0709e1c753
Model: fpadovani/eng-latn-10mb-after-ppt-sort-curriculum_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": 0.5408328826392644,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 7.820888805389404,
|
|
"epoch": 0.005408328826392645,
|
|
"grad_norm": 7.65625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 13.4867,
|
|
"mean_token_accuracy": 0.011032939096912742,
|
|
"num_tokens": 10942.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 7.835813283920288,
|
|
"epoch": 0.01081665765278529,
|
|
"grad_norm": 7.8125,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 13.4729,
|
|
"mean_token_accuracy": 0.010750519391149283,
|
|
"num_tokens": 21887.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 7.866043424606323,
|
|
"epoch": 0.016224986479177934,
|
|
"grad_norm": 8.375,
|
|
"learning_rate": 7e-06,
|
|
"loss": 13.4343,
|
|
"mean_token_accuracy": 0.010895622195675969,
|
|
"num_tokens": 34345.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 7.880505704879761,
|
|
"epoch": 0.02163331530557058,
|
|
"grad_norm": 8.75,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 13.3238,
|
|
"mean_token_accuracy": 0.01247827559709549,
|
|
"num_tokens": 45445.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 7.962141990661621,
|
|
"epoch": 0.02704164413196322,
|
|
"grad_norm": 8.75,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 13.2274,
|
|
"mean_token_accuracy": 0.011504577286541463,
|
|
"num_tokens": 58237.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 8.106424951553345,
|
|
"epoch": 0.03244997295835587,
|
|
"grad_norm": 9.625,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 12.88,
|
|
"mean_token_accuracy": 0.01291654328815639,
|
|
"num_tokens": 68098.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 8.39170503616333,
|
|
"epoch": 0.03785830178474851,
|
|
"grad_norm": 9.4375,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 12.4981,
|
|
"mean_token_accuracy": 0.01724564200267196,
|
|
"num_tokens": 79867.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 8.898138904571534,
|
|
"epoch": 0.04326663061114116,
|
|
"grad_norm": 6.75,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 11.9446,
|
|
"mean_token_accuracy": 0.019018208887428046,
|
|
"num_tokens": 91257.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 9.481475162506104,
|
|
"epoch": 0.048674959437533805,
|
|
"grad_norm": 4.21875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 11.4652,
|
|
"mean_token_accuracy": 0.025027596950531007,
|
|
"num_tokens": 103474.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 9.992423725128173,
|
|
"epoch": 0.05408328826392644,
|
|
"grad_norm": 2.8125,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 11.0587,
|
|
"mean_token_accuracy": 0.024848397076129913,
|
|
"num_tokens": 114830.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.25519962310791,
|
|
"epoch": 0.05949161709031909,
|
|
"grad_norm": 2.15625,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 10.8236,
|
|
"mean_token_accuracy": 0.024381351098418234,
|
|
"num_tokens": 126017.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.374946212768554,
|
|
"epoch": 0.06489994591671173,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 10.7281,
|
|
"mean_token_accuracy": 0.028067474253475666,
|
|
"num_tokens": 138509.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.438200187683105,
|
|
"epoch": 0.07030827474310439,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 10.5835,
|
|
"mean_token_accuracy": 0.031993019953370096,
|
|
"num_tokens": 148353.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.488709259033204,
|
|
"epoch": 0.07571660356949703,
|
|
"grad_norm": 1.90625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 10.5393,
|
|
"mean_token_accuracy": 0.036837692372500896,
|
|
"num_tokens": 160269.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.523839092254638,
|
|
"epoch": 0.08112493239588967,
|
|
"grad_norm": 2.0,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 10.4312,
|
|
"mean_token_accuracy": 0.03589091971516609,
|
|
"num_tokens": 171568.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.546321773529053,
|
|
"epoch": 0.08653326122228232,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 10.3481,
|
|
"mean_token_accuracy": 0.0360051903873682,
|
|
"num_tokens": 184243.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.563430118560792,
|
|
"epoch": 0.09194159004867496,
|
|
"grad_norm": 2.0,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 10.2337,
|
|
"mean_token_accuracy": 0.03699877411127091,
|
|
"num_tokens": 194887.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.580722332000732,
|
|
"epoch": 0.09734991887506761,
|
|
"grad_norm": 1.953125,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 10.1385,
|
|
"mean_token_accuracy": 0.03550931885838508,
|
|
"num_tokens": 206476.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.59913902282715,
|
|
"epoch": 0.10275824770146025,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.9926,
|
|
"mean_token_accuracy": 0.039441929385066035,
|
|
"num_tokens": 216578.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.6052095413208,
|
|
"epoch": 0.10816657652785289,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.8908,
|
|
"mean_token_accuracy": 0.03636195547878742,
|
|
"num_tokens": 226941.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.607266902923584,
|
|
"epoch": 0.11357490535424554,
|
|
"grad_norm": 1.890625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.7695,
|
|
"mean_token_accuracy": 0.04152456596493721,
|
|
"num_tokens": 238060.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.605393600463866,
|
|
"epoch": 0.11898323418063818,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.6438,
|
|
"mean_token_accuracy": 0.0380807850509882,
|
|
"num_tokens": 249137.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.596975135803223,
|
|
"epoch": 0.12439156300703083,
|
|
"grad_norm": 1.96875,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.5365,
|
|
"mean_token_accuracy": 0.039401752315461634,
|
|
"num_tokens": 260712.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.588339138031007,
|
|
"epoch": 0.12979989183342347,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 9.3761,
|
|
"mean_token_accuracy": 0.03739211615175009,
|
|
"num_tokens": 271039.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.577256965637208,
|
|
"epoch": 0.1352082206598161,
|
|
"grad_norm": 2.125,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 9.2529,
|
|
"mean_token_accuracy": 0.04276929795742035,
|
|
"num_tokens": 281353.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.553887462615966,
|
|
"epoch": 0.14061654948620878,
|
|
"grad_norm": 1.8671875,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 9.1552,
|
|
"mean_token_accuracy": 0.039082702808082105,
|
|
"num_tokens": 292221.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.525830459594726,
|
|
"epoch": 0.1460248783126014,
|
|
"grad_norm": 1.875,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 9.0356,
|
|
"mean_token_accuracy": 0.04222068060189486,
|
|
"num_tokens": 303353.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.492454433441162,
|
|
"epoch": 0.15143320713899405,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.9125,
|
|
"mean_token_accuracy": 0.04995499067008495,
|
|
"num_tokens": 314165.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.450113677978516,
|
|
"epoch": 0.1568415359653867,
|
|
"grad_norm": 1.90625,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.7544,
|
|
"mean_token_accuracy": 0.04123007263988256,
|
|
"num_tokens": 326387.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 10.41439380645752,
|
|
"epoch": 0.16224986479177933,
|
|
"grad_norm": 1.8671875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.6823,
|
|
"mean_token_accuracy": 0.04394615534693003,
|
|
"num_tokens": 339155.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 10.365491104125976,
|
|
"epoch": 0.167658193618172,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.5153,
|
|
"mean_token_accuracy": 0.03965858686715364,
|
|
"num_tokens": 350643.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 10.289417934417724,
|
|
"epoch": 0.17306652244456464,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.3842,
|
|
"mean_token_accuracy": 0.043310442194342616,
|
|
"num_tokens": 361977.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 10.205756950378419,
|
|
"epoch": 0.17847485127095727,
|
|
"grad_norm": 1.703125,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.3159,
|
|
"mean_token_accuracy": 0.04538454748690128,
|
|
"num_tokens": 373604.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 10.099522399902344,
|
|
"epoch": 0.1838831800973499,
|
|
"grad_norm": 1.5625,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.1886,
|
|
"mean_token_accuracy": 0.04867874421179295,
|
|
"num_tokens": 384306.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.964338874816894,
|
|
"epoch": 0.18929150892374255,
|
|
"grad_norm": 1.5234375,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.0329,
|
|
"mean_token_accuracy": 0.049391471594572064,
|
|
"num_tokens": 396479.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 9.800485134124756,
|
|
"epoch": 0.19469983775013522,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.9052,
|
|
"mean_token_accuracy": 0.049093520268797874,
|
|
"num_tokens": 408489.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 9.617319297790527,
|
|
"epoch": 0.20010816657652786,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.8279,
|
|
"mean_token_accuracy": 0.04878813587129116,
|
|
"num_tokens": 419339.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 9.397418594360351,
|
|
"epoch": 0.2055164954029205,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.7213,
|
|
"mean_token_accuracy": 0.048774123191833496,
|
|
"num_tokens": 430486.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 9.130149364471436,
|
|
"epoch": 0.21092482422931313,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.6542,
|
|
"mean_token_accuracy": 0.05405276417732239,
|
|
"num_tokens": 440414.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.82462682723999,
|
|
"epoch": 0.21633315305570577,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.5581,
|
|
"mean_token_accuracy": 0.05878009013831616,
|
|
"num_tokens": 451475.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.563184356689453,
|
|
"epoch": 0.22174148188209844,
|
|
"grad_norm": 0.765625,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.5674,
|
|
"mean_token_accuracy": 0.06661942265927792,
|
|
"num_tokens": 462053.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.337231540679932,
|
|
"epoch": 0.22714981070849108,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.5121,
|
|
"mean_token_accuracy": 0.06138431690633297,
|
|
"num_tokens": 473207.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.17379379272461,
|
|
"epoch": 0.23255813953488372,
|
|
"grad_norm": 0.73046875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.4845,
|
|
"mean_token_accuracy": 0.06530819572508335,
|
|
"num_tokens": 484999.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.07261848449707,
|
|
"epoch": 0.23796646836127636,
|
|
"grad_norm": 0.734375,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.4822,
|
|
"mean_token_accuracy": 0.0644452515989542,
|
|
"num_tokens": 496772.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.936752462387085,
|
|
"epoch": 0.24337479718766902,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.4461,
|
|
"mean_token_accuracy": 0.06566701531410217,
|
|
"num_tokens": 507782.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.9283324718475345,
|
|
"epoch": 0.24878312601406166,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.366,
|
|
"mean_token_accuracy": 0.06866429261863231,
|
|
"num_tokens": 518950.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.8786633014678955,
|
|
"epoch": 0.2541914548404543,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.4323,
|
|
"mean_token_accuracy": 0.06760875023901462,
|
|
"num_tokens": 530795.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.858989715576172,
|
|
"epoch": 0.25959978366684694,
|
|
"grad_norm": 0.8203125,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.4452,
|
|
"mean_token_accuracy": 0.06729580424726009,
|
|
"num_tokens": 542878.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.774307203292847,
|
|
"epoch": 0.2650081124932396,
|
|
"grad_norm": 0.796875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.3401,
|
|
"mean_token_accuracy": 0.07460085228085518,
|
|
"num_tokens": 553682.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.707247972488403,
|
|
"epoch": 0.2704164413196322,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.3454,
|
|
"mean_token_accuracy": 0.07869893908500672,
|
|
"num_tokens": 564433.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.7594318866729735,
|
|
"epoch": 0.27582477014602486,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.3708,
|
|
"mean_token_accuracy": 0.07336842231452465,
|
|
"num_tokens": 576001.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.717825031280517,
|
|
"epoch": 0.28123309897241755,
|
|
"grad_norm": 0.66015625,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.3119,
|
|
"mean_token_accuracy": 0.07627563402056695,
|
|
"num_tokens": 587138.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.747607707977295,
|
|
"epoch": 0.2866414277988102,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.2671,
|
|
"mean_token_accuracy": 0.08127573654055595,
|
|
"num_tokens": 597572.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.623431205749512,
|
|
"epoch": 0.2920497566252028,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.2359,
|
|
"mean_token_accuracy": 0.08385064378380776,
|
|
"num_tokens": 608879.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.603728199005127,
|
|
"epoch": 0.29745808545159547,
|
|
"grad_norm": 0.79296875,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.2864,
|
|
"mean_token_accuracy": 0.0743965484201908,
|
|
"num_tokens": 620439.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.6806840896606445,
|
|
"epoch": 0.3028664142779881,
|
|
"grad_norm": 0.78515625,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.3331,
|
|
"mean_token_accuracy": 0.0795133963227272,
|
|
"num_tokens": 632543.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.567217206954956,
|
|
"epoch": 0.30827474310438074,
|
|
"grad_norm": 0.83203125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.3707,
|
|
"mean_token_accuracy": 0.0822471871972084,
|
|
"num_tokens": 643369.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.696122837066651,
|
|
"epoch": 0.3136830719307734,
|
|
"grad_norm": 0.69140625,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.2433,
|
|
"mean_token_accuracy": 0.07976862639188767,
|
|
"num_tokens": 655328.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.673062705993653,
|
|
"epoch": 0.319091400757166,
|
|
"grad_norm": 0.9609375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.3085,
|
|
"mean_token_accuracy": 0.08202499970793724,
|
|
"num_tokens": 665780.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.630768585205078,
|
|
"epoch": 0.32449972958355866,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.2183,
|
|
"mean_token_accuracy": 0.08841075897216796,
|
|
"num_tokens": 676566.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.61181788444519,
|
|
"epoch": 0.3299080584099513,
|
|
"grad_norm": 0.79296875,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.2092,
|
|
"mean_token_accuracy": 0.08559464290738106,
|
|
"num_tokens": 687407.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.709430265426636,
|
|
"epoch": 0.335316387236344,
|
|
"grad_norm": 0.73828125,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.1314,
|
|
"mean_token_accuracy": 0.08758908063173294,
|
|
"num_tokens": 698162.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.624366712570191,
|
|
"epoch": 0.34072471606273663,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.1345,
|
|
"mean_token_accuracy": 0.08050442710518838,
|
|
"num_tokens": 708743.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.5649957180023195,
|
|
"epoch": 0.34613304488912927,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.3075,
|
|
"mean_token_accuracy": 0.08924563452601433,
|
|
"num_tokens": 719828.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.668677186965942,
|
|
"epoch": 0.3515413737155219,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.275,
|
|
"mean_token_accuracy": 0.0808012131601572,
|
|
"num_tokens": 732514.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.571430635452271,
|
|
"epoch": 0.35694970254191455,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.1181,
|
|
"mean_token_accuracy": 0.08540427088737487,
|
|
"num_tokens": 742391.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.561294841766357,
|
|
"epoch": 0.3623580313683072,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.1192,
|
|
"mean_token_accuracy": 0.07976256236433983,
|
|
"num_tokens": 753638.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.643744564056396,
|
|
"epoch": 0.3677663601946998,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.3106,
|
|
"mean_token_accuracy": 0.08193527534604073,
|
|
"num_tokens": 764850.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.528887176513672,
|
|
"epoch": 0.37317468902109246,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.0978,
|
|
"mean_token_accuracy": 0.0877677209675312,
|
|
"num_tokens": 776192.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.488804197311401,
|
|
"epoch": 0.3785830178474851,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.0452,
|
|
"mean_token_accuracy": 0.09009440615773201,
|
|
"num_tokens": 787051.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.453413486480713,
|
|
"epoch": 0.3839913466738778,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.0818,
|
|
"mean_token_accuracy": 0.08464901968836784,
|
|
"num_tokens": 799281.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.47227840423584,
|
|
"epoch": 0.38939967550027044,
|
|
"grad_norm": 0.80078125,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.0947,
|
|
"mean_token_accuracy": 0.08406926393508911,
|
|
"num_tokens": 810003.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.510661792755127,
|
|
"epoch": 0.3948080043266631,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.0577,
|
|
"mean_token_accuracy": 0.08685395568609237,
|
|
"num_tokens": 819749.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.452503824234009,
|
|
"epoch": 0.4002163331530557,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.0694,
|
|
"mean_token_accuracy": 0.08692100048065185,
|
|
"num_tokens": 830552.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.486446285247803,
|
|
"epoch": 0.40562466197944835,
|
|
"grad_norm": 0.875,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.1663,
|
|
"mean_token_accuracy": 0.09074072763323784,
|
|
"num_tokens": 840495.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.408175134658814,
|
|
"epoch": 0.411032990805841,
|
|
"grad_norm": 0.73828125,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.0813,
|
|
"mean_token_accuracy": 0.0877058044075966,
|
|
"num_tokens": 852915.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.418009948730469,
|
|
"epoch": 0.41644131963223363,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.0365,
|
|
"mean_token_accuracy": 0.0944710485637188,
|
|
"num_tokens": 862762.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.437512016296386,
|
|
"epoch": 0.42184964845862627,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.9974,
|
|
"mean_token_accuracy": 0.09396275207400322,
|
|
"num_tokens": 873023.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.443737745285034,
|
|
"epoch": 0.4272579772850189,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.0967,
|
|
"mean_token_accuracy": 0.08894659876823426,
|
|
"num_tokens": 885281.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.369609642028808,
|
|
"epoch": 0.43266630611141155,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.0939,
|
|
"mean_token_accuracy": 0.09273864775896072,
|
|
"num_tokens": 896210.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.422218275070191,
|
|
"epoch": 0.43807463493780424,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 0.000202,
|
|
"loss": 7.0119,
|
|
"mean_token_accuracy": 0.09136949926614761,
|
|
"num_tokens": 909028.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.3562903881073,
|
|
"epoch": 0.4434829637641969,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.9403,
|
|
"mean_token_accuracy": 0.09579638168215751,
|
|
"num_tokens": 919710.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.465255784988403,
|
|
"epoch": 0.4488912925905895,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.0033,
|
|
"mean_token_accuracy": 0.09164737239480018,
|
|
"num_tokens": 930019.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.330646514892578,
|
|
"epoch": 0.45429962141698216,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.9277,
|
|
"mean_token_accuracy": 0.09849832803010941,
|
|
"num_tokens": 940535.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.225619316101074,
|
|
"epoch": 0.4597079502433748,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.8877,
|
|
"mean_token_accuracy": 0.10022106692194939,
|
|
"num_tokens": 951390.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.338269996643066,
|
|
"epoch": 0.46511627906976744,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.9181,
|
|
"mean_token_accuracy": 0.10022875815629959,
|
|
"num_tokens": 963010.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.250108957290649,
|
|
"epoch": 0.4705246078961601,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.8709,
|
|
"mean_token_accuracy": 0.10117963030934334,
|
|
"num_tokens": 974032.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.42537899017334,
|
|
"epoch": 0.4759329367225527,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.9201,
|
|
"mean_token_accuracy": 0.10263018682599068,
|
|
"num_tokens": 983954.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.279693603515625,
|
|
"epoch": 0.48134126554894535,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.9332,
|
|
"mean_token_accuracy": 0.10275022611021996,
|
|
"num_tokens": 994135.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.299053049087524,
|
|
"epoch": 0.48674959437533805,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.8981,
|
|
"mean_token_accuracy": 0.10349252223968505,
|
|
"num_tokens": 1006275.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.163119125366211,
|
|
"epoch": 0.4921579232017307,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.7321,
|
|
"mean_token_accuracy": 0.10733722224831581,
|
|
"num_tokens": 1017767.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.236297607421875,
|
|
"epoch": 0.4975662520281233,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.8507,
|
|
"mean_token_accuracy": 0.1048464372754097,
|
|
"num_tokens": 1028286.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.231361341476441,
|
|
"epoch": 0.502974580854516,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.8625,
|
|
"mean_token_accuracy": 0.1016843743622303,
|
|
"num_tokens": 1040555.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.289519309997559,
|
|
"epoch": 0.5083829096809086,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.9481,
|
|
"mean_token_accuracy": 0.09957458972930908,
|
|
"num_tokens": 1053668.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.211695528030395,
|
|
"epoch": 0.5137912385073012,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.891,
|
|
"mean_token_accuracy": 0.10144442990422249,
|
|
"num_tokens": 1066042.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.23263463973999,
|
|
"epoch": 0.5191995673336939,
|
|
"grad_norm": 0.9609375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.8999,
|
|
"mean_token_accuracy": 0.10248073190450668,
|
|
"num_tokens": 1077843.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.1748435497283936,
|
|
"epoch": 0.5246078961600865,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.794,
|
|
"mean_token_accuracy": 0.10380105674266815,
|
|
"num_tokens": 1089448.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.180902671813965,
|
|
"epoch": 0.5300162249864792,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.8816,
|
|
"mean_token_accuracy": 0.1034023180603981,
|
|
"num_tokens": 1100989.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.378252983093262,
|
|
"epoch": 0.5354245538128718,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.9212,
|
|
"mean_token_accuracy": 0.10497874468564987,
|
|
"num_tokens": 1113012.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.092680644989014,
|
|
"epoch": 0.5408328826392644,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.8695,
|
|
"mean_token_accuracy": 0.10232841521501541,
|
|
"num_tokens": 1123551.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.5408328826392644,
|
|
"eval_entropy": 7.271889178319411,
|
|
"eval_loss": 7.0051140785217285,
|
|
"eval_mean_token_accuracy": 0.10177409166300838,
|
|
"eval_num_tokens": 1123551.0,
|
|
"eval_runtime": 4.518,
|
|
"eval_samples_per_second": 778.002,
|
|
"eval_steps_per_second": 97.389,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 9240,
|
|
"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": 259611278376960.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|