a8cd0c01a6
Model: fpadovani/heb-hebr-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": 0.5339028296849974,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 4.68021821975708,
|
|
"epoch": 0.005339028296849973,
|
|
"grad_norm": 31.625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 13.9711,
|
|
"mean_token_accuracy": 0.009074668399989605,
|
|
"num_tokens": 10658.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 4.704092264175415,
|
|
"epoch": 0.010678056593699947,
|
|
"grad_norm": 32.75,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 13.9298,
|
|
"mean_token_accuracy": 0.00866024480201304,
|
|
"num_tokens": 21327.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 4.7672216415405275,
|
|
"epoch": 0.01601708489054992,
|
|
"grad_norm": 36.5,
|
|
"learning_rate": 7e-06,
|
|
"loss": 13.7823,
|
|
"mean_token_accuracy": 0.007704685814678669,
|
|
"num_tokens": 32403.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 4.958004379272461,
|
|
"epoch": 0.021356113187399894,
|
|
"grad_norm": 40.5,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 13.3459,
|
|
"mean_token_accuracy": 0.006649625021964312,
|
|
"num_tokens": 43471.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 5.477138376235962,
|
|
"epoch": 0.026695141484249868,
|
|
"grad_norm": 38.75,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 12.4632,
|
|
"mean_token_accuracy": 0.0020860590098891406,
|
|
"num_tokens": 53724.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 6.8245813846588135,
|
|
"epoch": 0.03203416978109984,
|
|
"grad_norm": 30.375,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 11.6429,
|
|
"mean_token_accuracy": 0.0007934643130283803,
|
|
"num_tokens": 65115.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.22093849182129,
|
|
"epoch": 0.03737319807794982,
|
|
"grad_norm": 5.46875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.8521,
|
|
"mean_token_accuracy": 0.0002344618726056069,
|
|
"num_tokens": 75902.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.684853076934814,
|
|
"epoch": 0.04271222637479979,
|
|
"grad_norm": 4.0625,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.7564,
|
|
"mean_token_accuracy": 0.0011873395298607648,
|
|
"num_tokens": 87051.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.69392204284668,
|
|
"epoch": 0.04805125467164976,
|
|
"grad_norm": 2.84375,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.5885,
|
|
"mean_token_accuracy": 0.006381610780954361,
|
|
"num_tokens": 98998.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.600316715240478,
|
|
"epoch": 0.053390282968499736,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 10.4445,
|
|
"mean_token_accuracy": 0.011505381856113672,
|
|
"num_tokens": 111207.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.673304462432862,
|
|
"epoch": 0.05872931126534971,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 10.2638,
|
|
"mean_token_accuracy": 0.016619019117206334,
|
|
"num_tokens": 121039.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.670036125183106,
|
|
"epoch": 0.06406833956219968,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 10.1533,
|
|
"mean_token_accuracy": 0.015064884908497334,
|
|
"num_tokens": 132604.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.671291160583497,
|
|
"epoch": 0.06940736785904965,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 10.0181,
|
|
"mean_token_accuracy": 0.03150726184248924,
|
|
"num_tokens": 142825.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.688156223297119,
|
|
"epoch": 0.07474639615589963,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.9346,
|
|
"mean_token_accuracy": 0.056240572780370715,
|
|
"num_tokens": 153121.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.670577144622802,
|
|
"epoch": 0.0800854244527496,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.8809,
|
|
"mean_token_accuracy": 0.05582823716104031,
|
|
"num_tokens": 164675.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.664248943328857,
|
|
"epoch": 0.08542445274959957,
|
|
"grad_norm": 1.5703125,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.8352,
|
|
"mean_token_accuracy": 0.04962811656296253,
|
|
"num_tokens": 175146.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.648977661132813,
|
|
"epoch": 0.09076348104644955,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.8124,
|
|
"mean_token_accuracy": 0.04548483826220036,
|
|
"num_tokens": 186177.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.642164325714111,
|
|
"epoch": 0.09610250934329952,
|
|
"grad_norm": 1.6640625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.685,
|
|
"mean_token_accuracy": 0.048678891733288765,
|
|
"num_tokens": 197143.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.636518096923828,
|
|
"epoch": 0.10144153764014949,
|
|
"grad_norm": 1.625,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.6182,
|
|
"mean_token_accuracy": 0.052110004052519796,
|
|
"num_tokens": 207979.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.610146522521973,
|
|
"epoch": 0.10678056593699947,
|
|
"grad_norm": 1.640625,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.5135,
|
|
"mean_token_accuracy": 0.0568073820322752,
|
|
"num_tokens": 218733.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.569573020935058,
|
|
"epoch": 0.11211959423384944,
|
|
"grad_norm": 1.625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.4144,
|
|
"mean_token_accuracy": 0.05693732798099518,
|
|
"num_tokens": 228887.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.517056083679199,
|
|
"epoch": 0.11745862253069941,
|
|
"grad_norm": 1.5390625,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.3275,
|
|
"mean_token_accuracy": 0.05374570451676845,
|
|
"num_tokens": 240289.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.480204105377197,
|
|
"epoch": 0.12279765082754938,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.2538,
|
|
"mean_token_accuracy": 0.05631314069032669,
|
|
"num_tokens": 250548.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.422281837463379,
|
|
"epoch": 0.12813667912439936,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 9.1616,
|
|
"mean_token_accuracy": 0.05645494796335697,
|
|
"num_tokens": 261669.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.368197917938232,
|
|
"epoch": 0.13347570742124934,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 9.0945,
|
|
"mean_token_accuracy": 0.05810702405869961,
|
|
"num_tokens": 272926.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.291021633148194,
|
|
"epoch": 0.1388147357180993,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 9.0261,
|
|
"mean_token_accuracy": 0.05110162906348705,
|
|
"num_tokens": 284173.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.235691547393799,
|
|
"epoch": 0.14415376401494928,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.9153,
|
|
"mean_token_accuracy": 0.05686397515237331,
|
|
"num_tokens": 294850.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.190592193603516,
|
|
"epoch": 0.14949279231179927,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.8582,
|
|
"mean_token_accuracy": 0.05346960201859474,
|
|
"num_tokens": 305727.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.10048542022705,
|
|
"epoch": 0.15483182060864922,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.7462,
|
|
"mean_token_accuracy": 0.058417992666363716,
|
|
"num_tokens": 317243.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.970914077758788,
|
|
"epoch": 0.1601708489054992,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.7387,
|
|
"mean_token_accuracy": 0.05338601246476173,
|
|
"num_tokens": 328334.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.846445274353027,
|
|
"epoch": 0.16550987720234916,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.6482,
|
|
"mean_token_accuracy": 0.054720787703990935,
|
|
"num_tokens": 339774.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.73973569869995,
|
|
"epoch": 0.17084890549919915,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.5543,
|
|
"mean_token_accuracy": 0.05573372282087803,
|
|
"num_tokens": 350211.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.550614070892333,
|
|
"epoch": 0.1761879337960491,
|
|
"grad_norm": 0.78125,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.4684,
|
|
"mean_token_accuracy": 0.054642071947455406,
|
|
"num_tokens": 360109.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.380690956115723,
|
|
"epoch": 0.1815269620928991,
|
|
"grad_norm": 0.8359375,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.4131,
|
|
"mean_token_accuracy": 0.054827965795993805,
|
|
"num_tokens": 370911.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.282039070129395,
|
|
"epoch": 0.18686599038974908,
|
|
"grad_norm": 0.76953125,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.3991,
|
|
"mean_token_accuracy": 0.060196897760033605,
|
|
"num_tokens": 381808.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 9.13407211303711,
|
|
"epoch": 0.19220501868659903,
|
|
"grad_norm": 0.86328125,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 8.3091,
|
|
"mean_token_accuracy": 0.06106638610363006,
|
|
"num_tokens": 392175.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 9.000042152404784,
|
|
"epoch": 0.19754404698344902,
|
|
"grad_norm": 0.78125,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 8.3874,
|
|
"mean_token_accuracy": 0.049684269726276396,
|
|
"num_tokens": 403261.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.931573295593262,
|
|
"epoch": 0.20288307528029897,
|
|
"grad_norm": 0.7265625,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 8.2853,
|
|
"mean_token_accuracy": 0.05462104417383671,
|
|
"num_tokens": 414992.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.791546726226807,
|
|
"epoch": 0.20822210357714896,
|
|
"grad_norm": 0.6328125,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 8.2732,
|
|
"mean_token_accuracy": 0.05595530979335308,
|
|
"num_tokens": 425316.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.729821968078614,
|
|
"epoch": 0.21356113187399894,
|
|
"grad_norm": 0.75,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 8.344,
|
|
"mean_token_accuracy": 0.05437092892825603,
|
|
"num_tokens": 437759.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.712938117980958,
|
|
"epoch": 0.2189001601708489,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.000102,
|
|
"loss": 8.2844,
|
|
"mean_token_accuracy": 0.05534115061163902,
|
|
"num_tokens": 448066.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.651909160614014,
|
|
"epoch": 0.22423918846769889,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 8.2206,
|
|
"mean_token_accuracy": 0.055615832656621934,
|
|
"num_tokens": 458632.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.63439655303955,
|
|
"epoch": 0.22957821676454884,
|
|
"grad_norm": 0.76171875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 8.2606,
|
|
"mean_token_accuracy": 0.054173830151557925,
|
|
"num_tokens": 469224.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.596088218688966,
|
|
"epoch": 0.23491724506139883,
|
|
"grad_norm": 0.875,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 8.2823,
|
|
"mean_token_accuracy": 0.05661116987466812,
|
|
"num_tokens": 480492.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.625334930419921,
|
|
"epoch": 0.2402562733582488,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.000112,
|
|
"loss": 8.2398,
|
|
"mean_token_accuracy": 0.057639367505908014,
|
|
"num_tokens": 491325.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.569988918304443,
|
|
"epoch": 0.24559530165509877,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 8.2465,
|
|
"mean_token_accuracy": 0.0603976845741272,
|
|
"num_tokens": 502240.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.592293834686279,
|
|
"epoch": 0.25093432995194875,
|
|
"grad_norm": 0.84375,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 8.2035,
|
|
"mean_token_accuracy": 0.06263627484440804,
|
|
"num_tokens": 512710.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.58193883895874,
|
|
"epoch": 0.2562733582487987,
|
|
"grad_norm": 0.71484375,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 8.2513,
|
|
"mean_token_accuracy": 0.05789706408977509,
|
|
"num_tokens": 524023.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.547989177703858,
|
|
"epoch": 0.26161238654564867,
|
|
"grad_norm": 3.234375,
|
|
"learning_rate": 0.000122,
|
|
"loss": 8.1422,
|
|
"mean_token_accuracy": 0.06462252773344516,
|
|
"num_tokens": 534612.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.475037765502929,
|
|
"epoch": 0.2669514148424987,
|
|
"grad_norm": 0.68359375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 8.2036,
|
|
"mean_token_accuracy": 0.05881951674818993,
|
|
"num_tokens": 544866.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.527704906463622,
|
|
"epoch": 0.27229044313934864,
|
|
"grad_norm": 0.84375,
|
|
"learning_rate": 0.000127,
|
|
"loss": 8.2523,
|
|
"mean_token_accuracy": 0.055548783764243125,
|
|
"num_tokens": 556091.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.524747467041015,
|
|
"epoch": 0.2776294714361986,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 8.2536,
|
|
"mean_token_accuracy": 0.05479630120098591,
|
|
"num_tokens": 566946.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 8.54754581451416,
|
|
"epoch": 0.2829684997330486,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.000132,
|
|
"loss": 8.1371,
|
|
"mean_token_accuracy": 0.060486416518688205,
|
|
"num_tokens": 577164.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.525337314605713,
|
|
"epoch": 0.28830752802989856,
|
|
"grad_norm": 0.671875,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 8.1768,
|
|
"mean_token_accuracy": 0.05711319632828236,
|
|
"num_tokens": 588744.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.505339241027832,
|
|
"epoch": 0.2936465563267485,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 8.1727,
|
|
"mean_token_accuracy": 0.055232659354805945,
|
|
"num_tokens": 600943.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.492932891845703,
|
|
"epoch": 0.29898558462359853,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 8.1726,
|
|
"mean_token_accuracy": 0.06307518370449543,
|
|
"num_tokens": 611864.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.451044940948487,
|
|
"epoch": 0.3043246129204485,
|
|
"grad_norm": 0.75,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 8.152,
|
|
"mean_token_accuracy": 0.05931071788072586,
|
|
"num_tokens": 623314.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.428306388854981,
|
|
"epoch": 0.30966364121729845,
|
|
"grad_norm": 0.69921875,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 8.1763,
|
|
"mean_token_accuracy": 0.06025952883064747,
|
|
"num_tokens": 633509.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 8.451824760437011,
|
|
"epoch": 0.3150026695141484,
|
|
"grad_norm": 0.83984375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 8.0983,
|
|
"mean_token_accuracy": 0.06582859642803669,
|
|
"num_tokens": 644655.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.374590492248535,
|
|
"epoch": 0.3203416978109984,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 8.135,
|
|
"mean_token_accuracy": 0.05644083172082901,
|
|
"num_tokens": 655940.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 8.408124351501465,
|
|
"epoch": 0.3256807261078484,
|
|
"grad_norm": 0.765625,
|
|
"learning_rate": 0.000152,
|
|
"loss": 8.1809,
|
|
"mean_token_accuracy": 0.056555739417672155,
|
|
"num_tokens": 667074.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 8.498903751373291,
|
|
"epoch": 0.33101975440469833,
|
|
"grad_norm": 0.703125,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 8.2309,
|
|
"mean_token_accuracy": 0.06025502309203148,
|
|
"num_tokens": 677942.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.455632591247559,
|
|
"epoch": 0.33635878270154834,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.000157,
|
|
"loss": 8.1642,
|
|
"mean_token_accuracy": 0.06168182455003261,
|
|
"num_tokens": 689226.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 8.403513431549072,
|
|
"epoch": 0.3416978109983983,
|
|
"grad_norm": 0.76171875,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 8.1786,
|
|
"mean_token_accuracy": 0.059801208600401876,
|
|
"num_tokens": 700830.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.456379795074463,
|
|
"epoch": 0.34703683929524826,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.000162,
|
|
"loss": 8.1377,
|
|
"mean_token_accuracy": 0.06001129299402237,
|
|
"num_tokens": 712677.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 8.363616943359375,
|
|
"epoch": 0.3523758675920982,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 8.2395,
|
|
"mean_token_accuracy": 0.05579938478767872,
|
|
"num_tokens": 723833.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 8.497436428070069,
|
|
"epoch": 0.3577148958889482,
|
|
"grad_norm": 0.828125,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 8.1771,
|
|
"mean_token_accuracy": 0.059333324804902075,
|
|
"num_tokens": 735526.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 8.438710403442382,
|
|
"epoch": 0.3630539241857982,
|
|
"grad_norm": 0.69140625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 8.1223,
|
|
"mean_token_accuracy": 0.06136825978755951,
|
|
"num_tokens": 745934.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.391829586029052,
|
|
"epoch": 0.36839295248264814,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 8.0837,
|
|
"mean_token_accuracy": 0.06251248680055141,
|
|
"num_tokens": 756693.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 8.393558979034424,
|
|
"epoch": 0.37373198077949815,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 8.1418,
|
|
"mean_token_accuracy": 0.05746800377964974,
|
|
"num_tokens": 767236.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 8.491841030120849,
|
|
"epoch": 0.3790710090763481,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000177,
|
|
"loss": 8.1254,
|
|
"mean_token_accuracy": 0.06065751984715462,
|
|
"num_tokens": 778552.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 8.428674793243408,
|
|
"epoch": 0.38441003737319807,
|
|
"grad_norm": 0.79296875,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 8.1788,
|
|
"mean_token_accuracy": 0.05917464569211006,
|
|
"num_tokens": 789615.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 8.48459882736206,
|
|
"epoch": 0.3897490656700481,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 0.000182,
|
|
"loss": 8.1355,
|
|
"mean_token_accuracy": 0.06367055512964725,
|
|
"num_tokens": 799998.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 8.322556686401366,
|
|
"epoch": 0.39508809396689804,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 8.0679,
|
|
"mean_token_accuracy": 0.062298642471432686,
|
|
"num_tokens": 810042.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 8.422898483276366,
|
|
"epoch": 0.400427122263748,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 8.0755,
|
|
"mean_token_accuracy": 0.05906784273684025,
|
|
"num_tokens": 821692.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 8.36357364654541,
|
|
"epoch": 0.40576615056059795,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 8.1164,
|
|
"mean_token_accuracy": 0.06748287416994572,
|
|
"num_tokens": 832262.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 8.333745193481445,
|
|
"epoch": 0.41110517885744796,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 0.000192,
|
|
"loss": 8.04,
|
|
"mean_token_accuracy": 0.06151870898902416,
|
|
"num_tokens": 843006.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 8.353145503997803,
|
|
"epoch": 0.4164442071542979,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 8.0332,
|
|
"mean_token_accuracy": 0.062286530062556265,
|
|
"num_tokens": 853978.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 8.357436656951904,
|
|
"epoch": 0.4217832354511479,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 8.0743,
|
|
"mean_token_accuracy": 0.05977092795073986,
|
|
"num_tokens": 864634.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.423241233825683,
|
|
"epoch": 0.4271222637479979,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 8.0568,
|
|
"mean_token_accuracy": 0.06662314645946026,
|
|
"num_tokens": 875379.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 8.398690319061279,
|
|
"epoch": 0.43246129204484784,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 8.0674,
|
|
"mean_token_accuracy": 0.06020735241472721,
|
|
"num_tokens": 886428.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.425256729125977,
|
|
"epoch": 0.4378003203416978,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 8.0827,
|
|
"mean_token_accuracy": 0.0645828690379858,
|
|
"num_tokens": 896700.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 8.369815444946289,
|
|
"epoch": 0.44313934863854776,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 8.0577,
|
|
"mean_token_accuracy": 0.06499057859182358,
|
|
"num_tokens": 907478.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 8.39303207397461,
|
|
"epoch": 0.44847837693539777,
|
|
"grad_norm": 1.5234375,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.9707,
|
|
"mean_token_accuracy": 0.07010989226400852,
|
|
"num_tokens": 918170.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 8.28305778503418,
|
|
"epoch": 0.45381740523224773,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 0.000212,
|
|
"loss": 8.0345,
|
|
"mean_token_accuracy": 0.06685064136981964,
|
|
"num_tokens": 928866.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 8.279892349243164,
|
|
"epoch": 0.4591564335290977,
|
|
"grad_norm": 0.80859375,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 8.0205,
|
|
"mean_token_accuracy": 0.06419647410511971,
|
|
"num_tokens": 940395.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 8.475063610076905,
|
|
"epoch": 0.4644954618259477,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 8.0365,
|
|
"mean_token_accuracy": 0.062472108751535416,
|
|
"num_tokens": 950814.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 8.192297077178955,
|
|
"epoch": 0.46983449012279765,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 8.0225,
|
|
"mean_token_accuracy": 0.0661535993218422,
|
|
"num_tokens": 961658.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 8.43046703338623,
|
|
"epoch": 0.4751735184196476,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.000222,
|
|
"loss": 8.0993,
|
|
"mean_token_accuracy": 0.06110986284911633,
|
|
"num_tokens": 974790.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 8.331879615783691,
|
|
"epoch": 0.4805125467164976,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.9772,
|
|
"mean_token_accuracy": 0.06222324036061764,
|
|
"num_tokens": 986599.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 8.243362045288086,
|
|
"epoch": 0.4858515750133476,
|
|
"grad_norm": 0.84375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.9787,
|
|
"mean_token_accuracy": 0.06861450895667076,
|
|
"num_tokens": 997731.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 8.290034008026122,
|
|
"epoch": 0.49119060331019754,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 8.0374,
|
|
"mean_token_accuracy": 0.06714494079351425,
|
|
"num_tokens": 1009343.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 8.351651859283447,
|
|
"epoch": 0.4965296316070475,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.9541,
|
|
"mean_token_accuracy": 0.0648241501301527,
|
|
"num_tokens": 1019878.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 8.248888206481933,
|
|
"epoch": 0.5018686599038975,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 7.9553,
|
|
"mean_token_accuracy": 0.07038948647677898,
|
|
"num_tokens": 1030551.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 8.30935525894165,
|
|
"epoch": 0.5072076882007475,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.000237,
|
|
"loss": 7.9893,
|
|
"mean_token_accuracy": 0.06879530549049377,
|
|
"num_tokens": 1041432.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 8.283312702178955,
|
|
"epoch": 0.5125467164975974,
|
|
"grad_norm": 0.84375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.9704,
|
|
"mean_token_accuracy": 0.06568947955965995,
|
|
"num_tokens": 1051769.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 8.256877326965332,
|
|
"epoch": 0.5178857447944474,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000242,
|
|
"loss": 7.9282,
|
|
"mean_token_accuracy": 0.06400737687945365,
|
|
"num_tokens": 1062334.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 8.27625093460083,
|
|
"epoch": 0.5232247730912973,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 7.921,
|
|
"mean_token_accuracy": 0.06769514791667461,
|
|
"num_tokens": 1072792.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 8.173498821258544,
|
|
"epoch": 0.5285638013881474,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000247,
|
|
"loss": 7.8148,
|
|
"mean_token_accuracy": 0.07047795243561268,
|
|
"num_tokens": 1083642.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 8.192062282562256,
|
|
"epoch": 0.5339028296849974,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.9111,
|
|
"mean_token_accuracy": 0.07375151067972183,
|
|
"num_tokens": 1094454.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.5339028296849974,
|
|
"eval_entropy": 8.031556909734553,
|
|
"eval_loss": 8.007415771484375,
|
|
"eval_mean_token_accuracy": 0.06165650523715205,
|
|
"eval_num_tokens": 1094454.0,
|
|
"eval_runtime": 1.8278,
|
|
"eval_samples_per_second": 1685.073,
|
|
"eval_steps_per_second": 210.634,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 9360,
|
|
"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": 236538190725120.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|