bd86e71c20
Model: fpadovani/dan-latn-10mb-after-ppt-shuff-dyck-10mb-ckpt500_seed3407 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": 0.41841004184100417,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 4.891409587860108,
|
|
"epoch": 0.0041841004184100415,
|
|
"grad_norm": 24.125,
|
|
"learning_rate": 2e-06,
|
|
"loss": 14.1621,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 9848.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 4.9063023090362545,
|
|
"epoch": 0.008368200836820083,
|
|
"grad_norm": 25.0,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 14.0066,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 18196.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 4.956460618972779,
|
|
"epoch": 0.012552301255230125,
|
|
"grad_norm": 27.25,
|
|
"learning_rate": 7e-06,
|
|
"loss": 13.9152,
|
|
"mean_token_accuracy": 0.00022191319148987532,
|
|
"num_tokens": 26927.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 5.071391153335571,
|
|
"epoch": 0.016736401673640166,
|
|
"grad_norm": 34.5,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 13.5124,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 34987.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 5.681827545166016,
|
|
"epoch": 0.02092050209205021,
|
|
"grad_norm": 42.5,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 12.7729,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 44936.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 7.654202032089233,
|
|
"epoch": 0.02510460251046025,
|
|
"grad_norm": 24.5,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 11.6009,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 54562.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.472756958007812,
|
|
"epoch": 0.029288702928870293,
|
|
"grad_norm": 3.703125,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.6978,
|
|
"mean_token_accuracy": 0.005889179778750986,
|
|
"num_tokens": 63844.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.730471515655518,
|
|
"epoch": 0.03347280334728033,
|
|
"grad_norm": 3.296875,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.5735,
|
|
"mean_token_accuracy": 0.015836346428841352,
|
|
"num_tokens": 74331.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.730430603027344,
|
|
"epoch": 0.03765690376569038,
|
|
"grad_norm": 2.90625,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.3807,
|
|
"mean_token_accuracy": 0.019579107593744993,
|
|
"num_tokens": 83028.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.661893463134765,
|
|
"epoch": 0.04184100418410042,
|
|
"grad_norm": 2.265625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 10.2168,
|
|
"mean_token_accuracy": 0.020131182763725518,
|
|
"num_tokens": 91395.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.666141891479493,
|
|
"epoch": 0.04602510460251046,
|
|
"grad_norm": 2.578125,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 10.0907,
|
|
"mean_token_accuracy": 0.016700492519885302,
|
|
"num_tokens": 101868.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.728130531311034,
|
|
"epoch": 0.0502092050209205,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.9346,
|
|
"mean_token_accuracy": 0.02238125018775463,
|
|
"num_tokens": 110403.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.730528163909913,
|
|
"epoch": 0.05439330543933055,
|
|
"grad_norm": 1.8828125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.9073,
|
|
"mean_token_accuracy": 0.02675023004412651,
|
|
"num_tokens": 119557.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.718873310089112,
|
|
"epoch": 0.058577405857740586,
|
|
"grad_norm": 1.9765625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.7675,
|
|
"mean_token_accuracy": 0.024789097160100936,
|
|
"num_tokens": 128903.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.718735504150391,
|
|
"epoch": 0.06276150627615062,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.7466,
|
|
"mean_token_accuracy": 0.03036366552114487,
|
|
"num_tokens": 137630.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.720433235168457,
|
|
"epoch": 0.06694560669456066,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.6143,
|
|
"mean_token_accuracy": 0.03396541755646467,
|
|
"num_tokens": 146335.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.716510105133057,
|
|
"epoch": 0.07112970711297072,
|
|
"grad_norm": 1.9765625,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.5501,
|
|
"mean_token_accuracy": 0.03693135865032673,
|
|
"num_tokens": 155766.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.7073184967041,
|
|
"epoch": 0.07531380753138076,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.4819,
|
|
"mean_token_accuracy": 0.029630407132208346,
|
|
"num_tokens": 165844.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.698113632202148,
|
|
"epoch": 0.0794979079497908,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.3976,
|
|
"mean_token_accuracy": 0.040011851117014885,
|
|
"num_tokens": 174858.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.68637809753418,
|
|
"epoch": 0.08368200836820083,
|
|
"grad_norm": 1.859375,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.2692,
|
|
"mean_token_accuracy": 0.04454572442919016,
|
|
"num_tokens": 184020.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.661810874938965,
|
|
"epoch": 0.08786610878661087,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.1619,
|
|
"mean_token_accuracy": 0.044953469187021255,
|
|
"num_tokens": 193714.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.638691425323486,
|
|
"epoch": 0.09205020920502092,
|
|
"grad_norm": 1.875,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.0919,
|
|
"mean_token_accuracy": 0.04682178944349289,
|
|
"num_tokens": 203139.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.608509731292724,
|
|
"epoch": 0.09623430962343096,
|
|
"grad_norm": 1.859375,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.935,
|
|
"mean_token_accuracy": 0.05361910648643971,
|
|
"num_tokens": 212446.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.535148620605469,
|
|
"epoch": 0.100418410041841,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.7879,
|
|
"mean_token_accuracy": 0.05296131558716297,
|
|
"num_tokens": 220910.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.479767608642579,
|
|
"epoch": 0.10460251046025104,
|
|
"grad_norm": 1.6953125,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.7201,
|
|
"mean_token_accuracy": 0.05455400906503201,
|
|
"num_tokens": 230522.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.398798561096191,
|
|
"epoch": 0.1087866108786611,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.5632,
|
|
"mean_token_accuracy": 0.05519989021122455,
|
|
"num_tokens": 240092.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.31678285598755,
|
|
"epoch": 0.11297071129707113,
|
|
"grad_norm": 1.578125,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.469,
|
|
"mean_token_accuracy": 0.056049537286162375,
|
|
"num_tokens": 249478.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.232394123077393,
|
|
"epoch": 0.11715481171548117,
|
|
"grad_norm": 1.875,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.2797,
|
|
"mean_token_accuracy": 0.05960724465548992,
|
|
"num_tokens": 257714.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.128987789154053,
|
|
"epoch": 0.12133891213389121,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.2716,
|
|
"mean_token_accuracy": 0.054746852815151216,
|
|
"num_tokens": 266569.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 10.016991806030273,
|
|
"epoch": 0.12552301255230125,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.1522,
|
|
"mean_token_accuracy": 0.05324812866747379,
|
|
"num_tokens": 276688.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.930090236663819,
|
|
"epoch": 0.1297071129707113,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.969,
|
|
"mean_token_accuracy": 0.05717452354729176,
|
|
"num_tokens": 285430.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.747330951690675,
|
|
"epoch": 0.13389121338912133,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.8918,
|
|
"mean_token_accuracy": 0.05585745945572853,
|
|
"num_tokens": 295727.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.500588703155518,
|
|
"epoch": 0.13807531380753138,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.7962,
|
|
"mean_token_accuracy": 0.05966520681977272,
|
|
"num_tokens": 304854.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.284924411773682,
|
|
"epoch": 0.14225941422594143,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.7152,
|
|
"mean_token_accuracy": 0.06037968210875988,
|
|
"num_tokens": 313579.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.0167311668396,
|
|
"epoch": 0.14644351464435146,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.7018,
|
|
"mean_token_accuracy": 0.05947771333158016,
|
|
"num_tokens": 323310.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.793067455291748,
|
|
"epoch": 0.1506276150627615,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.5597,
|
|
"mean_token_accuracy": 0.054980911687016486,
|
|
"num_tokens": 331720.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.578739547729493,
|
|
"epoch": 0.15481171548117154,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.4836,
|
|
"mean_token_accuracy": 0.05964849032461643,
|
|
"num_tokens": 340331.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.347243404388427,
|
|
"epoch": 0.1589958158995816,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.4873,
|
|
"mean_token_accuracy": 0.05853373594582081,
|
|
"num_tokens": 349677.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.249125862121582,
|
|
"epoch": 0.16317991631799164,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.4457,
|
|
"mean_token_accuracy": 0.05640666857361794,
|
|
"num_tokens": 358674.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.108446025848389,
|
|
"epoch": 0.16736401673640167,
|
|
"grad_norm": 0.83203125,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.4564,
|
|
"mean_token_accuracy": 0.05889376923441887,
|
|
"num_tokens": 368258.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.07214674949646,
|
|
"epoch": 0.17154811715481172,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.3668,
|
|
"mean_token_accuracy": 0.06832478605210782,
|
|
"num_tokens": 377083.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.066781759262085,
|
|
"epoch": 0.17573221757322174,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.4131,
|
|
"mean_token_accuracy": 0.05864137038588524,
|
|
"num_tokens": 385456.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.966042947769165,
|
|
"epoch": 0.1799163179916318,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.3069,
|
|
"mean_token_accuracy": 0.06798864156007767,
|
|
"num_tokens": 395295.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.028728580474853,
|
|
"epoch": 0.18410041841004185,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.3528,
|
|
"mean_token_accuracy": 0.06413010321557522,
|
|
"num_tokens": 405023.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.929877233505249,
|
|
"epoch": 0.18828451882845187,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.3556,
|
|
"mean_token_accuracy": 0.06588337235152722,
|
|
"num_tokens": 414224.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.9531862258911135,
|
|
"epoch": 0.19246861924686193,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.2507,
|
|
"mean_token_accuracy": 0.06776628717780113,
|
|
"num_tokens": 423951.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.959342336654663,
|
|
"epoch": 0.19665271966527198,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.2897,
|
|
"mean_token_accuracy": 0.0697428148239851,
|
|
"num_tokens": 432427.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.9486246585845945,
|
|
"epoch": 0.200836820083682,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.1966,
|
|
"mean_token_accuracy": 0.07294891811907292,
|
|
"num_tokens": 441560.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.970178651809692,
|
|
"epoch": 0.20502092050209206,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.2917,
|
|
"mean_token_accuracy": 0.06312177963554859,
|
|
"num_tokens": 450364.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.8968164920806885,
|
|
"epoch": 0.20920502092050208,
|
|
"grad_norm": 0.86328125,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.2249,
|
|
"mean_token_accuracy": 0.07028925977647305,
|
|
"num_tokens": 460076.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.926723957061768,
|
|
"epoch": 0.21338912133891214,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.1878,
|
|
"mean_token_accuracy": 0.07347588390111923,
|
|
"num_tokens": 468718.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.886239528656006,
|
|
"epoch": 0.2175732217573222,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.2306,
|
|
"mean_token_accuracy": 0.07341902256011963,
|
|
"num_tokens": 478436.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.901343250274659,
|
|
"epoch": 0.2217573221757322,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.2425,
|
|
"mean_token_accuracy": 0.07287401147186756,
|
|
"num_tokens": 486833.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.8618532657623295,
|
|
"epoch": 0.22594142259414227,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.1112,
|
|
"mean_token_accuracy": 0.08160877451300622,
|
|
"num_tokens": 495119.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.881023740768432,
|
|
"epoch": 0.2301255230125523,
|
|
"grad_norm": 2.8125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.2378,
|
|
"mean_token_accuracy": 0.07865233719348907,
|
|
"num_tokens": 505396.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.713179016113282,
|
|
"epoch": 0.23430962343096234,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.1126,
|
|
"mean_token_accuracy": 0.07563550472259521,
|
|
"num_tokens": 515259.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.866283369064331,
|
|
"epoch": 0.2384937238493724,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.1237,
|
|
"mean_token_accuracy": 0.07651542238891125,
|
|
"num_tokens": 523608.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.718380069732666,
|
|
"epoch": 0.24267782426778242,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.123,
|
|
"mean_token_accuracy": 0.07430066615343094,
|
|
"num_tokens": 534613.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.648086595535278,
|
|
"epoch": 0.24686192468619247,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.1218,
|
|
"mean_token_accuracy": 0.08086465299129486,
|
|
"num_tokens": 543347.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.774254369735718,
|
|
"epoch": 0.2510460251046025,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.096,
|
|
"mean_token_accuracy": 0.07647029384970665,
|
|
"num_tokens": 552371.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.655633068084716,
|
|
"epoch": 0.25523012552301255,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.0231,
|
|
"mean_token_accuracy": 0.08150922022759914,
|
|
"num_tokens": 560742.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.669692802429199,
|
|
"epoch": 0.2594142259414226,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.1089,
|
|
"mean_token_accuracy": 0.08333353251218796,
|
|
"num_tokens": 569960.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.660837411880493,
|
|
"epoch": 0.26359832635983266,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.0674,
|
|
"mean_token_accuracy": 0.07987545132637024,
|
|
"num_tokens": 578859.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.71635251045227,
|
|
"epoch": 0.26778242677824265,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.0662,
|
|
"mean_token_accuracy": 0.08209685683250427,
|
|
"num_tokens": 588146.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.540524530410766,
|
|
"epoch": 0.2719665271966527,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.0169,
|
|
"mean_token_accuracy": 0.08419515639543533,
|
|
"num_tokens": 597268.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.600224113464355,
|
|
"epoch": 0.27615062761506276,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 6.9492,
|
|
"mean_token_accuracy": 0.08558865413069724,
|
|
"num_tokens": 606199.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.614625310897827,
|
|
"epoch": 0.2803347280334728,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.0511,
|
|
"mean_token_accuracy": 0.07545995935797692,
|
|
"num_tokens": 615966.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.659131479263306,
|
|
"epoch": 0.28451882845188287,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.026,
|
|
"mean_token_accuracy": 0.08085056543350219,
|
|
"num_tokens": 625689.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.566850900650024,
|
|
"epoch": 0.28870292887029286,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.061,
|
|
"mean_token_accuracy": 0.08259228393435478,
|
|
"num_tokens": 634877.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.583298397064209,
|
|
"epoch": 0.2928870292887029,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 6.9891,
|
|
"mean_token_accuracy": 0.08486707210540771,
|
|
"num_tokens": 644042.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.560001087188721,
|
|
"epoch": 0.29707112970711297,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.959,
|
|
"mean_token_accuracy": 0.08440612144768238,
|
|
"num_tokens": 653072.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.451096248626709,
|
|
"epoch": 0.301255230125523,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.9641,
|
|
"mean_token_accuracy": 0.08682873398065567,
|
|
"num_tokens": 663686.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.523260593414307,
|
|
"epoch": 0.3054393305439331,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.9034,
|
|
"mean_token_accuracy": 0.08619396463036537,
|
|
"num_tokens": 672293.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.444020557403564,
|
|
"epoch": 0.30962343096234307,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.867,
|
|
"mean_token_accuracy": 0.08871779963374138,
|
|
"num_tokens": 681758.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.431160640716553,
|
|
"epoch": 0.3138075313807531,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.8923,
|
|
"mean_token_accuracy": 0.08616945594549179,
|
|
"num_tokens": 690727.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.471702432632446,
|
|
"epoch": 0.3179916317991632,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.8543,
|
|
"mean_token_accuracy": 0.08554501757025719,
|
|
"num_tokens": 701210.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.374780702590942,
|
|
"epoch": 0.32217573221757323,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.8415,
|
|
"mean_token_accuracy": 0.08815910518169404,
|
|
"num_tokens": 710869.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.413737440109253,
|
|
"epoch": 0.3263598326359833,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.9429,
|
|
"mean_token_accuracy": 0.08294537365436554,
|
|
"num_tokens": 721455.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.397308969497681,
|
|
"epoch": 0.3305439330543933,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.8361,
|
|
"mean_token_accuracy": 0.09202742874622345,
|
|
"num_tokens": 730917.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.375412464141846,
|
|
"epoch": 0.33472803347280333,
|
|
"grad_norm": 1.71875,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.8569,
|
|
"mean_token_accuracy": 0.09109241589903831,
|
|
"num_tokens": 739302.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.446383142471314,
|
|
"epoch": 0.3389121338912134,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.8401,
|
|
"mean_token_accuracy": 0.08787635415792465,
|
|
"num_tokens": 748785.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.324094533920288,
|
|
"epoch": 0.34309623430962344,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.7754,
|
|
"mean_token_accuracy": 0.09026021733880044,
|
|
"num_tokens": 757215.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.308840560913086,
|
|
"epoch": 0.3472803347280335,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.7076,
|
|
"mean_token_accuracy": 0.09466831386089325,
|
|
"num_tokens": 765682.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.378483152389526,
|
|
"epoch": 0.3514644351464435,
|
|
"grad_norm": 1.828125,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.9007,
|
|
"mean_token_accuracy": 0.08436162248253823,
|
|
"num_tokens": 774814.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.392928743362427,
|
|
"epoch": 0.35564853556485354,
|
|
"grad_norm": 1.7578125,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.8148,
|
|
"mean_token_accuracy": 0.091542549431324,
|
|
"num_tokens": 784019.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.225587844848633,
|
|
"epoch": 0.3598326359832636,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.6954,
|
|
"mean_token_accuracy": 0.0938286691904068,
|
|
"num_tokens": 793237.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.288524055480957,
|
|
"epoch": 0.36401673640167365,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.7272,
|
|
"mean_token_accuracy": 0.09245252907276154,
|
|
"num_tokens": 801548.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.260332345962524,
|
|
"epoch": 0.3682008368200837,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.7725,
|
|
"mean_token_accuracy": 0.09467287361621857,
|
|
"num_tokens": 810974.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.250288343429565,
|
|
"epoch": 0.3723849372384937,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.6705,
|
|
"mean_token_accuracy": 0.09519676417112351,
|
|
"num_tokens": 820023.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.248160552978516,
|
|
"epoch": 0.37656903765690375,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.7607,
|
|
"mean_token_accuracy": 0.0901703730225563,
|
|
"num_tokens": 828297.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.292386770248413,
|
|
"epoch": 0.3807531380753138,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.807,
|
|
"mean_token_accuracy": 0.08642732873558998,
|
|
"num_tokens": 838268.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.304117202758789,
|
|
"epoch": 0.38493723849372385,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.7244,
|
|
"mean_token_accuracy": 0.09284975379705429,
|
|
"num_tokens": 846491.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.163941526412964,
|
|
"epoch": 0.3891213389121339,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.6779,
|
|
"mean_token_accuracy": 0.09622682482004166,
|
|
"num_tokens": 856501.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.226222562789917,
|
|
"epoch": 0.39330543933054396,
|
|
"grad_norm": 1.625,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.7058,
|
|
"mean_token_accuracy": 0.09297544807195664,
|
|
"num_tokens": 867054.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.1907635688781735,
|
|
"epoch": 0.39748953974895396,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.7284,
|
|
"mean_token_accuracy": 0.0978313222527504,
|
|
"num_tokens": 875956.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.051689147949219,
|
|
"epoch": 0.401673640167364,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.6452,
|
|
"mean_token_accuracy": 0.09968722686171531,
|
|
"num_tokens": 884427.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.282907915115357,
|
|
"epoch": 0.40585774058577406,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.7298,
|
|
"mean_token_accuracy": 0.09669127762317657,
|
|
"num_tokens": 894181.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.166344404220581,
|
|
"epoch": 0.4100418410041841,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.5834,
|
|
"mean_token_accuracy": 0.10224275141954423,
|
|
"num_tokens": 904076.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.067501544952393,
|
|
"epoch": 0.41422594142259417,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.7161,
|
|
"mean_token_accuracy": 0.09675642251968383,
|
|
"num_tokens": 913915.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.267646312713623,
|
|
"epoch": 0.41841004184100417,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.7693,
|
|
"mean_token_accuracy": 0.0949087992310524,
|
|
"num_tokens": 922876.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.41841004184100417,
|
|
"eval_entropy": 7.11538662354639,
|
|
"eval_loss": 6.712912559509277,
|
|
"eval_mean_token_accuracy": 0.1014816391915997,
|
|
"eval_num_tokens": 922876.0,
|
|
"eval_runtime": 2.2023,
|
|
"eval_samples_per_second": 1774.078,
|
|
"eval_steps_per_second": 222.044,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 11950,
|
|
"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": 200351734824960.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|