1785 lines
52 KiB
JSON
1785 lines
52 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 2.0,
|
|
"eval_steps": 500,
|
|
"global_step": 1750,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 1.6696195736527444,
|
|
"epoch": 0.011428571428571429,
|
|
"grad_norm": 3.7833805084228516,
|
|
"learning_rate": 1.9931428571428572e-05,
|
|
"loss": 1.3203315734863281,
|
|
"mean_token_accuracy": 0.7056225836277008,
|
|
"num_tokens": 157266.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 1.7283091336488723,
|
|
"epoch": 0.022857142857142857,
|
|
"grad_norm": 3.4964022636413574,
|
|
"learning_rate": 1.9855238095238097e-05,
|
|
"loss": 1.4036809921264648,
|
|
"mean_token_accuracy": 0.6812524735927582,
|
|
"num_tokens": 300430.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 1.7663277059793472,
|
|
"epoch": 0.03428571428571429,
|
|
"grad_norm": 4.590687274932861,
|
|
"learning_rate": 1.977904761904762e-05,
|
|
"loss": 1.319040012359619,
|
|
"mean_token_accuracy": 0.6928375385701656,
|
|
"num_tokens": 456129.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 1.7699106186628342,
|
|
"epoch": 0.045714285714285714,
|
|
"grad_norm": 3.2531967163085938,
|
|
"learning_rate": 1.9702857142857144e-05,
|
|
"loss": 1.3776089668273925,
|
|
"mean_token_accuracy": 0.686311125010252,
|
|
"num_tokens": 615179.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 1.7856220349669456,
|
|
"epoch": 0.05714285714285714,
|
|
"grad_norm": 3.356832981109619,
|
|
"learning_rate": 1.9626666666666666e-05,
|
|
"loss": 1.32177791595459,
|
|
"mean_token_accuracy": 0.6952089086174965,
|
|
"num_tokens": 769672.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 1.7628433778882027,
|
|
"epoch": 0.06857142857142857,
|
|
"grad_norm": 3.965970754623413,
|
|
"learning_rate": 1.955047619047619e-05,
|
|
"loss": 1.338463306427002,
|
|
"mean_token_accuracy": 0.6953431364148855,
|
|
"num_tokens": 918402.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 1.7276440173387528,
|
|
"epoch": 0.08,
|
|
"grad_norm": 3.4119951725006104,
|
|
"learning_rate": 1.9474285714285717e-05,
|
|
"loss": 1.2838379859924316,
|
|
"mean_token_accuracy": 0.6987539391964674,
|
|
"num_tokens": 1076598.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 1.748648078739643,
|
|
"epoch": 0.09142857142857143,
|
|
"grad_norm": 3.6497740745544434,
|
|
"learning_rate": 1.9398095238095242e-05,
|
|
"loss": 1.3314701080322267,
|
|
"mean_token_accuracy": 0.695335328578949,
|
|
"num_tokens": 1228084.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 1.7111000671982766,
|
|
"epoch": 0.10285714285714286,
|
|
"grad_norm": 3.3122711181640625,
|
|
"learning_rate": 1.9321904761904764e-05,
|
|
"loss": 1.3027889251708984,
|
|
"mean_token_accuracy": 0.6987856134772301,
|
|
"num_tokens": 1386707.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 1.7278845444321633,
|
|
"epoch": 0.11428571428571428,
|
|
"grad_norm": 3.183835744857788,
|
|
"learning_rate": 1.924571428571429e-05,
|
|
"loss": 1.282834243774414,
|
|
"mean_token_accuracy": 0.6966647863388061,
|
|
"num_tokens": 1535406.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 1.7132800757884978,
|
|
"epoch": 0.12571428571428572,
|
|
"grad_norm": 3.4031126499176025,
|
|
"learning_rate": 1.916952380952381e-05,
|
|
"loss": 1.2409746170043945,
|
|
"mean_token_accuracy": 0.7101543735712766,
|
|
"num_tokens": 1680743.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 1.7373003125190736,
|
|
"epoch": 0.13714285714285715,
|
|
"grad_norm": 3.4207799434661865,
|
|
"learning_rate": 1.9093333333333336e-05,
|
|
"loss": 1.2909335136413573,
|
|
"mean_token_accuracy": 0.701117354631424,
|
|
"num_tokens": 1834014.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 1.7092810094356536,
|
|
"epoch": 0.14857142857142858,
|
|
"grad_norm": 3.3323278427124023,
|
|
"learning_rate": 1.9017142857142858e-05,
|
|
"loss": 1.3009337425231933,
|
|
"mean_token_accuracy": 0.6907773412764072,
|
|
"num_tokens": 1989349.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 1.7240261450409888,
|
|
"epoch": 0.16,
|
|
"grad_norm": 2.9499671459198,
|
|
"learning_rate": 1.8940952380952383e-05,
|
|
"loss": 1.2700564384460449,
|
|
"mean_token_accuracy": 0.7077916815876961,
|
|
"num_tokens": 2139614.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 1.655298152565956,
|
|
"epoch": 0.17142857142857143,
|
|
"grad_norm": 3.1871607303619385,
|
|
"learning_rate": 1.8864761904761905e-05,
|
|
"loss": 1.2251598358154296,
|
|
"mean_token_accuracy": 0.7074991308152676,
|
|
"num_tokens": 2293879.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 1.665707252919674,
|
|
"epoch": 0.18285714285714286,
|
|
"grad_norm": 3.290358304977417,
|
|
"learning_rate": 1.878857142857143e-05,
|
|
"loss": 1.2641545295715333,
|
|
"mean_token_accuracy": 0.7021077655255794,
|
|
"num_tokens": 2446447.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 1.7259689077734948,
|
|
"epoch": 0.19428571428571428,
|
|
"grad_norm": 3.1401703357696533,
|
|
"learning_rate": 1.871238095238095e-05,
|
|
"loss": 1.315788745880127,
|
|
"mean_token_accuracy": 0.6926948539912701,
|
|
"num_tokens": 2595389.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 1.7142695009708404,
|
|
"epoch": 0.2057142857142857,
|
|
"grad_norm": 3.4904985427856445,
|
|
"learning_rate": 1.8636190476190477e-05,
|
|
"loss": 1.2755393028259276,
|
|
"mean_token_accuracy": 0.7040164832025766,
|
|
"num_tokens": 2745608.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 1.6242914512753486,
|
|
"epoch": 0.21714285714285714,
|
|
"grad_norm": 2.949880361557007,
|
|
"learning_rate": 1.8560000000000002e-05,
|
|
"loss": 1.190676498413086,
|
|
"mean_token_accuracy": 0.7185754984617233,
|
|
"num_tokens": 2894625.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 1.7364101350307464,
|
|
"epoch": 0.22857142857142856,
|
|
"grad_norm": 3.125112533569336,
|
|
"learning_rate": 1.8483809523809527e-05,
|
|
"loss": 1.2999603271484375,
|
|
"mean_token_accuracy": 0.6976742841303348,
|
|
"num_tokens": 3046487.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 1.6772031649947166,
|
|
"epoch": 0.24,
|
|
"grad_norm": 2.813882350921631,
|
|
"learning_rate": 1.840761904761905e-05,
|
|
"loss": 1.2316917419433593,
|
|
"mean_token_accuracy": 0.7120772495865821,
|
|
"num_tokens": 3196704.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 1.6737122178077697,
|
|
"epoch": 0.25142857142857145,
|
|
"grad_norm": 2.8218936920166016,
|
|
"learning_rate": 1.8331428571428574e-05,
|
|
"loss": 1.274891471862793,
|
|
"mean_token_accuracy": 0.6991564333438873,
|
|
"num_tokens": 3348365.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 1.6344200998544693,
|
|
"epoch": 0.26285714285714284,
|
|
"grad_norm": 2.9688668251037598,
|
|
"learning_rate": 1.8255238095238096e-05,
|
|
"loss": 1.2082077026367188,
|
|
"mean_token_accuracy": 0.7141494430601597,
|
|
"num_tokens": 3509464.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 1.6166698306798934,
|
|
"epoch": 0.2742857142857143,
|
|
"grad_norm": 3.007690668106079,
|
|
"learning_rate": 1.817904761904762e-05,
|
|
"loss": 1.1915247917175293,
|
|
"mean_token_accuracy": 0.7145515654236079,
|
|
"num_tokens": 3661779.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 1.6418054014444352,
|
|
"epoch": 0.2857142857142857,
|
|
"grad_norm": 2.9565091133117676,
|
|
"learning_rate": 1.8102857142857143e-05,
|
|
"loss": 1.2041254043579102,
|
|
"mean_token_accuracy": 0.7189058735966682,
|
|
"num_tokens": 3809611.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 1.6628487035632133,
|
|
"epoch": 0.29714285714285715,
|
|
"grad_norm": 2.7851827144622803,
|
|
"learning_rate": 1.8026666666666668e-05,
|
|
"loss": 1.2013505935668944,
|
|
"mean_token_accuracy": 0.7156483136117459,
|
|
"num_tokens": 3960501.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 1.6495753720402717,
|
|
"epoch": 0.30857142857142855,
|
|
"grad_norm": 3.0212292671203613,
|
|
"learning_rate": 1.7950476190476193e-05,
|
|
"loss": 1.2240056991577148,
|
|
"mean_token_accuracy": 0.7174453124403953,
|
|
"num_tokens": 4107783.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 1.693277756869793,
|
|
"epoch": 0.32,
|
|
"grad_norm": 2.9716391563415527,
|
|
"learning_rate": 1.7874285714285715e-05,
|
|
"loss": 1.207960319519043,
|
|
"mean_token_accuracy": 0.7157351300120354,
|
|
"num_tokens": 4252335.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 1.6619624510407447,
|
|
"epoch": 0.3314285714285714,
|
|
"grad_norm": 2.851775646209717,
|
|
"learning_rate": 1.779809523809524e-05,
|
|
"loss": 1.248055648803711,
|
|
"mean_token_accuracy": 0.7128231346607208,
|
|
"num_tokens": 4407579.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 1.6617254123091698,
|
|
"epoch": 0.34285714285714286,
|
|
"grad_norm": 3.254999876022339,
|
|
"learning_rate": 1.7721904761904762e-05,
|
|
"loss": 1.184535598754883,
|
|
"mean_token_accuracy": 0.7231362499296665,
|
|
"num_tokens": 4551517.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 1.5988039791584014,
|
|
"epoch": 0.35428571428571426,
|
|
"grad_norm": 3.110564708709717,
|
|
"learning_rate": 1.7645714285714287e-05,
|
|
"loss": 1.148481559753418,
|
|
"mean_token_accuracy": 0.7262889519333839,
|
|
"num_tokens": 4708257.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 1.6530713498592378,
|
|
"epoch": 0.3657142857142857,
|
|
"grad_norm": 3.1247048377990723,
|
|
"learning_rate": 1.7569523809523812e-05,
|
|
"loss": 1.2834845542907716,
|
|
"mean_token_accuracy": 0.702876628190279,
|
|
"num_tokens": 4859075.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 1.6123583316802979,
|
|
"epoch": 0.37714285714285717,
|
|
"grad_norm": 2.90073823928833,
|
|
"learning_rate": 1.7493333333333334e-05,
|
|
"loss": 1.1132370948791503,
|
|
"mean_token_accuracy": 0.7313019946217537,
|
|
"num_tokens": 5011395.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 1.618579277396202,
|
|
"epoch": 0.38857142857142857,
|
|
"grad_norm": 2.936490058898926,
|
|
"learning_rate": 1.741714285714286e-05,
|
|
"loss": 1.2263901710510254,
|
|
"mean_token_accuracy": 0.7069775387644768,
|
|
"num_tokens": 5174272.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 1.663301782310009,
|
|
"epoch": 0.4,
|
|
"grad_norm": 3.241185426712036,
|
|
"learning_rate": 1.734095238095238e-05,
|
|
"loss": 1.2628487586975097,
|
|
"mean_token_accuracy": 0.7058808147907257,
|
|
"num_tokens": 5327670.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 1.6857714764773846,
|
|
"epoch": 0.4114285714285714,
|
|
"grad_norm": 2.9059698581695557,
|
|
"learning_rate": 1.7264761904761906e-05,
|
|
"loss": 1.2412065505981444,
|
|
"mean_token_accuracy": 0.7062458634376526,
|
|
"num_tokens": 5476064.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 1.6253925159573555,
|
|
"epoch": 0.4228571428571429,
|
|
"grad_norm": 2.9899251461029053,
|
|
"learning_rate": 1.718857142857143e-05,
|
|
"loss": 1.210451316833496,
|
|
"mean_token_accuracy": 0.7089586913585663,
|
|
"num_tokens": 5636828.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 1.5880890995264054,
|
|
"epoch": 0.4342857142857143,
|
|
"grad_norm": 3.0461251735687256,
|
|
"learning_rate": 1.7112380952380953e-05,
|
|
"loss": 1.1992740631103516,
|
|
"mean_token_accuracy": 0.7144776731729507,
|
|
"num_tokens": 5788462.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 1.6384857930243015,
|
|
"epoch": 0.44571428571428573,
|
|
"grad_norm": 2.8842580318450928,
|
|
"learning_rate": 1.703619047619048e-05,
|
|
"loss": 1.2124631881713868,
|
|
"mean_token_accuracy": 0.7115042388439179,
|
|
"num_tokens": 5943108.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 1.5625910967588426,
|
|
"epoch": 0.45714285714285713,
|
|
"grad_norm": 3.0432047843933105,
|
|
"learning_rate": 1.696e-05,
|
|
"loss": 1.1515995025634767,
|
|
"mean_token_accuracy": 0.716732618585229,
|
|
"num_tokens": 6098679.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 1.6026007533073425,
|
|
"epoch": 0.4685714285714286,
|
|
"grad_norm": 2.781804323196411,
|
|
"learning_rate": 1.6883809523809525e-05,
|
|
"loss": 1.180698299407959,
|
|
"mean_token_accuracy": 0.7132757268846035,
|
|
"num_tokens": 6250301.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 1.5140968441963196,
|
|
"epoch": 0.48,
|
|
"grad_norm": 2.9592671394348145,
|
|
"learning_rate": 1.6807619047619047e-05,
|
|
"loss": 1.1643908500671387,
|
|
"mean_token_accuracy": 0.7223096333444119,
|
|
"num_tokens": 6407356.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 1.6122839108109475,
|
|
"epoch": 0.49142857142857144,
|
|
"grad_norm": 3.2809886932373047,
|
|
"learning_rate": 1.6731428571428572e-05,
|
|
"loss": 1.2397551536560059,
|
|
"mean_token_accuracy": 0.707809828966856,
|
|
"num_tokens": 6559819.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 1.603949698805809,
|
|
"epoch": 0.5028571428571429,
|
|
"grad_norm": 2.721527338027954,
|
|
"learning_rate": 1.6655238095238098e-05,
|
|
"loss": 1.1895122528076172,
|
|
"mean_token_accuracy": 0.7227885220199823,
|
|
"num_tokens": 6716227.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 1.582472313940525,
|
|
"epoch": 0.5142857142857142,
|
|
"grad_norm": 2.8015570640563965,
|
|
"learning_rate": 1.6579047619047623e-05,
|
|
"loss": 1.1144920349121095,
|
|
"mean_token_accuracy": 0.7284132238477469,
|
|
"num_tokens": 6867486.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 1.6025653272867202,
|
|
"epoch": 0.5257142857142857,
|
|
"grad_norm": 3.1553401947021484,
|
|
"learning_rate": 1.6502857142857145e-05,
|
|
"loss": 1.1947656631469727,
|
|
"mean_token_accuracy": 0.7160489454865455,
|
|
"num_tokens": 7012179.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 1.5280416823923588,
|
|
"epoch": 0.5371428571428571,
|
|
"grad_norm": 2.97623348236084,
|
|
"learning_rate": 1.642666666666667e-05,
|
|
"loss": 1.149429702758789,
|
|
"mean_token_accuracy": 0.7256167627871036,
|
|
"num_tokens": 7171638.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 1.6620354026556015,
|
|
"epoch": 0.5485714285714286,
|
|
"grad_norm": 3.1873958110809326,
|
|
"learning_rate": 1.635047619047619e-05,
|
|
"loss": 1.2588795661926269,
|
|
"mean_token_accuracy": 0.707894566655159,
|
|
"num_tokens": 7322167.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 1.581991147994995,
|
|
"epoch": 0.56,
|
|
"grad_norm": 3.385500907897949,
|
|
"learning_rate": 1.6274285714285717e-05,
|
|
"loss": 1.2005508422851563,
|
|
"mean_token_accuracy": 0.7142980419099331,
|
|
"num_tokens": 7466661.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 1.6193878114223481,
|
|
"epoch": 0.5714285714285714,
|
|
"grad_norm": 2.9256980419158936,
|
|
"learning_rate": 1.619809523809524e-05,
|
|
"loss": 1.1964147567749024,
|
|
"mean_token_accuracy": 0.715008432790637,
|
|
"num_tokens": 7620522.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"entropy": 1.560711294412613,
|
|
"epoch": 0.5828571428571429,
|
|
"grad_norm": 2.548938035964966,
|
|
"learning_rate": 1.6121904761904764e-05,
|
|
"loss": 1.174423599243164,
|
|
"mean_token_accuracy": 0.7205559216439724,
|
|
"num_tokens": 7772467.0,
|
|
"step": 510
|
|
},
|
|
{
|
|
"entropy": 1.6210713252425193,
|
|
"epoch": 0.5942857142857143,
|
|
"grad_norm": 2.6339938640594482,
|
|
"learning_rate": 1.6045714285714286e-05,
|
|
"loss": 1.153661060333252,
|
|
"mean_token_accuracy": 0.7224138833582401,
|
|
"num_tokens": 7922030.0,
|
|
"step": 520
|
|
},
|
|
{
|
|
"entropy": 1.5822248332202435,
|
|
"epoch": 0.6057142857142858,
|
|
"grad_norm": 2.4934332370758057,
|
|
"learning_rate": 1.596952380952381e-05,
|
|
"loss": 1.1586255073547362,
|
|
"mean_token_accuracy": 0.7230928719043732,
|
|
"num_tokens": 8067955.0,
|
|
"step": 530
|
|
},
|
|
{
|
|
"entropy": 1.5642932564020158,
|
|
"epoch": 0.6171428571428571,
|
|
"grad_norm": 2.985651969909668,
|
|
"learning_rate": 1.5893333333333333e-05,
|
|
"loss": 1.1338199615478515,
|
|
"mean_token_accuracy": 0.7240421302616596,
|
|
"num_tokens": 8226369.0,
|
|
"step": 540
|
|
},
|
|
{
|
|
"entropy": 1.6006409794092178,
|
|
"epoch": 0.6285714285714286,
|
|
"grad_norm": 2.6047215461730957,
|
|
"learning_rate": 1.5817142857142858e-05,
|
|
"loss": 1.155271625518799,
|
|
"mean_token_accuracy": 0.7230217695236206,
|
|
"num_tokens": 8376285.0,
|
|
"step": 550
|
|
},
|
|
{
|
|
"entropy": 1.5791691079735757,
|
|
"epoch": 0.64,
|
|
"grad_norm": 2.855867385864258,
|
|
"learning_rate": 1.5740952380952383e-05,
|
|
"loss": 1.1597973823547363,
|
|
"mean_token_accuracy": 0.7209223508834839,
|
|
"num_tokens": 8530322.0,
|
|
"step": 560
|
|
},
|
|
{
|
|
"entropy": 1.6016968488693237,
|
|
"epoch": 0.6514285714285715,
|
|
"grad_norm": 3.205124855041504,
|
|
"learning_rate": 1.5664761904761908e-05,
|
|
"loss": 1.1405302047729493,
|
|
"mean_token_accuracy": 0.7231174889951945,
|
|
"num_tokens": 8676991.0,
|
|
"step": 570
|
|
},
|
|
{
|
|
"entropy": 1.5366713739931583,
|
|
"epoch": 0.6628571428571428,
|
|
"grad_norm": 3.1047000885009766,
|
|
"learning_rate": 1.558857142857143e-05,
|
|
"loss": 1.1293843269348145,
|
|
"mean_token_accuracy": 0.7271328650414943,
|
|
"num_tokens": 8833861.0,
|
|
"step": 580
|
|
},
|
|
{
|
|
"entropy": 1.5866449058055878,
|
|
"epoch": 0.6742857142857143,
|
|
"grad_norm": 2.884098529815674,
|
|
"learning_rate": 1.5512380952380955e-05,
|
|
"loss": 1.2017478942871094,
|
|
"mean_token_accuracy": 0.716648469120264,
|
|
"num_tokens": 8983438.0,
|
|
"step": 590
|
|
},
|
|
{
|
|
"entropy": 1.610584917664528,
|
|
"epoch": 0.6857142857142857,
|
|
"grad_norm": 2.7602009773254395,
|
|
"learning_rate": 1.5436190476190477e-05,
|
|
"loss": 1.1553681373596192,
|
|
"mean_token_accuracy": 0.72271120660007,
|
|
"num_tokens": 9138958.0,
|
|
"step": 600
|
|
},
|
|
{
|
|
"entropy": 1.5292311601340771,
|
|
"epoch": 0.6971428571428572,
|
|
"grad_norm": 2.8975393772125244,
|
|
"learning_rate": 1.5360000000000002e-05,
|
|
"loss": 1.1547229766845704,
|
|
"mean_token_accuracy": 0.7226926662027836,
|
|
"num_tokens": 9290314.0,
|
|
"step": 610
|
|
},
|
|
{
|
|
"entropy": 1.5577459454536438,
|
|
"epoch": 0.7085714285714285,
|
|
"grad_norm": 2.5054473876953125,
|
|
"learning_rate": 1.5283809523809524e-05,
|
|
"loss": 1.1239413261413573,
|
|
"mean_token_accuracy": 0.7288053043186664,
|
|
"num_tokens": 9442584.0,
|
|
"step": 620
|
|
},
|
|
{
|
|
"entropy": 1.6545007571578025,
|
|
"epoch": 0.72,
|
|
"grad_norm": 3.156757354736328,
|
|
"learning_rate": 1.5207619047619049e-05,
|
|
"loss": 1.2162582397460937,
|
|
"mean_token_accuracy": 0.7100601557642221,
|
|
"num_tokens": 9589146.0,
|
|
"step": 630
|
|
},
|
|
{
|
|
"entropy": 1.6293067336082458,
|
|
"epoch": 0.7314285714285714,
|
|
"grad_norm": 2.7666168212890625,
|
|
"learning_rate": 1.5131428571428572e-05,
|
|
"loss": 1.1888948440551759,
|
|
"mean_token_accuracy": 0.7159062504768372,
|
|
"num_tokens": 9736056.0,
|
|
"step": 640
|
|
},
|
|
{
|
|
"entropy": 1.5786321811378001,
|
|
"epoch": 0.7428571428571429,
|
|
"grad_norm": 2.6019463539123535,
|
|
"learning_rate": 1.5055238095238096e-05,
|
|
"loss": 1.0923637390136718,
|
|
"mean_token_accuracy": 0.7382036492228508,
|
|
"num_tokens": 9898410.0,
|
|
"step": 650
|
|
},
|
|
{
|
|
"entropy": 1.574275678396225,
|
|
"epoch": 0.7542857142857143,
|
|
"grad_norm": 2.7830586433410645,
|
|
"learning_rate": 1.497904761904762e-05,
|
|
"loss": 1.1398842811584473,
|
|
"mean_token_accuracy": 0.7246641829609871,
|
|
"num_tokens": 10051154.0,
|
|
"step": 660
|
|
},
|
|
{
|
|
"entropy": 1.5554455459117889,
|
|
"epoch": 0.7657142857142857,
|
|
"grad_norm": 2.8458149433135986,
|
|
"learning_rate": 1.4902857142857143e-05,
|
|
"loss": 1.139054298400879,
|
|
"mean_token_accuracy": 0.726585789769888,
|
|
"num_tokens": 10199110.0,
|
|
"step": 670
|
|
},
|
|
{
|
|
"entropy": 1.6094272822141646,
|
|
"epoch": 0.7771428571428571,
|
|
"grad_norm": 2.9835047721862793,
|
|
"learning_rate": 1.4826666666666666e-05,
|
|
"loss": 1.1978397369384766,
|
|
"mean_token_accuracy": 0.7123216778039932,
|
|
"num_tokens": 10344927.0,
|
|
"step": 680
|
|
},
|
|
{
|
|
"entropy": 1.5142188727855683,
|
|
"epoch": 0.7885714285714286,
|
|
"grad_norm": 2.7492172718048096,
|
|
"learning_rate": 1.475047619047619e-05,
|
|
"loss": 1.048453140258789,
|
|
"mean_token_accuracy": 0.7415693461894989,
|
|
"num_tokens": 10505422.0,
|
|
"step": 690
|
|
},
|
|
{
|
|
"entropy": 1.633454330265522,
|
|
"epoch": 0.8,
|
|
"grad_norm": 2.645536184310913,
|
|
"learning_rate": 1.4674285714285717e-05,
|
|
"loss": 1.1423879623413087,
|
|
"mean_token_accuracy": 0.7272885538637638,
|
|
"num_tokens": 10652367.0,
|
|
"step": 700
|
|
},
|
|
{
|
|
"entropy": 1.4862130410969256,
|
|
"epoch": 0.8114285714285714,
|
|
"grad_norm": 2.473473310470581,
|
|
"learning_rate": 1.459809523809524e-05,
|
|
"loss": 1.019435214996338,
|
|
"mean_token_accuracy": 0.7454098269343377,
|
|
"num_tokens": 10812282.0,
|
|
"step": 710
|
|
},
|
|
{
|
|
"entropy": 1.541966749727726,
|
|
"epoch": 0.8228571428571428,
|
|
"grad_norm": 2.5247135162353516,
|
|
"learning_rate": 1.4521904761904764e-05,
|
|
"loss": 1.0726137161254883,
|
|
"mean_token_accuracy": 0.7371681958436966,
|
|
"num_tokens": 10967866.0,
|
|
"step": 720
|
|
},
|
|
{
|
|
"entropy": 1.5219792507588863,
|
|
"epoch": 0.8342857142857143,
|
|
"grad_norm": 2.35876202583313,
|
|
"learning_rate": 1.4445714285714287e-05,
|
|
"loss": 1.0680880546569824,
|
|
"mean_token_accuracy": 0.7386457361280918,
|
|
"num_tokens": 11120412.0,
|
|
"step": 730
|
|
},
|
|
{
|
|
"entropy": 1.506186830997467,
|
|
"epoch": 0.8457142857142858,
|
|
"grad_norm": 2.9589080810546875,
|
|
"learning_rate": 1.436952380952381e-05,
|
|
"loss": 1.0974149703979492,
|
|
"mean_token_accuracy": 0.7378807857632637,
|
|
"num_tokens": 11277244.0,
|
|
"step": 740
|
|
},
|
|
{
|
|
"entropy": 1.569691103696823,
|
|
"epoch": 0.8571428571428571,
|
|
"grad_norm": 2.5306074619293213,
|
|
"learning_rate": 1.4293333333333334e-05,
|
|
"loss": 1.177268123626709,
|
|
"mean_token_accuracy": 0.7250622183084487,
|
|
"num_tokens": 11436007.0,
|
|
"step": 750
|
|
},
|
|
{
|
|
"entropy": 1.5297324053943158,
|
|
"epoch": 0.8685714285714285,
|
|
"grad_norm": 2.7663497924804688,
|
|
"learning_rate": 1.4217142857142858e-05,
|
|
"loss": 1.1369609832763672,
|
|
"mean_token_accuracy": 0.7221393447369338,
|
|
"num_tokens": 11587615.0,
|
|
"step": 760
|
|
},
|
|
{
|
|
"entropy": 1.4932486459612846,
|
|
"epoch": 0.88,
|
|
"grad_norm": 2.5511255264282227,
|
|
"learning_rate": 1.4140952380952381e-05,
|
|
"loss": 1.1008304595947265,
|
|
"mean_token_accuracy": 0.7303555272519588,
|
|
"num_tokens": 11747686.0,
|
|
"step": 770
|
|
},
|
|
{
|
|
"entropy": 1.5960537254810334,
|
|
"epoch": 0.8914285714285715,
|
|
"grad_norm": 2.450721025466919,
|
|
"learning_rate": 1.4064761904761905e-05,
|
|
"loss": 1.1708695411682128,
|
|
"mean_token_accuracy": 0.7219238199293614,
|
|
"num_tokens": 11904676.0,
|
|
"step": 780
|
|
},
|
|
{
|
|
"entropy": 1.5482905827462674,
|
|
"epoch": 0.9028571428571428,
|
|
"grad_norm": 3.1114189624786377,
|
|
"learning_rate": 1.398857142857143e-05,
|
|
"loss": 1.184025478363037,
|
|
"mean_token_accuracy": 0.720519745349884,
|
|
"num_tokens": 12059825.0,
|
|
"step": 790
|
|
},
|
|
{
|
|
"entropy": 1.5766600042581558,
|
|
"epoch": 0.9142857142857143,
|
|
"grad_norm": 3.0541417598724365,
|
|
"learning_rate": 1.3912380952380953e-05,
|
|
"loss": 1.2167075157165528,
|
|
"mean_token_accuracy": 0.7153365164995193,
|
|
"num_tokens": 12218813.0,
|
|
"step": 800
|
|
},
|
|
{
|
|
"entropy": 1.5715911239385605,
|
|
"epoch": 0.9257142857142857,
|
|
"grad_norm": 2.4761741161346436,
|
|
"learning_rate": 1.3836190476190477e-05,
|
|
"loss": 1.0981733322143554,
|
|
"mean_token_accuracy": 0.7329004533588886,
|
|
"num_tokens": 12366807.0,
|
|
"step": 810
|
|
},
|
|
{
|
|
"entropy": 1.5434480041265488,
|
|
"epoch": 0.9371428571428572,
|
|
"grad_norm": 3.157036304473877,
|
|
"learning_rate": 1.376e-05,
|
|
"loss": 1.0953783988952637,
|
|
"mean_token_accuracy": 0.7374757573008537,
|
|
"num_tokens": 12519277.0,
|
|
"step": 820
|
|
},
|
|
{
|
|
"entropy": 1.5731646195054054,
|
|
"epoch": 0.9485714285714286,
|
|
"grad_norm": 2.9288859367370605,
|
|
"learning_rate": 1.3683809523809526e-05,
|
|
"loss": 1.1861271858215332,
|
|
"mean_token_accuracy": 0.7202281527221203,
|
|
"num_tokens": 12668916.0,
|
|
"step": 830
|
|
},
|
|
{
|
|
"entropy": 1.5766675099730492,
|
|
"epoch": 0.96,
|
|
"grad_norm": 2.3330445289611816,
|
|
"learning_rate": 1.3607619047619049e-05,
|
|
"loss": 1.1634528160095214,
|
|
"mean_token_accuracy": 0.7218175616115332,
|
|
"num_tokens": 12824349.0,
|
|
"step": 840
|
|
},
|
|
{
|
|
"entropy": 1.5784073442220687,
|
|
"epoch": 0.9714285714285714,
|
|
"grad_norm": 2.8480734825134277,
|
|
"learning_rate": 1.3531428571428573e-05,
|
|
"loss": 1.1402252197265625,
|
|
"mean_token_accuracy": 0.7281969003379345,
|
|
"num_tokens": 12974198.0,
|
|
"step": 850
|
|
},
|
|
{
|
|
"entropy": 1.5565635196864605,
|
|
"epoch": 0.9828571428571429,
|
|
"grad_norm": 3.2169361114501953,
|
|
"learning_rate": 1.3455238095238096e-05,
|
|
"loss": 1.0919082641601563,
|
|
"mean_token_accuracy": 0.7359561800956727,
|
|
"num_tokens": 13125877.0,
|
|
"step": 860
|
|
},
|
|
{
|
|
"entropy": 1.6499854236841203,
|
|
"epoch": 0.9942857142857143,
|
|
"grad_norm": 2.59342622756958,
|
|
"learning_rate": 1.3379047619047621e-05,
|
|
"loss": 1.1824271202087402,
|
|
"mean_token_accuracy": 0.7206701695919037,
|
|
"num_tokens": 13270434.0,
|
|
"step": 870
|
|
},
|
|
{
|
|
"entropy": 1.5392458021640778,
|
|
"epoch": 1.0057142857142858,
|
|
"grad_norm": 2.1802525520324707,
|
|
"learning_rate": 1.3302857142857145e-05,
|
|
"loss": 0.9705311775207519,
|
|
"mean_token_accuracy": 0.7628285072743892,
|
|
"num_tokens": 13419321.0,
|
|
"step": 880
|
|
},
|
|
{
|
|
"entropy": 1.3396786622703076,
|
|
"epoch": 1.0171428571428571,
|
|
"grad_norm": 2.7976434230804443,
|
|
"learning_rate": 1.3226666666666668e-05,
|
|
"loss": 0.753804874420166,
|
|
"mean_token_accuracy": 0.8103815123438836,
|
|
"num_tokens": 13564219.0,
|
|
"step": 890
|
|
},
|
|
{
|
|
"entropy": 1.263124795258045,
|
|
"epoch": 1.0285714285714285,
|
|
"grad_norm": 2.680875062942505,
|
|
"learning_rate": 1.3150476190476192e-05,
|
|
"loss": 0.8032585144042969,
|
|
"mean_token_accuracy": 0.8010107159614563,
|
|
"num_tokens": 13716646.0,
|
|
"step": 900
|
|
},
|
|
{
|
|
"entropy": 1.2998421356081962,
|
|
"epoch": 1.04,
|
|
"grad_norm": 3.176466941833496,
|
|
"learning_rate": 1.3074285714285715e-05,
|
|
"loss": 0.7720170497894288,
|
|
"mean_token_accuracy": 0.8062204010784626,
|
|
"num_tokens": 13865045.0,
|
|
"step": 910
|
|
},
|
|
{
|
|
"entropy": 1.268384473770857,
|
|
"epoch": 1.0514285714285714,
|
|
"grad_norm": 3.2034316062927246,
|
|
"learning_rate": 1.2998095238095239e-05,
|
|
"loss": 0.7246803283691406,
|
|
"mean_token_accuracy": 0.8114085927605629,
|
|
"num_tokens": 14019242.0,
|
|
"step": 920
|
|
},
|
|
{
|
|
"entropy": 1.2792991541326046,
|
|
"epoch": 1.062857142857143,
|
|
"grad_norm": 2.799799919128418,
|
|
"learning_rate": 1.2921904761904762e-05,
|
|
"loss": 0.7459653377532959,
|
|
"mean_token_accuracy": 0.8081640765070915,
|
|
"num_tokens": 14170551.0,
|
|
"step": 930
|
|
},
|
|
{
|
|
"entropy": 1.2749989867210387,
|
|
"epoch": 1.0742857142857143,
|
|
"grad_norm": 2.315112352371216,
|
|
"learning_rate": 1.2845714285714286e-05,
|
|
"loss": 0.7502207279205322,
|
|
"mean_token_accuracy": 0.8043648317456246,
|
|
"num_tokens": 14320423.0,
|
|
"step": 940
|
|
},
|
|
{
|
|
"entropy": 1.3014213733375073,
|
|
"epoch": 1.0857142857142856,
|
|
"grad_norm": 2.8643505573272705,
|
|
"learning_rate": 1.2769523809523811e-05,
|
|
"loss": 0.7972130298614502,
|
|
"mean_token_accuracy": 0.798911166191101,
|
|
"num_tokens": 14472454.0,
|
|
"step": 950
|
|
},
|
|
{
|
|
"entropy": 1.309807014465332,
|
|
"epoch": 1.0971428571428572,
|
|
"grad_norm": 2.9528074264526367,
|
|
"learning_rate": 1.2693333333333336e-05,
|
|
"loss": 0.762692928314209,
|
|
"mean_token_accuracy": 0.7993374645709992,
|
|
"num_tokens": 14621285.0,
|
|
"step": 960
|
|
},
|
|
{
|
|
"entropy": 1.2928462289273739,
|
|
"epoch": 1.1085714285714285,
|
|
"grad_norm": 3.0441250801086426,
|
|
"learning_rate": 1.261714285714286e-05,
|
|
"loss": 0.8246999740600586,
|
|
"mean_token_accuracy": 0.7930732540786266,
|
|
"num_tokens": 14776609.0,
|
|
"step": 970
|
|
},
|
|
{
|
|
"entropy": 1.2388829864561557,
|
|
"epoch": 1.12,
|
|
"grad_norm": 2.831936836242676,
|
|
"learning_rate": 1.2540952380952383e-05,
|
|
"loss": 0.7267831802368164,
|
|
"mean_token_accuracy": 0.8151453614234925,
|
|
"num_tokens": 14935547.0,
|
|
"step": 980
|
|
},
|
|
{
|
|
"entropy": 1.2560633935034275,
|
|
"epoch": 1.1314285714285715,
|
|
"grad_norm": 2.852947473526001,
|
|
"learning_rate": 1.2464761904761907e-05,
|
|
"loss": 0.7547249794006348,
|
|
"mean_token_accuracy": 0.805939918756485,
|
|
"num_tokens": 15090483.0,
|
|
"step": 990
|
|
},
|
|
{
|
|
"entropy": 1.241182105243206,
|
|
"epoch": 1.1428571428571428,
|
|
"grad_norm": 3.028001546859741,
|
|
"learning_rate": 1.238857142857143e-05,
|
|
"loss": 0.7496088027954102,
|
|
"mean_token_accuracy": 0.8107783786952496,
|
|
"num_tokens": 15242584.0,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"entropy": 1.348277000337839,
|
|
"epoch": 1.1542857142857144,
|
|
"grad_norm": 2.8477680683135986,
|
|
"learning_rate": 1.2312380952380953e-05,
|
|
"loss": 0.8604803085327148,
|
|
"mean_token_accuracy": 0.7804459646344185,
|
|
"num_tokens": 15391767.0,
|
|
"step": 1010
|
|
},
|
|
{
|
|
"entropy": 1.250984299182892,
|
|
"epoch": 1.1657142857142857,
|
|
"grad_norm": 2.5354862213134766,
|
|
"learning_rate": 1.2236190476190477e-05,
|
|
"loss": 0.7215850353240967,
|
|
"mean_token_accuracy": 0.811321247369051,
|
|
"num_tokens": 15543892.0,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"entropy": 1.2813407756388187,
|
|
"epoch": 1.177142857142857,
|
|
"grad_norm": 2.451107978820801,
|
|
"learning_rate": 1.216e-05,
|
|
"loss": 0.7772514820098877,
|
|
"mean_token_accuracy": 0.8003417395055294,
|
|
"num_tokens": 15692671.0,
|
|
"step": 1030
|
|
},
|
|
{
|
|
"entropy": 1.2703118346631528,
|
|
"epoch": 1.1885714285714286,
|
|
"grad_norm": 2.9708104133605957,
|
|
"learning_rate": 1.2083809523809524e-05,
|
|
"loss": 0.8101517677307128,
|
|
"mean_token_accuracy": 0.7938590474426747,
|
|
"num_tokens": 15849058.0,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"entropy": 1.2321988992393016,
|
|
"epoch": 1.2,
|
|
"grad_norm": 2.5784900188446045,
|
|
"learning_rate": 1.2007619047619047e-05,
|
|
"loss": 0.7412232398986817,
|
|
"mean_token_accuracy": 0.8034046024084092,
|
|
"num_tokens": 16006785.0,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"entropy": 1.2155711680650712,
|
|
"epoch": 1.2114285714285715,
|
|
"grad_norm": 2.337512731552124,
|
|
"learning_rate": 1.1931428571428571e-05,
|
|
"loss": 0.7488626003265381,
|
|
"mean_token_accuracy": 0.8063951514661312,
|
|
"num_tokens": 16158185.0,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"entropy": 1.2575144089758397,
|
|
"epoch": 1.2228571428571429,
|
|
"grad_norm": 2.3885722160339355,
|
|
"learning_rate": 1.1855238095238094e-05,
|
|
"loss": 0.7211280345916748,
|
|
"mean_token_accuracy": 0.8121630363166332,
|
|
"num_tokens": 16316647.0,
|
|
"step": 1070
|
|
},
|
|
{
|
|
"entropy": 1.2669233828783035,
|
|
"epoch": 1.2342857142857142,
|
|
"grad_norm": 3.1025376319885254,
|
|
"learning_rate": 1.1779047619047621e-05,
|
|
"loss": 0.7715346813201904,
|
|
"mean_token_accuracy": 0.7969170480966568,
|
|
"num_tokens": 16471457.0,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"entropy": 1.2340376809239388,
|
|
"epoch": 1.2457142857142858,
|
|
"grad_norm": 2.8538618087768555,
|
|
"learning_rate": 1.1702857142857145e-05,
|
|
"loss": 0.7577665328979493,
|
|
"mean_token_accuracy": 0.8027225121855736,
|
|
"num_tokens": 16631017.0,
|
|
"step": 1090
|
|
},
|
|
{
|
|
"entropy": 1.2692770406603813,
|
|
"epoch": 1.2571428571428571,
|
|
"grad_norm": 3.1801536083221436,
|
|
"learning_rate": 1.1626666666666668e-05,
|
|
"loss": 0.7788435935974121,
|
|
"mean_token_accuracy": 0.8010938227176666,
|
|
"num_tokens": 16786752.0,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"entropy": 1.2486526250839234,
|
|
"epoch": 1.2685714285714287,
|
|
"grad_norm": 2.5821192264556885,
|
|
"learning_rate": 1.1550476190476192e-05,
|
|
"loss": 0.762976884841919,
|
|
"mean_token_accuracy": 0.8053238935768604,
|
|
"num_tokens": 16943519.0,
|
|
"step": 1110
|
|
},
|
|
{
|
|
"entropy": 1.3036663837730884,
|
|
"epoch": 1.28,
|
|
"grad_norm": 2.6454882621765137,
|
|
"learning_rate": 1.1474285714285715e-05,
|
|
"loss": 0.7800922393798828,
|
|
"mean_token_accuracy": 0.7931916259229184,
|
|
"num_tokens": 17098242.0,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"entropy": 1.2358268782496453,
|
|
"epoch": 1.2914285714285714,
|
|
"grad_norm": 2.732434034347534,
|
|
"learning_rate": 1.1398095238095239e-05,
|
|
"loss": 0.7546168804168701,
|
|
"mean_token_accuracy": 0.8025222033262253,
|
|
"num_tokens": 17254880.0,
|
|
"step": 1130
|
|
},
|
|
{
|
|
"entropy": 1.2517999060451985,
|
|
"epoch": 1.302857142857143,
|
|
"grad_norm": 2.6394054889678955,
|
|
"learning_rate": 1.1321904761904762e-05,
|
|
"loss": 0.7450598239898681,
|
|
"mean_token_accuracy": 0.8091251827776432,
|
|
"num_tokens": 17404975.0,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"entropy": 1.2538614131510257,
|
|
"epoch": 1.3142857142857143,
|
|
"grad_norm": 3.0480637550354004,
|
|
"learning_rate": 1.1245714285714286e-05,
|
|
"loss": 0.7634236335754394,
|
|
"mean_token_accuracy": 0.8049138620495796,
|
|
"num_tokens": 17557114.0,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"entropy": 1.2000257663428784,
|
|
"epoch": 1.3257142857142856,
|
|
"grad_norm": 3.085055351257324,
|
|
"learning_rate": 1.1169523809523811e-05,
|
|
"loss": 0.7426016807556153,
|
|
"mean_token_accuracy": 0.8077944576740265,
|
|
"num_tokens": 17704128.0,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"entropy": 1.2293061904609204,
|
|
"epoch": 1.3371428571428572,
|
|
"grad_norm": 2.3842477798461914,
|
|
"learning_rate": 1.1093333333333334e-05,
|
|
"loss": 0.7181476593017578,
|
|
"mean_token_accuracy": 0.8123824037611485,
|
|
"num_tokens": 17865916.0,
|
|
"step": 1170
|
|
},
|
|
{
|
|
"entropy": 1.2635096937417984,
|
|
"epoch": 1.3485714285714285,
|
|
"grad_norm": 2.483147382736206,
|
|
"learning_rate": 1.1017142857142858e-05,
|
|
"loss": 0.7818619251251221,
|
|
"mean_token_accuracy": 0.8041326075792312,
|
|
"num_tokens": 18023210.0,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"entropy": 1.3030154168605805,
|
|
"epoch": 1.3599999999999999,
|
|
"grad_norm": 3.100837469100952,
|
|
"learning_rate": 1.0940952380952381e-05,
|
|
"loss": 0.7691655158996582,
|
|
"mean_token_accuracy": 0.8024266973137856,
|
|
"num_tokens": 18172271.0,
|
|
"step": 1190
|
|
},
|
|
{
|
|
"entropy": 1.2880703262984752,
|
|
"epoch": 1.3714285714285714,
|
|
"grad_norm": 3.068323850631714,
|
|
"learning_rate": 1.0864761904761905e-05,
|
|
"loss": 0.8048301696777344,
|
|
"mean_token_accuracy": 0.7959965549409389,
|
|
"num_tokens": 18325046.0,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"entropy": 1.3046514227986337,
|
|
"epoch": 1.3828571428571428,
|
|
"grad_norm": 3.33621883392334,
|
|
"learning_rate": 1.078857142857143e-05,
|
|
"loss": 0.8266304016113282,
|
|
"mean_token_accuracy": 0.7895134054124355,
|
|
"num_tokens": 18476571.0,
|
|
"step": 1210
|
|
},
|
|
{
|
|
"entropy": 1.2605851158499717,
|
|
"epoch": 1.3942857142857144,
|
|
"grad_norm": 3.5944457054138184,
|
|
"learning_rate": 1.0712380952380954e-05,
|
|
"loss": 0.8084980010986328,
|
|
"mean_token_accuracy": 0.7934398144483567,
|
|
"num_tokens": 18622849.0,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"entropy": 1.2391214780509472,
|
|
"epoch": 1.4057142857142857,
|
|
"grad_norm": 2.6554486751556396,
|
|
"learning_rate": 1.0636190476190477e-05,
|
|
"loss": 0.7550242900848388,
|
|
"mean_token_accuracy": 0.8065047845244407,
|
|
"num_tokens": 18768698.0,
|
|
"step": 1230
|
|
},
|
|
{
|
|
"entropy": 1.2420047149062157,
|
|
"epoch": 1.4171428571428573,
|
|
"grad_norm": 2.5860495567321777,
|
|
"learning_rate": 1.056e-05,
|
|
"loss": 0.7261144161224365,
|
|
"mean_token_accuracy": 0.8117271907627582,
|
|
"num_tokens": 18923547.0,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"entropy": 1.246000237762928,
|
|
"epoch": 1.4285714285714286,
|
|
"grad_norm": 3.1218044757843018,
|
|
"learning_rate": 1.0483809523809526e-05,
|
|
"loss": 0.7216352939605712,
|
|
"mean_token_accuracy": 0.8120975896716118,
|
|
"num_tokens": 19079479.0,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"entropy": 1.207762571424246,
|
|
"epoch": 1.44,
|
|
"grad_norm": 2.393723726272583,
|
|
"learning_rate": 1.040761904761905e-05,
|
|
"loss": 0.7272539138793945,
|
|
"mean_token_accuracy": 0.8084844700992108,
|
|
"num_tokens": 19236273.0,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"entropy": 1.2830321080982685,
|
|
"epoch": 1.4514285714285715,
|
|
"grad_norm": 3.0418026447296143,
|
|
"learning_rate": 1.0331428571428573e-05,
|
|
"loss": 0.7889233589172363,
|
|
"mean_token_accuracy": 0.7998958893120289,
|
|
"num_tokens": 19385795.0,
|
|
"step": 1270
|
|
},
|
|
{
|
|
"entropy": 1.2588836170732975,
|
|
"epoch": 1.4628571428571429,
|
|
"grad_norm": 2.668825626373291,
|
|
"learning_rate": 1.0255238095238096e-05,
|
|
"loss": 0.7696958065032959,
|
|
"mean_token_accuracy": 0.8028367407619953,
|
|
"num_tokens": 19540370.0,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"entropy": 1.2221255615353583,
|
|
"epoch": 1.4742857142857142,
|
|
"grad_norm": 3.121307611465454,
|
|
"learning_rate": 1.017904761904762e-05,
|
|
"loss": 0.7629749774932861,
|
|
"mean_token_accuracy": 0.8041565544903279,
|
|
"num_tokens": 19695452.0,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"entropy": 1.2615755833685398,
|
|
"epoch": 1.4857142857142858,
|
|
"grad_norm": 2.600470781326294,
|
|
"learning_rate": 1.0102857142857143e-05,
|
|
"loss": 0.8250988960266114,
|
|
"mean_token_accuracy": 0.7851416207849979,
|
|
"num_tokens": 19850924.0,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"entropy": 1.2793317660689354,
|
|
"epoch": 1.497142857142857,
|
|
"grad_norm": 2.874340295791626,
|
|
"learning_rate": 1.0026666666666667e-05,
|
|
"loss": 0.771949577331543,
|
|
"mean_token_accuracy": 0.8040384240448475,
|
|
"num_tokens": 20014514.0,
|
|
"step": 1310
|
|
},
|
|
{
|
|
"entropy": 1.1907848060131072,
|
|
"epoch": 1.5085714285714285,
|
|
"grad_norm": 2.6014015674591064,
|
|
"learning_rate": 9.950476190476192e-06,
|
|
"loss": 0.7447234630584717,
|
|
"mean_token_accuracy": 0.8111775033175945,
|
|
"num_tokens": 20173798.0,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"entropy": 1.2612031117081641,
|
|
"epoch": 1.52,
|
|
"grad_norm": 2.809692144393921,
|
|
"learning_rate": 9.874285714285715e-06,
|
|
"loss": 0.7573285102844238,
|
|
"mean_token_accuracy": 0.8068460263311863,
|
|
"num_tokens": 20318447.0,
|
|
"step": 1330
|
|
},
|
|
{
|
|
"entropy": 1.2434154883027078,
|
|
"epoch": 1.5314285714285716,
|
|
"grad_norm": 2.9109153747558594,
|
|
"learning_rate": 9.798095238095239e-06,
|
|
"loss": 0.7508068561553956,
|
|
"mean_token_accuracy": 0.8049707144498826,
|
|
"num_tokens": 20472944.0,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"entropy": 1.2488966889679431,
|
|
"epoch": 1.5428571428571427,
|
|
"grad_norm": 3.0063366889953613,
|
|
"learning_rate": 9.721904761904762e-06,
|
|
"loss": 0.7871552467346191,
|
|
"mean_token_accuracy": 0.8020347572863102,
|
|
"num_tokens": 20625826.0,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"entropy": 1.2348267659544945,
|
|
"epoch": 1.5542857142857143,
|
|
"grad_norm": 2.792886734008789,
|
|
"learning_rate": 9.645714285714286e-06,
|
|
"loss": 0.7784818172454834,
|
|
"mean_token_accuracy": 0.8054304026067257,
|
|
"num_tokens": 20774171.0,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"entropy": 1.2300232581794261,
|
|
"epoch": 1.5657142857142858,
|
|
"grad_norm": 2.861610174179077,
|
|
"learning_rate": 9.569523809523811e-06,
|
|
"loss": 0.7702634811401368,
|
|
"mean_token_accuracy": 0.8006386004388333,
|
|
"num_tokens": 20922487.0,
|
|
"step": 1370
|
|
},
|
|
{
|
|
"entropy": 1.24477871209383,
|
|
"epoch": 1.5771428571428572,
|
|
"grad_norm": 3.1448233127593994,
|
|
"learning_rate": 9.493333333333334e-06,
|
|
"loss": 0.7313426494598388,
|
|
"mean_token_accuracy": 0.812004954367876,
|
|
"num_tokens": 21075326.0,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"entropy": 1.2703936398029327,
|
|
"epoch": 1.5885714285714285,
|
|
"grad_norm": 2.5799062252044678,
|
|
"learning_rate": 9.417142857142858e-06,
|
|
"loss": 0.7714487552642822,
|
|
"mean_token_accuracy": 0.799166913330555,
|
|
"num_tokens": 21224257.0,
|
|
"step": 1390
|
|
},
|
|
{
|
|
"entropy": 1.2416778221726417,
|
|
"epoch": 1.6,
|
|
"grad_norm": 2.992957353591919,
|
|
"learning_rate": 9.340952380952381e-06,
|
|
"loss": 0.7809468746185303,
|
|
"mean_token_accuracy": 0.7984777301549911,
|
|
"num_tokens": 21373259.0,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"entropy": 1.2557578206062316,
|
|
"epoch": 1.6114285714285714,
|
|
"grad_norm": 2.7290945053100586,
|
|
"learning_rate": 9.264761904761905e-06,
|
|
"loss": 0.7919119834899903,
|
|
"mean_token_accuracy": 0.7954873479902744,
|
|
"num_tokens": 21522253.0,
|
|
"step": 1410
|
|
},
|
|
{
|
|
"entropy": 1.2285942271351815,
|
|
"epoch": 1.6228571428571428,
|
|
"grad_norm": 2.8919999599456787,
|
|
"learning_rate": 9.188571428571428e-06,
|
|
"loss": 0.7635447978973389,
|
|
"mean_token_accuracy": 0.8025862380862236,
|
|
"num_tokens": 21671374.0,
|
|
"step": 1420
|
|
},
|
|
{
|
|
"entropy": 1.3119461573660374,
|
|
"epoch": 1.6342857142857143,
|
|
"grad_norm": 2.7214455604553223,
|
|
"learning_rate": 9.112380952380954e-06,
|
|
"loss": 0.7926105976104736,
|
|
"mean_token_accuracy": 0.7944708272814751,
|
|
"num_tokens": 21822289.0,
|
|
"step": 1430
|
|
},
|
|
{
|
|
"entropy": 1.2220344461500645,
|
|
"epoch": 1.6457142857142857,
|
|
"grad_norm": 2.327855348587036,
|
|
"learning_rate": 9.036190476190477e-06,
|
|
"loss": 0.7583229064941406,
|
|
"mean_token_accuracy": 0.8038429766893387,
|
|
"num_tokens": 21978538.0,
|
|
"step": 1440
|
|
},
|
|
{
|
|
"entropy": 1.266616266965866,
|
|
"epoch": 1.657142857142857,
|
|
"grad_norm": 2.7889652252197266,
|
|
"learning_rate": 8.96e-06,
|
|
"loss": 0.7689779281616211,
|
|
"mean_token_accuracy": 0.8045293688774109,
|
|
"num_tokens": 22133231.0,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"entropy": 1.3114601746201515,
|
|
"epoch": 1.6685714285714286,
|
|
"grad_norm": 3.2477846145629883,
|
|
"learning_rate": 8.883809523809524e-06,
|
|
"loss": 0.7963542461395263,
|
|
"mean_token_accuracy": 0.8005617402493954,
|
|
"num_tokens": 22280643.0,
|
|
"step": 1460
|
|
},
|
|
{
|
|
"entropy": 1.2778782725334168,
|
|
"epoch": 1.6800000000000002,
|
|
"grad_norm": 3.1621718406677246,
|
|
"learning_rate": 8.807619047619048e-06,
|
|
"loss": 0.801125431060791,
|
|
"mean_token_accuracy": 0.7964559637010098,
|
|
"num_tokens": 22430891.0,
|
|
"step": 1470
|
|
},
|
|
{
|
|
"entropy": 1.2413391053676606,
|
|
"epoch": 1.6914285714285713,
|
|
"grad_norm": 2.561835527420044,
|
|
"learning_rate": 8.731428571428571e-06,
|
|
"loss": 0.7200276374816894,
|
|
"mean_token_accuracy": 0.8137163199484349,
|
|
"num_tokens": 22581471.0,
|
|
"step": 1480
|
|
},
|
|
{
|
|
"entropy": 1.2620924778282643,
|
|
"epoch": 1.7028571428571428,
|
|
"grad_norm": 2.8986573219299316,
|
|
"learning_rate": 8.655238095238096e-06,
|
|
"loss": 0.7427529335021973,
|
|
"mean_token_accuracy": 0.803333692252636,
|
|
"num_tokens": 22732244.0,
|
|
"step": 1490
|
|
},
|
|
{
|
|
"entropy": 1.2637714959681035,
|
|
"epoch": 1.7142857142857144,
|
|
"grad_norm": 2.8469996452331543,
|
|
"learning_rate": 8.57904761904762e-06,
|
|
"loss": 0.782107925415039,
|
|
"mean_token_accuracy": 0.8046764589846134,
|
|
"num_tokens": 22884081.0,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"entropy": 1.2661903858184815,
|
|
"epoch": 1.7257142857142858,
|
|
"grad_norm": 3.0864884853363037,
|
|
"learning_rate": 8.502857142857143e-06,
|
|
"loss": 0.7898892879486084,
|
|
"mean_token_accuracy": 0.7921097949147224,
|
|
"num_tokens": 23034449.0,
|
|
"step": 1510
|
|
},
|
|
{
|
|
"entropy": 1.250692894309759,
|
|
"epoch": 1.737142857142857,
|
|
"grad_norm": 2.6670966148376465,
|
|
"learning_rate": 8.426666666666667e-06,
|
|
"loss": 0.7688261508941651,
|
|
"mean_token_accuracy": 0.8037325166165828,
|
|
"num_tokens": 23188136.0,
|
|
"step": 1520
|
|
},
|
|
{
|
|
"entropy": 1.215890847891569,
|
|
"epoch": 1.7485714285714287,
|
|
"grad_norm": 2.3385775089263916,
|
|
"learning_rate": 8.35047619047619e-06,
|
|
"loss": 0.7499132633209229,
|
|
"mean_token_accuracy": 0.8074461244046688,
|
|
"num_tokens": 23343060.0,
|
|
"step": 1530
|
|
},
|
|
{
|
|
"entropy": 1.242823239415884,
|
|
"epoch": 1.76,
|
|
"grad_norm": 2.630499839782715,
|
|
"learning_rate": 8.274285714285715e-06,
|
|
"loss": 0.7520731925964356,
|
|
"mean_token_accuracy": 0.8076161026954651,
|
|
"num_tokens": 23487902.0,
|
|
"step": 1540
|
|
},
|
|
{
|
|
"entropy": 1.2385052397847176,
|
|
"epoch": 1.7714285714285714,
|
|
"grad_norm": 2.76127552986145,
|
|
"learning_rate": 8.198095238095239e-06,
|
|
"loss": 0.7834786415100098,
|
|
"mean_token_accuracy": 0.7965474635362625,
|
|
"num_tokens": 23635117.0,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"entropy": 1.1929515838623046,
|
|
"epoch": 1.782857142857143,
|
|
"grad_norm": 2.612879991531372,
|
|
"learning_rate": 8.121904761904762e-06,
|
|
"loss": 0.7256328582763671,
|
|
"mean_token_accuracy": 0.8150137662887573,
|
|
"num_tokens": 23795433.0,
|
|
"step": 1560
|
|
},
|
|
{
|
|
"entropy": 1.194692898541689,
|
|
"epoch": 1.7942857142857143,
|
|
"grad_norm": 2.6522936820983887,
|
|
"learning_rate": 8.045714285714286e-06,
|
|
"loss": 0.7276816368103027,
|
|
"mean_token_accuracy": 0.8094555623829365,
|
|
"num_tokens": 23950370.0,
|
|
"step": 1570
|
|
},
|
|
{
|
|
"entropy": 1.2585241250693797,
|
|
"epoch": 1.8057142857142856,
|
|
"grad_norm": 2.611926794052124,
|
|
"learning_rate": 7.969523809523811e-06,
|
|
"loss": 0.7796142578125,
|
|
"mean_token_accuracy": 0.8012560330331325,
|
|
"num_tokens": 24103914.0,
|
|
"step": 1580
|
|
},
|
|
{
|
|
"entropy": 1.2676213465631008,
|
|
"epoch": 1.8171428571428572,
|
|
"grad_norm": 2.5129570960998535,
|
|
"learning_rate": 7.893333333333335e-06,
|
|
"loss": 0.7887630462646484,
|
|
"mean_token_accuracy": 0.7965363718569278,
|
|
"num_tokens": 24258081.0,
|
|
"step": 1590
|
|
},
|
|
{
|
|
"entropy": 1.2579274140298367,
|
|
"epoch": 1.8285714285714287,
|
|
"grad_norm": 2.86506724357605,
|
|
"learning_rate": 7.817142857142858e-06,
|
|
"loss": 0.7497401714324952,
|
|
"mean_token_accuracy": 0.8082531146705151,
|
|
"num_tokens": 24409168.0,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"entropy": 1.2318767301738263,
|
|
"epoch": 1.8399999999999999,
|
|
"grad_norm": 2.7256534099578857,
|
|
"learning_rate": 7.740952380952382e-06,
|
|
"loss": 0.7274055480957031,
|
|
"mean_token_accuracy": 0.8118016675114632,
|
|
"num_tokens": 24566252.0,
|
|
"step": 1610
|
|
},
|
|
{
|
|
"entropy": 1.2056004114449024,
|
|
"epoch": 1.8514285714285714,
|
|
"grad_norm": 2.6447949409484863,
|
|
"learning_rate": 7.664761904761905e-06,
|
|
"loss": 0.7131245613098145,
|
|
"mean_token_accuracy": 0.8102400578558445,
|
|
"num_tokens": 24721993.0,
|
|
"step": 1620
|
|
},
|
|
{
|
|
"entropy": 1.245617663115263,
|
|
"epoch": 1.862857142857143,
|
|
"grad_norm": 2.835374116897583,
|
|
"learning_rate": 7.588571428571429e-06,
|
|
"loss": 0.783673095703125,
|
|
"mean_token_accuracy": 0.8022521927952766,
|
|
"num_tokens": 24868978.0,
|
|
"step": 1630
|
|
},
|
|
{
|
|
"entropy": 1.240978506952524,
|
|
"epoch": 1.8742857142857143,
|
|
"grad_norm": 2.6456923484802246,
|
|
"learning_rate": 7.512380952380953e-06,
|
|
"loss": 0.7802390098571778,
|
|
"mean_token_accuracy": 0.7991835795342922,
|
|
"num_tokens": 25026573.0,
|
|
"step": 1640
|
|
},
|
|
{
|
|
"entropy": 1.2280743092298507,
|
|
"epoch": 1.8857142857142857,
|
|
"grad_norm": 2.847167491912842,
|
|
"learning_rate": 7.436190476190477e-06,
|
|
"loss": 0.8147419929504395,
|
|
"mean_token_accuracy": 0.7919134043157101,
|
|
"num_tokens": 25182791.0,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"entropy": 1.216932328045368,
|
|
"epoch": 1.8971428571428572,
|
|
"grad_norm": 2.861248731613159,
|
|
"learning_rate": 7.360000000000001e-06,
|
|
"loss": 0.7451428890228271,
|
|
"mean_token_accuracy": 0.8088080659508705,
|
|
"num_tokens": 25326320.0,
|
|
"step": 1660
|
|
},
|
|
{
|
|
"entropy": 1.2367267780005933,
|
|
"epoch": 1.9085714285714286,
|
|
"grad_norm": 2.702206611633301,
|
|
"learning_rate": 7.283809523809524e-06,
|
|
"loss": 0.7687783241271973,
|
|
"mean_token_accuracy": 0.8046720631420612,
|
|
"num_tokens": 25481482.0,
|
|
"step": 1670
|
|
},
|
|
{
|
|
"entropy": 1.2137604996562004,
|
|
"epoch": 1.92,
|
|
"grad_norm": 2.446934938430786,
|
|
"learning_rate": 7.207619047619048e-06,
|
|
"loss": 0.7720860958099365,
|
|
"mean_token_accuracy": 0.8001258425414562,
|
|
"num_tokens": 25627745.0,
|
|
"step": 1680
|
|
},
|
|
{
|
|
"entropy": 1.2351884521543979,
|
|
"epoch": 1.9314285714285715,
|
|
"grad_norm": 2.682955026626587,
|
|
"learning_rate": 7.131428571428573e-06,
|
|
"loss": 0.7554657936096192,
|
|
"mean_token_accuracy": 0.8067528575658798,
|
|
"num_tokens": 25778484.0,
|
|
"step": 1690
|
|
},
|
|
{
|
|
"entropy": 1.2149645321071147,
|
|
"epoch": 1.9428571428571428,
|
|
"grad_norm": 3.032944679260254,
|
|
"learning_rate": 7.055238095238096e-06,
|
|
"loss": 0.6940214157104492,
|
|
"mean_token_accuracy": 0.8167023897171021,
|
|
"num_tokens": 25932265.0,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"entropy": 1.205052537471056,
|
|
"epoch": 1.9542857142857142,
|
|
"grad_norm": 2.7198259830474854,
|
|
"learning_rate": 6.97904761904762e-06,
|
|
"loss": 0.7255962371826172,
|
|
"mean_token_accuracy": 0.8064791038632393,
|
|
"num_tokens": 26077725.0,
|
|
"step": 1710
|
|
},
|
|
{
|
|
"entropy": 1.1943283297121525,
|
|
"epoch": 1.9657142857142857,
|
|
"grad_norm": 2.6816186904907227,
|
|
"learning_rate": 6.902857142857143e-06,
|
|
"loss": 0.7239448547363281,
|
|
"mean_token_accuracy": 0.809258659183979,
|
|
"num_tokens": 26233086.0,
|
|
"step": 1720
|
|
},
|
|
{
|
|
"entropy": 1.2314366959035397,
|
|
"epoch": 1.977142857142857,
|
|
"grad_norm": 3.0669050216674805,
|
|
"learning_rate": 6.826666666666667e-06,
|
|
"loss": 0.8160985946655274,
|
|
"mean_token_accuracy": 0.7934448584914208,
|
|
"num_tokens": 26383494.0,
|
|
"step": 1730
|
|
},
|
|
{
|
|
"entropy": 1.201389105618,
|
|
"epoch": 1.9885714285714284,
|
|
"grad_norm": 2.6934287548065186,
|
|
"learning_rate": 6.75047619047619e-06,
|
|
"loss": 0.7483705043792724,
|
|
"mean_token_accuracy": 0.8099792324006557,
|
|
"num_tokens": 26534600.0,
|
|
"step": 1740
|
|
},
|
|
{
|
|
"entropy": 1.2289574392139913,
|
|
"epoch": 2.0,
|
|
"grad_norm": 2.9322891235351562,
|
|
"learning_rate": 6.6742857142857155e-06,
|
|
"loss": 0.7264975070953369,
|
|
"mean_token_accuracy": 0.8107106305658818,
|
|
"num_tokens": 26690084.0,
|
|
"step": 1750
|
|
}
|
|
],
|
|
"logging_steps": 10,
|
|
"max_steps": 2625,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 3,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 3.268869501104333e+17,
|
|
"train_batch_size": 4,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|