2035 lines
54 KiB
JSON
2035 lines
54 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.05135051864023826,
|
|
"eval_steps": 500,
|
|
"global_step": 1000,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.742582702636719,
|
|
"epoch": 0.00025675259320119136,
|
|
"grad_norm": 5.15625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.7385,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 8466.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.742582893371582,
|
|
"epoch": 0.0005135051864023827,
|
|
"grad_norm": 5.375,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7383,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 17819.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.742578315734864,
|
|
"epoch": 0.000770257779603574,
|
|
"grad_norm": 5.15625,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.7074,
|
|
"mean_token_accuracy": 8.474576170556248e-05,
|
|
"num_tokens": 26701.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.7426007270813,
|
|
"epoch": 0.0010270103728047654,
|
|
"grad_norm": 4.96875,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.6576,
|
|
"mean_token_accuracy": 0.0011235648184083403,
|
|
"num_tokens": 35386.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.742522716522217,
|
|
"epoch": 0.0012837629660059566,
|
|
"grad_norm": 4.8125,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.5556,
|
|
"mean_token_accuracy": 0.009902489895466715,
|
|
"num_tokens": 44013.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.742104721069335,
|
|
"epoch": 0.001540515559207148,
|
|
"grad_norm": 4.03125,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.4463,
|
|
"mean_token_accuracy": 0.03455784637480974,
|
|
"num_tokens": 52692.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.740787982940674,
|
|
"epoch": 0.0017972681524083393,
|
|
"grad_norm": 3.6875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.2846,
|
|
"mean_token_accuracy": 0.04273700974881649,
|
|
"num_tokens": 60741.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.736649703979491,
|
|
"epoch": 0.002054020745609531,
|
|
"grad_norm": 2.5,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.1801,
|
|
"mean_token_accuracy": 0.04606366418302059,
|
|
"num_tokens": 69580.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.729912948608398,
|
|
"epoch": 0.002310773338810722,
|
|
"grad_norm": 2.421875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.036,
|
|
"mean_token_accuracy": 0.04552499800920486,
|
|
"num_tokens": 77450.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.72356538772583,
|
|
"epoch": 0.002567525932011913,
|
|
"grad_norm": 2.25,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.9714,
|
|
"mean_token_accuracy": 0.04347024708986282,
|
|
"num_tokens": 85519.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.719997215270997,
|
|
"epoch": 0.0028242785252131047,
|
|
"grad_norm": 2.125,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.8927,
|
|
"mean_token_accuracy": 0.04672105684876442,
|
|
"num_tokens": 93628.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.717273998260499,
|
|
"epoch": 0.003081031118414296,
|
|
"grad_norm": 1.984375,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.8626,
|
|
"mean_token_accuracy": 0.04250718653202057,
|
|
"num_tokens": 102441.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.714400959014892,
|
|
"epoch": 0.0033377837116154874,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.7868,
|
|
"mean_token_accuracy": 0.04955284185707569,
|
|
"num_tokens": 111166.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.709486484527588,
|
|
"epoch": 0.0035945363048166786,
|
|
"grad_norm": 1.8984375,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.7258,
|
|
"mean_token_accuracy": 0.04404304698109627,
|
|
"num_tokens": 119300.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.703517818450928,
|
|
"epoch": 0.00385128889801787,
|
|
"grad_norm": 1.8046875,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.6956,
|
|
"mean_token_accuracy": 0.04594614543020725,
|
|
"num_tokens": 127959.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.697421169281006,
|
|
"epoch": 0.004108041491219062,
|
|
"grad_norm": 1.84375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.6217,
|
|
"mean_token_accuracy": 0.047571105882525444,
|
|
"num_tokens": 136631.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.68674201965332,
|
|
"epoch": 0.004364794084420253,
|
|
"grad_norm": 1.90625,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.5111,
|
|
"mean_token_accuracy": 0.057168309763073924,
|
|
"num_tokens": 145023.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.670315933227538,
|
|
"epoch": 0.004621546677621444,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.4523,
|
|
"mean_token_accuracy": 0.05657172314822674,
|
|
"num_tokens": 154075.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.65313549041748,
|
|
"epoch": 0.004878299270822635,
|
|
"grad_norm": 1.921875,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.3482,
|
|
"mean_token_accuracy": 0.06485468745231629,
|
|
"num_tokens": 162006.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.625462627410888,
|
|
"epoch": 0.005135051864023826,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.2717,
|
|
"mean_token_accuracy": 0.06588986031711101,
|
|
"num_tokens": 170230.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.582051181793213,
|
|
"epoch": 0.005391804457225018,
|
|
"grad_norm": 1.6796875,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.2423,
|
|
"mean_token_accuracy": 0.05897444784641266,
|
|
"num_tokens": 179215.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.52384090423584,
|
|
"epoch": 0.0056485570504262094,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.1095,
|
|
"mean_token_accuracy": 0.06802426725625992,
|
|
"num_tokens": 187858.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.461370754241944,
|
|
"epoch": 0.005905309643627401,
|
|
"grad_norm": 1.6171875,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.0136,
|
|
"mean_token_accuracy": 0.06227913834154606,
|
|
"num_tokens": 196198.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.423934841156006,
|
|
"epoch": 0.006162062236828592,
|
|
"grad_norm": 1.7890625,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.8603,
|
|
"mean_token_accuracy": 0.06186963692307472,
|
|
"num_tokens": 204940.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.381671047210693,
|
|
"epoch": 0.006418814830029783,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.7743,
|
|
"mean_token_accuracy": 0.0610227856785059,
|
|
"num_tokens": 213092.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.294575500488282,
|
|
"epoch": 0.006675567423230975,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.5688,
|
|
"mean_token_accuracy": 0.06969367675483226,
|
|
"num_tokens": 220978.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.18680181503296,
|
|
"epoch": 0.006932320016432166,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.5845,
|
|
"mean_token_accuracy": 0.06069661751389503,
|
|
"num_tokens": 229852.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.096609592437744,
|
|
"epoch": 0.007189072609633357,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.4392,
|
|
"mean_token_accuracy": 0.06324404701590539,
|
|
"num_tokens": 238908.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 9.960357856750488,
|
|
"epoch": 0.007445825202834548,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.3607,
|
|
"mean_token_accuracy": 0.06009394600987435,
|
|
"num_tokens": 248473.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.86625270843506,
|
|
"epoch": 0.00770257779603574,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.2554,
|
|
"mean_token_accuracy": 0.06259206980466843,
|
|
"num_tokens": 257210.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.693745231628418,
|
|
"epoch": 0.007959330389236931,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.1019,
|
|
"mean_token_accuracy": 0.06411121524870396,
|
|
"num_tokens": 266677.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.491777038574218,
|
|
"epoch": 0.008216082982438123,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.9959,
|
|
"mean_token_accuracy": 0.06184820123016834,
|
|
"num_tokens": 275216.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.286455821990966,
|
|
"epoch": 0.008472835575639314,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.9137,
|
|
"mean_token_accuracy": 0.06522414833307266,
|
|
"num_tokens": 283770.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.112856197357178,
|
|
"epoch": 0.008729588168840506,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.8781,
|
|
"mean_token_accuracy": 0.06161203049123287,
|
|
"num_tokens": 292414.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.901847839355469,
|
|
"epoch": 0.008986340762041696,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.7992,
|
|
"mean_token_accuracy": 0.06540008261799812,
|
|
"num_tokens": 300569.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.703197193145751,
|
|
"epoch": 0.009243093355242888,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.6644,
|
|
"mean_token_accuracy": 0.06989255100488663,
|
|
"num_tokens": 308285.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.601799583435058,
|
|
"epoch": 0.00949984594844408,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.7802,
|
|
"mean_token_accuracy": 0.06652609743177891,
|
|
"num_tokens": 317558.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.49561529159546,
|
|
"epoch": 0.00975659854164527,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.6158,
|
|
"mean_token_accuracy": 0.06920971758663655,
|
|
"num_tokens": 326377.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.364750003814697,
|
|
"epoch": 0.010013351134846462,
|
|
"grad_norm": 0.734375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.7265,
|
|
"mean_token_accuracy": 0.06648653373122215,
|
|
"num_tokens": 334669.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.294552993774413,
|
|
"epoch": 0.010270103728047653,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.5956,
|
|
"mean_token_accuracy": 0.07071267962455749,
|
|
"num_tokens": 343100.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.269872093200684,
|
|
"epoch": 0.010526856321248845,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.6108,
|
|
"mean_token_accuracy": 0.07263344749808312,
|
|
"num_tokens": 351447.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.213550281524657,
|
|
"epoch": 0.010783608914450037,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.5708,
|
|
"mean_token_accuracy": 0.06988951861858368,
|
|
"num_tokens": 360908.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.182147026062012,
|
|
"epoch": 0.011040361507651227,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.5837,
|
|
"mean_token_accuracy": 0.06750233620405197,
|
|
"num_tokens": 370075.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.132856750488282,
|
|
"epoch": 0.011297114100852419,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.4565,
|
|
"mean_token_accuracy": 0.07424333952367305,
|
|
"num_tokens": 378598.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.098768615722657,
|
|
"epoch": 0.01155386669405361,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.5533,
|
|
"mean_token_accuracy": 0.07423756159842014,
|
|
"num_tokens": 387177.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.190262031555175,
|
|
"epoch": 0.011810619287254801,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.6521,
|
|
"mean_token_accuracy": 0.06974112130701542,
|
|
"num_tokens": 396017.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.12903299331665,
|
|
"epoch": 0.012067371880455993,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.5765,
|
|
"mean_token_accuracy": 0.07110274769365788,
|
|
"num_tokens": 405324.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.074048614501953,
|
|
"epoch": 0.012324124473657183,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.5065,
|
|
"mean_token_accuracy": 0.07165832705795765,
|
|
"num_tokens": 414633.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.055518102645873,
|
|
"epoch": 0.012580877066858375,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.558,
|
|
"mean_token_accuracy": 0.06795871779322624,
|
|
"num_tokens": 424088.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.072611761093139,
|
|
"epoch": 0.012837629660059566,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.5031,
|
|
"mean_token_accuracy": 0.0740334302186966,
|
|
"num_tokens": 432868.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.065638113021851,
|
|
"epoch": 0.013094382253260758,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.4989,
|
|
"mean_token_accuracy": 0.07697901576757431,
|
|
"num_tokens": 442282.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.062051105499268,
|
|
"epoch": 0.01335113484646195,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.4639,
|
|
"mean_token_accuracy": 0.07005516327917576,
|
|
"num_tokens": 451600.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.881219291687012,
|
|
"epoch": 0.01360788743966314,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.4421,
|
|
"mean_token_accuracy": 0.07474162802100182,
|
|
"num_tokens": 461072.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.112028455734253,
|
|
"epoch": 0.013864640032864332,
|
|
"grad_norm": 0.91015625,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.5392,
|
|
"mean_token_accuracy": 0.07459446638822556,
|
|
"num_tokens": 469793.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.97756552696228,
|
|
"epoch": 0.014121392626065524,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.4663,
|
|
"mean_token_accuracy": 0.07501543574035167,
|
|
"num_tokens": 479147.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.046127462387085,
|
|
"epoch": 0.014378145219266714,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.491,
|
|
"mean_token_accuracy": 0.07903510183095933,
|
|
"num_tokens": 487684.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.984290552139282,
|
|
"epoch": 0.014634897812467906,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.4167,
|
|
"mean_token_accuracy": 0.07433984354138375,
|
|
"num_tokens": 496806.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.933219766616821,
|
|
"epoch": 0.014891650405669097,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.4893,
|
|
"mean_token_accuracy": 0.07593215778470039,
|
|
"num_tokens": 505937.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.9124115943908695,
|
|
"epoch": 0.015148402998870289,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.3239,
|
|
"mean_token_accuracy": 0.07792473435401917,
|
|
"num_tokens": 515503.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.034483766555786,
|
|
"epoch": 0.01540515559207148,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.3811,
|
|
"mean_token_accuracy": 0.07625493034720421,
|
|
"num_tokens": 524506.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.836033296585083,
|
|
"epoch": 0.01566190818527267,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.3815,
|
|
"mean_token_accuracy": 0.0733937170356512,
|
|
"num_tokens": 534009.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.863447093963623,
|
|
"epoch": 0.015918660778473863,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.266,
|
|
"mean_token_accuracy": 0.07975648492574691,
|
|
"num_tokens": 543104.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.837308692932129,
|
|
"epoch": 0.016175413371675055,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.2217,
|
|
"mean_token_accuracy": 0.0892428256571293,
|
|
"num_tokens": 551303.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.922349071502685,
|
|
"epoch": 0.016432165964876247,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.4125,
|
|
"mean_token_accuracy": 0.08186463341116905,
|
|
"num_tokens": 559361.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.849794530868531,
|
|
"epoch": 0.016688918558077435,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.333,
|
|
"mean_token_accuracy": 0.08158622235059738,
|
|
"num_tokens": 568629.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.817071199417114,
|
|
"epoch": 0.016945671151278627,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.2763,
|
|
"mean_token_accuracy": 0.08513593450188636,
|
|
"num_tokens": 576776.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.845745038986206,
|
|
"epoch": 0.01720242374447982,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.3776,
|
|
"mean_token_accuracy": 0.07514606490731239,
|
|
"num_tokens": 585970.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.911917018890381,
|
|
"epoch": 0.01745917633768101,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.4101,
|
|
"mean_token_accuracy": 0.07996492646634579,
|
|
"num_tokens": 594837.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.7797877311706545,
|
|
"epoch": 0.017715928930882203,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.3007,
|
|
"mean_token_accuracy": 0.08453992381691933,
|
|
"num_tokens": 602584.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.790009212493897,
|
|
"epoch": 0.017972681524083392,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.3112,
|
|
"mean_token_accuracy": 0.08209161311388016,
|
|
"num_tokens": 611461.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.816317987442017,
|
|
"epoch": 0.018229434117284584,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.2513,
|
|
"mean_token_accuracy": 0.0919821560382843,
|
|
"num_tokens": 619873.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.7200432300567625,
|
|
"epoch": 0.018486186710485776,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.2821,
|
|
"mean_token_accuracy": 0.08631467223167419,
|
|
"num_tokens": 628487.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.797741556167603,
|
|
"epoch": 0.018742939303686968,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.2589,
|
|
"mean_token_accuracy": 0.08670427426695823,
|
|
"num_tokens": 638162.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.808457946777343,
|
|
"epoch": 0.01899969189688816,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.153,
|
|
"mean_token_accuracy": 0.08735950589179993,
|
|
"num_tokens": 646802.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.710161352157593,
|
|
"epoch": 0.01925644449008935,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.2292,
|
|
"mean_token_accuracy": 0.0810220967978239,
|
|
"num_tokens": 655527.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.7011494636535645,
|
|
"epoch": 0.01951319708329054,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.2127,
|
|
"mean_token_accuracy": 0.08935043439269066,
|
|
"num_tokens": 664001.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.72194185256958,
|
|
"epoch": 0.019769949676491733,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.1847,
|
|
"mean_token_accuracy": 0.09060925990343094,
|
|
"num_tokens": 672972.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.752814483642578,
|
|
"epoch": 0.020026702269692925,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.1859,
|
|
"mean_token_accuracy": 0.08640384674072266,
|
|
"num_tokens": 681841.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.733091974258423,
|
|
"epoch": 0.020283454862894117,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.2036,
|
|
"mean_token_accuracy": 0.08569250404834747,
|
|
"num_tokens": 690903.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.624167728424072,
|
|
"epoch": 0.020540207456095305,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.208,
|
|
"mean_token_accuracy": 0.08355596587061882,
|
|
"num_tokens": 700411.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.67492446899414,
|
|
"epoch": 0.020796960049296497,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 7.107,
|
|
"mean_token_accuracy": 0.08860503733158112,
|
|
"num_tokens": 709062.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.748419570922851,
|
|
"epoch": 0.02105371264249769,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 7.2563,
|
|
"mean_token_accuracy": 0.0841764748096466,
|
|
"num_tokens": 718726.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.661073493957519,
|
|
"epoch": 0.02131046523569888,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.1345,
|
|
"mean_token_accuracy": 0.0915317751467228,
|
|
"num_tokens": 726710.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.460268259048462,
|
|
"epoch": 0.021567217828900073,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.1529,
|
|
"mean_token_accuracy": 0.09219859838485718,
|
|
"num_tokens": 735145.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.778389644622803,
|
|
"epoch": 0.021823970422101262,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.000212,
|
|
"loss": 7.2522,
|
|
"mean_token_accuracy": 0.0854285180568695,
|
|
"num_tokens": 744267.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.646036958694458,
|
|
"epoch": 0.022080723015302454,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.1583,
|
|
"mean_token_accuracy": 0.08593368083238602,
|
|
"num_tokens": 753253.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.666396236419677,
|
|
"epoch": 0.022337475608503646,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 7.1065,
|
|
"mean_token_accuracy": 0.09268246218562126,
|
|
"num_tokens": 761697.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.581840705871582,
|
|
"epoch": 0.022594228201704838,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 7.0921,
|
|
"mean_token_accuracy": 0.09590916931629181,
|
|
"num_tokens": 769745.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.527474021911621,
|
|
"epoch": 0.02285098079490603,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.0425,
|
|
"mean_token_accuracy": 0.0901702880859375,
|
|
"num_tokens": 778474.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.621454095840454,
|
|
"epoch": 0.02310773338810722,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.0576,
|
|
"mean_token_accuracy": 0.09153895452618599,
|
|
"num_tokens": 786581.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.626477289199829,
|
|
"epoch": 0.02336448598130841,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.1844,
|
|
"mean_token_accuracy": 0.08645047172904015,
|
|
"num_tokens": 795359.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.494252109527588,
|
|
"epoch": 0.023621238574509602,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 7.031,
|
|
"mean_token_accuracy": 0.09843903109431267,
|
|
"num_tokens": 802836.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.607794904708863,
|
|
"epoch": 0.023877991167710794,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.0743,
|
|
"mean_token_accuracy": 0.09356487020850182,
|
|
"num_tokens": 811749.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.594829416275024,
|
|
"epoch": 0.024134743760911986,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 7.0839,
|
|
"mean_token_accuracy": 0.09629017040133477,
|
|
"num_tokens": 820315.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.577248907089233,
|
|
"epoch": 0.024391496354113175,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000237,
|
|
"loss": 7.0196,
|
|
"mean_token_accuracy": 0.09056874141097068,
|
|
"num_tokens": 828983.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.597412490844727,
|
|
"epoch": 0.024648248947314367,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.1117,
|
|
"mean_token_accuracy": 0.09410159885883332,
|
|
"num_tokens": 838622.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.545994901657105,
|
|
"epoch": 0.02490500154051556,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 0.000242,
|
|
"loss": 7.0061,
|
|
"mean_token_accuracy": 0.09935126826167107,
|
|
"num_tokens": 848037.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.420415019989013,
|
|
"epoch": 0.02516175413371675,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.9918,
|
|
"mean_token_accuracy": 0.09526313245296478,
|
|
"num_tokens": 857048.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.4686620235443115,
|
|
"epoch": 0.025418506726917943,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.907,
|
|
"mean_token_accuracy": 0.09526302069425582,
|
|
"num_tokens": 865120.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.445110416412353,
|
|
"epoch": 0.02567525932011913,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.0197,
|
|
"mean_token_accuracy": 0.0922734022140503,
|
|
"num_tokens": 874061.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"entropy": 7.49036660194397,
|
|
"epoch": 0.025932011913320323,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000252,
|
|
"loss": 6.969,
|
|
"mean_token_accuracy": 0.10023247674107552,
|
|
"num_tokens": 883216.0,
|
|
"step": 505
|
|
},
|
|
{
|
|
"entropy": 7.455933570861816,
|
|
"epoch": 0.026188764506521515,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0002545,
|
|
"loss": 7.0138,
|
|
"mean_token_accuracy": 0.09378844946622848,
|
|
"num_tokens": 892350.0,
|
|
"step": 510
|
|
},
|
|
{
|
|
"entropy": 7.489746046066284,
|
|
"epoch": 0.026445517099722708,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000257,
|
|
"loss": 6.9823,
|
|
"mean_token_accuracy": 0.09784873127937317,
|
|
"num_tokens": 900558.0,
|
|
"step": 515
|
|
},
|
|
{
|
|
"entropy": 7.49373345375061,
|
|
"epoch": 0.0267022696929239,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0002595,
|
|
"loss": 6.9777,
|
|
"mean_token_accuracy": 0.0942293494939804,
|
|
"num_tokens": 909947.0,
|
|
"step": 520
|
|
},
|
|
{
|
|
"entropy": 7.483462619781494,
|
|
"epoch": 0.02695902228612509,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000262,
|
|
"loss": 7.051,
|
|
"mean_token_accuracy": 0.09402686804533004,
|
|
"num_tokens": 918739.0,
|
|
"step": 525
|
|
},
|
|
{
|
|
"entropy": 7.447958755493164,
|
|
"epoch": 0.02721577487932628,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00026450000000000003,
|
|
"loss": 6.9791,
|
|
"mean_token_accuracy": 0.09842836782336235,
|
|
"num_tokens": 927362.0,
|
|
"step": 530
|
|
},
|
|
{
|
|
"entropy": 7.415892648696899,
|
|
"epoch": 0.027472527472527472,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.00026700000000000004,
|
|
"loss": 7.0044,
|
|
"mean_token_accuracy": 0.10114640071988105,
|
|
"num_tokens": 936383.0,
|
|
"step": 535
|
|
},
|
|
{
|
|
"entropy": 7.461966943740845,
|
|
"epoch": 0.027729280065728664,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.00026950000000000005,
|
|
"loss": 6.8998,
|
|
"mean_token_accuracy": 0.10296619608998299,
|
|
"num_tokens": 944997.0,
|
|
"step": 540
|
|
},
|
|
{
|
|
"entropy": 7.497038841247559,
|
|
"epoch": 0.027986032658929856,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.00027200000000000005,
|
|
"loss": 7.059,
|
|
"mean_token_accuracy": 0.09460930153727531,
|
|
"num_tokens": 954077.0,
|
|
"step": 545
|
|
},
|
|
{
|
|
"entropy": 7.522246599197388,
|
|
"epoch": 0.028242785252131048,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0002745,
|
|
"loss": 7.0347,
|
|
"mean_token_accuracy": 0.09197534620761871,
|
|
"num_tokens": 964255.0,
|
|
"step": 550
|
|
},
|
|
{
|
|
"entropy": 7.422016334533692,
|
|
"epoch": 0.028499537845332237,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000277,
|
|
"loss": 6.9696,
|
|
"mean_token_accuracy": 0.09874921962618828,
|
|
"num_tokens": 973018.0,
|
|
"step": 555
|
|
},
|
|
{
|
|
"entropy": 7.3553742408752445,
|
|
"epoch": 0.02875629043853343,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.0002795,
|
|
"loss": 6.9592,
|
|
"mean_token_accuracy": 0.09615055844187737,
|
|
"num_tokens": 981991.0,
|
|
"step": 560
|
|
},
|
|
{
|
|
"entropy": 7.295046949386597,
|
|
"epoch": 0.02901304303173462,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00028199999999999997,
|
|
"loss": 6.8329,
|
|
"mean_token_accuracy": 0.10283219069242477,
|
|
"num_tokens": 990627.0,
|
|
"step": 565
|
|
},
|
|
{
|
|
"entropy": 7.459612798690796,
|
|
"epoch": 0.029269795624935813,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0002845,
|
|
"loss": 6.8763,
|
|
"mean_token_accuracy": 0.09862258359789848,
|
|
"num_tokens": 998596.0,
|
|
"step": 570
|
|
},
|
|
{
|
|
"entropy": 7.385720443725586,
|
|
"epoch": 0.029526548218137005,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.000287,
|
|
"loss": 6.973,
|
|
"mean_token_accuracy": 0.09788134768605232,
|
|
"num_tokens": 1007175.0,
|
|
"step": 575
|
|
},
|
|
{
|
|
"entropy": 7.259206295013428,
|
|
"epoch": 0.029783300811338193,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0002895,
|
|
"loss": 6.9401,
|
|
"mean_token_accuracy": 0.09557250440120697,
|
|
"num_tokens": 1015721.0,
|
|
"step": 580
|
|
},
|
|
{
|
|
"entropy": 7.413903093338012,
|
|
"epoch": 0.030040053404539385,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000292,
|
|
"loss": 6.993,
|
|
"mean_token_accuracy": 0.09768485501408578,
|
|
"num_tokens": 1024636.0,
|
|
"step": 585
|
|
},
|
|
{
|
|
"entropy": 7.447688007354737,
|
|
"epoch": 0.030296805997740577,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0002945,
|
|
"loss": 6.9748,
|
|
"mean_token_accuracy": 0.09863441213965415,
|
|
"num_tokens": 1033345.0,
|
|
"step": 590
|
|
},
|
|
{
|
|
"entropy": 7.267191457748413,
|
|
"epoch": 0.03055355859094177,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.000297,
|
|
"loss": 6.8654,
|
|
"mean_token_accuracy": 0.10642952471971512,
|
|
"num_tokens": 1042145.0,
|
|
"step": 595
|
|
},
|
|
{
|
|
"entropy": 7.3619465827941895,
|
|
"epoch": 0.03081031118414296,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0002995,
|
|
"loss": 6.881,
|
|
"mean_token_accuracy": 0.10219442248344421,
|
|
"num_tokens": 1050989.0,
|
|
"step": 600
|
|
},
|
|
{
|
|
"entropy": 7.355409574508667,
|
|
"epoch": 0.03106706377734415,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000302,
|
|
"loss": 6.8932,
|
|
"mean_token_accuracy": 0.10146859213709832,
|
|
"num_tokens": 1059523.0,
|
|
"step": 605
|
|
},
|
|
{
|
|
"entropy": 7.273333930969239,
|
|
"epoch": 0.03132381637054534,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0003045,
|
|
"loss": 6.9519,
|
|
"mean_token_accuracy": 0.09742964655160904,
|
|
"num_tokens": 1068597.0,
|
|
"step": 610
|
|
},
|
|
{
|
|
"entropy": 7.319173955917359,
|
|
"epoch": 0.03158056896374654,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.000307,
|
|
"loss": 6.8821,
|
|
"mean_token_accuracy": 0.1009488008916378,
|
|
"num_tokens": 1076687.0,
|
|
"step": 615
|
|
},
|
|
{
|
|
"entropy": 7.361953592300415,
|
|
"epoch": 0.031837321556947726,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0003095,
|
|
"loss": 6.8987,
|
|
"mean_token_accuracy": 0.10232065096497536,
|
|
"num_tokens": 1085473.0,
|
|
"step": 620
|
|
},
|
|
{
|
|
"entropy": 7.328115177154541,
|
|
"epoch": 0.032094074150148914,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.000312,
|
|
"loss": 6.9003,
|
|
"mean_token_accuracy": 0.10789056047797203,
|
|
"num_tokens": 1093755.0,
|
|
"step": 625
|
|
},
|
|
{
|
|
"entropy": 7.14094033241272,
|
|
"epoch": 0.03235082674335011,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0003145,
|
|
"loss": 6.7555,
|
|
"mean_token_accuracy": 0.10103646516799927,
|
|
"num_tokens": 1103555.0,
|
|
"step": 630
|
|
},
|
|
{
|
|
"entropy": 7.497932863235474,
|
|
"epoch": 0.0326075793365513,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000317,
|
|
"loss": 6.8988,
|
|
"mean_token_accuracy": 0.10374749824404716,
|
|
"num_tokens": 1112158.0,
|
|
"step": 635
|
|
},
|
|
{
|
|
"entropy": 7.159998369216919,
|
|
"epoch": 0.032864331929752494,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.0003195,
|
|
"loss": 6.7945,
|
|
"mean_token_accuracy": 0.10642132759094239,
|
|
"num_tokens": 1120512.0,
|
|
"step": 640
|
|
},
|
|
{
|
|
"entropy": 7.276958751678467,
|
|
"epoch": 0.03312108452295368,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000322,
|
|
"loss": 6.9121,
|
|
"mean_token_accuracy": 0.1034335196018219,
|
|
"num_tokens": 1128879.0,
|
|
"step": 645
|
|
},
|
|
{
|
|
"entropy": 7.273133230209351,
|
|
"epoch": 0.03337783711615487,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00032450000000000003,
|
|
"loss": 6.9686,
|
|
"mean_token_accuracy": 0.09752831980586052,
|
|
"num_tokens": 1137366.0,
|
|
"step": 650
|
|
},
|
|
{
|
|
"entropy": 7.269780445098877,
|
|
"epoch": 0.033634589709356066,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.00032700000000000003,
|
|
"loss": 6.7687,
|
|
"mean_token_accuracy": 0.10863290056586265,
|
|
"num_tokens": 1146535.0,
|
|
"step": 655
|
|
},
|
|
{
|
|
"entropy": 7.186447620391846,
|
|
"epoch": 0.033891342302557255,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.00032950000000000004,
|
|
"loss": 6.8351,
|
|
"mean_token_accuracy": 0.10125730335712432,
|
|
"num_tokens": 1155043.0,
|
|
"step": 660
|
|
},
|
|
{
|
|
"entropy": 7.206981992721557,
|
|
"epoch": 0.03414809489575845,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00033200000000000005,
|
|
"loss": 6.8185,
|
|
"mean_token_accuracy": 0.10186112374067306,
|
|
"num_tokens": 1163703.0,
|
|
"step": 665
|
|
},
|
|
{
|
|
"entropy": 7.267451667785645,
|
|
"epoch": 0.03440484748895964,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.00033450000000000005,
|
|
"loss": 6.8533,
|
|
"mean_token_accuracy": 0.10344401970505715,
|
|
"num_tokens": 1171770.0,
|
|
"step": 670
|
|
},
|
|
{
|
|
"entropy": 7.236594724655151,
|
|
"epoch": 0.03466160008216083,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000337,
|
|
"loss": 6.7989,
|
|
"mean_token_accuracy": 0.10623008012771606,
|
|
"num_tokens": 1180717.0,
|
|
"step": 675
|
|
},
|
|
{
|
|
"entropy": 7.21160855293274,
|
|
"epoch": 0.03491835267536202,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.0003395,
|
|
"loss": 6.7809,
|
|
"mean_token_accuracy": 0.10603002682328225,
|
|
"num_tokens": 1188913.0,
|
|
"step": 680
|
|
},
|
|
{
|
|
"entropy": 7.069119596481324,
|
|
"epoch": 0.03517510526856321,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000342,
|
|
"loss": 6.7651,
|
|
"mean_token_accuracy": 0.10698620229959488,
|
|
"num_tokens": 1197572.0,
|
|
"step": 685
|
|
},
|
|
{
|
|
"entropy": 7.312754535675049,
|
|
"epoch": 0.03543185786176441,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.00034449999999999997,
|
|
"loss": 6.7779,
|
|
"mean_token_accuracy": 0.10408765450119972,
|
|
"num_tokens": 1206248.0,
|
|
"step": 690
|
|
},
|
|
{
|
|
"entropy": 7.221107006072998,
|
|
"epoch": 0.035688610454965596,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000347,
|
|
"loss": 6.8614,
|
|
"mean_token_accuracy": 0.10458310097455978,
|
|
"num_tokens": 1214834.0,
|
|
"step": 695
|
|
},
|
|
{
|
|
"entropy": 7.180823087692261,
|
|
"epoch": 0.035945363048166784,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0003495,
|
|
"loss": 6.8282,
|
|
"mean_token_accuracy": 0.10219828709959984,
|
|
"num_tokens": 1224493.0,
|
|
"step": 700
|
|
},
|
|
{
|
|
"entropy": 7.227695560455322,
|
|
"epoch": 0.03620211564136798,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000352,
|
|
"loss": 6.7849,
|
|
"mean_token_accuracy": 0.10087160542607307,
|
|
"num_tokens": 1233706.0,
|
|
"step": 705
|
|
},
|
|
{
|
|
"entropy": 7.126480054855347,
|
|
"epoch": 0.03645886823456917,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0003545,
|
|
"loss": 6.8267,
|
|
"mean_token_accuracy": 0.10416910126805305,
|
|
"num_tokens": 1242897.0,
|
|
"step": 710
|
|
},
|
|
{
|
|
"entropy": 7.2293792247772215,
|
|
"epoch": 0.036715620827770364,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000357,
|
|
"loss": 6.7923,
|
|
"mean_token_accuracy": 0.1014643669128418,
|
|
"num_tokens": 1250920.0,
|
|
"step": 715
|
|
},
|
|
{
|
|
"entropy": 7.1777020454406735,
|
|
"epoch": 0.03697237342097155,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0003595,
|
|
"loss": 6.7983,
|
|
"mean_token_accuracy": 0.10312379300594329,
|
|
"num_tokens": 1258978.0,
|
|
"step": 720
|
|
},
|
|
{
|
|
"entropy": 7.225130081176758,
|
|
"epoch": 0.03722912601417274,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000362,
|
|
"loss": 6.7935,
|
|
"mean_token_accuracy": 0.10582225993275643,
|
|
"num_tokens": 1267302.0,
|
|
"step": 725
|
|
},
|
|
{
|
|
"entropy": 7.090721082687378,
|
|
"epoch": 0.037485878607373936,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0003645,
|
|
"loss": 6.6958,
|
|
"mean_token_accuracy": 0.10774607136845589,
|
|
"num_tokens": 1275469.0,
|
|
"step": 730
|
|
},
|
|
{
|
|
"entropy": 7.144723510742187,
|
|
"epoch": 0.037742631200575125,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000367,
|
|
"loss": 6.7851,
|
|
"mean_token_accuracy": 0.10810397937893867,
|
|
"num_tokens": 1283247.0,
|
|
"step": 735
|
|
},
|
|
{
|
|
"entropy": 7.263766145706176,
|
|
"epoch": 0.03799938379377632,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0003695,
|
|
"loss": 6.772,
|
|
"mean_token_accuracy": 0.1056891031563282,
|
|
"num_tokens": 1291501.0,
|
|
"step": 740
|
|
},
|
|
{
|
|
"entropy": 7.1570065975189205,
|
|
"epoch": 0.03825613638697751,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000372,
|
|
"loss": 6.8218,
|
|
"mean_token_accuracy": 0.10072930157184601,
|
|
"num_tokens": 1300677.0,
|
|
"step": 745
|
|
},
|
|
{
|
|
"entropy": 7.055219650268555,
|
|
"epoch": 0.0385128889801787,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.0003745,
|
|
"loss": 6.7675,
|
|
"mean_token_accuracy": 0.10408244505524636,
|
|
"num_tokens": 1309499.0,
|
|
"step": 750
|
|
},
|
|
{
|
|
"entropy": 7.173019313812256,
|
|
"epoch": 0.03876964157337989,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.000377,
|
|
"loss": 6.6451,
|
|
"mean_token_accuracy": 0.10914059057831764,
|
|
"num_tokens": 1317929.0,
|
|
"step": 755
|
|
},
|
|
{
|
|
"entropy": 7.017005825042725,
|
|
"epoch": 0.03902639416658108,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0003795,
|
|
"loss": 6.6827,
|
|
"mean_token_accuracy": 0.11490903943777084,
|
|
"num_tokens": 1326707.0,
|
|
"step": 760
|
|
},
|
|
{
|
|
"entropy": 7.176887035369873,
|
|
"epoch": 0.03928314675978228,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000382,
|
|
"loss": 6.7504,
|
|
"mean_token_accuracy": 0.1038425773382187,
|
|
"num_tokens": 1335481.0,
|
|
"step": 765
|
|
},
|
|
{
|
|
"entropy": 7.086644124984741,
|
|
"epoch": 0.039539899352983465,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0003845,
|
|
"loss": 6.6711,
|
|
"mean_token_accuracy": 0.10754814520478248,
|
|
"num_tokens": 1344880.0,
|
|
"step": 770
|
|
},
|
|
{
|
|
"entropy": 7.088477706909179,
|
|
"epoch": 0.039796651946184654,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.00038700000000000003,
|
|
"loss": 6.7537,
|
|
"mean_token_accuracy": 0.1024178683757782,
|
|
"num_tokens": 1353711.0,
|
|
"step": 775
|
|
},
|
|
{
|
|
"entropy": 7.03897066116333,
|
|
"epoch": 0.04005340453938585,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00038950000000000003,
|
|
"loss": 6.6974,
|
|
"mean_token_accuracy": 0.10224836692214012,
|
|
"num_tokens": 1363159.0,
|
|
"step": 780
|
|
},
|
|
{
|
|
"entropy": 6.968532657623291,
|
|
"epoch": 0.04031015713258704,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.00039200000000000004,
|
|
"loss": 6.594,
|
|
"mean_token_accuracy": 0.11433639004826546,
|
|
"num_tokens": 1371396.0,
|
|
"step": 785
|
|
},
|
|
{
|
|
"entropy": 7.022575330734253,
|
|
"epoch": 0.04056690972578823,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.00039450000000000005,
|
|
"loss": 6.5903,
|
|
"mean_token_accuracy": 0.11093990504741669,
|
|
"num_tokens": 1379862.0,
|
|
"step": 790
|
|
},
|
|
{
|
|
"entropy": 7.060764265060425,
|
|
"epoch": 0.04082366231898942,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.00039700000000000005,
|
|
"loss": 6.6821,
|
|
"mean_token_accuracy": 0.11073108166456222,
|
|
"num_tokens": 1388220.0,
|
|
"step": 795
|
|
},
|
|
{
|
|
"entropy": 7.100030374526978,
|
|
"epoch": 0.04108041491219061,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 0.0003995,
|
|
"loss": 6.834,
|
|
"mean_token_accuracy": 0.10184895843267441,
|
|
"num_tokens": 1397409.0,
|
|
"step": 800
|
|
},
|
|
{
|
|
"entropy": 7.060917663574219,
|
|
"epoch": 0.041337167505391806,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000402,
|
|
"loss": 6.659,
|
|
"mean_token_accuracy": 0.11565897613763809,
|
|
"num_tokens": 1405967.0,
|
|
"step": 805
|
|
},
|
|
{
|
|
"entropy": 7.036740016937256,
|
|
"epoch": 0.041593920098592994,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.0004045,
|
|
"loss": 6.7056,
|
|
"mean_token_accuracy": 0.10764685571193695,
|
|
"num_tokens": 1414107.0,
|
|
"step": 810
|
|
},
|
|
{
|
|
"entropy": 7.039219760894776,
|
|
"epoch": 0.04185067269179419,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00040699999999999997,
|
|
"loss": 6.6238,
|
|
"mean_token_accuracy": 0.10947153940796853,
|
|
"num_tokens": 1422834.0,
|
|
"step": 815
|
|
},
|
|
{
|
|
"entropy": 7.04460973739624,
|
|
"epoch": 0.04210742528499538,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0004095,
|
|
"loss": 6.7356,
|
|
"mean_token_accuracy": 0.10661220997571945,
|
|
"num_tokens": 1432053.0,
|
|
"step": 820
|
|
},
|
|
{
|
|
"entropy": 7.139359998703003,
|
|
"epoch": 0.04236417787819657,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000412,
|
|
"loss": 6.6804,
|
|
"mean_token_accuracy": 0.10453324764966965,
|
|
"num_tokens": 1440562.0,
|
|
"step": 825
|
|
},
|
|
{
|
|
"entropy": 6.938592052459716,
|
|
"epoch": 0.04262093047139776,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0004145,
|
|
"loss": 6.6066,
|
|
"mean_token_accuracy": 0.11469024196267127,
|
|
"num_tokens": 1449202.0,
|
|
"step": 830
|
|
},
|
|
{
|
|
"entropy": 7.033561897277832,
|
|
"epoch": 0.04287768306459895,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000417,
|
|
"loss": 6.6811,
|
|
"mean_token_accuracy": 0.10556919798254967,
|
|
"num_tokens": 1457994.0,
|
|
"step": 835
|
|
},
|
|
{
|
|
"entropy": 7.035749006271362,
|
|
"epoch": 0.043134435657800146,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.0004195,
|
|
"loss": 6.5594,
|
|
"mean_token_accuracy": 0.11265857443213463,
|
|
"num_tokens": 1467610.0,
|
|
"step": 840
|
|
},
|
|
{
|
|
"entropy": 6.930410718917846,
|
|
"epoch": 0.043391188251001335,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.000422,
|
|
"loss": 6.6115,
|
|
"mean_token_accuracy": 0.11150972470641136,
|
|
"num_tokens": 1475070.0,
|
|
"step": 845
|
|
},
|
|
{
|
|
"entropy": 7.036338949203492,
|
|
"epoch": 0.043647940844202524,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0004245,
|
|
"loss": 6.6788,
|
|
"mean_token_accuracy": 0.10261262133717537,
|
|
"num_tokens": 1483813.0,
|
|
"step": 850
|
|
},
|
|
{
|
|
"entropy": 7.053100967407227,
|
|
"epoch": 0.04390469343740372,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000427,
|
|
"loss": 6.6764,
|
|
"mean_token_accuracy": 0.10689589381217957,
|
|
"num_tokens": 1492261.0,
|
|
"step": 855
|
|
},
|
|
{
|
|
"entropy": 6.972715282440186,
|
|
"epoch": 0.04416144603060491,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0004295,
|
|
"loss": 6.6628,
|
|
"mean_token_accuracy": 0.1079073965549469,
|
|
"num_tokens": 1501036.0,
|
|
"step": 860
|
|
},
|
|
{
|
|
"entropy": 6.978304815292359,
|
|
"epoch": 0.0444181986238061,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000432,
|
|
"loss": 6.6607,
|
|
"mean_token_accuracy": 0.1067041240632534,
|
|
"num_tokens": 1509755.0,
|
|
"step": 865
|
|
},
|
|
{
|
|
"entropy": 7.03783950805664,
|
|
"epoch": 0.04467495121700729,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0004345,
|
|
"loss": 6.662,
|
|
"mean_token_accuracy": 0.10334146693348885,
|
|
"num_tokens": 1518433.0,
|
|
"step": 870
|
|
},
|
|
{
|
|
"entropy": 7.073136329650879,
|
|
"epoch": 0.04493170381020848,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000437,
|
|
"loss": 6.7077,
|
|
"mean_token_accuracy": 0.1048253409564495,
|
|
"num_tokens": 1527501.0,
|
|
"step": 875
|
|
},
|
|
{
|
|
"entropy": 6.931898355484009,
|
|
"epoch": 0.045188456403409676,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0004395,
|
|
"loss": 6.5618,
|
|
"mean_token_accuracy": 0.11353855803608895,
|
|
"num_tokens": 1536299.0,
|
|
"step": 880
|
|
},
|
|
{
|
|
"entropy": 6.8974997997283936,
|
|
"epoch": 0.045445208996610864,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.000442,
|
|
"loss": 6.6404,
|
|
"mean_token_accuracy": 0.10956602469086647,
|
|
"num_tokens": 1544408.0,
|
|
"step": 885
|
|
},
|
|
{
|
|
"entropy": 6.950430822372437,
|
|
"epoch": 0.04570196158981206,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0004445,
|
|
"loss": 6.5004,
|
|
"mean_token_accuracy": 0.11693055927753448,
|
|
"num_tokens": 1552564.0,
|
|
"step": 890
|
|
},
|
|
{
|
|
"entropy": 6.900494909286499,
|
|
"epoch": 0.04595871418301325,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000447,
|
|
"loss": 6.5678,
|
|
"mean_token_accuracy": 0.11564396619796753,
|
|
"num_tokens": 1561378.0,
|
|
"step": 895
|
|
},
|
|
{
|
|
"entropy": 6.850385761260986,
|
|
"epoch": 0.04621546677621444,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.00044950000000000003,
|
|
"loss": 6.5308,
|
|
"mean_token_accuracy": 0.11212190836668015,
|
|
"num_tokens": 1570663.0,
|
|
"step": 900
|
|
},
|
|
{
|
|
"entropy": 6.92402753829956,
|
|
"epoch": 0.04647221936941563,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00045200000000000004,
|
|
"loss": 6.629,
|
|
"mean_token_accuracy": 0.11024776846170425,
|
|
"num_tokens": 1579681.0,
|
|
"step": 905
|
|
},
|
|
{
|
|
"entropy": 7.058414793014526,
|
|
"epoch": 0.04672897196261682,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.00045450000000000004,
|
|
"loss": 6.6784,
|
|
"mean_token_accuracy": 0.10603147745132446,
|
|
"num_tokens": 1588680.0,
|
|
"step": 910
|
|
},
|
|
{
|
|
"entropy": 6.741462564468383,
|
|
"epoch": 0.046985724555818016,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00045700000000000005,
|
|
"loss": 6.4856,
|
|
"mean_token_accuracy": 0.11496321633458137,
|
|
"num_tokens": 1597141.0,
|
|
"step": 915
|
|
},
|
|
{
|
|
"entropy": 7.009930515289307,
|
|
"epoch": 0.047242477149019205,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00045950000000000006,
|
|
"loss": 6.5692,
|
|
"mean_token_accuracy": 0.11231728866696358,
|
|
"num_tokens": 1606314.0,
|
|
"step": 920
|
|
},
|
|
{
|
|
"entropy": 6.935445070266724,
|
|
"epoch": 0.04749922974222039,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000462,
|
|
"loss": 6.705,
|
|
"mean_token_accuracy": 0.10736741870641708,
|
|
"num_tokens": 1615822.0,
|
|
"step": 925
|
|
},
|
|
{
|
|
"entropy": 7.000515937805176,
|
|
"epoch": 0.04775598233542159,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0004645,
|
|
"loss": 6.6088,
|
|
"mean_token_accuracy": 0.10640651285648346,
|
|
"num_tokens": 1624630.0,
|
|
"step": 930
|
|
},
|
|
{
|
|
"entropy": 6.91909384727478,
|
|
"epoch": 0.04801273492862278,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000467,
|
|
"loss": 6.5477,
|
|
"mean_token_accuracy": 0.11673416793346406,
|
|
"num_tokens": 1632416.0,
|
|
"step": 935
|
|
},
|
|
{
|
|
"entropy": 6.916115617752075,
|
|
"epoch": 0.04826948752182397,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0004695,
|
|
"loss": 6.5873,
|
|
"mean_token_accuracy": 0.10929609909653663,
|
|
"num_tokens": 1641619.0,
|
|
"step": 940
|
|
},
|
|
{
|
|
"entropy": 6.8772941589355465,
|
|
"epoch": 0.04852624011502516,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000472,
|
|
"loss": 6.5304,
|
|
"mean_token_accuracy": 0.11052975505590439,
|
|
"num_tokens": 1649752.0,
|
|
"step": 945
|
|
},
|
|
{
|
|
"entropy": 6.90340223312378,
|
|
"epoch": 0.04878299270822635,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0004745,
|
|
"loss": 6.5347,
|
|
"mean_token_accuracy": 0.11766225099563599,
|
|
"num_tokens": 1658074.0,
|
|
"step": 950
|
|
},
|
|
{
|
|
"entropy": 6.747793102264405,
|
|
"epoch": 0.049039745301427545,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000477,
|
|
"loss": 6.4964,
|
|
"mean_token_accuracy": 0.11006889939308166,
|
|
"num_tokens": 1666767.0,
|
|
"step": 955
|
|
},
|
|
{
|
|
"entropy": 6.807773399353027,
|
|
"epoch": 0.049296497894628734,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0004795,
|
|
"loss": 6.4424,
|
|
"mean_token_accuracy": 0.1198193997144699,
|
|
"num_tokens": 1674952.0,
|
|
"step": 960
|
|
},
|
|
{
|
|
"entropy": 6.811061191558838,
|
|
"epoch": 0.04955325048782993,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000482,
|
|
"loss": 6.5249,
|
|
"mean_token_accuracy": 0.11521950364112854,
|
|
"num_tokens": 1683642.0,
|
|
"step": 965
|
|
},
|
|
{
|
|
"entropy": 6.836861085891724,
|
|
"epoch": 0.04981000308103112,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0004845,
|
|
"loss": 6.5997,
|
|
"mean_token_accuracy": 0.1094896785914898,
|
|
"num_tokens": 1692501.0,
|
|
"step": 970
|
|
},
|
|
{
|
|
"entropy": 6.962518167495728,
|
|
"epoch": 0.050066755674232306,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000487,
|
|
"loss": 6.5998,
|
|
"mean_token_accuracy": 0.109241334348917,
|
|
"num_tokens": 1702210.0,
|
|
"step": 975
|
|
},
|
|
{
|
|
"entropy": 6.7919018268585205,
|
|
"epoch": 0.0503235082674335,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0004895,
|
|
"loss": 6.5486,
|
|
"mean_token_accuracy": 0.10999819412827491,
|
|
"num_tokens": 1710942.0,
|
|
"step": 980
|
|
},
|
|
{
|
|
"entropy": 6.740083837509156,
|
|
"epoch": 0.05058026086063469,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000492,
|
|
"loss": 6.4648,
|
|
"mean_token_accuracy": 0.11652028635144233,
|
|
"num_tokens": 1719876.0,
|
|
"step": 985
|
|
},
|
|
{
|
|
"entropy": 6.961598348617554,
|
|
"epoch": 0.050837013453835886,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0004945,
|
|
"loss": 6.5059,
|
|
"mean_token_accuracy": 0.11878032237291336,
|
|
"num_tokens": 1728259.0,
|
|
"step": 990
|
|
},
|
|
{
|
|
"entropy": 6.795040130615234,
|
|
"epoch": 0.051093766047037074,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000497,
|
|
"loss": 6.5472,
|
|
"mean_token_accuracy": 0.10858382433652877,
|
|
"num_tokens": 1737165.0,
|
|
"step": 995
|
|
},
|
|
{
|
|
"entropy": 6.943842172622681,
|
|
"epoch": 0.05135051864023826,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0004995,
|
|
"loss": 6.6405,
|
|
"mean_token_accuracy": 0.10827328190207482,
|
|
"num_tokens": 1746090.0,
|
|
"step": 1000
|
|
}
|
|
],
|
|
"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": 351960527831040.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|