1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.032988058322887114,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.74233112335205,
|
|
"epoch": 0.00032988058322887113,
|
|
"grad_norm": 4.78125,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.8047,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 11578.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.742351722717284,
|
|
"epoch": 0.0006597611664577423,
|
|
"grad_norm": 4.53125,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7851,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 22232.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.742332267761231,
|
|
"epoch": 0.0009896417496866134,
|
|
"grad_norm": 4.65625,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.8042,
|
|
"mean_token_accuracy": 7.412898703478277e-05,
|
|
"num_tokens": 35495.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.74229564666748,
|
|
"epoch": 0.0013195223329154845,
|
|
"grad_norm": 5.03125,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.7299,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 47005.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.742135810852051,
|
|
"epoch": 0.0016494029161443558,
|
|
"grad_norm": 4.21875,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.6387,
|
|
"mean_token_accuracy": 0.00025946117821149526,
|
|
"num_tokens": 57675.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.741827201843261,
|
|
"epoch": 0.001979283499373227,
|
|
"grad_norm": 3.9375,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.5537,
|
|
"mean_token_accuracy": 0.008250601799227297,
|
|
"num_tokens": 70455.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.7412504196167,
|
|
"epoch": 0.002309164082602098,
|
|
"grad_norm": 3.046875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.4147,
|
|
"mean_token_accuracy": 0.03285279367119074,
|
|
"num_tokens": 82422.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.7402099609375,
|
|
"epoch": 0.002639044665830969,
|
|
"grad_norm": 2.46875,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.2672,
|
|
"mean_token_accuracy": 0.036276179179549214,
|
|
"num_tokens": 92109.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.738385868072509,
|
|
"epoch": 0.0029689252490598403,
|
|
"grad_norm": 2.265625,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.18,
|
|
"mean_token_accuracy": 0.04042476825416088,
|
|
"num_tokens": 103440.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.736727714538574,
|
|
"epoch": 0.0032988058322887116,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 10.1099,
|
|
"mean_token_accuracy": 0.04008350595831871,
|
|
"num_tokens": 114742.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.73554048538208,
|
|
"epoch": 0.0036286864155175825,
|
|
"grad_norm": 1.984375,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 10.0092,
|
|
"mean_token_accuracy": 0.043487312830984594,
|
|
"num_tokens": 125626.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.733216762542725,
|
|
"epoch": 0.003958566998746454,
|
|
"grad_norm": 1.9765625,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.976,
|
|
"mean_token_accuracy": 0.043048874288797376,
|
|
"num_tokens": 137957.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.730610179901124,
|
|
"epoch": 0.004288447581975325,
|
|
"grad_norm": 2.3125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.9097,
|
|
"mean_token_accuracy": 0.036802740022540095,
|
|
"num_tokens": 148277.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.729512119293213,
|
|
"epoch": 0.004618328165204196,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.8853,
|
|
"mean_token_accuracy": 0.04078914504498243,
|
|
"num_tokens": 159878.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.727227592468262,
|
|
"epoch": 0.004948208748433067,
|
|
"grad_norm": 1.84375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.7999,
|
|
"mean_token_accuracy": 0.04081474579870701,
|
|
"num_tokens": 172156.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.724674892425536,
|
|
"epoch": 0.005278089331661938,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.7314,
|
|
"mean_token_accuracy": 0.0359017526730895,
|
|
"num_tokens": 183913.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.720765209197998,
|
|
"epoch": 0.00560796991489081,
|
|
"grad_norm": 1.6796875,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.6786,
|
|
"mean_token_accuracy": 0.041428479738533495,
|
|
"num_tokens": 195817.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.713240337371825,
|
|
"epoch": 0.005937850498119681,
|
|
"grad_norm": 1.75,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.6194,
|
|
"mean_token_accuracy": 0.0394784739241004,
|
|
"num_tokens": 207962.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.704511070251465,
|
|
"epoch": 0.0062677310813485515,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.5465,
|
|
"mean_token_accuracy": 0.03562910072505474,
|
|
"num_tokens": 219781.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.693966960906982,
|
|
"epoch": 0.006597611664577423,
|
|
"grad_norm": 1.6953125,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.4197,
|
|
"mean_token_accuracy": 0.037568676844239236,
|
|
"num_tokens": 231163.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.679013442993163,
|
|
"epoch": 0.006927492247806294,
|
|
"grad_norm": 1.671875,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.326,
|
|
"mean_token_accuracy": 0.03616092000156641,
|
|
"num_tokens": 242141.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.657819652557373,
|
|
"epoch": 0.007257372831035165,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.2134,
|
|
"mean_token_accuracy": 0.03960070498287678,
|
|
"num_tokens": 253879.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.626339530944824,
|
|
"epoch": 0.007587253414264037,
|
|
"grad_norm": 1.7578125,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.0565,
|
|
"mean_token_accuracy": 0.044448528438806534,
|
|
"num_tokens": 264486.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.585787487030029,
|
|
"epoch": 0.007917133997492908,
|
|
"grad_norm": 1.640625,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.9692,
|
|
"mean_token_accuracy": 0.037905681505799295,
|
|
"num_tokens": 275750.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.545704460144043,
|
|
"epoch": 0.008247014580721778,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.894,
|
|
"mean_token_accuracy": 0.042156589217484,
|
|
"num_tokens": 287746.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.492252826690674,
|
|
"epoch": 0.00857689516395065,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.7427,
|
|
"mean_token_accuracy": 0.044464115798473355,
|
|
"num_tokens": 300383.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.41234998703003,
|
|
"epoch": 0.008906775747179522,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.6308,
|
|
"mean_token_accuracy": 0.05071255713701248,
|
|
"num_tokens": 313091.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.307254981994628,
|
|
"epoch": 0.009236656330408393,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.4857,
|
|
"mean_token_accuracy": 0.04862055517733097,
|
|
"num_tokens": 323251.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.19880542755127,
|
|
"epoch": 0.009566536913637264,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.3857,
|
|
"mean_token_accuracy": 0.04940937012434006,
|
|
"num_tokens": 333707.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 10.061772537231445,
|
|
"epoch": 0.009896417496866134,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.3134,
|
|
"mean_token_accuracy": 0.04397304020822048,
|
|
"num_tokens": 345426.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.90941858291626,
|
|
"epoch": 0.010226298080095005,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.2127,
|
|
"mean_token_accuracy": 0.04301855657249689,
|
|
"num_tokens": 358876.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.72692346572876,
|
|
"epoch": 0.010556178663323876,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.0847,
|
|
"mean_token_accuracy": 0.04769652187824249,
|
|
"num_tokens": 370955.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.496499633789062,
|
|
"epoch": 0.010886059246552749,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.9552,
|
|
"mean_token_accuracy": 0.04982128106057644,
|
|
"num_tokens": 383667.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.298235988616943,
|
|
"epoch": 0.01121593982978162,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.8896,
|
|
"mean_token_accuracy": 0.047108123824000356,
|
|
"num_tokens": 394347.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.061514949798584,
|
|
"epoch": 0.01154582041301049,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.7922,
|
|
"mean_token_accuracy": 0.05443224273622036,
|
|
"num_tokens": 406589.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.849029541015625,
|
|
"epoch": 0.011875700996239361,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.7205,
|
|
"mean_token_accuracy": 0.06005025506019592,
|
|
"num_tokens": 417462.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.702647686004639,
|
|
"epoch": 0.012205581579468232,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.7819,
|
|
"mean_token_accuracy": 0.05638592652976513,
|
|
"num_tokens": 428076.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.6129864692688,
|
|
"epoch": 0.012535462162697103,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.6524,
|
|
"mean_token_accuracy": 0.059817750751972196,
|
|
"num_tokens": 438714.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.444343852996827,
|
|
"epoch": 0.012865342745925976,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.6408,
|
|
"mean_token_accuracy": 0.06158410720527172,
|
|
"num_tokens": 450373.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.418584156036378,
|
|
"epoch": 0.013195223329154846,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.5929,
|
|
"mean_token_accuracy": 0.057680726051330566,
|
|
"num_tokens": 461201.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.309362125396728,
|
|
"epoch": 0.013525103912383717,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.5542,
|
|
"mean_token_accuracy": 0.06454258635640145,
|
|
"num_tokens": 471519.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.26671142578125,
|
|
"epoch": 0.013854984495612588,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.5927,
|
|
"mean_token_accuracy": 0.05961795505136251,
|
|
"num_tokens": 483245.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.234414291381835,
|
|
"epoch": 0.014184865078841459,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.4793,
|
|
"mean_token_accuracy": 0.06982782371342182,
|
|
"num_tokens": 494398.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.201716423034668,
|
|
"epoch": 0.01451474566207033,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.4962,
|
|
"mean_token_accuracy": 0.06742062233388424,
|
|
"num_tokens": 505550.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.13823642730713,
|
|
"epoch": 0.014844626245299202,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.4418,
|
|
"mean_token_accuracy": 0.06280249543488026,
|
|
"num_tokens": 515824.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.11952476501465,
|
|
"epoch": 0.015174506828528073,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.4716,
|
|
"mean_token_accuracy": 0.06296740882098675,
|
|
"num_tokens": 527272.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.108990573883057,
|
|
"epoch": 0.015504387411756944,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.4469,
|
|
"mean_token_accuracy": 0.06231762431561947,
|
|
"num_tokens": 537970.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.067597770690918,
|
|
"epoch": 0.015834267994985815,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.3392,
|
|
"mean_token_accuracy": 0.06825867146253586,
|
|
"num_tokens": 549026.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.033491706848144,
|
|
"epoch": 0.016164148578214688,
|
|
"grad_norm": 0.84765625,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.4567,
|
|
"mean_token_accuracy": 0.06924333348870278,
|
|
"num_tokens": 561001.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.013448429107665,
|
|
"epoch": 0.016494029161443557,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.3986,
|
|
"mean_token_accuracy": 0.07052349597215653,
|
|
"num_tokens": 571910.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.004467582702636,
|
|
"epoch": 0.01682390974467243,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.3889,
|
|
"mean_token_accuracy": 0.0650747399777174,
|
|
"num_tokens": 583073.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.977169990539551,
|
|
"epoch": 0.0171537903279013,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.3194,
|
|
"mean_token_accuracy": 0.07421667948365211,
|
|
"num_tokens": 593909.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.92542724609375,
|
|
"epoch": 0.01748367091113017,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.347,
|
|
"mean_token_accuracy": 0.07075322493910789,
|
|
"num_tokens": 605235.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.959033679962158,
|
|
"epoch": 0.017813551494359044,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.2928,
|
|
"mean_token_accuracy": 0.07352816425263882,
|
|
"num_tokens": 616491.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.854368066787719,
|
|
"epoch": 0.018143432077587913,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.301,
|
|
"mean_token_accuracy": 0.07566386498510838,
|
|
"num_tokens": 627576.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.89704909324646,
|
|
"epoch": 0.018473312660816785,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.2279,
|
|
"mean_token_accuracy": 0.07298410087823867,
|
|
"num_tokens": 638833.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.945636034011841,
|
|
"epoch": 0.018803193244045655,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.3253,
|
|
"mean_token_accuracy": 0.06714979782700539,
|
|
"num_tokens": 648954.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.888553476333618,
|
|
"epoch": 0.019133073827274527,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.2291,
|
|
"mean_token_accuracy": 0.07276964522898197,
|
|
"num_tokens": 659653.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.827844858169556,
|
|
"epoch": 0.019462954410503396,
|
|
"grad_norm": 0.875,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.3204,
|
|
"mean_token_accuracy": 0.0698035504668951,
|
|
"num_tokens": 671282.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.846695137023926,
|
|
"epoch": 0.01979283499373227,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.2734,
|
|
"mean_token_accuracy": 0.06930247582495212,
|
|
"num_tokens": 683096.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.817320489883423,
|
|
"epoch": 0.02012271557696114,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.2386,
|
|
"mean_token_accuracy": 0.07485006004571915,
|
|
"num_tokens": 694278.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.90717658996582,
|
|
"epoch": 0.02045259616019001,
|
|
"grad_norm": 0.76953125,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.2358,
|
|
"mean_token_accuracy": 0.0718875303864479,
|
|
"num_tokens": 705878.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.742885494232178,
|
|
"epoch": 0.020782476743418883,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.1725,
|
|
"mean_token_accuracy": 0.0757001593708992,
|
|
"num_tokens": 717459.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.795225667953491,
|
|
"epoch": 0.021112357326647752,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.1761,
|
|
"mean_token_accuracy": 0.07542249858379364,
|
|
"num_tokens": 726708.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.819357681274414,
|
|
"epoch": 0.021442237909876625,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.1691,
|
|
"mean_token_accuracy": 0.07912719771265983,
|
|
"num_tokens": 736799.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.716309642791748,
|
|
"epoch": 0.021772118493105497,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.1528,
|
|
"mean_token_accuracy": 0.07632440999150276,
|
|
"num_tokens": 747820.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.650016450881958,
|
|
"epoch": 0.022101999076334367,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.1569,
|
|
"mean_token_accuracy": 0.07818866521120071,
|
|
"num_tokens": 757938.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.751991653442383,
|
|
"epoch": 0.02243187965956324,
|
|
"grad_norm": 0.84765625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.1513,
|
|
"mean_token_accuracy": 0.07743561789393424,
|
|
"num_tokens": 768900.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.693883180618286,
|
|
"epoch": 0.02276176024279211,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.218,
|
|
"mean_token_accuracy": 0.07597828209400177,
|
|
"num_tokens": 780839.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.725241041183471,
|
|
"epoch": 0.02309164082602098,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.1578,
|
|
"mean_token_accuracy": 0.07896500080823898,
|
|
"num_tokens": 793144.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.7039953708648685,
|
|
"epoch": 0.02342152140924985,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.1385,
|
|
"mean_token_accuracy": 0.0761062603443861,
|
|
"num_tokens": 805293.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.650927495956421,
|
|
"epoch": 0.023751401992478723,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.0847,
|
|
"mean_token_accuracy": 0.07881326377391815,
|
|
"num_tokens": 817313.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.617914915084839,
|
|
"epoch": 0.024081282575707595,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.039,
|
|
"mean_token_accuracy": 0.07709232643246651,
|
|
"num_tokens": 827823.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.612017059326172,
|
|
"epoch": 0.024411163158936464,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.0581,
|
|
"mean_token_accuracy": 0.07990295961499214,
|
|
"num_tokens": 839221.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.634074735641479,
|
|
"epoch": 0.024741043742165337,
|
|
"grad_norm": 0.88671875,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.0299,
|
|
"mean_token_accuracy": 0.0763829343020916,
|
|
"num_tokens": 850616.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.552822065353394,
|
|
"epoch": 0.025070924325394206,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.0531,
|
|
"mean_token_accuracy": 0.07905212491750717,
|
|
"num_tokens": 862467.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.6426232814788815,
|
|
"epoch": 0.02540080490862308,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.0412,
|
|
"mean_token_accuracy": 0.07432318702340127,
|
|
"num_tokens": 872130.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.561140012741089,
|
|
"epoch": 0.02573068549185195,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.0034,
|
|
"mean_token_accuracy": 0.08546995297074318,
|
|
"num_tokens": 883598.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.582135248184204,
|
|
"epoch": 0.02606056607508082,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.0003,
|
|
"mean_token_accuracy": 0.07829435840249062,
|
|
"num_tokens": 896478.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.555583810806274,
|
|
"epoch": 0.026390446658309693,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.0712,
|
|
"mean_token_accuracy": 0.07842531427741051,
|
|
"num_tokens": 909351.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.550145435333252,
|
|
"epoch": 0.026720327241538562,
|
|
"grad_norm": 0.88671875,
|
|
"learning_rate": 0.000202,
|
|
"loss": 7.0363,
|
|
"mean_token_accuracy": 0.08111794963479042,
|
|
"num_tokens": 920170.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.451640224456787,
|
|
"epoch": 0.027050207824767435,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.978,
|
|
"mean_token_accuracy": 0.0850048691034317,
|
|
"num_tokens": 929498.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.552112293243408,
|
|
"epoch": 0.027380088407996304,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.9678,
|
|
"mean_token_accuracy": 0.07753312811255456,
|
|
"num_tokens": 939975.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.444381189346314,
|
|
"epoch": 0.027709968991225176,
|
|
"grad_norm": 0.9609375,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.9504,
|
|
"mean_token_accuracy": 0.0817045159637928,
|
|
"num_tokens": 950593.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.561855220794678,
|
|
"epoch": 0.02803984957445405,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.9561,
|
|
"mean_token_accuracy": 0.08070995509624482,
|
|
"num_tokens": 962281.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.47752628326416,
|
|
"epoch": 0.028369730157682918,
|
|
"grad_norm": 0.9609375,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.8784,
|
|
"mean_token_accuracy": 0.08914879560470582,
|
|
"num_tokens": 972717.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.474664211273193,
|
|
"epoch": 0.02869961074091179,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.8833,
|
|
"mean_token_accuracy": 0.08555072918534279,
|
|
"num_tokens": 983669.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.426136112213134,
|
|
"epoch": 0.02902949132414066,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.8871,
|
|
"mean_token_accuracy": 0.08220960162580013,
|
|
"num_tokens": 994528.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.392468643188477,
|
|
"epoch": 0.029359371907369532,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.8482,
|
|
"mean_token_accuracy": 0.0958639606833458,
|
|
"num_tokens": 1005445.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.446725940704345,
|
|
"epoch": 0.029689252490598405,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.0157,
|
|
"mean_token_accuracy": 0.0827529065310955,
|
|
"num_tokens": 1018747.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.441160249710083,
|
|
"epoch": 0.030019133073827274,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.8981,
|
|
"mean_token_accuracy": 0.08528381213545799,
|
|
"num_tokens": 1028956.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.43569803237915,
|
|
"epoch": 0.030349013657056147,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 7.0367,
|
|
"mean_token_accuracy": 0.07986071296036243,
|
|
"num_tokens": 1041227.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.422861957550049,
|
|
"epoch": 0.030678894240285016,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.9143,
|
|
"mean_token_accuracy": 0.07918849810957909,
|
|
"num_tokens": 1053002.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.4423298835754395,
|
|
"epoch": 0.03100877482351389,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.9095,
|
|
"mean_token_accuracy": 0.08558695167303085,
|
|
"num_tokens": 1063693.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.416697883605957,
|
|
"epoch": 0.03133865540674276,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.88,
|
|
"mean_token_accuracy": 0.09078652560710906,
|
|
"num_tokens": 1074245.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.32618203163147,
|
|
"epoch": 0.03166853598997163,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.8789,
|
|
"mean_token_accuracy": 0.08656272403895855,
|
|
"num_tokens": 1084663.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.368553400039673,
|
|
"epoch": 0.0319984165732005,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.8227,
|
|
"mean_token_accuracy": 0.0852760024368763,
|
|
"num_tokens": 1094777.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.348766326904297,
|
|
"epoch": 0.032328297156429375,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.8406,
|
|
"mean_token_accuracy": 0.08816310912370681,
|
|
"num_tokens": 1106712.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.286499500274658,
|
|
"epoch": 0.03265817773965824,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.8402,
|
|
"mean_token_accuracy": 0.09093867242336273,
|
|
"num_tokens": 1118488.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.348550319671631,
|
|
"epoch": 0.032988058322887114,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.8115,
|
|
"mean_token_accuracy": 0.08822314143180847,
|
|
"num_tokens": 1128803.0,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 4000,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"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": 277649423400960.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|