1046 lines
28 KiB
JSON
1046 lines
28 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.42973785990545765,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.69186782836914,
|
|
"epoch": 0.004297378599054577,
|
|
"grad_norm": 14.0625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.767,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 8559.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.691853904724121,
|
|
"epoch": 0.008594757198109154,
|
|
"grad_norm": 13.4375,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7343,
|
|
"mean_token_accuracy": 0.0006453584705013782,
|
|
"num_tokens": 18235.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.68972110748291,
|
|
"epoch": 0.01289213579716373,
|
|
"grad_norm": 10.375,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.4481,
|
|
"mean_token_accuracy": 0.04634292460978031,
|
|
"num_tokens": 28257.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.653952026367188,
|
|
"epoch": 0.017189514396218308,
|
|
"grad_norm": 6.28125,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.1016,
|
|
"mean_token_accuracy": 0.05583016015589237,
|
|
"num_tokens": 38142.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.492306423187255,
|
|
"epoch": 0.021486892995272882,
|
|
"grad_norm": 3.90625,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 9.8388,
|
|
"mean_token_accuracy": 0.051388037577271464,
|
|
"num_tokens": 47363.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.393956565856934,
|
|
"epoch": 0.02578427159432746,
|
|
"grad_norm": 3.09375,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 9.6855,
|
|
"mean_token_accuracy": 0.05032752677798271,
|
|
"num_tokens": 57121.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.352927017211915,
|
|
"epoch": 0.030081650193382038,
|
|
"grad_norm": 2.71875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 9.59,
|
|
"mean_token_accuracy": 0.056775762140750884,
|
|
"num_tokens": 66785.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.365454959869385,
|
|
"epoch": 0.034379028792436615,
|
|
"grad_norm": 2.671875,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.4974,
|
|
"mean_token_accuracy": 0.05595241449773312,
|
|
"num_tokens": 76313.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.268372726440429,
|
|
"epoch": 0.03867640739149119,
|
|
"grad_norm": 2.25,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.546,
|
|
"mean_token_accuracy": 0.05818333774805069,
|
|
"num_tokens": 86151.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.402025985717774,
|
|
"epoch": 0.042973785990545764,
|
|
"grad_norm": 2.484375,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.4872,
|
|
"mean_token_accuracy": 0.06535410135984421,
|
|
"num_tokens": 94916.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.304343509674073,
|
|
"epoch": 0.047271164589600345,
|
|
"grad_norm": 2.265625,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.4098,
|
|
"mean_token_accuracy": 0.06619026586413383,
|
|
"num_tokens": 104592.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.089207649230957,
|
|
"epoch": 0.05156854318865492,
|
|
"grad_norm": 2.28125,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.317,
|
|
"mean_token_accuracy": 0.07636640071868897,
|
|
"num_tokens": 113942.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.178072547912597,
|
|
"epoch": 0.055865921787709494,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.329,
|
|
"mean_token_accuracy": 0.07130967006087303,
|
|
"num_tokens": 123190.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.223089408874511,
|
|
"epoch": 0.060163300386764075,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.3231,
|
|
"mean_token_accuracy": 0.06454960964620113,
|
|
"num_tokens": 132367.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.157601451873779,
|
|
"epoch": 0.06446067898581866,
|
|
"grad_norm": 1.8828125,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.1389,
|
|
"mean_token_accuracy": 0.07756163887679576,
|
|
"num_tokens": 141375.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.095236587524415,
|
|
"epoch": 0.06875805758487323,
|
|
"grad_norm": 1.859375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.0625,
|
|
"mean_token_accuracy": 0.07905065715312957,
|
|
"num_tokens": 150791.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.089058017730713,
|
|
"epoch": 0.0730554361839278,
|
|
"grad_norm": 1.96875,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.0294,
|
|
"mean_token_accuracy": 0.07317423708736896,
|
|
"num_tokens": 158686.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.112849235534668,
|
|
"epoch": 0.07735281478298238,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 8.9757,
|
|
"mean_token_accuracy": 0.06819208525121212,
|
|
"num_tokens": 167482.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 9.973779201507568,
|
|
"epoch": 0.08165019338203695,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.802,
|
|
"mean_token_accuracy": 0.08643188402056694,
|
|
"num_tokens": 176598.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 9.932288265228271,
|
|
"epoch": 0.08594757198109153,
|
|
"grad_norm": 1.8984375,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.7014,
|
|
"mean_token_accuracy": 0.08371126018464566,
|
|
"num_tokens": 185883.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 9.903890705108642,
|
|
"epoch": 0.09024495058014612,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.672,
|
|
"mean_token_accuracy": 0.0767141304910183,
|
|
"num_tokens": 194820.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 9.905889225006103,
|
|
"epoch": 0.09454232917920069,
|
|
"grad_norm": 1.8671875,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.5994,
|
|
"mean_token_accuracy": 0.07663930654525757,
|
|
"num_tokens": 203710.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 9.7015700340271,
|
|
"epoch": 0.09883970777825526,
|
|
"grad_norm": 1.640625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.4883,
|
|
"mean_token_accuracy": 0.08735396713018417,
|
|
"num_tokens": 212642.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.712564849853516,
|
|
"epoch": 0.10313708637730984,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.4829,
|
|
"mean_token_accuracy": 0.07560812048614025,
|
|
"num_tokens": 221411.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.618709659576416,
|
|
"epoch": 0.10743446497636441,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.276,
|
|
"mean_token_accuracy": 0.07886423245072365,
|
|
"num_tokens": 230454.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.445035266876221,
|
|
"epoch": 0.11173184357541899,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.2553,
|
|
"mean_token_accuracy": 0.08224238455295563,
|
|
"num_tokens": 240272.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.343283653259277,
|
|
"epoch": 0.11602922217447358,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.1773,
|
|
"mean_token_accuracy": 0.07860668338835239,
|
|
"num_tokens": 249652.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.141777420043946,
|
|
"epoch": 0.12032660077352815,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.0586,
|
|
"mean_token_accuracy": 0.08432014659047127,
|
|
"num_tokens": 258359.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 9.028361320495605,
|
|
"epoch": 0.12462397937258272,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.965,
|
|
"mean_token_accuracy": 0.08845296204090118,
|
|
"num_tokens": 266863.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.872007942199707,
|
|
"epoch": 0.1289213579716373,
|
|
"grad_norm": 1.703125,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.9485,
|
|
"mean_token_accuracy": 0.08544271066784859,
|
|
"num_tokens": 274694.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.746376705169677,
|
|
"epoch": 0.1332187365706919,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.998,
|
|
"mean_token_accuracy": 0.07424286939203739,
|
|
"num_tokens": 285038.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.652623462677003,
|
|
"epoch": 0.13751611516974646,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.9434,
|
|
"mean_token_accuracy": 0.07819288298487663,
|
|
"num_tokens": 293451.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.498311519622803,
|
|
"epoch": 0.14181349376880104,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.7897,
|
|
"mean_token_accuracy": 0.08914674445986748,
|
|
"num_tokens": 302184.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.397662544250489,
|
|
"epoch": 0.1461108723678556,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.8596,
|
|
"mean_token_accuracy": 0.08310422226786614,
|
|
"num_tokens": 311069.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.409068965911866,
|
|
"epoch": 0.15040825096691018,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.8454,
|
|
"mean_token_accuracy": 0.07755065970122814,
|
|
"num_tokens": 319841.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.261888885498047,
|
|
"epoch": 0.15470562956596476,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.7995,
|
|
"mean_token_accuracy": 0.0760527990758419,
|
|
"num_tokens": 328621.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.25572681427002,
|
|
"epoch": 0.15900300816501933,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.7913,
|
|
"mean_token_accuracy": 0.07687021419405937,
|
|
"num_tokens": 336469.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.273696994781494,
|
|
"epoch": 0.1633003867640739,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.8148,
|
|
"mean_token_accuracy": 0.07753114774823189,
|
|
"num_tokens": 344859.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.1952486038208,
|
|
"epoch": 0.16759776536312848,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.8017,
|
|
"mean_token_accuracy": 0.08184871897101402,
|
|
"num_tokens": 354536.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.130875205993652,
|
|
"epoch": 0.17189514396218306,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.782,
|
|
"mean_token_accuracy": 0.08327300399541855,
|
|
"num_tokens": 363592.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.071674633026124,
|
|
"epoch": 0.17619252256123766,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.7545,
|
|
"mean_token_accuracy": 0.08299842327833176,
|
|
"num_tokens": 373278.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.152774953842163,
|
|
"epoch": 0.18048990116029223,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.8002,
|
|
"mean_token_accuracy": 0.08091530725359916,
|
|
"num_tokens": 382498.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.122348976135253,
|
|
"epoch": 0.1847872797593468,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.7393,
|
|
"mean_token_accuracy": 0.09220762960612774,
|
|
"num_tokens": 391273.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.141310453414917,
|
|
"epoch": 0.18908465835840138,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.6711,
|
|
"mean_token_accuracy": 0.07944022566080093,
|
|
"num_tokens": 400340.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.08761625289917,
|
|
"epoch": 0.19338203695745596,
|
|
"grad_norm": 1.6171875,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.7203,
|
|
"mean_token_accuracy": 0.07989789322018623,
|
|
"num_tokens": 409162.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.068865013122558,
|
|
"epoch": 0.19767941555651053,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.6955,
|
|
"mean_token_accuracy": 0.08263300582766533,
|
|
"num_tokens": 417345.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.020773887634277,
|
|
"epoch": 0.2019767941555651,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.6885,
|
|
"mean_token_accuracy": 0.08577688783407211,
|
|
"num_tokens": 427488.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.00288724899292,
|
|
"epoch": 0.20627417275461968,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.6381,
|
|
"mean_token_accuracy": 0.08082681894302368,
|
|
"num_tokens": 435496.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.024074983596801,
|
|
"epoch": 0.21057155135367425,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.592,
|
|
"mean_token_accuracy": 0.08299721553921699,
|
|
"num_tokens": 444290.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.002217769622803,
|
|
"epoch": 0.21486892995272883,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.7047,
|
|
"mean_token_accuracy": 0.0830993577837944,
|
|
"num_tokens": 452704.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.036895895004273,
|
|
"epoch": 0.2191663085517834,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.6246,
|
|
"mean_token_accuracy": 0.08470720425248146,
|
|
"num_tokens": 461445.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.038204240798951,
|
|
"epoch": 0.22346368715083798,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.6401,
|
|
"mean_token_accuracy": 0.08966793864965439,
|
|
"num_tokens": 470644.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.892484998703003,
|
|
"epoch": 0.22776106574989258,
|
|
"grad_norm": 1.671875,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.5766,
|
|
"mean_token_accuracy": 0.08981397226452828,
|
|
"num_tokens": 479212.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.9553313732147215,
|
|
"epoch": 0.23205844434894715,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.6849,
|
|
"mean_token_accuracy": 0.07991519421339036,
|
|
"num_tokens": 489027.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.001228666305542,
|
|
"epoch": 0.23635582294800173,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.7078,
|
|
"mean_token_accuracy": 0.0864037737250328,
|
|
"num_tokens": 497548.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.95409255027771,
|
|
"epoch": 0.2406532015470563,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.523,
|
|
"mean_token_accuracy": 0.09064189270138741,
|
|
"num_tokens": 505836.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.008459424972534,
|
|
"epoch": 0.24495058014611087,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.6857,
|
|
"mean_token_accuracy": 0.08223827481269837,
|
|
"num_tokens": 515463.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.042573165893554,
|
|
"epoch": 0.24924795874516545,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.5642,
|
|
"mean_token_accuracy": 0.08510619327425957,
|
|
"num_tokens": 524882.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.878521060943603,
|
|
"epoch": 0.25354533734422,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.5627,
|
|
"mean_token_accuracy": 0.08412178531289101,
|
|
"num_tokens": 534523.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.9167151927948,
|
|
"epoch": 0.2578427159432746,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.5608,
|
|
"mean_token_accuracy": 0.08494351357221604,
|
|
"num_tokens": 543772.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.85640025138855,
|
|
"epoch": 0.26214009454232917,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.5374,
|
|
"mean_token_accuracy": 0.0830413281917572,
|
|
"num_tokens": 552024.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.888205242156983,
|
|
"epoch": 0.2664374731413838,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.5269,
|
|
"mean_token_accuracy": 0.08928481489419937,
|
|
"num_tokens": 561114.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.020092916488647,
|
|
"epoch": 0.2707348517404383,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.5728,
|
|
"mean_token_accuracy": 0.08862031325697899,
|
|
"num_tokens": 569655.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.867624950408936,
|
|
"epoch": 0.2750322303394929,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.5604,
|
|
"mean_token_accuracy": 0.08833226263523102,
|
|
"num_tokens": 578701.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.848216819763183,
|
|
"epoch": 0.27932960893854747,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.5257,
|
|
"mean_token_accuracy": 0.08904304206371308,
|
|
"num_tokens": 588166.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.775031900405883,
|
|
"epoch": 0.28362698753760207,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.5406,
|
|
"mean_token_accuracy": 0.08560147807002068,
|
|
"num_tokens": 598862.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.946314001083374,
|
|
"epoch": 0.2879243661366566,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.5054,
|
|
"mean_token_accuracy": 0.08752046972513199,
|
|
"num_tokens": 607680.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.798824310302734,
|
|
"epoch": 0.2922217447357112,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.5409,
|
|
"mean_token_accuracy": 0.08527569100260735,
|
|
"num_tokens": 617194.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.840095901489258,
|
|
"epoch": 0.29651912333476577,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.4681,
|
|
"mean_token_accuracy": 0.09199495688080787,
|
|
"num_tokens": 625946.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.863882684707642,
|
|
"epoch": 0.30081650193382037,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.4424,
|
|
"mean_token_accuracy": 0.09303685426712036,
|
|
"num_tokens": 634293.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.814202308654785,
|
|
"epoch": 0.30511388053287497,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.4365,
|
|
"mean_token_accuracy": 0.08837181255221367,
|
|
"num_tokens": 643479.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.780666065216065,
|
|
"epoch": 0.3094112591319295,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.4565,
|
|
"mean_token_accuracy": 0.0916847325861454,
|
|
"num_tokens": 653599.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.727724599838257,
|
|
"epoch": 0.3137086377309841,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.4396,
|
|
"mean_token_accuracy": 0.0935065358877182,
|
|
"num_tokens": 663042.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.7599810600280765,
|
|
"epoch": 0.31800601633003867,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.3799,
|
|
"mean_token_accuracy": 0.09194350391626357,
|
|
"num_tokens": 672254.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.850320053100586,
|
|
"epoch": 0.32230339492909327,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.4208,
|
|
"mean_token_accuracy": 0.09203241541981697,
|
|
"num_tokens": 681648.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.66558690071106,
|
|
"epoch": 0.3266007735281478,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.3756,
|
|
"mean_token_accuracy": 0.09696918800473213,
|
|
"num_tokens": 691080.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.681433391571045,
|
|
"epoch": 0.3308981521272024,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.4656,
|
|
"mean_token_accuracy": 0.08918402977287769,
|
|
"num_tokens": 700549.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.880253696441651,
|
|
"epoch": 0.33519553072625696,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.5014,
|
|
"mean_token_accuracy": 0.09039791896939278,
|
|
"num_tokens": 710623.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.726108407974243,
|
|
"epoch": 0.33949290932531156,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.4269,
|
|
"mean_token_accuracy": 0.09598973840475082,
|
|
"num_tokens": 719695.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.658704042434692,
|
|
"epoch": 0.3437902879243661,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.3817,
|
|
"mean_token_accuracy": 0.09576035216450692,
|
|
"num_tokens": 728998.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.817303800582886,
|
|
"epoch": 0.3480876665234207,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.000202,
|
|
"loss": 7.4597,
|
|
"mean_token_accuracy": 0.08954662233591079,
|
|
"num_tokens": 738537.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.673436260223388,
|
|
"epoch": 0.3523850451224753,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 7.4039,
|
|
"mean_token_accuracy": 0.092615757137537,
|
|
"num_tokens": 747555.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.643343782424926,
|
|
"epoch": 0.35668242372152986,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.3747,
|
|
"mean_token_accuracy": 0.09669034704566001,
|
|
"num_tokens": 756857.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.74475588798523,
|
|
"epoch": 0.36097980232058446,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.3802,
|
|
"mean_token_accuracy": 0.10021467804908753,
|
|
"num_tokens": 766181.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.649684762954712,
|
|
"epoch": 0.365277180919639,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.000212,
|
|
"loss": 7.4102,
|
|
"mean_token_accuracy": 0.09460910633206368,
|
|
"num_tokens": 774894.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.730149173736573,
|
|
"epoch": 0.3695745595186936,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.4062,
|
|
"mean_token_accuracy": 0.090732641518116,
|
|
"num_tokens": 784601.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.6217389583587645,
|
|
"epoch": 0.37387193811774816,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 7.4417,
|
|
"mean_token_accuracy": 0.09168312624096871,
|
|
"num_tokens": 794373.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.600076055526733,
|
|
"epoch": 0.37816931671680276,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 7.3455,
|
|
"mean_token_accuracy": 0.09434943050146102,
|
|
"num_tokens": 803608.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.752554702758789,
|
|
"epoch": 0.3824666953158573,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.3726,
|
|
"mean_token_accuracy": 0.09299151077866555,
|
|
"num_tokens": 812287.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.518798732757569,
|
|
"epoch": 0.3867640739149119,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.2985,
|
|
"mean_token_accuracy": 0.0987668864428997,
|
|
"num_tokens": 821049.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.701573610305786,
|
|
"epoch": 0.39106145251396646,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.3033,
|
|
"mean_token_accuracy": 0.10087452456355095,
|
|
"num_tokens": 829148.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.614718294143676,
|
|
"epoch": 0.39535883111302106,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 7.338,
|
|
"mean_token_accuracy": 0.09538405686616898,
|
|
"num_tokens": 838228.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.615123701095581,
|
|
"epoch": 0.39965620971207566,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.3246,
|
|
"mean_token_accuracy": 0.09597658589482308,
|
|
"num_tokens": 846059.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.705140447616577,
|
|
"epoch": 0.4039535883111302,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 7.4473,
|
|
"mean_token_accuracy": 0.09011114612221718,
|
|
"num_tokens": 856309.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.4943718910217285,
|
|
"epoch": 0.4082509669101848,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000237,
|
|
"loss": 7.2909,
|
|
"mean_token_accuracy": 0.1010310634970665,
|
|
"num_tokens": 865921.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.549269151687622,
|
|
"epoch": 0.41254834550923936,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.2654,
|
|
"mean_token_accuracy": 0.10411950796842576,
|
|
"num_tokens": 875076.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.629985094070435,
|
|
"epoch": 0.41684572410829396,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000242,
|
|
"loss": 7.3165,
|
|
"mean_token_accuracy": 0.0936420701444149,
|
|
"num_tokens": 884586.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.624928522109985,
|
|
"epoch": 0.4211431027073485,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 7.4262,
|
|
"mean_token_accuracy": 0.09749620333313942,
|
|
"num_tokens": 894192.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.579270219802856,
|
|
"epoch": 0.4254404813064031,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000247,
|
|
"loss": 7.3251,
|
|
"mean_token_accuracy": 0.09591478407382965,
|
|
"num_tokens": 903542.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.532130336761474,
|
|
"epoch": 0.42973785990545765,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.259,
|
|
"mean_token_accuracy": 0.10112758129835128,
|
|
"num_tokens": 912625.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.42973785990545765,
|
|
"eval_entropy": 7.499067693143277,
|
|
"eval_loss": 7.302952289581299,
|
|
"eval_mean_token_accuracy": 0.1013092825793334,
|
|
"eval_num_tokens": 912625.0,
|
|
"eval_runtime": 2.6557,
|
|
"eval_samples_per_second": 1336.368,
|
|
"eval_steps_per_second": 167.187,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 11630,
|
|
"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": 1381344823296000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|