1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.03269469692015955,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.691958713531495,
|
|
"epoch": 0.0003269469692015955,
|
|
"grad_norm": 11.125,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.8332,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 9940.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.691952419281005,
|
|
"epoch": 0.000653893938403191,
|
|
"grad_norm": 11.125,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7742,
|
|
"mean_token_accuracy": 0.0001418439671397209,
|
|
"num_tokens": 19783.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.6908465385437,
|
|
"epoch": 0.0009808409076047864,
|
|
"grad_norm": 8.9375,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.5776,
|
|
"mean_token_accuracy": 0.023957793507725,
|
|
"num_tokens": 30189.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.673107433319093,
|
|
"epoch": 0.001307787876806382,
|
|
"grad_norm": 5.5,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.3372,
|
|
"mean_token_accuracy": 0.036642605997622014,
|
|
"num_tokens": 40463.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.544582939147949,
|
|
"epoch": 0.0016347348460079775,
|
|
"grad_norm": 3.90625,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.1584,
|
|
"mean_token_accuracy": 0.03861969262361527,
|
|
"num_tokens": 50919.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.540802764892579,
|
|
"epoch": 0.001961681815209573,
|
|
"grad_norm": 3.078125,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.0293,
|
|
"mean_token_accuracy": 0.043454410135746004,
|
|
"num_tokens": 61272.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.524931526184082,
|
|
"epoch": 0.0022886287844111684,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 9.989,
|
|
"mean_token_accuracy": 0.040898256562650205,
|
|
"num_tokens": 70877.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.534891700744629,
|
|
"epoch": 0.002615575753612764,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.9194,
|
|
"mean_token_accuracy": 0.04178280588239432,
|
|
"num_tokens": 80600.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.505007362365722,
|
|
"epoch": 0.0029425227228143595,
|
|
"grad_norm": 2.171875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.9154,
|
|
"mean_token_accuracy": 0.042212999984622,
|
|
"num_tokens": 91002.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.510712146759033,
|
|
"epoch": 0.003269469692015955,
|
|
"grad_norm": 2.203125,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.8381,
|
|
"mean_token_accuracy": 0.04490188360214233,
|
|
"num_tokens": 101156.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.485781574249268,
|
|
"epoch": 0.0035964166612175506,
|
|
"grad_norm": 2.1875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.7976,
|
|
"mean_token_accuracy": 0.053155021369457246,
|
|
"num_tokens": 111896.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.422010612487792,
|
|
"epoch": 0.003923363630419146,
|
|
"grad_norm": 2.234375,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.6372,
|
|
"mean_token_accuracy": 0.06276935711503029,
|
|
"num_tokens": 120899.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.29280891418457,
|
|
"epoch": 0.004250310599620741,
|
|
"grad_norm": 1.875,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.694,
|
|
"mean_token_accuracy": 0.05080278925597668,
|
|
"num_tokens": 131504.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.418154048919678,
|
|
"epoch": 0.004577257568822337,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.5967,
|
|
"mean_token_accuracy": 0.0578270398080349,
|
|
"num_tokens": 142365.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.359278392791747,
|
|
"epoch": 0.004904204538023932,
|
|
"grad_norm": 1.6875,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.5648,
|
|
"mean_token_accuracy": 0.05457159243524075,
|
|
"num_tokens": 153481.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.342585277557372,
|
|
"epoch": 0.005231151507225528,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.4747,
|
|
"mean_token_accuracy": 0.056017789617180826,
|
|
"num_tokens": 163941.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.287201023101806,
|
|
"epoch": 0.0055580984764271235,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.3941,
|
|
"mean_token_accuracy": 0.057008447125554085,
|
|
"num_tokens": 173992.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.303301525115966,
|
|
"epoch": 0.005885045445628719,
|
|
"grad_norm": 1.6484375,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.3741,
|
|
"mean_token_accuracy": 0.05458107776939869,
|
|
"num_tokens": 184474.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.273002815246581,
|
|
"epoch": 0.006211992414830315,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.2908,
|
|
"mean_token_accuracy": 0.055625283345580104,
|
|
"num_tokens": 194998.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.184755516052245,
|
|
"epoch": 0.00653893938403191,
|
|
"grad_norm": 1.6875,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.1048,
|
|
"mean_token_accuracy": 0.06638492792844772,
|
|
"num_tokens": 203327.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.124643421173095,
|
|
"epoch": 0.006865886353233506,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.0818,
|
|
"mean_token_accuracy": 0.056092340126633645,
|
|
"num_tokens": 213921.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.116442012786866,
|
|
"epoch": 0.007192833322435101,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.0058,
|
|
"mean_token_accuracy": 0.06091702468693257,
|
|
"num_tokens": 225790.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.03341302871704,
|
|
"epoch": 0.007519780291636697,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.8789,
|
|
"mean_token_accuracy": 0.057968761026859286,
|
|
"num_tokens": 235689.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.888718128204346,
|
|
"epoch": 0.007846727260838291,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.81,
|
|
"mean_token_accuracy": 0.05678555406630039,
|
|
"num_tokens": 246525.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.8173038482666,
|
|
"epoch": 0.008173674230039887,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.7335,
|
|
"mean_token_accuracy": 0.05988549403846264,
|
|
"num_tokens": 256642.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.75066261291504,
|
|
"epoch": 0.008500621199241483,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.702,
|
|
"mean_token_accuracy": 0.05682391189038753,
|
|
"num_tokens": 268145.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.635010147094727,
|
|
"epoch": 0.008827568168443078,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.6022,
|
|
"mean_token_accuracy": 0.06169581487774849,
|
|
"num_tokens": 279783.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.380872344970703,
|
|
"epoch": 0.009154515137644674,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.4967,
|
|
"mean_token_accuracy": 0.059948813170194626,
|
|
"num_tokens": 289723.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 9.371010780334473,
|
|
"epoch": 0.00948146210684627,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.415,
|
|
"mean_token_accuracy": 0.06255604662001132,
|
|
"num_tokens": 300002.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.191289615631103,
|
|
"epoch": 0.009808409076047865,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.4676,
|
|
"mean_token_accuracy": 0.0592008963227272,
|
|
"num_tokens": 310041.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.172502422332764,
|
|
"epoch": 0.01013535604524946,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.3962,
|
|
"mean_token_accuracy": 0.0683298297226429,
|
|
"num_tokens": 321469.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.02190465927124,
|
|
"epoch": 0.010462303014451056,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.3063,
|
|
"mean_token_accuracy": 0.07015354111790657,
|
|
"num_tokens": 330796.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.985146427154541,
|
|
"epoch": 0.010789249983652651,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.3555,
|
|
"mean_token_accuracy": 0.065846299380064,
|
|
"num_tokens": 341402.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.864819717407226,
|
|
"epoch": 0.011116196952854247,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.394,
|
|
"mean_token_accuracy": 0.06491452008485794,
|
|
"num_tokens": 351717.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.891894435882568,
|
|
"epoch": 0.011443143922055843,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.3088,
|
|
"mean_token_accuracy": 0.062341967970132826,
|
|
"num_tokens": 361787.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.835786247253418,
|
|
"epoch": 0.011770090891257438,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 8.2624,
|
|
"mean_token_accuracy": 0.06518663801252841,
|
|
"num_tokens": 371827.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.719056224822998,
|
|
"epoch": 0.012097037860459034,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 8.2741,
|
|
"mean_token_accuracy": 0.06445901691913605,
|
|
"num_tokens": 382150.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.604046440124511,
|
|
"epoch": 0.01242398482966063,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 8.2087,
|
|
"mean_token_accuracy": 0.06914092674851417,
|
|
"num_tokens": 392151.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.6498384475708,
|
|
"epoch": 0.012750931798862225,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 8.2286,
|
|
"mean_token_accuracy": 0.06609960272908211,
|
|
"num_tokens": 402779.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.605819988250733,
|
|
"epoch": 0.01307787876806382,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 8.2307,
|
|
"mean_token_accuracy": 0.07222314588725567,
|
|
"num_tokens": 413500.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.616723346710206,
|
|
"epoch": 0.013404825737265416,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000102,
|
|
"loss": 8.1735,
|
|
"mean_token_accuracy": 0.06858775578439236,
|
|
"num_tokens": 423054.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.610979270935058,
|
|
"epoch": 0.013731772706467011,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 8.2071,
|
|
"mean_token_accuracy": 0.06590324304997922,
|
|
"num_tokens": 433002.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.533074569702148,
|
|
"epoch": 0.014058719675668607,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000107,
|
|
"loss": 8.1663,
|
|
"mean_token_accuracy": 0.07016990669071674,
|
|
"num_tokens": 443492.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.550239372253419,
|
|
"epoch": 0.014385666644870202,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 8.2172,
|
|
"mean_token_accuracy": 0.06546101979911327,
|
|
"num_tokens": 453001.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.485262393951416,
|
|
"epoch": 0.014712613614071798,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 8.1183,
|
|
"mean_token_accuracy": 0.07102874703705311,
|
|
"num_tokens": 462916.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.503191184997558,
|
|
"epoch": 0.015039560583273394,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 8.1605,
|
|
"mean_token_accuracy": 0.06698749326169491,
|
|
"num_tokens": 473549.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.51591625213623,
|
|
"epoch": 0.015366507552474989,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 8.1877,
|
|
"mean_token_accuracy": 0.0670891109853983,
|
|
"num_tokens": 484472.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.499968433380127,
|
|
"epoch": 0.015693454521676583,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 8.1136,
|
|
"mean_token_accuracy": 0.07003521285951138,
|
|
"num_tokens": 494202.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.394450187683105,
|
|
"epoch": 0.01602040149087818,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000122,
|
|
"loss": 8.0364,
|
|
"mean_token_accuracy": 0.07509022057056428,
|
|
"num_tokens": 504635.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.355302047729491,
|
|
"epoch": 0.016347348460079774,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 8.0779,
|
|
"mean_token_accuracy": 0.07193736173212528,
|
|
"num_tokens": 514423.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.388672256469727,
|
|
"epoch": 0.01667429542928137,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000127,
|
|
"loss": 8.1229,
|
|
"mean_token_accuracy": 0.07487528696656227,
|
|
"num_tokens": 525820.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.406553173065186,
|
|
"epoch": 0.017001242398482965,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 8.0896,
|
|
"mean_token_accuracy": 0.07299871928989887,
|
|
"num_tokens": 536732.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 8.397539806365966,
|
|
"epoch": 0.01732818936768456,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.000132,
|
|
"loss": 8.0803,
|
|
"mean_token_accuracy": 0.07458276934921741,
|
|
"num_tokens": 546230.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.383978271484375,
|
|
"epoch": 0.017655136336886156,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 8.0957,
|
|
"mean_token_accuracy": 0.07023126147687435,
|
|
"num_tokens": 556078.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.382139301300048,
|
|
"epoch": 0.017982083306087752,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 8.0756,
|
|
"mean_token_accuracy": 0.0762918971478939,
|
|
"num_tokens": 565385.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.379422664642334,
|
|
"epoch": 0.018309030275289347,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 8.1261,
|
|
"mean_token_accuracy": 0.0720827341079712,
|
|
"num_tokens": 576066.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.449523544311523,
|
|
"epoch": 0.018635977244490943,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 8.0576,
|
|
"mean_token_accuracy": 0.06984641179442405,
|
|
"num_tokens": 587472.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.372716808319092,
|
|
"epoch": 0.01896292421369254,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 8.0499,
|
|
"mean_token_accuracy": 0.07255740985274314,
|
|
"num_tokens": 597400.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 8.317928695678711,
|
|
"epoch": 0.019289871182894134,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.000147,
|
|
"loss": 8.1228,
|
|
"mean_token_accuracy": 0.07397399172186851,
|
|
"num_tokens": 606856.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.43379259109497,
|
|
"epoch": 0.01961681815209573,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 8.0314,
|
|
"mean_token_accuracy": 0.07692977488040924,
|
|
"num_tokens": 617209.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 8.34858226776123,
|
|
"epoch": 0.019943765121297325,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 0.000152,
|
|
"loss": 8.0346,
|
|
"mean_token_accuracy": 0.0782922476530075,
|
|
"num_tokens": 626826.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 8.350518417358398,
|
|
"epoch": 0.02027071209049892,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 8.0169,
|
|
"mean_token_accuracy": 0.08028958663344384,
|
|
"num_tokens": 636836.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.336189079284669,
|
|
"epoch": 0.020597659059700516,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 8.0964,
|
|
"mean_token_accuracy": 0.07140604555606841,
|
|
"num_tokens": 648177.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 8.35500545501709,
|
|
"epoch": 0.02092460602890211,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 8.0817,
|
|
"mean_token_accuracy": 0.07423182912170886,
|
|
"num_tokens": 658365.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.270884132385254,
|
|
"epoch": 0.021251552998103707,
|
|
"grad_norm": 2.171875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.9958,
|
|
"mean_token_accuracy": 0.07361883968114853,
|
|
"num_tokens": 668980.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 8.373892879486084,
|
|
"epoch": 0.021578499967305303,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 8.0238,
|
|
"mean_token_accuracy": 0.07899413257837296,
|
|
"num_tokens": 679406.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 8.297880268096923,
|
|
"epoch": 0.0219054469365069,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.9837,
|
|
"mean_token_accuracy": 0.07990795224905015,
|
|
"num_tokens": 689871.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 8.258869075775147,
|
|
"epoch": 0.022232393905708494,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.9343,
|
|
"mean_token_accuracy": 0.08032900020480156,
|
|
"num_tokens": 700969.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.323282146453858,
|
|
"epoch": 0.02255934087491009,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.9559,
|
|
"mean_token_accuracy": 0.07624984867870807,
|
|
"num_tokens": 711153.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 8.160166168212891,
|
|
"epoch": 0.022886287844111685,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.9377,
|
|
"mean_token_accuracy": 0.07862194329500198,
|
|
"num_tokens": 721475.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 8.269203758239746,
|
|
"epoch": 0.02321323481331328,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.9963,
|
|
"mean_token_accuracy": 0.07536103650927543,
|
|
"num_tokens": 732823.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 8.250344467163085,
|
|
"epoch": 0.023540181782514876,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.9424,
|
|
"mean_token_accuracy": 0.08161943927407264,
|
|
"num_tokens": 743853.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 8.223281860351562,
|
|
"epoch": 0.02386712875171647,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.000182,
|
|
"loss": 8.0159,
|
|
"mean_token_accuracy": 0.0797721978276968,
|
|
"num_tokens": 754579.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 8.255972194671632,
|
|
"epoch": 0.024194075720918067,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.9773,
|
|
"mean_token_accuracy": 0.07408010140061379,
|
|
"num_tokens": 764836.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 8.241560363769532,
|
|
"epoch": 0.024521022690119663,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.9208,
|
|
"mean_token_accuracy": 0.07477326951920986,
|
|
"num_tokens": 773961.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 8.172150325775146,
|
|
"epoch": 0.02484796965932126,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.885,
|
|
"mean_token_accuracy": 0.0803740806877613,
|
|
"num_tokens": 784331.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 8.211754417419433,
|
|
"epoch": 0.025174916628522854,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.8592,
|
|
"mean_token_accuracy": 0.08332195580005645,
|
|
"num_tokens": 794724.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 8.202788066864013,
|
|
"epoch": 0.02550186359772445,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.8739,
|
|
"mean_token_accuracy": 0.08039483428001404,
|
|
"num_tokens": 804680.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 8.0855233669281,
|
|
"epoch": 0.025828810566926045,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.8878,
|
|
"mean_token_accuracy": 0.07802873328328133,
|
|
"num_tokens": 815066.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.215076637268066,
|
|
"epoch": 0.02615575753612764,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.8892,
|
|
"mean_token_accuracy": 0.07889655604958534,
|
|
"num_tokens": 824958.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 8.19112491607666,
|
|
"epoch": 0.026482704505329236,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000202,
|
|
"loss": 7.938,
|
|
"mean_token_accuracy": 0.08244092017412186,
|
|
"num_tokens": 835128.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.075815010070801,
|
|
"epoch": 0.02680965147453083,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 7.8864,
|
|
"mean_token_accuracy": 0.07554915137588977,
|
|
"num_tokens": 844590.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 8.19192361831665,
|
|
"epoch": 0.027136598443732427,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.8498,
|
|
"mean_token_accuracy": 0.08013993203639984,
|
|
"num_tokens": 854914.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 8.128836631774902,
|
|
"epoch": 0.027463545412934023,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.904,
|
|
"mean_token_accuracy": 0.07867389619350433,
|
|
"num_tokens": 866171.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 8.131003665924073,
|
|
"epoch": 0.027790492382135618,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000212,
|
|
"loss": 7.8898,
|
|
"mean_token_accuracy": 0.07613531462848186,
|
|
"num_tokens": 876603.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 8.116858911514282,
|
|
"epoch": 0.028117439351337214,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.812,
|
|
"mean_token_accuracy": 0.08004510477185249,
|
|
"num_tokens": 887124.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 8.1330491065979,
|
|
"epoch": 0.02844438632053881,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 7.8043,
|
|
"mean_token_accuracy": 0.0826384425163269,
|
|
"num_tokens": 897275.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 8.032530117034913,
|
|
"epoch": 0.028771333289740405,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 7.7923,
|
|
"mean_token_accuracy": 0.08558057546615601,
|
|
"num_tokens": 907174.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 8.115189123153687,
|
|
"epoch": 0.029098280258942,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.7937,
|
|
"mean_token_accuracy": 0.08341945856809616,
|
|
"num_tokens": 917126.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 8.031993722915649,
|
|
"epoch": 0.029425227228143596,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.816,
|
|
"mean_token_accuracy": 0.0821719765663147,
|
|
"num_tokens": 926758.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 8.04862174987793,
|
|
"epoch": 0.02975217419734519,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.8059,
|
|
"mean_token_accuracy": 0.08259067237377167,
|
|
"num_tokens": 936973.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 8.002158880233765,
|
|
"epoch": 0.030079121166546787,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 7.7631,
|
|
"mean_token_accuracy": 0.07943878024816513,
|
|
"num_tokens": 946845.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 8.083532094955444,
|
|
"epoch": 0.030406068135748383,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.7913,
|
|
"mean_token_accuracy": 0.08536312282085419,
|
|
"num_tokens": 956904.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.990770101547241,
|
|
"epoch": 0.030733015104949978,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 7.7994,
|
|
"mean_token_accuracy": 0.07841084524989128,
|
|
"num_tokens": 966133.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 8.080002784729004,
|
|
"epoch": 0.031059962074151574,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000237,
|
|
"loss": 7.8442,
|
|
"mean_token_accuracy": 0.07818925343453884,
|
|
"num_tokens": 976889.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 8.10468544960022,
|
|
"epoch": 0.031386909043353166,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.8343,
|
|
"mean_token_accuracy": 0.07947736643254757,
|
|
"num_tokens": 987385.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 8.022674083709717,
|
|
"epoch": 0.03171385601255476,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000242,
|
|
"loss": 7.8081,
|
|
"mean_token_accuracy": 0.08132115826010704,
|
|
"num_tokens": 998116.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.999949407577515,
|
|
"epoch": 0.03204080298175636,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 7.78,
|
|
"mean_token_accuracy": 0.08014609068632125,
|
|
"num_tokens": 1007412.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 8.02022352218628,
|
|
"epoch": 0.03236774995095795,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000247,
|
|
"loss": 7.818,
|
|
"mean_token_accuracy": 0.07570124790072441,
|
|
"num_tokens": 1016843.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 8.02644681930542,
|
|
"epoch": 0.03269469692015955,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.8097,
|
|
"mean_token_accuracy": 0.08114490360021591,
|
|
"num_tokens": 1027285.0,
|
|
"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": 1615421657088000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|