1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.02567525932011913,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.691911220550537,
|
|
"epoch": 0.00025675259320119136,
|
|
"grad_norm": 12.8125,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.799,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 9317.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.69188060760498,
|
|
"epoch": 0.0005135051864023827,
|
|
"grad_norm": 12.8125,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7416,
|
|
"mean_token_accuracy": 0.00035366101656109095,
|
|
"num_tokens": 19116.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.690849304199219,
|
|
"epoch": 0.000770257779603574,
|
|
"grad_norm": 10.3125,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.4704,
|
|
"mean_token_accuracy": 0.028299430618062615,
|
|
"num_tokens": 27735.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.677557849884034,
|
|
"epoch": 0.0010270103728047654,
|
|
"grad_norm": 6.3125,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.1572,
|
|
"mean_token_accuracy": 0.04511106200516224,
|
|
"num_tokens": 36118.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.60137643814087,
|
|
"epoch": 0.0012837629660059566,
|
|
"grad_norm": 4.15625,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 9.9161,
|
|
"mean_token_accuracy": 0.04636201411485672,
|
|
"num_tokens": 45475.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.534404182434082,
|
|
"epoch": 0.001540515559207148,
|
|
"grad_norm": 3.375,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 9.7714,
|
|
"mean_token_accuracy": 0.043980153650045394,
|
|
"num_tokens": 54903.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.49902982711792,
|
|
"epoch": 0.0017972681524083393,
|
|
"grad_norm": 2.9375,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 9.7268,
|
|
"mean_token_accuracy": 0.040795256942510606,
|
|
"num_tokens": 64161.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.505692100524902,
|
|
"epoch": 0.002054020745609531,
|
|
"grad_norm": 2.71875,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.6231,
|
|
"mean_token_accuracy": 0.04894539378583431,
|
|
"num_tokens": 72522.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.48769702911377,
|
|
"epoch": 0.002310773338810722,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.5734,
|
|
"mean_token_accuracy": 0.05449059717357159,
|
|
"num_tokens": 80969.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.460338306427001,
|
|
"epoch": 0.002567525932011913,
|
|
"grad_norm": 2.40625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.5207,
|
|
"mean_token_accuracy": 0.059795507788658143,
|
|
"num_tokens": 89180.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.396354579925537,
|
|
"epoch": 0.0028242785252131047,
|
|
"grad_norm": 2.375,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.489,
|
|
"mean_token_accuracy": 0.06651670336723328,
|
|
"num_tokens": 98187.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.27314863204956,
|
|
"epoch": 0.003081031118414296,
|
|
"grad_norm": 2.203125,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.4142,
|
|
"mean_token_accuracy": 0.06263468749821186,
|
|
"num_tokens": 107276.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.325285243988038,
|
|
"epoch": 0.0033377837116154874,
|
|
"grad_norm": 2.296875,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.2979,
|
|
"mean_token_accuracy": 0.06050475314259529,
|
|
"num_tokens": 115585.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.303994750976562,
|
|
"epoch": 0.0035945363048166786,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.3171,
|
|
"mean_token_accuracy": 0.06043886691331864,
|
|
"num_tokens": 125353.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.29463996887207,
|
|
"epoch": 0.00385128889801787,
|
|
"grad_norm": 2.15625,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.0823,
|
|
"mean_token_accuracy": 0.06879212968051433,
|
|
"num_tokens": 133313.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.185758209228515,
|
|
"epoch": 0.004108041491219062,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.0795,
|
|
"mean_token_accuracy": 0.06610373817384244,
|
|
"num_tokens": 141741.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.269254875183105,
|
|
"epoch": 0.004364794084420253,
|
|
"grad_norm": 1.9921875,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.0745,
|
|
"mean_token_accuracy": 0.05969331823289394,
|
|
"num_tokens": 152093.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.147056770324706,
|
|
"epoch": 0.004621546677621444,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 8.8707,
|
|
"mean_token_accuracy": 0.06663852371275425,
|
|
"num_tokens": 161064.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.158480739593506,
|
|
"epoch": 0.004878299270822635,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.8128,
|
|
"mean_token_accuracy": 0.05744527503848076,
|
|
"num_tokens": 169534.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.072422409057618,
|
|
"epoch": 0.005135051864023826,
|
|
"grad_norm": 1.625,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.7571,
|
|
"mean_token_accuracy": 0.06482505239546299,
|
|
"num_tokens": 178090.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.049930477142334,
|
|
"epoch": 0.005391804457225018,
|
|
"grad_norm": 1.6640625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.6771,
|
|
"mean_token_accuracy": 0.05720922462642193,
|
|
"num_tokens": 187191.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 9.983645057678222,
|
|
"epoch": 0.0056485570504262094,
|
|
"grad_norm": 1.6171875,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.4955,
|
|
"mean_token_accuracy": 0.06595060117542743,
|
|
"num_tokens": 196206.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 9.82193603515625,
|
|
"epoch": 0.005905309643627401,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.3509,
|
|
"mean_token_accuracy": 0.0692622035741806,
|
|
"num_tokens": 205427.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.779964828491211,
|
|
"epoch": 0.006162062236828592,
|
|
"grad_norm": 1.53125,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.3154,
|
|
"mean_token_accuracy": 0.065452441573143,
|
|
"num_tokens": 214221.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.568517971038819,
|
|
"epoch": 0.006418814830029783,
|
|
"grad_norm": 1.84375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.131,
|
|
"mean_token_accuracy": 0.0722281850874424,
|
|
"num_tokens": 222254.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.459226322174072,
|
|
"epoch": 0.006675567423230975,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 7.9852,
|
|
"mean_token_accuracy": 0.07301772199571133,
|
|
"num_tokens": 230150.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.283203601837158,
|
|
"epoch": 0.006932320016432166,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 7.9713,
|
|
"mean_token_accuracy": 0.07257844731211663,
|
|
"num_tokens": 238453.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.199154663085938,
|
|
"epoch": 0.007189072609633357,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.9177,
|
|
"mean_token_accuracy": 0.06936928369104862,
|
|
"num_tokens": 246878.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 8.980141639709473,
|
|
"epoch": 0.007445825202834548,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.817,
|
|
"mean_token_accuracy": 0.06908788979053497,
|
|
"num_tokens": 255094.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.778532218933105,
|
|
"epoch": 0.00770257779603574,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.7953,
|
|
"mean_token_accuracy": 0.07565106488764287,
|
|
"num_tokens": 263358.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.707105541229248,
|
|
"epoch": 0.007959330389236931,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.7452,
|
|
"mean_token_accuracy": 0.07268079333007335,
|
|
"num_tokens": 272609.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.517619037628174,
|
|
"epoch": 0.008216082982438123,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.6494,
|
|
"mean_token_accuracy": 0.07468854188919068,
|
|
"num_tokens": 280983.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.38097620010376,
|
|
"epoch": 0.008472835575639314,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.5879,
|
|
"mean_token_accuracy": 0.07265269383788109,
|
|
"num_tokens": 290838.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.340629863739014,
|
|
"epoch": 0.008729588168840506,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.5351,
|
|
"mean_token_accuracy": 0.06626910269260407,
|
|
"num_tokens": 300411.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.264066886901855,
|
|
"epoch": 0.008986340762041696,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.6098,
|
|
"mean_token_accuracy": 0.07406073287129403,
|
|
"num_tokens": 308918.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.195462322235107,
|
|
"epoch": 0.009243093355242888,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.484,
|
|
"mean_token_accuracy": 0.07268239781260491,
|
|
"num_tokens": 318420.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.154305839538575,
|
|
"epoch": 0.00949984594844408,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.5116,
|
|
"mean_token_accuracy": 0.0776434913277626,
|
|
"num_tokens": 327055.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.08856725692749,
|
|
"epoch": 0.00975659854164527,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.3616,
|
|
"mean_token_accuracy": 0.07818598598241806,
|
|
"num_tokens": 335915.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.033003520965575,
|
|
"epoch": 0.010013351134846462,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.4749,
|
|
"mean_token_accuracy": 0.07401596866548062,
|
|
"num_tokens": 345106.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.030381727218629,
|
|
"epoch": 0.010270103728047653,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.4917,
|
|
"mean_token_accuracy": 0.07544633224606515,
|
|
"num_tokens": 353935.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.991475439071655,
|
|
"epoch": 0.010526856321248845,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.3455,
|
|
"mean_token_accuracy": 0.08079211339354515,
|
|
"num_tokens": 362605.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.986777877807617,
|
|
"epoch": 0.010783608914450037,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.471,
|
|
"mean_token_accuracy": 0.06931170262396336,
|
|
"num_tokens": 372168.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.995200538635254,
|
|
"epoch": 0.011040361507651227,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.536,
|
|
"mean_token_accuracy": 0.07623609974980354,
|
|
"num_tokens": 381468.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.939433813095093,
|
|
"epoch": 0.011297114100852419,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.3218,
|
|
"mean_token_accuracy": 0.08064079508185387,
|
|
"num_tokens": 389561.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.908961153030395,
|
|
"epoch": 0.01155386669405361,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.5704,
|
|
"mean_token_accuracy": 0.0783368855714798,
|
|
"num_tokens": 397805.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.937159585952759,
|
|
"epoch": 0.011810619287254801,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.3989,
|
|
"mean_token_accuracy": 0.08215266242623329,
|
|
"num_tokens": 406896.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.882169580459594,
|
|
"epoch": 0.012067371880455993,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.2734,
|
|
"mean_token_accuracy": 0.08328061774373055,
|
|
"num_tokens": 414860.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.8939135551452635,
|
|
"epoch": 0.012324124473657183,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.355,
|
|
"mean_token_accuracy": 0.08430732116103172,
|
|
"num_tokens": 423443.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.78408784866333,
|
|
"epoch": 0.012580877066858375,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.2922,
|
|
"mean_token_accuracy": 0.08125164359807968,
|
|
"num_tokens": 432117.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.77522349357605,
|
|
"epoch": 0.012837629660059566,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.3195,
|
|
"mean_token_accuracy": 0.08494714722037315,
|
|
"num_tokens": 440840.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.8891784191131595,
|
|
"epoch": 0.013094382253260758,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.3992,
|
|
"mean_token_accuracy": 0.08097291365265846,
|
|
"num_tokens": 449360.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.78587646484375,
|
|
"epoch": 0.01335113484646195,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.3074,
|
|
"mean_token_accuracy": 0.081997200101614,
|
|
"num_tokens": 458876.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.7375373363494875,
|
|
"epoch": 0.01360788743966314,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.1912,
|
|
"mean_token_accuracy": 0.08457952067255974,
|
|
"num_tokens": 467969.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.766811180114746,
|
|
"epoch": 0.013864640032864332,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.3171,
|
|
"mean_token_accuracy": 0.08289734944701195,
|
|
"num_tokens": 476704.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.914922475814819,
|
|
"epoch": 0.014121392626065524,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.3412,
|
|
"mean_token_accuracy": 0.08541710525751114,
|
|
"num_tokens": 485195.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.65853967666626,
|
|
"epoch": 0.014378145219266714,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.2956,
|
|
"mean_token_accuracy": 0.08422470390796662,
|
|
"num_tokens": 494407.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.690006875991822,
|
|
"epoch": 0.014634897812467906,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.3619,
|
|
"mean_token_accuracy": 0.07757975198328496,
|
|
"num_tokens": 503531.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.814142894744873,
|
|
"epoch": 0.014891650405669097,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.2233,
|
|
"mean_token_accuracy": 0.08842376917600632,
|
|
"num_tokens": 512305.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.703822946548462,
|
|
"epoch": 0.015148402998870289,
|
|
"grad_norm": 1.5703125,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.282,
|
|
"mean_token_accuracy": 0.08247971273958683,
|
|
"num_tokens": 520948.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.692700719833374,
|
|
"epoch": 0.01540515559207148,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.2455,
|
|
"mean_token_accuracy": 0.08176964893937111,
|
|
"num_tokens": 529977.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.672420310974121,
|
|
"epoch": 0.01566190818527267,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.2191,
|
|
"mean_token_accuracy": 0.08188581019639969,
|
|
"num_tokens": 539228.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.733105230331421,
|
|
"epoch": 0.015918660778473863,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.2403,
|
|
"mean_token_accuracy": 0.08645093813538551,
|
|
"num_tokens": 548090.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.628307485580445,
|
|
"epoch": 0.016175413371675055,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.2116,
|
|
"mean_token_accuracy": 0.08736620321869851,
|
|
"num_tokens": 556418.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.579625797271729,
|
|
"epoch": 0.016432165964876247,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.2116,
|
|
"mean_token_accuracy": 0.09048386812210082,
|
|
"num_tokens": 565429.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.6401910305023195,
|
|
"epoch": 0.016688918558077435,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.2101,
|
|
"mean_token_accuracy": 0.08711659237742424,
|
|
"num_tokens": 573920.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.615418386459351,
|
|
"epoch": 0.016945671151278627,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.2057,
|
|
"mean_token_accuracy": 0.0882373534142971,
|
|
"num_tokens": 582270.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.543057155609131,
|
|
"epoch": 0.01720242374447982,
|
|
"grad_norm": 1.6328125,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.2358,
|
|
"mean_token_accuracy": 0.08880284726619721,
|
|
"num_tokens": 590930.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.635246229171753,
|
|
"epoch": 0.01745917633768101,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.1773,
|
|
"mean_token_accuracy": 0.08663938567042351,
|
|
"num_tokens": 599572.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.654460287094116,
|
|
"epoch": 0.017715928930882203,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.1733,
|
|
"mean_token_accuracy": 0.09477491900324822,
|
|
"num_tokens": 607911.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.545411205291748,
|
|
"epoch": 0.017972681524083392,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.1219,
|
|
"mean_token_accuracy": 0.08721931949257851,
|
|
"num_tokens": 616444.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.5561144828796385,
|
|
"epoch": 0.018229434117284584,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.0951,
|
|
"mean_token_accuracy": 0.08625085130333901,
|
|
"num_tokens": 624714.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.543462324142456,
|
|
"epoch": 0.018486186710485776,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.2033,
|
|
"mean_token_accuracy": 0.08905729576945305,
|
|
"num_tokens": 633837.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.518378496170044,
|
|
"epoch": 0.018742939303686968,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.0503,
|
|
"mean_token_accuracy": 0.08793065920472146,
|
|
"num_tokens": 642651.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.5862819194793705,
|
|
"epoch": 0.01899969189688816,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.1634,
|
|
"mean_token_accuracy": 0.08841664642095566,
|
|
"num_tokens": 651313.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.547570848464966,
|
|
"epoch": 0.01925644449008935,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.1722,
|
|
"mean_token_accuracy": 0.08741778731346131,
|
|
"num_tokens": 659564.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.563963174819946,
|
|
"epoch": 0.01951319708329054,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.1386,
|
|
"mean_token_accuracy": 0.08939097225666046,
|
|
"num_tokens": 668499.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.461315679550171,
|
|
"epoch": 0.019769949676491733,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.151,
|
|
"mean_token_accuracy": 0.08465680629014968,
|
|
"num_tokens": 677675.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.548288154602051,
|
|
"epoch": 0.020026702269692925,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.1371,
|
|
"mean_token_accuracy": 0.09032455012202263,
|
|
"num_tokens": 686555.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.353896141052246,
|
|
"epoch": 0.020283454862894117,
|
|
"grad_norm": 1.5546875,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.9751,
|
|
"mean_token_accuracy": 0.09022012501955032,
|
|
"num_tokens": 694682.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.487498664855957,
|
|
"epoch": 0.020540207456095305,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.0423,
|
|
"mean_token_accuracy": 0.09877757728099823,
|
|
"num_tokens": 702503.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.513206958770752,
|
|
"epoch": 0.020796960049296497,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000202,
|
|
"loss": 7.1483,
|
|
"mean_token_accuracy": 0.08845404237508774,
|
|
"num_tokens": 711877.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.403651523590088,
|
|
"epoch": 0.02105371264249769,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.9786,
|
|
"mean_token_accuracy": 0.09564082995057106,
|
|
"num_tokens": 720561.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.373950386047364,
|
|
"epoch": 0.02131046523569888,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.9904,
|
|
"mean_token_accuracy": 0.08903604373335838,
|
|
"num_tokens": 730025.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.472009801864624,
|
|
"epoch": 0.021567217828900073,
|
|
"grad_norm": 1.5625,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.0876,
|
|
"mean_token_accuracy": 0.08974287211894989,
|
|
"num_tokens": 739548.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.318749809265137,
|
|
"epoch": 0.021823970422101262,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.9577,
|
|
"mean_token_accuracy": 0.09518374651670455,
|
|
"num_tokens": 747650.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.377895450592041,
|
|
"epoch": 0.022080723015302454,
|
|
"grad_norm": 1.6640625,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.0842,
|
|
"mean_token_accuracy": 0.0949647605419159,
|
|
"num_tokens": 756268.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.38172607421875,
|
|
"epoch": 0.022337475608503646,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 7.0908,
|
|
"mean_token_accuracy": 0.09451524391770363,
|
|
"num_tokens": 764547.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.389048147201538,
|
|
"epoch": 0.022594228201704838,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.979,
|
|
"mean_token_accuracy": 0.09728751182556153,
|
|
"num_tokens": 772770.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.495685863494873,
|
|
"epoch": 0.02285098079490603,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.033,
|
|
"mean_token_accuracy": 0.08979090973734856,
|
|
"num_tokens": 781929.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.204121541976929,
|
|
"epoch": 0.02310773338810722,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.9463,
|
|
"mean_token_accuracy": 0.09810084253549575,
|
|
"num_tokens": 790347.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.445066928863525,
|
|
"epoch": 0.02336448598130841,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.0875,
|
|
"mean_token_accuracy": 0.08975477069616318,
|
|
"num_tokens": 799459.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.3512537479400635,
|
|
"epoch": 0.023621238574509602,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.9551,
|
|
"mean_token_accuracy": 0.08982622176408768,
|
|
"num_tokens": 808794.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.38092827796936,
|
|
"epoch": 0.023877991167710794,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.0366,
|
|
"mean_token_accuracy": 0.09353781640529632,
|
|
"num_tokens": 816847.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.3289604663848875,
|
|
"epoch": 0.024134743760911986,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 7.0519,
|
|
"mean_token_accuracy": 0.09150565192103385,
|
|
"num_tokens": 826265.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.382613706588745,
|
|
"epoch": 0.024391496354113175,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.9688,
|
|
"mean_token_accuracy": 0.09861691147089005,
|
|
"num_tokens": 834946.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.206914281845092,
|
|
"epoch": 0.024648248947314367,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.8336,
|
|
"mean_token_accuracy": 0.097340989112854,
|
|
"num_tokens": 843314.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.18478798866272,
|
|
"epoch": 0.02490500154051556,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.8258,
|
|
"mean_token_accuracy": 0.10658634081482887,
|
|
"num_tokens": 851461.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.266985607147217,
|
|
"epoch": 0.02516175413371675,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.9496,
|
|
"mean_token_accuracy": 0.09785787239670754,
|
|
"num_tokens": 859940.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.2672138690948485,
|
|
"epoch": 0.025418506726917943,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.9646,
|
|
"mean_token_accuracy": 0.09655739292502404,
|
|
"num_tokens": 868934.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.28968653678894,
|
|
"epoch": 0.02567525932011913,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.009,
|
|
"mean_token_accuracy": 0.094332055747509,
|
|
"num_tokens": 877417.0,
|
|
"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": 1196856317952000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|