1046 lines
28 KiB
JSON
1046 lines
28 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.999000999000999,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.74254903793335,
|
|
"epoch": 0.00999000999000999,
|
|
"grad_norm": 4.6875,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.8409,
|
|
"mean_token_accuracy": 0.00010789126390591264,
|
|
"num_tokens": 17348.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.742575931549073,
|
|
"epoch": 0.01998001998001998,
|
|
"grad_norm": 5.28125,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.8325,
|
|
"mean_token_accuracy": 6.410256610251963e-05,
|
|
"num_tokens": 33613.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.742553329467773,
|
|
"epoch": 0.029970029970029972,
|
|
"grad_norm": 5.0625,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.8012,
|
|
"mean_token_accuracy": 0.00016108142444863914,
|
|
"num_tokens": 49416.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.742562675476075,
|
|
"epoch": 0.03996003996003996,
|
|
"grad_norm": 5.0625,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.7438,
|
|
"mean_token_accuracy": 0.0010525182529818266,
|
|
"num_tokens": 64663.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.742408943176269,
|
|
"epoch": 0.04995004995004995,
|
|
"grad_norm": 4.75,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.6283,
|
|
"mean_token_accuracy": 0.03072175462730229,
|
|
"num_tokens": 79972.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.74140043258667,
|
|
"epoch": 0.059940059940059943,
|
|
"grad_norm": 4.0,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.5246,
|
|
"mean_token_accuracy": 0.05747625604271889,
|
|
"num_tokens": 96254.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.735374069213867,
|
|
"epoch": 0.06993006993006994,
|
|
"grad_norm": 2.875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.3755,
|
|
"mean_token_accuracy": 0.06859578937292099,
|
|
"num_tokens": 112713.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.707574653625489,
|
|
"epoch": 0.07992007992007992,
|
|
"grad_norm": 2.28125,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.1924,
|
|
"mean_token_accuracy": 0.06884920224547386,
|
|
"num_tokens": 128259.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.654945564270019,
|
|
"epoch": 0.0899100899100899,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.1434,
|
|
"mean_token_accuracy": 0.06479340717196465,
|
|
"num_tokens": 145521.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.624096393585205,
|
|
"epoch": 0.0999000999000999,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 10.0779,
|
|
"mean_token_accuracy": 0.06704385466873646,
|
|
"num_tokens": 162665.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.62326774597168,
|
|
"epoch": 0.10989010989010989,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 10.0119,
|
|
"mean_token_accuracy": 0.06695662550628186,
|
|
"num_tokens": 178901.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.61907787322998,
|
|
"epoch": 0.11988011988011989,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.9679,
|
|
"mean_token_accuracy": 0.06775063388049603,
|
|
"num_tokens": 194617.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.605632114410401,
|
|
"epoch": 0.12987012987012986,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.8939,
|
|
"mean_token_accuracy": 0.07534115239977837,
|
|
"num_tokens": 209675.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.59527235031128,
|
|
"epoch": 0.13986013986013987,
|
|
"grad_norm": 2.0,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.8733,
|
|
"mean_token_accuracy": 0.07048867233097553,
|
|
"num_tokens": 225848.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.599591922760009,
|
|
"epoch": 0.14985014985014986,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.8167,
|
|
"mean_token_accuracy": 0.07247593812644482,
|
|
"num_tokens": 242981.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.579588508605957,
|
|
"epoch": 0.15984015984015984,
|
|
"grad_norm": 1.546875,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.7625,
|
|
"mean_token_accuracy": 0.07427209392189979,
|
|
"num_tokens": 258173.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.544640350341798,
|
|
"epoch": 0.16983016983016982,
|
|
"grad_norm": 1.6484375,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.69,
|
|
"mean_token_accuracy": 0.07245487086474896,
|
|
"num_tokens": 273570.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.5300874710083,
|
|
"epoch": 0.1798201798201798,
|
|
"grad_norm": 1.5625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.6529,
|
|
"mean_token_accuracy": 0.07408605217933655,
|
|
"num_tokens": 290549.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.528420066833496,
|
|
"epoch": 0.18981018981018982,
|
|
"grad_norm": 1.5390625,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.6194,
|
|
"mean_token_accuracy": 0.07625656872987747,
|
|
"num_tokens": 306856.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.485299491882325,
|
|
"epoch": 0.1998001998001998,
|
|
"grad_norm": 1.5234375,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.5443,
|
|
"mean_token_accuracy": 0.08020635470747947,
|
|
"num_tokens": 322515.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.434676933288575,
|
|
"epoch": 0.2097902097902098,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.4966,
|
|
"mean_token_accuracy": 0.07586914226412773,
|
|
"num_tokens": 337941.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.45520372390747,
|
|
"epoch": 0.21978021978021978,
|
|
"grad_norm": 1.546875,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.4232,
|
|
"mean_token_accuracy": 0.08006052449345588,
|
|
"num_tokens": 353788.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.380233097076417,
|
|
"epoch": 0.22977022977022976,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.3782,
|
|
"mean_token_accuracy": 0.08523525297641754,
|
|
"num_tokens": 371012.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.3286771774292,
|
|
"epoch": 0.23976023976023977,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 9.2336,
|
|
"mean_token_accuracy": 0.08941392749547958,
|
|
"num_tokens": 387063.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.305299186706543,
|
|
"epoch": 0.24975024975024976,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 9.2413,
|
|
"mean_token_accuracy": 0.08080004155635834,
|
|
"num_tokens": 404360.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.245867156982422,
|
|
"epoch": 0.2597402597402597,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 9.0529,
|
|
"mean_token_accuracy": 0.09887080416083335,
|
|
"num_tokens": 421267.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.113175201416016,
|
|
"epoch": 0.26973026973026976,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 9.0745,
|
|
"mean_token_accuracy": 0.0869515560567379,
|
|
"num_tokens": 438900.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.11645679473877,
|
|
"epoch": 0.27972027972027974,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.9962,
|
|
"mean_token_accuracy": 0.08159712329506874,
|
|
"num_tokens": 454580.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.033879661560059,
|
|
"epoch": 0.2897102897102897,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.9364,
|
|
"mean_token_accuracy": 0.082859006524086,
|
|
"num_tokens": 470894.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.949287033081054,
|
|
"epoch": 0.2997002997002997,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.8864,
|
|
"mean_token_accuracy": 0.0842631220817566,
|
|
"num_tokens": 487872.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.799868774414062,
|
|
"epoch": 0.3096903096903097,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.7141,
|
|
"mean_token_accuracy": 0.09437366873025894,
|
|
"num_tokens": 503899.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.570561790466309,
|
|
"epoch": 0.3196803196803197,
|
|
"grad_norm": 0.83203125,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.6153,
|
|
"mean_token_accuracy": 0.09862063378095627,
|
|
"num_tokens": 519023.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.580165386199951,
|
|
"epoch": 0.32967032967032966,
|
|
"grad_norm": 0.796875,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.6321,
|
|
"mean_token_accuracy": 0.087054193764925,
|
|
"num_tokens": 535424.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.45827875137329,
|
|
"epoch": 0.33966033966033965,
|
|
"grad_norm": 0.83203125,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.6716,
|
|
"mean_token_accuracy": 0.08567041307687759,
|
|
"num_tokens": 552528.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.326940536499023,
|
|
"epoch": 0.34965034965034963,
|
|
"grad_norm": 0.70703125,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.5523,
|
|
"mean_token_accuracy": 0.08963110744953155,
|
|
"num_tokens": 567149.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 9.181477642059326,
|
|
"epoch": 0.3596403596403596,
|
|
"grad_norm": 0.8984375,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 8.5255,
|
|
"mean_token_accuracy": 0.08932835683226585,
|
|
"num_tokens": 582033.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 9.184029483795166,
|
|
"epoch": 0.3696303696303696,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 8.492,
|
|
"mean_token_accuracy": 0.09364427700638771,
|
|
"num_tokens": 598930.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 9.105390739440917,
|
|
"epoch": 0.37962037962037964,
|
|
"grad_norm": 0.91015625,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 8.4764,
|
|
"mean_token_accuracy": 0.08496234193444252,
|
|
"num_tokens": 615859.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 9.057480430603027,
|
|
"epoch": 0.38961038961038963,
|
|
"grad_norm": 0.66015625,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 8.4525,
|
|
"mean_token_accuracy": 0.09097891747951507,
|
|
"num_tokens": 631984.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.931283473968506,
|
|
"epoch": 0.3996003996003996,
|
|
"grad_norm": 0.76171875,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 8.4712,
|
|
"mean_token_accuracy": 0.09022903516888618,
|
|
"num_tokens": 648599.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.999445819854737,
|
|
"epoch": 0.4095904095904096,
|
|
"grad_norm": 0.75,
|
|
"learning_rate": 0.000102,
|
|
"loss": 8.5152,
|
|
"mean_token_accuracy": 0.08909457549452782,
|
|
"num_tokens": 666196.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.872494792938232,
|
|
"epoch": 0.4195804195804196,
|
|
"grad_norm": 0.66015625,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 8.4462,
|
|
"mean_token_accuracy": 0.09361587092280388,
|
|
"num_tokens": 682848.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.852845001220704,
|
|
"epoch": 0.42957042957042957,
|
|
"grad_norm": 0.5703125,
|
|
"learning_rate": 0.000107,
|
|
"loss": 8.3932,
|
|
"mean_token_accuracy": 0.0924599327147007,
|
|
"num_tokens": 700324.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.826492881774902,
|
|
"epoch": 0.43956043956043955,
|
|
"grad_norm": 0.6796875,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 8.438,
|
|
"mean_token_accuracy": 0.0906071975827217,
|
|
"num_tokens": 715401.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.716631221771241,
|
|
"epoch": 0.44955044955044954,
|
|
"grad_norm": 0.68359375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 8.4143,
|
|
"mean_token_accuracy": 0.09328261092305183,
|
|
"num_tokens": 730372.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.727209281921386,
|
|
"epoch": 0.4595404595404595,
|
|
"grad_norm": 0.796875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 8.3391,
|
|
"mean_token_accuracy": 0.09173915013670922,
|
|
"num_tokens": 745880.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.743009567260742,
|
|
"epoch": 0.4695304695304695,
|
|
"grad_norm": 0.7265625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 8.3552,
|
|
"mean_token_accuracy": 0.09261953458189964,
|
|
"num_tokens": 761474.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.669536972045899,
|
|
"epoch": 0.47952047952047955,
|
|
"grad_norm": 0.62890625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 8.3798,
|
|
"mean_token_accuracy": 0.09077078998088836,
|
|
"num_tokens": 776871.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.689266204833984,
|
|
"epoch": 0.48951048951048953,
|
|
"grad_norm": 0.70703125,
|
|
"learning_rate": 0.000122,
|
|
"loss": 8.3659,
|
|
"mean_token_accuracy": 0.09065382108092308,
|
|
"num_tokens": 793506.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.591617012023926,
|
|
"epoch": 0.4995004995004995,
|
|
"grad_norm": 0.68359375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 8.3325,
|
|
"mean_token_accuracy": 0.09529750868678093,
|
|
"num_tokens": 808366.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.718113708496094,
|
|
"epoch": 0.5094905094905094,
|
|
"grad_norm": 0.765625,
|
|
"learning_rate": 0.000127,
|
|
"loss": 8.2646,
|
|
"mean_token_accuracy": 0.10416125357151032,
|
|
"num_tokens": 825461.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.644307136535645,
|
|
"epoch": 0.5194805194805194,
|
|
"grad_norm": 0.65625,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 8.3552,
|
|
"mean_token_accuracy": 0.08870847970247268,
|
|
"num_tokens": 840998.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 8.638798427581786,
|
|
"epoch": 0.5294705294705294,
|
|
"grad_norm": 0.70703125,
|
|
"learning_rate": 0.000132,
|
|
"loss": 8.4149,
|
|
"mean_token_accuracy": 0.08986097797751427,
|
|
"num_tokens": 858383.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.622299480438233,
|
|
"epoch": 0.5394605394605395,
|
|
"grad_norm": 0.703125,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 8.3192,
|
|
"mean_token_accuracy": 0.08967429846525192,
|
|
"num_tokens": 873611.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.645909023284911,
|
|
"epoch": 0.5494505494505495,
|
|
"grad_norm": 0.8046875,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 8.2318,
|
|
"mean_token_accuracy": 0.10625835433602333,
|
|
"num_tokens": 889918.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.576266384124756,
|
|
"epoch": 0.5594405594405595,
|
|
"grad_norm": 0.80859375,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 8.2966,
|
|
"mean_token_accuracy": 0.1017056480050087,
|
|
"num_tokens": 906967.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.678112125396728,
|
|
"epoch": 0.5694305694305695,
|
|
"grad_norm": 0.73828125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 8.3848,
|
|
"mean_token_accuracy": 0.09247232899069786,
|
|
"num_tokens": 922954.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.666139221191406,
|
|
"epoch": 0.5794205794205795,
|
|
"grad_norm": 0.7734375,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 8.3586,
|
|
"mean_token_accuracy": 0.0921535387635231,
|
|
"num_tokens": 938920.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 8.631329917907715,
|
|
"epoch": 0.5894105894105894,
|
|
"grad_norm": 0.66796875,
|
|
"learning_rate": 0.000147,
|
|
"loss": 8.3734,
|
|
"mean_token_accuracy": 0.0920196034014225,
|
|
"num_tokens": 956301.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.548529148101807,
|
|
"epoch": 0.5994005994005994,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 8.2358,
|
|
"mean_token_accuracy": 0.09958092793822289,
|
|
"num_tokens": 972454.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 8.61522102355957,
|
|
"epoch": 0.6093906093906094,
|
|
"grad_norm": 0.703125,
|
|
"learning_rate": 0.000152,
|
|
"loss": 8.333,
|
|
"mean_token_accuracy": 0.09208913743495942,
|
|
"num_tokens": 987326.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 8.584405994415283,
|
|
"epoch": 0.6193806193806194,
|
|
"grad_norm": 0.69921875,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 8.2977,
|
|
"mean_token_accuracy": 0.09532473459839821,
|
|
"num_tokens": 1002448.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.536137676239013,
|
|
"epoch": 0.6293706293706294,
|
|
"grad_norm": 0.69921875,
|
|
"learning_rate": 0.000157,
|
|
"loss": 8.3265,
|
|
"mean_token_accuracy": 0.09180266484618187,
|
|
"num_tokens": 1020348.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 8.623396396636963,
|
|
"epoch": 0.6393606393606394,
|
|
"grad_norm": 0.7265625,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 8.3499,
|
|
"mean_token_accuracy": 0.08997747674584389,
|
|
"num_tokens": 1036562.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.528428745269775,
|
|
"epoch": 0.6493506493506493,
|
|
"grad_norm": 0.7734375,
|
|
"learning_rate": 0.000162,
|
|
"loss": 8.3187,
|
|
"mean_token_accuracy": 0.09452675953507424,
|
|
"num_tokens": 1053106.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 8.570547008514405,
|
|
"epoch": 0.6593406593406593,
|
|
"grad_norm": 0.80078125,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 8.2708,
|
|
"mean_token_accuracy": 0.09287350997328758,
|
|
"num_tokens": 1068277.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 8.505113983154297,
|
|
"epoch": 0.6693306693306693,
|
|
"grad_norm": 0.7109375,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 8.212,
|
|
"mean_token_accuracy": 0.10472053438425064,
|
|
"num_tokens": 1083887.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 8.494814014434814,
|
|
"epoch": 0.6793206793206793,
|
|
"grad_norm": 0.65234375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 8.2264,
|
|
"mean_token_accuracy": 0.0991443045437336,
|
|
"num_tokens": 1099791.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.513353157043458,
|
|
"epoch": 0.6893106893106893,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 8.2843,
|
|
"mean_token_accuracy": 0.09210123345255852,
|
|
"num_tokens": 1116137.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 8.550718784332275,
|
|
"epoch": 0.6993006993006993,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 8.2293,
|
|
"mean_token_accuracy": 0.10072905272245407,
|
|
"num_tokens": 1132449.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 8.527479457855225,
|
|
"epoch": 0.7092907092907093,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000177,
|
|
"loss": 8.3552,
|
|
"mean_token_accuracy": 0.09086157828569412,
|
|
"num_tokens": 1148316.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 8.505389308929443,
|
|
"epoch": 0.7192807192807192,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 8.2142,
|
|
"mean_token_accuracy": 0.09256454855203629,
|
|
"num_tokens": 1164030.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 8.580228328704834,
|
|
"epoch": 0.7292707292707292,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 0.000182,
|
|
"loss": 8.3049,
|
|
"mean_token_accuracy": 0.09622592777013779,
|
|
"num_tokens": 1180624.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 8.496176528930665,
|
|
"epoch": 0.7392607392607392,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 8.2815,
|
|
"mean_token_accuracy": 0.09714617729187011,
|
|
"num_tokens": 1196986.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 8.580214786529542,
|
|
"epoch": 0.7492507492507493,
|
|
"grad_norm": 0.72265625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 8.2919,
|
|
"mean_token_accuracy": 0.09614738449454308,
|
|
"num_tokens": 1212754.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 8.44631586074829,
|
|
"epoch": 0.7592407592407593,
|
|
"grad_norm": 0.91015625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 8.2324,
|
|
"mean_token_accuracy": 0.09536803364753724,
|
|
"num_tokens": 1229627.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 8.515226554870605,
|
|
"epoch": 0.7692307692307693,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 8.1816,
|
|
"mean_token_accuracy": 0.0967013455927372,
|
|
"num_tokens": 1245127.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 8.458242416381836,
|
|
"epoch": 0.7792207792207793,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 8.2483,
|
|
"mean_token_accuracy": 0.09477976039052009,
|
|
"num_tokens": 1261438.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 8.479850482940673,
|
|
"epoch": 0.7892107892107892,
|
|
"grad_norm": 0.734375,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 8.2152,
|
|
"mean_token_accuracy": 0.09673597514629365,
|
|
"num_tokens": 1277476.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.536440181732178,
|
|
"epoch": 0.7992007992007992,
|
|
"grad_norm": 0.76953125,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 8.1608,
|
|
"mean_token_accuracy": 0.09755991101264953,
|
|
"num_tokens": 1294659.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 8.371343421936036,
|
|
"epoch": 0.8091908091908092,
|
|
"grad_norm": 0.8203125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 8.1319,
|
|
"mean_token_accuracy": 0.10120925828814506,
|
|
"num_tokens": 1308829.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.464437866210938,
|
|
"epoch": 0.8191808191808192,
|
|
"grad_norm": 0.765625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 8.1896,
|
|
"mean_token_accuracy": 0.10006042197346687,
|
|
"num_tokens": 1325659.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 8.507107639312744,
|
|
"epoch": 0.8291708291708292,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 0.000207,
|
|
"loss": 8.2255,
|
|
"mean_token_accuracy": 0.09681524932384492,
|
|
"num_tokens": 1341643.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 8.446794795989991,
|
|
"epoch": 0.8391608391608392,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 8.197,
|
|
"mean_token_accuracy": 0.09735974669456482,
|
|
"num_tokens": 1359197.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 8.502279472351074,
|
|
"epoch": 0.8491508491508492,
|
|
"grad_norm": 0.6796875,
|
|
"learning_rate": 0.000212,
|
|
"loss": 8.1875,
|
|
"mean_token_accuracy": 0.09608993604779244,
|
|
"num_tokens": 1376107.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 8.477056503295898,
|
|
"epoch": 0.8591408591408591,
|
|
"grad_norm": 0.79296875,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 8.2042,
|
|
"mean_token_accuracy": 0.09094029515981675,
|
|
"num_tokens": 1392232.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 8.460865306854249,
|
|
"epoch": 0.8691308691308691,
|
|
"grad_norm": 0.83984375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 8.1631,
|
|
"mean_token_accuracy": 0.09813632071018219,
|
|
"num_tokens": 1406747.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 8.48235101699829,
|
|
"epoch": 0.8791208791208791,
|
|
"grad_norm": 0.7890625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 8.1609,
|
|
"mean_token_accuracy": 0.09457436800003052,
|
|
"num_tokens": 1423599.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 8.372987842559814,
|
|
"epoch": 0.8891108891108891,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000222,
|
|
"loss": 8.091,
|
|
"mean_token_accuracy": 0.1050879828631878,
|
|
"num_tokens": 1439330.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 8.440707111358643,
|
|
"epoch": 0.8991008991008991,
|
|
"grad_norm": 0.66015625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 8.0636,
|
|
"mean_token_accuracy": 0.10564030036330223,
|
|
"num_tokens": 1456660.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 8.463158130645752,
|
|
"epoch": 0.9090909090909091,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 8.1173,
|
|
"mean_token_accuracy": 0.10065716579556465,
|
|
"num_tokens": 1471619.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 8.29682149887085,
|
|
"epoch": 0.919080919080919,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 8.0752,
|
|
"mean_token_accuracy": 0.09712542369961738,
|
|
"num_tokens": 1486971.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 8.374618434906006,
|
|
"epoch": 0.929070929070929,
|
|
"grad_norm": 0.7890625,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 8.1102,
|
|
"mean_token_accuracy": 0.09521022215485572,
|
|
"num_tokens": 1502923.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 8.45193510055542,
|
|
"epoch": 0.939060939060939,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 8.1818,
|
|
"mean_token_accuracy": 0.09111250266432762,
|
|
"num_tokens": 1520214.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 8.413396072387695,
|
|
"epoch": 0.949050949050949,
|
|
"grad_norm": 0.875,
|
|
"learning_rate": 0.000237,
|
|
"loss": 8.0682,
|
|
"mean_token_accuracy": 0.10277181193232536,
|
|
"num_tokens": 1534946.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 8.293998527526856,
|
|
"epoch": 0.9590409590409591,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.9474,
|
|
"mean_token_accuracy": 0.11179379150271415,
|
|
"num_tokens": 1551823.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 8.366506767272949,
|
|
"epoch": 0.9690309690309691,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 0.000242,
|
|
"loss": 8.0529,
|
|
"mean_token_accuracy": 0.10311459228396416,
|
|
"num_tokens": 1567354.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 8.338885498046874,
|
|
"epoch": 0.9790209790209791,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 8.0225,
|
|
"mean_token_accuracy": 0.09862796664237976,
|
|
"num_tokens": 1583749.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 8.336707592010498,
|
|
"epoch": 0.989010989010989,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 0.000247,
|
|
"loss": 8.1258,
|
|
"mean_token_accuracy": 0.09763901010155678,
|
|
"num_tokens": 1599678.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 8.445067310333252,
|
|
"epoch": 0.999000999000999,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 8.1224,
|
|
"mean_token_accuracy": 0.0928703673183918,
|
|
"num_tokens": 1616737.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.999000999000999,
|
|
"eval_entropy": 8.140788088618098,
|
|
"eval_loss": 8.047406196594238,
|
|
"eval_mean_token_accuracy": 0.09197118636724111,
|
|
"eval_num_tokens": 1616737.0,
|
|
"eval_runtime": 1.1978,
|
|
"eval_samples_per_second": 1230.577,
|
|
"eval_steps_per_second": 154.448,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 5000,
|
|
"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": 325253177671680.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|