
"loss": 0.9988,
"step": 18250
},
{
"epoch": 0.7674298972210849,
"grad_norm": 0.578125,
"learning_rate": 5.755366581223858e-06,
"loss": 0.9656,
"step": 18275
},
{
"epoch": 0.7684797329217977,
"grad_norm": 0.4453125,
"learning_rate": 5.590789873531398e-06,
"loss": 1.2579,
"step": 18300
},
{
"epoch": 0.7695295686225105,
"grad_norm": 0.640625,
"learning_rate": 5.428555802014284e-06,
"loss": 0.915,
"step": 18325
},
{
"epoch": 0.7705794043232234,
"grad_norm": 0.7578125,
"learning_rate": 5.268666998467297e-06,
"loss": 1.2215,
"step": 18350
},
{
"epoch": 0.7716292400239363,
"grad_norm": 0.6328125,
"learning_rate": 5.11112605663977e-06,
"loss": 1.142,
"step": 18375
},
{
"epoch": 0.7726790757246491,
"grad_norm": 0.490234375,
"learning_rate": 4.9559355321935665e-06,
"loss": 0.9578,
"step": 18400
},
{
"epoch": 0.7737289114253619,
"grad_norm": 0.60546875,
"learning_rate": 4.803097942661588e-06,
"loss": 1.232,
"step": 18425
},
{
"epoch": 0.7747787471260748,
"grad_norm": 0.828125,
"learning_rate": 4.652615767406914e-06,
"loss": 0.9323,
"step": 18450
},
{
"epoch": 0.7758285828267876,
"grad_norm": 0.451171875,
"learning_rate": 4.5044914475826255e-06,
"loss": 0.9519,
"step": 18475
},
{
"epoch": 0.7768784185275004,
"grad_norm": 0.66015625,
"learning_rate": 4.358727386092198e-06,
"loss": 1.1287,
"step": 18500
},
{
"epoch": 0.7779282542282133,
"grad_norm": 0.60546875,
"learning_rate": 4.215325947550452e-06,
"loss": 1.0316,
"step": 18525
},
{
"epoch": 0.7789780899289261,
"grad_norm": 0.51953125,
"learning_rate": 4.0742894582453335e-06,
"loss": 0.9982,
"step": 18550
},
{
"epoch": 0.780027925629639,
"grad_norm": 0.3515625,
"learning_rate": 3.935620206099976e-06,
"loss": 1.1033,
"step": 18575
},
{
"epoch": 0.7810777613303518,
"grad_norm": 0.63671875,
"learning_rate": 3.7993204406358336e-06,
"loss": 1.0477,
"step": 18600
},
{
"epoch": 0.7821275970310646,
"grad_norm": 0.5078125,
"learning_rate": 3.6653923729359214e-06,
"loss": 1.1773,
"step": 18625
},
{
"epoch": 0.7831774327317774,
"grad_norm": 0.50390625,
"learning_rate": 3.5338381756091827e-06,
"loss": 1.0592,
"step": 18650
},
{
"epoch": 0.7842272684324904,
"grad_norm": 0.66015625,
"learning_rate": 3.4046599827551147e-06,
"loss": 1.1306,
"step": 18675
},
{
"epoch": 0.7852771041332032,
"grad_norm": 0.5,
"learning_rate": 3.2778598899291465e-06,
"loss": 1.0932,
"step": 18700
},
{
"epoch": 0.786326939833916,
"grad_norm": 0.73046875,
"learning_rate": 3.153439954108733e-06,
"loss": 1.1019,
"step": 18725
},
{
"epoch": 0.7873767755346288,
"grad_norm": 0.578125,
"learning_rate": 3.0314021936598485e-06,
"loss": 1.163,
"step": 18750
},
{
"epoch": 0.7884266112353416,
"grad_norm": 0.5,
"learning_rate": 2.9117485883043966e-06,
"loss": 1.0679,
"step": 18775
},
{
"epoch": 0.7894764469360545,
"grad_norm": 0.314453125,
"learning_rate": 2.7944810790879845e-06,
"loss": 1.1934,
"step": 18800
},
{
"epoch": 0.7905262826367674,
"grad_norm": 0.60546875,
"learning_rate": 2.679601568348483e-06,
"loss": 1.1539,
"step": 18825
},
{
"epoch": 0.7915761183374802,
"grad_norm": 0.49609375,
"learning_rate": 2.567111919685183e-06,
"loss": 0.9144,
"step": 18850
},
{
"epoch": 0.792625954038193,
"grad_norm": 0.25390625,
"learning_rate": 2.457013957928472e-06,
"loss": 0.9195,
"step": 18875
},
{
"epoch": 0.7936757897389058,
"grad_norm": 0.3828125,
"learning_rate": 2.349309469110339e-06,
"loss": 1.0865,
"step": 18900
},
{
"epoch": 0.7947256254396187,
"grad_norm": 0.396484375,
"learning_rate": 2.2440002004353808e-06,
"loss": 0.9601,
"step": 18925
},
{
"epoch": 0.7957754611403315,
"grad_norm": 0.375,
"learning_rate": 2.1410878602523774e-06,
"loss": 0.9969,
"step": 18950
},
{
"epoch": 0.7968252968410444,
"grad_norm": 0.42578125,
"learning_rate": 2.0405741180266776e-06,
"loss": 1.1202,
"step": 18975
},
{
"epoch": 0.7978751325417572,
"grad_norm": 0.61328125,
"learning_rate": 1.942460604313073e-06,
"loss": 1.1523,
"step": 19000
},
{
"epoch": 0.7989249682424701,
"grad_norm": 0.5546875,
"learning_rate": 1.8467489107293509e-06,
"loss": 0.9934,
"step": 19025
},
{
"epoch": 0.7999748039431829,
"grad_norm": 0.52734375,
"learning_rate": 1.7534405899304494e-06,
"loss": 1.1894,
"step": 19050
},
{
"epoch": 0.8010246396438957,
"grad_norm": 0.66015625,
"learning_rate": 1.66253715558331e-06,
"loss": 0.8799,
"step": 19075
},
{
"epoch": 0.8020744753446085,
"grad_norm": 0.53515625,
"learning_rate": 1.5740400823423316e-06,
"loss": 1.0298,
"step": 19100
},
{
"epoch": 0.8031243110453214,
"grad_norm": 0.5390625,
"learning_rate": 1.4879508058253886e-06,
"loss": 1.0023,
"step": 19125
},
{
"epoch": 0.8041741467460343,
"grad_norm": 0.6171875,
"learning_rate": 1.4042707225906002e-06,
"loss": 1.0075,
"step": 19150
},
{
"epoch": 0.8052239824467471,
"grad_norm": 0.5625,
"learning_rate": 1.3230011901136317e-06,
"loss": 1.0803,
"step": 19175
},
{
"epoch": 0.8062738181474599,
"grad_norm": 0.46484375,
"learning_rate": 1.2441435267656951e-06,
"loss": 1.1006,
"step": 19200
},
{
"epoch": 0.8073236538481727,
"grad_norm": 0.546875,
"learning_rate": 1.1676990117922002e-06,
"loss": 0.9258,
"step": 19225
},
{
"epoch": 0.8083734895488855,
"grad_norm": 0.56640625,
"learning_rate": 1.093668885291904e-06,
"loss": 0.9101,
"step": 19250
},
{
"epoch": 0.8094233252495985,
"grad_norm": 0.484375,
"learning_rate": 1.0220543481968936e-06,
"loss": 1.0387,
"step": 19275
},
{
"epoch": 0.8104731609503113,
"grad_norm": 0.48828125,
"learning_rate": 9.528565622530515e-07,
"loss": 1.0057,
"step": 19300
},
{
"epoch": 0.8115229966510241,
"grad_norm": 0.447265625,
"learning_rate": 8.860766500012217e-07,
"loss": 1.1879,
"step": 19325
},
{
"epoch": 0.8125728323517369,
"grad_norm": 0.328125,
"learning_rate": 8.217156947590064e-07,
"loss": 1.1634,
"step": 19350
},
{
"epoch": 0.8136226680524498,
"grad_norm": 0.185546875,
"learning_rate": 7.59774740603164e-07,
"loss": 0.8734,
"step": 19375
},
{
"epoch": 0.8146725037531626,
"grad_norm": 0.34375,
"learning_rate": 7.002547923527058e-07,
"loss": 0.9525,
"step": 19400
},
{
"epoch": 0.8157223394538755,
"grad_norm": 0.53515625,
"learning_rate": 6.43156815552609e-07,
"loss": 1.2432,
"step": 19425
},
{
"epoch": 0.8167721751545883,
"grad_norm": 0.6796875,
"learning_rate": 5.884817364580629e-07,
"loss": 1.076,
"step": 19450
},
{
"epoch": 0.8178220108553012,
"grad_norm": 0.5859375,
"learning_rate": 5.362304420195973e-07,
"loss": 0.9748,
"step": 19475
},
{
"epoch": 0.818871846556014,
"grad_norm": 0.59765625,
"learning_rate": 4.864037798685105e-07,
"loss": 1.061,
"step": 19500
},
{
"epoch": 0.8199216822567268,
"grad_norm": 0.57421875,
"learning_rate": 4.390025583032641e-07,
"loss": 1.1153,
"step": 19525
},
{
"epoch": 0.8209715179574396,
"grad_norm": 0.6015625,
"learning_rate": 3.9402754627634316e-07,
"loss": 1.1485,
"step": 19550
},
{
"epoch": 0.8220213536581525,
"grad_norm": 0.59765625,
"learning_rate": 3.514794733816828e-07,
"loss": 0.8816,
"step": 19575
},
{
"epoch": 0.8230711893588654,
"grad_norm": 0.58984375,
"learning_rate": 3.11359029842978e-07,
"loss": 1.2177,
"step": 19600
},
{
"epoch": 0.8241210250595782,
"grad_norm": 0.341796875,
"learning_rate": 2.7366686650237556e-07,
"loss": 1.029,
"step": 19625
},
{
"epoch": 0.825170860760291,
"grad_norm": 0.52734375,
"learning_rate": 2.3840359480999937e-07,
"loss": 1.152,
"step": 19650
},
{
"epoch": 0.8262206964610038,
"grad_norm": 0.0849609375,
"learning_rate": 2.05569786813925e-07,
"loss": 1.0786,
"step": 19675
},
{
"epoch": 0.8272705321617166,
"grad_norm": 0.404296875,
"learning_rate": 1.7516597515100372e-07,
"loss": 0.9784,
"step": 19700
},
{
"epoch": 0.8283203678624296,
"grad_norm": 0.283203125,
"learning_rate": 1.4719265303821948e-07,
"loss": 0.8756,
"step": 19725
},
{
"epoch": 0.8293702035631424,
"grad_norm": 0.6015625,
"learning_rate": 1.2165027426456198e-07,
"loss": 1.007,
"step": 19750
},
{
"epoch": 0.8304200392638552,
"grad_norm": 0.58203125,
"learning_rate": 9.853925318381584e-08,
"loss": 1.1715,
"step": 19775
},
{
"epoch": 0.831469874964568,
"grad_norm": 0.2236328125,
"learning_rate": 7.785996470776601e-08,
"loss": 1.0351,
"step": 19800
},
{
"epoch": 0.8325197106652809,
"grad_norm": 0.546875,
"learning_rate": 5.961274430010265e-08,
"loss": 1.1492,
"step": 19825
},
{
"epoch": 0.8335695463659937,
"grad_norm": 0.5078125,
"learning_rate": 4.379788797105877e-08,
"loss": 1.0075,
"step": 19850
},
{
"epoch": 0.8346193820667066,
"grad_norm": 0.345703125,
"learning_rate": 3.041565227248077e-08,
"loss": 1.0898,
"step": 19875
},
{
"epoch": 0.8356692177674194,
"grad_norm": 0.63671875,
"learning_rate": 1.9466254293798445e-08,
"loss": 1.1322,
"step": 19900
},
{
"epoch": 0.8367190534681322,
"grad_norm": 0.57421875,
"learning_rate": 1.0949871658394449e-08,
"loss": 1.0682,
"step": 19925
},
{
"epoch": 0.8377688891688451,
"grad_norm": 0.1103515625,
"learning_rate": 4.866642520806596e-09,
"loss": 1.1459,
"step": 19950
},
{
"epoch": 0.8388187248695579,
"grad_norm": 0.38671875,
"learning_rate": 1.2166655644463286e-09,
"loss": 1.0392,
"step": 19975
},
{
"epoch": 0.8398685605702707,
"grad_norm": 0.5234375,
"learning_rate": 0.0,
"loss": 1.0401,
"step": 20000
}
],
"logging_steps": 25,
"max_steps": 20000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.60095410782208e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}