0aeb1fb267
Model: fpadovani/urd-arab-10mb-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": 0.5931198102016607,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 4.882989406585693,
|
|
"epoch": 0.005931198102016607,
|
|
"grad_norm": 22.5,
|
|
"learning_rate": 2e-06,
|
|
"loss": 14.5486,
|
|
"mean_token_accuracy": 0.00012430351926013828,
|
|
"num_tokens": 14276.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 4.891684865951538,
|
|
"epoch": 0.011862396204033215,
|
|
"grad_norm": 22.875,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 14.4388,
|
|
"mean_token_accuracy": 7.288630004040897e-05,
|
|
"num_tokens": 26454.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 4.908612489700317,
|
|
"epoch": 0.017793594306049824,
|
|
"grad_norm": 25.75,
|
|
"learning_rate": 7e-06,
|
|
"loss": 14.3151,
|
|
"mean_token_accuracy": 5.268703680485487e-05,
|
|
"num_tokens": 38987.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 4.985533952713013,
|
|
"epoch": 0.02372479240806643,
|
|
"grad_norm": 31.375,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 14.0964,
|
|
"mean_token_accuracy": 4.755111876875162e-05,
|
|
"num_tokens": 53908.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 5.351950407028198,
|
|
"epoch": 0.02965599051008304,
|
|
"grad_norm": 36.5,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 13.2683,
|
|
"mean_token_accuracy": 0.00010129763977602124,
|
|
"num_tokens": 66630.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 6.508245420455933,
|
|
"epoch": 0.03558718861209965,
|
|
"grad_norm": 38.0,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 12.2847,
|
|
"mean_token_accuracy": 0.00019785047043114902,
|
|
"num_tokens": 80227.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 9.8948805809021,
|
|
"epoch": 0.04151838671411625,
|
|
"grad_norm": 3.265625,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.9874,
|
|
"mean_token_accuracy": 0.00011806375114247203,
|
|
"num_tokens": 93260.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.6931321144104,
|
|
"epoch": 0.04744958481613286,
|
|
"grad_norm": 2.734375,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.7316,
|
|
"mean_token_accuracy": 0.0010395277058705688,
|
|
"num_tokens": 107368.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.705201148986816,
|
|
"epoch": 0.05338078291814947,
|
|
"grad_norm": 2.515625,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.6199,
|
|
"mean_token_accuracy": 0.002701799588976428,
|
|
"num_tokens": 121910.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.703446388244629,
|
|
"epoch": 0.05931198102016608,
|
|
"grad_norm": 2.125,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 10.4434,
|
|
"mean_token_accuracy": 0.012030716100707651,
|
|
"num_tokens": 136756.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.688198661804199,
|
|
"epoch": 0.06524317912218268,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 10.3119,
|
|
"mean_token_accuracy": 0.009988007182255388,
|
|
"num_tokens": 152921.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.69539442062378,
|
|
"epoch": 0.0711743772241993,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 10.1697,
|
|
"mean_token_accuracy": 0.012235769163817167,
|
|
"num_tokens": 167253.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.70052261352539,
|
|
"epoch": 0.0771055753262159,
|
|
"grad_norm": 1.8828125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 10.0223,
|
|
"mean_token_accuracy": 0.028588037192821502,
|
|
"num_tokens": 181088.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.69765625,
|
|
"epoch": 0.0830367734282325,
|
|
"grad_norm": 1.71875,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.9461,
|
|
"mean_token_accuracy": 0.030834931693971158,
|
|
"num_tokens": 193906.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.691731262207032,
|
|
"epoch": 0.08896797153024912,
|
|
"grad_norm": 1.6640625,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.9097,
|
|
"mean_token_accuracy": 0.03433742746710777,
|
|
"num_tokens": 209520.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.694703388214112,
|
|
"epoch": 0.09489916963226572,
|
|
"grad_norm": 1.703125,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.8355,
|
|
"mean_token_accuracy": 0.03231241405010223,
|
|
"num_tokens": 223967.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.692516708374024,
|
|
"epoch": 0.10083036773428232,
|
|
"grad_norm": 1.6328125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.7472,
|
|
"mean_token_accuracy": 0.031197282299399377,
|
|
"num_tokens": 238991.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.681371212005615,
|
|
"epoch": 0.10676156583629894,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.635,
|
|
"mean_token_accuracy": 0.034548941254615786,
|
|
"num_tokens": 254281.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.67520875930786,
|
|
"epoch": 0.11269276393831554,
|
|
"grad_norm": 1.703125,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.4991,
|
|
"mean_token_accuracy": 0.03750690445303917,
|
|
"num_tokens": 268697.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.662280464172364,
|
|
"epoch": 0.11862396204033215,
|
|
"grad_norm": 1.6875,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.4058,
|
|
"mean_token_accuracy": 0.03157996293157339,
|
|
"num_tokens": 283026.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.645224285125732,
|
|
"epoch": 0.12455516014234876,
|
|
"grad_norm": 1.5234375,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.2969,
|
|
"mean_token_accuracy": 0.033328772895038127,
|
|
"num_tokens": 298802.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.620039367675782,
|
|
"epoch": 0.13048635824436536,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.1491,
|
|
"mean_token_accuracy": 0.031768145971000195,
|
|
"num_tokens": 312390.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.58546199798584,
|
|
"epoch": 0.13641755634638197,
|
|
"grad_norm": 1.5234375,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.0007,
|
|
"mean_token_accuracy": 0.033303617313504216,
|
|
"num_tokens": 325525.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.541824913024902,
|
|
"epoch": 0.1423487544483986,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.8754,
|
|
"mean_token_accuracy": 0.031658035889267924,
|
|
"num_tokens": 338168.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.4882963180542,
|
|
"epoch": 0.14827995255041518,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.7439,
|
|
"mean_token_accuracy": 0.034061219170689584,
|
|
"num_tokens": 351065.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.421833515167236,
|
|
"epoch": 0.1542111506524318,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.6351,
|
|
"mean_token_accuracy": 0.03405258022248745,
|
|
"num_tokens": 366974.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.340956115722657,
|
|
"epoch": 0.1601423487544484,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.4547,
|
|
"mean_token_accuracy": 0.03421926330775023,
|
|
"num_tokens": 380570.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.244134426116943,
|
|
"epoch": 0.166073546856465,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.3673,
|
|
"mean_token_accuracy": 0.03386142123490572,
|
|
"num_tokens": 395653.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.126684474945069,
|
|
"epoch": 0.17200474495848161,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.2663,
|
|
"mean_token_accuracy": 0.03231785073876381,
|
|
"num_tokens": 412138.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.98361644744873,
|
|
"epoch": 0.17793594306049823,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.1045,
|
|
"mean_token_accuracy": 0.03140157163143158,
|
|
"num_tokens": 425588.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.823878002166747,
|
|
"epoch": 0.18386714116251482,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.9299,
|
|
"mean_token_accuracy": 0.03208461999893188,
|
|
"num_tokens": 438907.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.627764892578124,
|
|
"epoch": 0.18979833926453143,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.8438,
|
|
"mean_token_accuracy": 0.032532901503145695,
|
|
"num_tokens": 454179.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.385260581970215,
|
|
"epoch": 0.19572953736654805,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.7521,
|
|
"mean_token_accuracy": 0.034452256932854655,
|
|
"num_tokens": 468367.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.124578285217286,
|
|
"epoch": 0.20166073546856464,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.7039,
|
|
"mean_token_accuracy": 0.03381109684705734,
|
|
"num_tokens": 483127.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.871719264984131,
|
|
"epoch": 0.20759193357058126,
|
|
"grad_norm": 0.76171875,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.6055,
|
|
"mean_token_accuracy": 0.031578476540744305,
|
|
"num_tokens": 498918.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.65745210647583,
|
|
"epoch": 0.21352313167259787,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.5238,
|
|
"mean_token_accuracy": 0.0336442444473505,
|
|
"num_tokens": 512890.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.445704460144043,
|
|
"epoch": 0.21945432977461446,
|
|
"grad_norm": 0.59375,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.4868,
|
|
"mean_token_accuracy": 0.03619574885815382,
|
|
"num_tokens": 528517.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.257825183868409,
|
|
"epoch": 0.22538552787663108,
|
|
"grad_norm": 0.66796875,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.4523,
|
|
"mean_token_accuracy": 0.03298983443528414,
|
|
"num_tokens": 542528.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.134656715393067,
|
|
"epoch": 0.2313167259786477,
|
|
"grad_norm": 0.59765625,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.4896,
|
|
"mean_token_accuracy": 0.0337029367685318,
|
|
"num_tokens": 557801.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.057815742492675,
|
|
"epoch": 0.2372479240806643,
|
|
"grad_norm": 0.52734375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.466,
|
|
"mean_token_accuracy": 0.03353662136942148,
|
|
"num_tokens": 575318.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.983192443847656,
|
|
"epoch": 0.2431791221826809,
|
|
"grad_norm": 0.58984375,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.5065,
|
|
"mean_token_accuracy": 0.03289526663720608,
|
|
"num_tokens": 589283.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.915411520004272,
|
|
"epoch": 0.2491103202846975,
|
|
"grad_norm": 0.60546875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.3919,
|
|
"mean_token_accuracy": 0.03649586420506239,
|
|
"num_tokens": 603112.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.879695129394531,
|
|
"epoch": 0.25504151838671413,
|
|
"grad_norm": 0.6015625,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.4059,
|
|
"mean_token_accuracy": 0.03437527269124985,
|
|
"num_tokens": 616993.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.861191368103027,
|
|
"epoch": 0.2609727164887307,
|
|
"grad_norm": 0.55859375,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.415,
|
|
"mean_token_accuracy": 0.032618867419660094,
|
|
"num_tokens": 630717.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.825675582885742,
|
|
"epoch": 0.2669039145907473,
|
|
"grad_norm": 0.55859375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.4089,
|
|
"mean_token_accuracy": 0.03296402785927057,
|
|
"num_tokens": 646826.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.793451166152954,
|
|
"epoch": 0.27283511269276395,
|
|
"grad_norm": 0.54296875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.4121,
|
|
"mean_token_accuracy": 0.03432573340833187,
|
|
"num_tokens": 662277.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.786031770706177,
|
|
"epoch": 0.27876631079478054,
|
|
"grad_norm": 0.60546875,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.4215,
|
|
"mean_token_accuracy": 0.03696324098855257,
|
|
"num_tokens": 676522.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.828532743453979,
|
|
"epoch": 0.2846975088967972,
|
|
"grad_norm": 0.59375,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.5228,
|
|
"mean_token_accuracy": 0.033165539428591725,
|
|
"num_tokens": 690425.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.870665788650513,
|
|
"epoch": 0.29062870699881377,
|
|
"grad_norm": 0.84765625,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.5003,
|
|
"mean_token_accuracy": 0.037998564168810846,
|
|
"num_tokens": 704629.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.887278270721436,
|
|
"epoch": 0.29655990510083036,
|
|
"grad_norm": 0.609375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.4081,
|
|
"mean_token_accuracy": 0.0402442242950201,
|
|
"num_tokens": 721777.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.88170485496521,
|
|
"epoch": 0.302491103202847,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.3309,
|
|
"mean_token_accuracy": 0.04053100124001503,
|
|
"num_tokens": 734712.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.850059461593628,
|
|
"epoch": 0.3084223013048636,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.2525,
|
|
"mean_token_accuracy": 0.044293657690286634,
|
|
"num_tokens": 750222.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.8487138748168945,
|
|
"epoch": 0.3143534994068802,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.4068,
|
|
"mean_token_accuracy": 0.038819624856114385,
|
|
"num_tokens": 764699.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.8781726360321045,
|
|
"epoch": 0.3202846975088968,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.3138,
|
|
"mean_token_accuracy": 0.04715476129204035,
|
|
"num_tokens": 778738.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.895059490203858,
|
|
"epoch": 0.3262158956109134,
|
|
"grad_norm": 0.75,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.3011,
|
|
"mean_token_accuracy": 0.04708148837089539,
|
|
"num_tokens": 793904.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.921955537796021,
|
|
"epoch": 0.33214709371293,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.3829,
|
|
"mean_token_accuracy": 0.05229228027164936,
|
|
"num_tokens": 808310.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.833415269851685,
|
|
"epoch": 0.33807829181494664,
|
|
"grad_norm": 0.8359375,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.235,
|
|
"mean_token_accuracy": 0.05603795982897282,
|
|
"num_tokens": 826253.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.822495317459106,
|
|
"epoch": 0.34400948991696323,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.1784,
|
|
"mean_token_accuracy": 0.05520997829735279,
|
|
"num_tokens": 839635.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.788468694686889,
|
|
"epoch": 0.3499406880189798,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.2809,
|
|
"mean_token_accuracy": 0.05334235914051533,
|
|
"num_tokens": 853870.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.822190046310425,
|
|
"epoch": 0.35587188612099646,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.2341,
|
|
"mean_token_accuracy": 0.05465950295329094,
|
|
"num_tokens": 868316.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.7704771041870115,
|
|
"epoch": 0.36180308422301305,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.1512,
|
|
"mean_token_accuracy": 0.05797473154962063,
|
|
"num_tokens": 881412.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.757186412811279,
|
|
"epoch": 0.36773428232502964,
|
|
"grad_norm": 0.88671875,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.1676,
|
|
"mean_token_accuracy": 0.06134983114898205,
|
|
"num_tokens": 895656.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.723504638671875,
|
|
"epoch": 0.3736654804270463,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.0343,
|
|
"mean_token_accuracy": 0.06331710368394852,
|
|
"num_tokens": 909339.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.670499658584594,
|
|
"epoch": 0.37959667852906287,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.136,
|
|
"mean_token_accuracy": 0.05801789090037346,
|
|
"num_tokens": 922838.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.7397764205932615,
|
|
"epoch": 0.38552787663107946,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.1359,
|
|
"mean_token_accuracy": 0.06155713014304638,
|
|
"num_tokens": 936767.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.730058193206787,
|
|
"epoch": 0.3914590747330961,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.1752,
|
|
"mean_token_accuracy": 0.05990114398300648,
|
|
"num_tokens": 954078.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.619611692428589,
|
|
"epoch": 0.3973902728351127,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.0933,
|
|
"mean_token_accuracy": 0.06247694194316864,
|
|
"num_tokens": 968714.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.707681131362915,
|
|
"epoch": 0.4033214709371293,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.0391,
|
|
"mean_token_accuracy": 0.07253243401646614,
|
|
"num_tokens": 983001.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.655651664733886,
|
|
"epoch": 0.4092526690391459,
|
|
"grad_norm": 0.78125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.0613,
|
|
"mean_token_accuracy": 0.06838090978562832,
|
|
"num_tokens": 997501.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.6562416553497314,
|
|
"epoch": 0.4151838671411625,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.0229,
|
|
"mean_token_accuracy": 0.07049233056604862,
|
|
"num_tokens": 1011683.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.5962440967559814,
|
|
"epoch": 0.4211150652431791,
|
|
"grad_norm": 1.5703125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.9731,
|
|
"mean_token_accuracy": 0.0693504311144352,
|
|
"num_tokens": 1024871.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.58204460144043,
|
|
"epoch": 0.42704626334519574,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.9777,
|
|
"mean_token_accuracy": 0.07172644883394241,
|
|
"num_tokens": 1038776.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.63423752784729,
|
|
"epoch": 0.43297746144721233,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.9462,
|
|
"mean_token_accuracy": 0.07613400146365165,
|
|
"num_tokens": 1052457.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.596765565872192,
|
|
"epoch": 0.4389086595492289,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.985,
|
|
"mean_token_accuracy": 0.06885807923972606,
|
|
"num_tokens": 1067566.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.509483337402344,
|
|
"epoch": 0.44483985765124556,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.8577,
|
|
"mean_token_accuracy": 0.0751548632979393,
|
|
"num_tokens": 1081285.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.488412714004516,
|
|
"epoch": 0.45077105575326215,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.8741,
|
|
"mean_token_accuracy": 0.08314598426222801,
|
|
"num_tokens": 1093944.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.485768795013428,
|
|
"epoch": 0.45670225385527874,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.8729,
|
|
"mean_token_accuracy": 0.08215565755963325,
|
|
"num_tokens": 1107210.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.5015223026275635,
|
|
"epoch": 0.4626334519572954,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.9484,
|
|
"mean_token_accuracy": 0.07512684762477875,
|
|
"num_tokens": 1120035.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.528394508361816,
|
|
"epoch": 0.46856465005931197,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.9877,
|
|
"mean_token_accuracy": 0.07711658254265785,
|
|
"num_tokens": 1132735.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.506278848648071,
|
|
"epoch": 0.4744958481613286,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.8883,
|
|
"mean_token_accuracy": 0.07609864845871925,
|
|
"num_tokens": 1146964.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.526645755767822,
|
|
"epoch": 0.4804270462633452,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.8871,
|
|
"mean_token_accuracy": 0.07854933217167855,
|
|
"num_tokens": 1161174.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.447538709640503,
|
|
"epoch": 0.4863582443653618,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.7717,
|
|
"mean_token_accuracy": 0.080623859167099,
|
|
"num_tokens": 1176833.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.474844169616699,
|
|
"epoch": 0.49228944246737844,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.9384,
|
|
"mean_token_accuracy": 0.07880584299564361,
|
|
"num_tokens": 1190845.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.3909095287322994,
|
|
"epoch": 0.498220640569395,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.8394,
|
|
"mean_token_accuracy": 0.07916948273777961,
|
|
"num_tokens": 1203878.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.414308786392212,
|
|
"epoch": 0.5041518386714117,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.8609,
|
|
"mean_token_accuracy": 0.07901249378919602,
|
|
"num_tokens": 1219135.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.502074193954468,
|
|
"epoch": 0.5100830367734283,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.8924,
|
|
"mean_token_accuracy": 0.08341019675135612,
|
|
"num_tokens": 1233167.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.4084759712219235,
|
|
"epoch": 0.5160142348754448,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.7389,
|
|
"mean_token_accuracy": 0.08747640326619148,
|
|
"num_tokens": 1246221.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.389473342895508,
|
|
"epoch": 0.5219454329774614,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.732,
|
|
"mean_token_accuracy": 0.08541311249136925,
|
|
"num_tokens": 1260856.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.329696989059448,
|
|
"epoch": 0.527876631079478,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.8368,
|
|
"mean_token_accuracy": 0.08163531869649887,
|
|
"num_tokens": 1276688.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.337572288513184,
|
|
"epoch": 0.5338078291814946,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.6987,
|
|
"mean_token_accuracy": 0.08712242096662522,
|
|
"num_tokens": 1291525.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.3344158172607425,
|
|
"epoch": 0.5397390272835113,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.7266,
|
|
"mean_token_accuracy": 0.08258413746953011,
|
|
"num_tokens": 1307769.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.2570515155792235,
|
|
"epoch": 0.5456702253855279,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.7115,
|
|
"mean_token_accuracy": 0.08097989335656167,
|
|
"num_tokens": 1322462.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.2473875999450685,
|
|
"epoch": 0.5516014234875445,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.6807,
|
|
"mean_token_accuracy": 0.08938496559858322,
|
|
"num_tokens": 1338063.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.260567998886108,
|
|
"epoch": 0.5575326215895611,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.6452,
|
|
"mean_token_accuracy": 0.09352756068110465,
|
|
"num_tokens": 1351384.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.236065340042114,
|
|
"epoch": 0.5634638196915777,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.6144,
|
|
"mean_token_accuracy": 0.08987314701080322,
|
|
"num_tokens": 1363781.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.16572265625,
|
|
"epoch": 0.5693950177935944,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.6055,
|
|
"mean_token_accuracy": 0.08821886777877808,
|
|
"num_tokens": 1379463.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.195259714126587,
|
|
"epoch": 0.575326215895611,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.6073,
|
|
"mean_token_accuracy": 0.09338206499814987,
|
|
"num_tokens": 1393631.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.307022905349731,
|
|
"epoch": 0.5812574139976275,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.6937,
|
|
"mean_token_accuracy": 0.08968389555811881,
|
|
"num_tokens": 1407401.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.144880056381226,
|
|
"epoch": 0.5871886120996441,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.5565,
|
|
"mean_token_accuracy": 0.09463186785578728,
|
|
"num_tokens": 1422167.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.16906509399414,
|
|
"epoch": 0.5931198102016607,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.6659,
|
|
"mean_token_accuracy": 0.08580891266465188,
|
|
"num_tokens": 1438662.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.5931198102016607,
|
|
"eval_entropy": 7.088789626864568,
|
|
"eval_loss": 6.709393501281738,
|
|
"eval_mean_token_accuracy": 0.09428181704362294,
|
|
"eval_num_tokens": 1438662.0,
|
|
"eval_runtime": 1.9148,
|
|
"eval_samples_per_second": 1362.036,
|
|
"eval_steps_per_second": 170.254,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 8420,
|
|
"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": 375536020193280.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|