bd27151216
Model: fpadovani/ita-latn-100mb-after-ppt-shuff-dyck-10mb-ckpt500_seed10 Source: Original Platform
1046 lines
28 KiB
JSON
1046 lines
28 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.5042864346949067,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 4.801114463806153,
|
|
"epoch": 0.005042864346949067,
|
|
"grad_norm": 14.3125,
|
|
"learning_rate": 2e-06,
|
|
"loss": 14.3768,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 11122.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 4.812651681900024,
|
|
"epoch": 0.010085728693898134,
|
|
"grad_norm": 15.5,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 14.2394,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 20499.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 4.861662435531616,
|
|
"epoch": 0.015128593040847202,
|
|
"grad_norm": 20.375,
|
|
"learning_rate": 7e-06,
|
|
"loss": 14.0486,
|
|
"mean_token_accuracy": 0.00012091898825019597,
|
|
"num_tokens": 30664.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 5.152417182922363,
|
|
"epoch": 0.020171457387796268,
|
|
"grad_norm": 32.25,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 13.3871,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 41767.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 7.587842893600464,
|
|
"epoch": 0.025214321734745335,
|
|
"grad_norm": 14.25,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 11.6896,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 53233.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.463200283050536,
|
|
"epoch": 0.030257186081694403,
|
|
"grad_norm": 3.828125,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.7364,
|
|
"mean_token_accuracy": 0.004124969011172652,
|
|
"num_tokens": 63591.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.693463897705078,
|
|
"epoch": 0.03530005042864347,
|
|
"grad_norm": 3.296875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.5492,
|
|
"mean_token_accuracy": 0.013820098433643579,
|
|
"num_tokens": 75211.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.683560562133788,
|
|
"epoch": 0.040342914775592535,
|
|
"grad_norm": 2.640625,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.2263,
|
|
"mean_token_accuracy": 0.015752001013606787,
|
|
"num_tokens": 86163.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.606955909729004,
|
|
"epoch": 0.0453857791225416,
|
|
"grad_norm": 2.609375,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.9408,
|
|
"mean_token_accuracy": 0.02095262976363301,
|
|
"num_tokens": 96812.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.570794868469239,
|
|
"epoch": 0.05042864346949067,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.767,
|
|
"mean_token_accuracy": 0.03581012934446335,
|
|
"num_tokens": 107246.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.550295162200928,
|
|
"epoch": 0.05547150781643974,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.6982,
|
|
"mean_token_accuracy": 0.035054446384310725,
|
|
"num_tokens": 118169.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.557003593444824,
|
|
"epoch": 0.060514372163388806,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.596,
|
|
"mean_token_accuracy": 0.03818601593375206,
|
|
"num_tokens": 129744.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.555067825317384,
|
|
"epoch": 0.06555723651033787,
|
|
"grad_norm": 1.9609375,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.4879,
|
|
"mean_token_accuracy": 0.030292158760130406,
|
|
"num_tokens": 139826.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.539035034179687,
|
|
"epoch": 0.07060010085728693,
|
|
"grad_norm": 1.890625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.4026,
|
|
"mean_token_accuracy": 0.029551995545625688,
|
|
"num_tokens": 149767.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.505377006530761,
|
|
"epoch": 0.07564296520423601,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.3373,
|
|
"mean_token_accuracy": 0.039892033115029334,
|
|
"num_tokens": 160895.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.469569301605224,
|
|
"epoch": 0.08068582955118507,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.2884,
|
|
"mean_token_accuracy": 0.03964783567935228,
|
|
"num_tokens": 172080.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.455696964263916,
|
|
"epoch": 0.08572869389813415,
|
|
"grad_norm": 1.640625,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.2276,
|
|
"mean_token_accuracy": 0.03601204827427864,
|
|
"num_tokens": 182836.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.45373363494873,
|
|
"epoch": 0.0907715582450832,
|
|
"grad_norm": 1.8828125,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.1377,
|
|
"mean_token_accuracy": 0.02935194130986929,
|
|
"num_tokens": 194781.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.433523464202882,
|
|
"epoch": 0.09581442259203228,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.9882,
|
|
"mean_token_accuracy": 0.03305415883660316,
|
|
"num_tokens": 206722.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.359123516082764,
|
|
"epoch": 0.10085728693898134,
|
|
"grad_norm": 1.8984375,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.8489,
|
|
"mean_token_accuracy": 0.04083778336644173,
|
|
"num_tokens": 217167.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.264568710327149,
|
|
"epoch": 0.1059001512859304,
|
|
"grad_norm": 1.5546875,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.7978,
|
|
"mean_token_accuracy": 0.04325296822935343,
|
|
"num_tokens": 228427.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.185066509246827,
|
|
"epoch": 0.11094301563287948,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.6675,
|
|
"mean_token_accuracy": 0.04498021528124809,
|
|
"num_tokens": 239169.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.122811222076416,
|
|
"epoch": 0.11598587997982854,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.5549,
|
|
"mean_token_accuracy": 0.039299645088613035,
|
|
"num_tokens": 250393.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.999498176574708,
|
|
"epoch": 0.12102874432677761,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.3439,
|
|
"mean_token_accuracy": 0.049582514539361,
|
|
"num_tokens": 260573.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.756069564819336,
|
|
"epoch": 0.1260716086737267,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.2337,
|
|
"mean_token_accuracy": 0.0456912275403738,
|
|
"num_tokens": 271537.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.617351722717284,
|
|
"epoch": 0.13111447302067575,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.1765,
|
|
"mean_token_accuracy": 0.04553677849471569,
|
|
"num_tokens": 282080.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.397808265686034,
|
|
"epoch": 0.1361573373676248,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.0293,
|
|
"mean_token_accuracy": 0.04926962368190289,
|
|
"num_tokens": 292509.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.158194255828857,
|
|
"epoch": 0.14120020171457387,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.9266,
|
|
"mean_token_accuracy": 0.04916851259768009,
|
|
"num_tokens": 303156.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 8.969250679016113,
|
|
"epoch": 0.14624306606152296,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.8665,
|
|
"mean_token_accuracy": 0.04366149418056011,
|
|
"num_tokens": 314398.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.68967514038086,
|
|
"epoch": 0.15128593040847202,
|
|
"grad_norm": 0.8359375,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.894,
|
|
"mean_token_accuracy": 0.04541506320238113,
|
|
"num_tokens": 324774.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.563091468811034,
|
|
"epoch": 0.15632879475542108,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.7359,
|
|
"mean_token_accuracy": 0.04416811689734459,
|
|
"num_tokens": 335154.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.345014476776123,
|
|
"epoch": 0.16137165910237014,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.8209,
|
|
"mean_token_accuracy": 0.05068969242274761,
|
|
"num_tokens": 346180.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.28213872909546,
|
|
"epoch": 0.1664145234493192,
|
|
"grad_norm": 0.8046875,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.7329,
|
|
"mean_token_accuracy": 0.04978744387626648,
|
|
"num_tokens": 356722.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.216850662231446,
|
|
"epoch": 0.1714573877962683,
|
|
"grad_norm": 0.80859375,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.6846,
|
|
"mean_token_accuracy": 0.05312272384762764,
|
|
"num_tokens": 367245.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.235016822814941,
|
|
"epoch": 0.17650025214321735,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.6401,
|
|
"mean_token_accuracy": 0.05696842148900032,
|
|
"num_tokens": 377434.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.149722671508789,
|
|
"epoch": 0.1815431164901664,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.5829,
|
|
"mean_token_accuracy": 0.06260009892284871,
|
|
"num_tokens": 388685.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.027026557922364,
|
|
"epoch": 0.18658598083711547,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.5612,
|
|
"mean_token_accuracy": 0.05986677259206772,
|
|
"num_tokens": 400559.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.110166931152344,
|
|
"epoch": 0.19162884518406456,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.6123,
|
|
"mean_token_accuracy": 0.06047765538096428,
|
|
"num_tokens": 413059.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.079530906677245,
|
|
"epoch": 0.19667170953101362,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.5223,
|
|
"mean_token_accuracy": 0.06776017844676971,
|
|
"num_tokens": 424073.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 7.999411535263062,
|
|
"epoch": 0.20171457387796268,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.4703,
|
|
"mean_token_accuracy": 0.06763602308928966,
|
|
"num_tokens": 435960.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.994884729385376,
|
|
"epoch": 0.20675743822491174,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.4259,
|
|
"mean_token_accuracy": 0.06415937095880508,
|
|
"num_tokens": 445458.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.940965747833252,
|
|
"epoch": 0.2118003025718608,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.4427,
|
|
"mean_token_accuracy": 0.06450643800199032,
|
|
"num_tokens": 456237.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.971080732345581,
|
|
"epoch": 0.2168431669188099,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.4559,
|
|
"mean_token_accuracy": 0.06802051141858101,
|
|
"num_tokens": 467210.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.95701413154602,
|
|
"epoch": 0.22188603126575895,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.4836,
|
|
"mean_token_accuracy": 0.06283098589628935,
|
|
"num_tokens": 478447.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.983369398117065,
|
|
"epoch": 0.22692889561270801,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.4325,
|
|
"mean_token_accuracy": 0.06579711437225341,
|
|
"num_tokens": 489108.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.921357250213623,
|
|
"epoch": 0.23197175995965708,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.4914,
|
|
"mean_token_accuracy": 0.06151033192873001,
|
|
"num_tokens": 500889.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.812901639938355,
|
|
"epoch": 0.23701462430660616,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.3437,
|
|
"mean_token_accuracy": 0.06879591457545757,
|
|
"num_tokens": 511038.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.84229211807251,
|
|
"epoch": 0.24205748865355523,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.4227,
|
|
"mean_token_accuracy": 0.06682286001741886,
|
|
"num_tokens": 522801.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.874321031570434,
|
|
"epoch": 0.24710035300050429,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.4278,
|
|
"mean_token_accuracy": 0.06629796028137207,
|
|
"num_tokens": 534795.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.927615451812744,
|
|
"epoch": 0.2521432173474534,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.4248,
|
|
"mean_token_accuracy": 0.06450821422040462,
|
|
"num_tokens": 546656.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.768723440170288,
|
|
"epoch": 0.25718608169440244,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.29,
|
|
"mean_token_accuracy": 0.07409121245145797,
|
|
"num_tokens": 556673.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.753383922576904,
|
|
"epoch": 0.2622289460413515,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.3377,
|
|
"mean_token_accuracy": 0.06640650182962418,
|
|
"num_tokens": 567044.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.754617547988891,
|
|
"epoch": 0.26727181038830056,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.2684,
|
|
"mean_token_accuracy": 0.07174375429749488,
|
|
"num_tokens": 577305.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.622322416305542,
|
|
"epoch": 0.2723146747352496,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.3679,
|
|
"mean_token_accuracy": 0.07188783846795559,
|
|
"num_tokens": 589285.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.7097578048706055,
|
|
"epoch": 0.2773575390821987,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.1943,
|
|
"mean_token_accuracy": 0.07407240830361843,
|
|
"num_tokens": 598609.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.717816877365112,
|
|
"epoch": 0.28240040342914774,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.2602,
|
|
"mean_token_accuracy": 0.07445880733430385,
|
|
"num_tokens": 609872.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.6907618045806885,
|
|
"epoch": 0.2874432677760968,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.2711,
|
|
"mean_token_accuracy": 0.08134151697158813,
|
|
"num_tokens": 622583.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.654936122894287,
|
|
"epoch": 0.2924861321230459,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.2412,
|
|
"mean_token_accuracy": 0.07294923588633537,
|
|
"num_tokens": 633592.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.543200588226318,
|
|
"epoch": 0.297528996469995,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.169,
|
|
"mean_token_accuracy": 0.07622305229306221,
|
|
"num_tokens": 644858.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.596815156936645,
|
|
"epoch": 0.30257186081694404,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.245,
|
|
"mean_token_accuracy": 0.07087541855871678,
|
|
"num_tokens": 656438.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.674440145492554,
|
|
"epoch": 0.3076147251638931,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.2518,
|
|
"mean_token_accuracy": 0.07656564638018608,
|
|
"num_tokens": 667942.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.500104904174805,
|
|
"epoch": 0.31265758951084216,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.1468,
|
|
"mean_token_accuracy": 0.07846823334693909,
|
|
"num_tokens": 678596.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.610447406768799,
|
|
"epoch": 0.3177004538577912,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.3024,
|
|
"mean_token_accuracy": 0.06737342216074467,
|
|
"num_tokens": 691014.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.612302160263061,
|
|
"epoch": 0.3227433182047403,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.2079,
|
|
"mean_token_accuracy": 0.07754847705364228,
|
|
"num_tokens": 701849.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.523909950256348,
|
|
"epoch": 0.32778618255168934,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.1103,
|
|
"mean_token_accuracy": 0.08044524192810058,
|
|
"num_tokens": 712652.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.591261148452759,
|
|
"epoch": 0.3328290468986384,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.1597,
|
|
"mean_token_accuracy": 0.07344280071556568,
|
|
"num_tokens": 723704.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.501336288452149,
|
|
"epoch": 0.3378719112455875,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.1081,
|
|
"mean_token_accuracy": 0.07664402537047862,
|
|
"num_tokens": 734715.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.462927675247192,
|
|
"epoch": 0.3429147755925366,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.1461,
|
|
"mean_token_accuracy": 0.07232181765139103,
|
|
"num_tokens": 746772.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.542045450210571,
|
|
"epoch": 0.34795763993948564,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.1028,
|
|
"mean_token_accuracy": 0.07609748989343643,
|
|
"num_tokens": 757356.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.513222789764404,
|
|
"epoch": 0.3530005042864347,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.1177,
|
|
"mean_token_accuracy": 0.07893777638673782,
|
|
"num_tokens": 768269.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.457395029067993,
|
|
"epoch": 0.35804336863338376,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.0621,
|
|
"mean_token_accuracy": 0.07398857697844505,
|
|
"num_tokens": 779553.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.4384712219238285,
|
|
"epoch": 0.3630862329803328,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.1477,
|
|
"mean_token_accuracy": 0.07859819531440734,
|
|
"num_tokens": 791466.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.519556713104248,
|
|
"epoch": 0.3681290973272819,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.0409,
|
|
"mean_token_accuracy": 0.07564883790910244,
|
|
"num_tokens": 802064.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.367311239242554,
|
|
"epoch": 0.37317196167423095,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.0721,
|
|
"mean_token_accuracy": 0.08020395338535309,
|
|
"num_tokens": 813534.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.481927680969238,
|
|
"epoch": 0.37821482602118,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.0958,
|
|
"mean_token_accuracy": 0.07429472133517265,
|
|
"num_tokens": 824783.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.396484994888306,
|
|
"epoch": 0.3832576903681291,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.0813,
|
|
"mean_token_accuracy": 0.0780322715640068,
|
|
"num_tokens": 835019.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.43559775352478,
|
|
"epoch": 0.3883005547150782,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.9778,
|
|
"mean_token_accuracy": 0.07786969803273677,
|
|
"num_tokens": 845954.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.349666690826416,
|
|
"epoch": 0.39334341906202724,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.9968,
|
|
"mean_token_accuracy": 0.08358137160539628,
|
|
"num_tokens": 856393.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.377739429473877,
|
|
"epoch": 0.3983862834089763,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.0058,
|
|
"mean_token_accuracy": 0.08096945993602275,
|
|
"num_tokens": 868079.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.420732831954956,
|
|
"epoch": 0.40342914775592537,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.0597,
|
|
"mean_token_accuracy": 0.08519057407975197,
|
|
"num_tokens": 878995.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.359525680541992,
|
|
"epoch": 0.4084720121028744,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.9767,
|
|
"mean_token_accuracy": 0.08249035105109215,
|
|
"num_tokens": 889159.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.397777795791626,
|
|
"epoch": 0.4135148764498235,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 7.0906,
|
|
"mean_token_accuracy": 0.07337038293480873,
|
|
"num_tokens": 900186.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.36064920425415,
|
|
"epoch": 0.41855774079677255,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.0054,
|
|
"mean_token_accuracy": 0.0781717125326395,
|
|
"num_tokens": 910713.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.512122440338135,
|
|
"epoch": 0.4236006051437216,
|
|
"grad_norm": 2.359375,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.0956,
|
|
"mean_token_accuracy": 0.07817071452736854,
|
|
"num_tokens": 922020.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.32919135093689,
|
|
"epoch": 0.4286434694906707,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.000212,
|
|
"loss": 7.0302,
|
|
"mean_token_accuracy": 0.07741615772247315,
|
|
"num_tokens": 933033.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.3811193943023685,
|
|
"epoch": 0.4336863338376198,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.0623,
|
|
"mean_token_accuracy": 0.07646705359220504,
|
|
"num_tokens": 944394.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.382322645187378,
|
|
"epoch": 0.43872919818456885,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.9714,
|
|
"mean_token_accuracy": 0.0816558338701725,
|
|
"num_tokens": 955337.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.19914608001709,
|
|
"epoch": 0.4437720625315179,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.9435,
|
|
"mean_token_accuracy": 0.08770947605371475,
|
|
"num_tokens": 967020.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.364274787902832,
|
|
"epoch": 0.44881492687846697,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.9686,
|
|
"mean_token_accuracy": 0.08233175501227379,
|
|
"num_tokens": 978223.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.2527648448944095,
|
|
"epoch": 0.45385779122541603,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.9751,
|
|
"mean_token_accuracy": 0.08211390711367131,
|
|
"num_tokens": 990239.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.344982099533081,
|
|
"epoch": 0.4589006555723651,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.8979,
|
|
"mean_token_accuracy": 0.08162759393453597,
|
|
"num_tokens": 1001834.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.236715698242188,
|
|
"epoch": 0.46394351991931415,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.8744,
|
|
"mean_token_accuracy": 0.09091370068490505,
|
|
"num_tokens": 1012321.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.2280114650726315,
|
|
"epoch": 0.4689863842662632,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.9796,
|
|
"mean_token_accuracy": 0.08122261688113212,
|
|
"num_tokens": 1024500.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.270635604858398,
|
|
"epoch": 0.47402924861321233,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.8926,
|
|
"mean_token_accuracy": 0.08805500157177448,
|
|
"num_tokens": 1034825.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.1545487403869625,
|
|
"epoch": 0.4790721129601614,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.8648,
|
|
"mean_token_accuracy": 0.08775893896818161,
|
|
"num_tokens": 1046174.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.252149486541748,
|
|
"epoch": 0.48411497730711045,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.7807,
|
|
"mean_token_accuracy": 0.09338528215885163,
|
|
"num_tokens": 1055930.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.144695091247558,
|
|
"epoch": 0.4891578416540595,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.8153,
|
|
"mean_token_accuracy": 0.09350402504205704,
|
|
"num_tokens": 1066433.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.173965978622436,
|
|
"epoch": 0.49420070600100857,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.9306,
|
|
"mean_token_accuracy": 0.08193769156932831,
|
|
"num_tokens": 1077303.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.231033754348755,
|
|
"epoch": 0.49924357034795763,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.8453,
|
|
"mean_token_accuracy": 0.08849839866161346,
|
|
"num_tokens": 1088053.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.107387685775757,
|
|
"epoch": 0.5042864346949067,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.8499,
|
|
"mean_token_accuracy": 0.09236783608794212,
|
|
"num_tokens": 1098518.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.5042864346949067,
|
|
"eval_entropy": 7.061949415667973,
|
|
"eval_loss": 6.937669277191162,
|
|
"eval_mean_token_accuracy": 0.09019645222915341,
|
|
"eval_num_tokens": 1098518.0,
|
|
"eval_runtime": 2.6772,
|
|
"eval_samples_per_second": 1172.119,
|
|
"eval_steps_per_second": 146.795,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 9910,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 10,
|
|
"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": 1804074504192000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|