4fb63ddfe1
Model: codingmonster1234/Llama-3.1-8B-Instruct-Chess-Reasoning-SFT-v2 Source: Original Platform
1727 lines
47 KiB
JSON
1727 lines
47 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.0,
|
|
"eval_steps": 500,
|
|
"global_step": 168,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 1.8498776704072952,
|
|
"epoch": 0.005956813104988831,
|
|
"grad_norm": 10.25,
|
|
"learning_rate": 2e-05,
|
|
"loss": 1.7708,
|
|
"mean_token_accuracy": 0.599120743572712,
|
|
"num_tokens": 14922.0,
|
|
"step": 1
|
|
},
|
|
{
|
|
"entropy": 0.9681647345423698,
|
|
"epoch": 0.011913626209977662,
|
|
"grad_norm": 31.5,
|
|
"learning_rate": 1.9999805729315383e-05,
|
|
"loss": 2.1657,
|
|
"mean_token_accuracy": 0.5869186520576477,
|
|
"num_tokens": 30082.0,
|
|
"step": 2
|
|
},
|
|
{
|
|
"entropy": 1.3554665446281433,
|
|
"epoch": 0.017870439314966492,
|
|
"grad_norm": 10.5,
|
|
"learning_rate": 1.999922292480975e-05,
|
|
"loss": 1.5454,
|
|
"mean_token_accuracy": 0.655014343559742,
|
|
"num_tokens": 44458.0,
|
|
"step": 3
|
|
},
|
|
{
|
|
"entropy": 1.4668754190206528,
|
|
"epoch": 0.023827252419955324,
|
|
"grad_norm": 5.21875,
|
|
"learning_rate": 1.9998251609127465e-05,
|
|
"loss": 1.4003,
|
|
"mean_token_accuracy": 0.6703809276223183,
|
|
"num_tokens": 59381.0,
|
|
"step": 4
|
|
},
|
|
{
|
|
"entropy": 1.472496047616005,
|
|
"epoch": 0.029784065524944156,
|
|
"grad_norm": 4.625,
|
|
"learning_rate": 1.9996891820008165e-05,
|
|
"loss": 1.366,
|
|
"mean_token_accuracy": 0.6713378205895424,
|
|
"num_tokens": 74182.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 1.2937641143798828,
|
|
"epoch": 0.035740878629932984,
|
|
"grad_norm": 4.4375,
|
|
"learning_rate": 1.9995143610285275e-05,
|
|
"loss": 1.2684,
|
|
"mean_token_accuracy": 0.6905824095010757,
|
|
"num_tokens": 88869.0,
|
|
"step": 6
|
|
},
|
|
{
|
|
"entropy": 1.1484729945659637,
|
|
"epoch": 0.04169769173492182,
|
|
"grad_norm": 4.15625,
|
|
"learning_rate": 1.9993007047883988e-05,
|
|
"loss": 1.1351,
|
|
"mean_token_accuracy": 0.7156129479408264,
|
|
"num_tokens": 103614.0,
|
|
"step": 7
|
|
},
|
|
{
|
|
"entropy": 1.10194443911314,
|
|
"epoch": 0.04765450483991065,
|
|
"grad_norm": 4.03125,
|
|
"learning_rate": 1.999048221581858e-05,
|
|
"loss": 1.1358,
|
|
"mean_token_accuracy": 0.7098847851157188,
|
|
"num_tokens": 118065.0,
|
|
"step": 8
|
|
},
|
|
{
|
|
"entropy": 1.08524502068758,
|
|
"epoch": 0.05361131794489948,
|
|
"grad_norm": 4.0625,
|
|
"learning_rate": 1.9987569212189224e-05,
|
|
"loss": 1.1283,
|
|
"mean_token_accuracy": 0.7138840556144714,
|
|
"num_tokens": 133029.0,
|
|
"step": 9
|
|
},
|
|
{
|
|
"entropy": 1.0641729906201363,
|
|
"epoch": 0.05956813104988831,
|
|
"grad_norm": 3.75,
|
|
"learning_rate": 1.998426815017817e-05,
|
|
"loss": 1.0785,
|
|
"mean_token_accuracy": 0.7212501168251038,
|
|
"num_tokens": 147810.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 1.105129525065422,
|
|
"epoch": 0.06552494415487714,
|
|
"grad_norm": 3.859375,
|
|
"learning_rate": 1.9980579158045322e-05,
|
|
"loss": 1.1286,
|
|
"mean_token_accuracy": 0.706063449382782,
|
|
"num_tokens": 162781.0,
|
|
"step": 11
|
|
},
|
|
{
|
|
"entropy": 1.0460990443825722,
|
|
"epoch": 0.07148175725986597,
|
|
"grad_norm": 3.546875,
|
|
"learning_rate": 1.997650237912329e-05,
|
|
"loss": 1.0538,
|
|
"mean_token_accuracy": 0.7267542257905006,
|
|
"num_tokens": 176562.0,
|
|
"step": 12
|
|
},
|
|
{
|
|
"entropy": 1.096757523715496,
|
|
"epoch": 0.0774385703648548,
|
|
"grad_norm": 3.921875,
|
|
"learning_rate": 1.9972037971811802e-05,
|
|
"loss": 1.1255,
|
|
"mean_token_accuracy": 0.7059001103043556,
|
|
"num_tokens": 190938.0,
|
|
"step": 13
|
|
},
|
|
{
|
|
"entropy": 1.0409524962306023,
|
|
"epoch": 0.08339538346984364,
|
|
"grad_norm": 3.28125,
|
|
"learning_rate": 1.996718610957155e-05,
|
|
"loss": 1.0051,
|
|
"mean_token_accuracy": 0.7358693853020668,
|
|
"num_tokens": 205420.0,
|
|
"step": 14
|
|
},
|
|
{
|
|
"entropy": 0.998079240322113,
|
|
"epoch": 0.08935219657483247,
|
|
"grad_norm": 3.203125,
|
|
"learning_rate": 1.9961946980917457e-05,
|
|
"loss": 0.9823,
|
|
"mean_token_accuracy": 0.7424589470028877,
|
|
"num_tokens": 219949.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 1.0189584344625473,
|
|
"epoch": 0.0953090096798213,
|
|
"grad_norm": 3.234375,
|
|
"learning_rate": 1.9956320789411338e-05,
|
|
"loss": 1.0227,
|
|
"mean_token_accuracy": 0.727337934076786,
|
|
"num_tokens": 235321.0,
|
|
"step": 16
|
|
},
|
|
{
|
|
"entropy": 1.0216224193572998,
|
|
"epoch": 0.10126582278481013,
|
|
"grad_norm": 3.421875,
|
|
"learning_rate": 1.9950307753654016e-05,
|
|
"loss": 1.058,
|
|
"mean_token_accuracy": 0.7266535609960556,
|
|
"num_tokens": 249201.0,
|
|
"step": 17
|
|
},
|
|
{
|
|
"entropy": 1.0217409878969193,
|
|
"epoch": 0.10722263588979895,
|
|
"grad_norm": 3.265625,
|
|
"learning_rate": 1.99439081072768e-05,
|
|
"loss": 1.0486,
|
|
"mean_token_accuracy": 0.7262789905071259,
|
|
"num_tokens": 263940.0,
|
|
"step": 18
|
|
},
|
|
{
|
|
"entropy": 0.971530094742775,
|
|
"epoch": 0.11317944899478778,
|
|
"grad_norm": 3.25,
|
|
"learning_rate": 1.9937122098932428e-05,
|
|
"loss": 0.9757,
|
|
"mean_token_accuracy": 0.7425716742873192,
|
|
"num_tokens": 279156.0,
|
|
"step": 19
|
|
},
|
|
{
|
|
"entropy": 0.9931668639183044,
|
|
"epoch": 0.11913626209977662,
|
|
"grad_norm": 3.21875,
|
|
"learning_rate": 1.9929949992285397e-05,
|
|
"loss": 1.0121,
|
|
"mean_token_accuracy": 0.7311301454901695,
|
|
"num_tokens": 293490.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 0.9790017381310463,
|
|
"epoch": 0.12509307520476545,
|
|
"grad_norm": 3.125,
|
|
"learning_rate": 1.9922392066001724e-05,
|
|
"loss": 0.9584,
|
|
"mean_token_accuracy": 0.742002509534359,
|
|
"num_tokens": 308809.0,
|
|
"step": 21
|
|
},
|
|
{
|
|
"entropy": 0.937186636030674,
|
|
"epoch": 0.13104988830975428,
|
|
"grad_norm": 3.171875,
|
|
"learning_rate": 1.9914448613738107e-05,
|
|
"loss": 0.9301,
|
|
"mean_token_accuracy": 0.7529638782143593,
|
|
"num_tokens": 323801.0,
|
|
"step": 22
|
|
},
|
|
{
|
|
"entropy": 0.9425854161381721,
|
|
"epoch": 0.1370067014147431,
|
|
"grad_norm": 2.859375,
|
|
"learning_rate": 1.9906119944130527e-05,
|
|
"loss": 0.8937,
|
|
"mean_token_accuracy": 0.7547886222600937,
|
|
"num_tokens": 338809.0,
|
|
"step": 23
|
|
},
|
|
{
|
|
"entropy": 0.9758685901761055,
|
|
"epoch": 0.14296351451973194,
|
|
"grad_norm": 2.9375,
|
|
"learning_rate": 1.9897406380782262e-05,
|
|
"loss": 0.972,
|
|
"mean_token_accuracy": 0.7386127412319183,
|
|
"num_tokens": 353678.0,
|
|
"step": 24
|
|
},
|
|
{
|
|
"entropy": 0.9490021914243698,
|
|
"epoch": 0.14892032762472077,
|
|
"grad_norm": 2.90625,
|
|
"learning_rate": 1.9888308262251286e-05,
|
|
"loss": 0.9421,
|
|
"mean_token_accuracy": 0.7473454773426056,
|
|
"num_tokens": 368974.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 0.9499934315681458,
|
|
"epoch": 0.1548771407297096,
|
|
"grad_norm": 3.078125,
|
|
"learning_rate": 1.9878825942037147e-05,
|
|
"loss": 0.9335,
|
|
"mean_token_accuracy": 0.7485107779502869,
|
|
"num_tokens": 383442.0,
|
|
"step": 26
|
|
},
|
|
{
|
|
"entropy": 0.9617905765771866,
|
|
"epoch": 0.16083395383469842,
|
|
"grad_norm": 3.28125,
|
|
"learning_rate": 1.9868959788567213e-05,
|
|
"loss": 0.9586,
|
|
"mean_token_accuracy": 0.7396569699048996,
|
|
"num_tokens": 397144.0,
|
|
"step": 27
|
|
},
|
|
{
|
|
"entropy": 0.9021903425455093,
|
|
"epoch": 0.16679076693968728,
|
|
"grad_norm": 3.09375,
|
|
"learning_rate": 1.985871018518236e-05,
|
|
"loss": 0.9265,
|
|
"mean_token_accuracy": 0.7496669217944145,
|
|
"num_tokens": 411072.0,
|
|
"step": 28
|
|
},
|
|
{
|
|
"entropy": 0.9532824382185936,
|
|
"epoch": 0.1727475800446761,
|
|
"grad_norm": 3.234375,
|
|
"learning_rate": 1.9848077530122083e-05,
|
|
"loss": 0.9562,
|
|
"mean_token_accuracy": 0.7413105815649033,
|
|
"num_tokens": 424984.0,
|
|
"step": 29
|
|
},
|
|
{
|
|
"entropy": 0.8909583687782288,
|
|
"epoch": 0.17870439314966494,
|
|
"grad_norm": 2.984375,
|
|
"learning_rate": 1.9837062236509013e-05,
|
|
"loss": 0.8659,
|
|
"mean_token_accuracy": 0.7643471732735634,
|
|
"num_tokens": 439653.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 0.9447234719991684,
|
|
"epoch": 0.18466120625465376,
|
|
"grad_norm": 3.25,
|
|
"learning_rate": 1.9825664732332886e-05,
|
|
"loss": 0.9754,
|
|
"mean_token_accuracy": 0.7384027242660522,
|
|
"num_tokens": 453859.0,
|
|
"step": 31
|
|
},
|
|
{
|
|
"entropy": 1.0093699097633362,
|
|
"epoch": 0.1906180193596426,
|
|
"grad_norm": 2.90625,
|
|
"learning_rate": 1.981388546043388e-05,
|
|
"loss": 1.0087,
|
|
"mean_token_accuracy": 0.728794701397419,
|
|
"num_tokens": 469134.0,
|
|
"step": 32
|
|
},
|
|
{
|
|
"entropy": 0.9534936100244522,
|
|
"epoch": 0.19657483246463142,
|
|
"grad_norm": 3.140625,
|
|
"learning_rate": 1.9801724878485438e-05,
|
|
"loss": 0.9662,
|
|
"mean_token_accuracy": 0.7423886656761169,
|
|
"num_tokens": 483181.0,
|
|
"step": 33
|
|
},
|
|
{
|
|
"entropy": 0.9070451781153679,
|
|
"epoch": 0.20253164556962025,
|
|
"grad_norm": 2.90625,
|
|
"learning_rate": 1.9789183458976485e-05,
|
|
"loss": 0.8896,
|
|
"mean_token_accuracy": 0.7565986886620522,
|
|
"num_tokens": 497611.0,
|
|
"step": 34
|
|
},
|
|
{
|
|
"entropy": 0.8499675318598747,
|
|
"epoch": 0.20848845867460908,
|
|
"grad_norm": 2.9375,
|
|
"learning_rate": 1.977626168919305e-05,
|
|
"loss": 0.8478,
|
|
"mean_token_accuracy": 0.7672473266720772,
|
|
"num_tokens": 511742.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 0.864535354077816,
|
|
"epoch": 0.2144452717795979,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 1.9762960071199334e-05,
|
|
"loss": 0.8922,
|
|
"mean_token_accuracy": 0.7637492269277573,
|
|
"num_tokens": 527336.0,
|
|
"step": 36
|
|
},
|
|
{
|
|
"entropy": 0.8508971408009529,
|
|
"epoch": 0.22040208488458674,
|
|
"grad_norm": 2.953125,
|
|
"learning_rate": 1.9749279121818235e-05,
|
|
"loss": 0.8842,
|
|
"mean_token_accuracy": 0.7573391646146774,
|
|
"num_tokens": 541854.0,
|
|
"step": 37
|
|
},
|
|
{
|
|
"entropy": 0.9110140576958656,
|
|
"epoch": 0.22635889798957556,
|
|
"grad_norm": 2.9375,
|
|
"learning_rate": 1.9735219372611232e-05,
|
|
"loss": 0.9444,
|
|
"mean_token_accuracy": 0.7438178360462189,
|
|
"num_tokens": 556835.0,
|
|
"step": 38
|
|
},
|
|
{
|
|
"entropy": 0.8960491791367531,
|
|
"epoch": 0.23231571109456442,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 1.9720781369857747e-05,
|
|
"loss": 0.8835,
|
|
"mean_token_accuracy": 0.7513013556599617,
|
|
"num_tokens": 571583.0,
|
|
"step": 39
|
|
},
|
|
{
|
|
"entropy": 0.9107659161090851,
|
|
"epoch": 0.23827252419955325,
|
|
"grad_norm": 2.828125,
|
|
"learning_rate": 1.970596567453391e-05,
|
|
"loss": 0.907,
|
|
"mean_token_accuracy": 0.7521690353751183,
|
|
"num_tokens": 585756.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 0.9291563332080841,
|
|
"epoch": 0.24422933730454208,
|
|
"grad_norm": 3.0,
|
|
"learning_rate": 1.969077286229078e-05,
|
|
"loss": 0.9122,
|
|
"mean_token_accuracy": 0.7513668984174728,
|
|
"num_tokens": 600414.0,
|
|
"step": 41
|
|
},
|
|
{
|
|
"entropy": 0.9352462142705917,
|
|
"epoch": 0.2501861504095309,
|
|
"grad_norm": 2.953125,
|
|
"learning_rate": 1.9675203523431964e-05,
|
|
"loss": 0.946,
|
|
"mean_token_accuracy": 0.7436617910861969,
|
|
"num_tokens": 615312.0,
|
|
"step": 42
|
|
},
|
|
{
|
|
"entropy": 0.8683317676186562,
|
|
"epoch": 0.2561429635145197,
|
|
"grad_norm": 2.875,
|
|
"learning_rate": 1.9659258262890683e-05,
|
|
"loss": 0.8753,
|
|
"mean_token_accuracy": 0.7570225298404694,
|
|
"num_tokens": 630044.0,
|
|
"step": 43
|
|
},
|
|
{
|
|
"entropy": 0.8450798913836479,
|
|
"epoch": 0.26209977661950856,
|
|
"grad_norm": 2.71875,
|
|
"learning_rate": 1.964293770020628e-05,
|
|
"loss": 0.8343,
|
|
"mean_token_accuracy": 0.7695427238941193,
|
|
"num_tokens": 644892.0,
|
|
"step": 44
|
|
},
|
|
{
|
|
"entropy": 0.8672548830509186,
|
|
"epoch": 0.2680565897244974,
|
|
"grad_norm": 3.046875,
|
|
"learning_rate": 1.962624246950012e-05,
|
|
"loss": 0.8707,
|
|
"mean_token_accuracy": 0.761024072766304,
|
|
"num_tokens": 659231.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 0.8786624893546104,
|
|
"epoch": 0.2740134028294862,
|
|
"grad_norm": 3.015625,
|
|
"learning_rate": 1.9609173219450998e-05,
|
|
"loss": 0.9213,
|
|
"mean_token_accuracy": 0.7514273300766945,
|
|
"num_tokens": 673897.0,
|
|
"step": 46
|
|
},
|
|
{
|
|
"entropy": 0.8764197081327438,
|
|
"epoch": 0.2799702159344751,
|
|
"grad_norm": 2.828125,
|
|
"learning_rate": 1.9591730613269878e-05,
|
|
"loss": 0.893,
|
|
"mean_token_accuracy": 0.7548534125089645,
|
|
"num_tokens": 689100.0,
|
|
"step": 47
|
|
},
|
|
{
|
|
"entropy": 0.8522048592567444,
|
|
"epoch": 0.2859270290394639,
|
|
"grad_norm": 3.265625,
|
|
"learning_rate": 1.957391532867418e-05,
|
|
"loss": 0.889,
|
|
"mean_token_accuracy": 0.7559506595134735,
|
|
"num_tokens": 703112.0,
|
|
"step": 48
|
|
},
|
|
{
|
|
"entropy": 0.8236794173717499,
|
|
"epoch": 0.29188384214445273,
|
|
"grad_norm": 3.0,
|
|
"learning_rate": 1.955572805786141e-05,
|
|
"loss": 0.8568,
|
|
"mean_token_accuracy": 0.76165372133255,
|
|
"num_tokens": 718088.0,
|
|
"step": 49
|
|
},
|
|
{
|
|
"entropy": 0.8609960898756981,
|
|
"epoch": 0.29784065524944153,
|
|
"grad_norm": 2.890625,
|
|
"learning_rate": 1.953716950748227e-05,
|
|
"loss": 0.8717,
|
|
"mean_token_accuracy": 0.7558733373880386,
|
|
"num_tokens": 732625.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 0.8724842816591263,
|
|
"epoch": 0.3037974683544304,
|
|
"grad_norm": 2.90625,
|
|
"learning_rate": 1.9518240398613226e-05,
|
|
"loss": 0.8882,
|
|
"mean_token_accuracy": 0.7598619237542152,
|
|
"num_tokens": 747023.0,
|
|
"step": 51
|
|
},
|
|
{
|
|
"entropy": 0.8647735044360161,
|
|
"epoch": 0.3097542814594192,
|
|
"grad_norm": 2.765625,
|
|
"learning_rate": 1.9498941466728462e-05,
|
|
"loss": 0.8551,
|
|
"mean_token_accuracy": 0.765186607837677,
|
|
"num_tokens": 762573.0,
|
|
"step": 52
|
|
},
|
|
{
|
|
"entropy": 0.8827357888221741,
|
|
"epoch": 0.31571109456440805,
|
|
"grad_norm": 2.890625,
|
|
"learning_rate": 1.947927346167132e-05,
|
|
"loss": 0.8665,
|
|
"mean_token_accuracy": 0.7552934139966965,
|
|
"num_tokens": 777221.0,
|
|
"step": 53
|
|
},
|
|
{
|
|
"entropy": 0.9407295510172844,
|
|
"epoch": 0.32166790766939685,
|
|
"grad_norm": 2.875,
|
|
"learning_rate": 1.945923714762516e-05,
|
|
"loss": 0.9366,
|
|
"mean_token_accuracy": 0.7403006702661514,
|
|
"num_tokens": 792127.0,
|
|
"step": 54
|
|
},
|
|
{
|
|
"entropy": 0.8792069405317307,
|
|
"epoch": 0.3276247207743857,
|
|
"grad_norm": 2.78125,
|
|
"learning_rate": 1.9438833303083677e-05,
|
|
"loss": 0.8368,
|
|
"mean_token_accuracy": 0.7654407322406769,
|
|
"num_tokens": 806760.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 0.8472597599029541,
|
|
"epoch": 0.33358153387937456,
|
|
"grad_norm": 2.84375,
|
|
"learning_rate": 1.9418062720820636e-05,
|
|
"loss": 0.8111,
|
|
"mean_token_accuracy": 0.7706683054566383,
|
|
"num_tokens": 821430.0,
|
|
"step": 56
|
|
},
|
|
{
|
|
"entropy": 0.9144669771194458,
|
|
"epoch": 0.33953834698436336,
|
|
"grad_norm": 2.96875,
|
|
"learning_rate": 1.9396926207859085e-05,
|
|
"loss": 0.9204,
|
|
"mean_token_accuracy": 0.7486886456608772,
|
|
"num_tokens": 835919.0,
|
|
"step": 57
|
|
},
|
|
{
|
|
"entropy": 0.8967088237404823,
|
|
"epoch": 0.3454951600893522,
|
|
"grad_norm": 2.9375,
|
|
"learning_rate": 1.9375424585439994e-05,
|
|
"loss": 0.9173,
|
|
"mean_token_accuracy": 0.7497560828924179,
|
|
"num_tokens": 851231.0,
|
|
"step": 58
|
|
},
|
|
{
|
|
"entropy": 0.7931167930364609,
|
|
"epoch": 0.351451973194341,
|
|
"grad_norm": 2.875,
|
|
"learning_rate": 1.935355868899034e-05,
|
|
"loss": 0.7795,
|
|
"mean_token_accuracy": 0.7818325236439705,
|
|
"num_tokens": 865741.0,
|
|
"step": 59
|
|
},
|
|
{
|
|
"entropy": 0.8725070431828499,
|
|
"epoch": 0.3574087862993299,
|
|
"grad_norm": 3.109375,
|
|
"learning_rate": 1.9331329368090664e-05,
|
|
"loss": 0.8915,
|
|
"mean_token_accuracy": 0.7528255730867386,
|
|
"num_tokens": 879839.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 0.8313445076346397,
|
|
"epoch": 0.3633655994043187,
|
|
"grad_norm": 2.765625,
|
|
"learning_rate": 1.9308737486442045e-05,
|
|
"loss": 0.8357,
|
|
"mean_token_accuracy": 0.7655151635408401,
|
|
"num_tokens": 894957.0,
|
|
"step": 61
|
|
},
|
|
{
|
|
"entropy": 0.9018023759126663,
|
|
"epoch": 0.36932241250930753,
|
|
"grad_norm": 3.0,
|
|
"learning_rate": 1.9285783921832537e-05,
|
|
"loss": 0.9242,
|
|
"mean_token_accuracy": 0.7491057440638542,
|
|
"num_tokens": 909035.0,
|
|
"step": 62
|
|
},
|
|
{
|
|
"entropy": 0.7900512292981148,
|
|
"epoch": 0.37527922561429633,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 1.926246956610309e-05,
|
|
"loss": 0.8221,
|
|
"mean_token_accuracy": 0.7712607458233833,
|
|
"num_tokens": 923699.0,
|
|
"step": 63
|
|
},
|
|
{
|
|
"entropy": 0.7777893543243408,
|
|
"epoch": 0.3812360387192852,
|
|
"grad_norm": 2.671875,
|
|
"learning_rate": 1.9238795325112867e-05,
|
|
"loss": 0.7841,
|
|
"mean_token_accuracy": 0.7807165011763573,
|
|
"num_tokens": 937978.0,
|
|
"step": 64
|
|
},
|
|
{
|
|
"entropy": 0.8521334230899811,
|
|
"epoch": 0.387192851824274,
|
|
"grad_norm": 2.765625,
|
|
"learning_rate": 1.921476211870408e-05,
|
|
"loss": 0.8709,
|
|
"mean_token_accuracy": 0.756039597094059,
|
|
"num_tokens": 953189.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 0.8546372950077057,
|
|
"epoch": 0.39314966492926284,
|
|
"grad_norm": 2.828125,
|
|
"learning_rate": 1.9190370880666206e-05,
|
|
"loss": 0.8477,
|
|
"mean_token_accuracy": 0.7626818343997002,
|
|
"num_tokens": 967694.0,
|
|
"step": 66
|
|
},
|
|
{
|
|
"entropy": 0.875230647623539,
|
|
"epoch": 0.3991064780342517,
|
|
"grad_norm": 2.9375,
|
|
"learning_rate": 1.9165622558699763e-05,
|
|
"loss": 0.9035,
|
|
"mean_token_accuracy": 0.7550581321120262,
|
|
"num_tokens": 981909.0,
|
|
"step": 67
|
|
},
|
|
{
|
|
"entropy": 0.8000878095626831,
|
|
"epoch": 0.4050632911392405,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 1.9140518114379433e-05,
|
|
"loss": 0.8157,
|
|
"mean_token_accuracy": 0.7727322801947594,
|
|
"num_tokens": 996738.0,
|
|
"step": 68
|
|
},
|
|
{
|
|
"entropy": 0.8553767129778862,
|
|
"epoch": 0.41102010424422936,
|
|
"grad_norm": 2.703125,
|
|
"learning_rate": 1.9115058523116734e-05,
|
|
"loss": 0.821,
|
|
"mean_token_accuracy": 0.7679566293954849,
|
|
"num_tokens": 1011216.0,
|
|
"step": 69
|
|
},
|
|
{
|
|
"entropy": 0.849766805768013,
|
|
"epoch": 0.41697691734921816,
|
|
"grad_norm": 2.703125,
|
|
"learning_rate": 1.908924477412211e-05,
|
|
"loss": 0.8123,
|
|
"mean_token_accuracy": 0.7721216231584549,
|
|
"num_tokens": 1026284.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 0.8852875977754593,
|
|
"epoch": 0.422933730454207,
|
|
"grad_norm": 2.703125,
|
|
"learning_rate": 1.9063077870366504e-05,
|
|
"loss": 0.8797,
|
|
"mean_token_accuracy": 0.7579413875937462,
|
|
"num_tokens": 1041027.0,
|
|
"step": 71
|
|
},
|
|
{
|
|
"entropy": 0.8613467961549759,
|
|
"epoch": 0.4288905435591958,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 1.903655882854237e-05,
|
|
"loss": 0.8642,
|
|
"mean_token_accuracy": 0.759636901319027,
|
|
"num_tokens": 1056203.0,
|
|
"step": 72
|
|
},
|
|
{
|
|
"entropy": 0.8282302543520927,
|
|
"epoch": 0.43484735666418467,
|
|
"grad_norm": 2.9375,
|
|
"learning_rate": 1.900968867902419e-05,
|
|
"loss": 0.8192,
|
|
"mean_token_accuracy": 0.7647311314940453,
|
|
"num_tokens": 1069994.0,
|
|
"step": 73
|
|
},
|
|
{
|
|
"entropy": 0.8411881923675537,
|
|
"epoch": 0.44080416976917347,
|
|
"grad_norm": 3.109375,
|
|
"learning_rate": 1.898246846582844e-05,
|
|
"loss": 0.8602,
|
|
"mean_token_accuracy": 0.7605989798903465,
|
|
"num_tokens": 1084280.0,
|
|
"step": 74
|
|
},
|
|
{
|
|
"entropy": 0.7777607515454292,
|
|
"epoch": 0.4467609828741623,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 1.895489924657301e-05,
|
|
"loss": 0.7905,
|
|
"mean_token_accuracy": 0.7830282896757126,
|
|
"num_tokens": 1099332.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 0.7751665562391281,
|
|
"epoch": 0.4527177959791511,
|
|
"grad_norm": 2.921875,
|
|
"learning_rate": 1.8926982092436117e-05,
|
|
"loss": 0.7942,
|
|
"mean_token_accuracy": 0.775122381746769,
|
|
"num_tokens": 1113758.0,
|
|
"step": 76
|
|
},
|
|
{
|
|
"entropy": 0.8039308115839958,
|
|
"epoch": 0.45867460908414,
|
|
"grad_norm": 2.84375,
|
|
"learning_rate": 1.8898718088114688e-05,
|
|
"loss": 0.842,
|
|
"mean_token_accuracy": 0.7672324255108833,
|
|
"num_tokens": 1128814.0,
|
|
"step": 77
|
|
},
|
|
{
|
|
"entropy": 0.7784704491496086,
|
|
"epoch": 0.46463142218912884,
|
|
"grad_norm": 2.84375,
|
|
"learning_rate": 1.887010833178222e-05,
|
|
"loss": 0.7914,
|
|
"mean_token_accuracy": 0.7708240523934364,
|
|
"num_tokens": 1143044.0,
|
|
"step": 78
|
|
},
|
|
{
|
|
"entropy": 0.7843014299869537,
|
|
"epoch": 0.47058823529411764,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 1.8841153935046098e-05,
|
|
"loss": 0.7906,
|
|
"mean_token_accuracy": 0.7736422047019005,
|
|
"num_tokens": 1158074.0,
|
|
"step": 79
|
|
},
|
|
{
|
|
"entropy": 0.8171004354953766,
|
|
"epoch": 0.4765450483991065,
|
|
"grad_norm": 2.859375,
|
|
"learning_rate": 1.8811856022904423e-05,
|
|
"loss": 0.8187,
|
|
"mean_token_accuracy": 0.7730057835578918,
|
|
"num_tokens": 1172247.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 0.7312511578202248,
|
|
"epoch": 0.4825018615040953,
|
|
"grad_norm": 2.609375,
|
|
"learning_rate": 1.8782215733702286e-05,
|
|
"loss": 0.7357,
|
|
"mean_token_accuracy": 0.7900804728269577,
|
|
"num_tokens": 1187054.0,
|
|
"step": 81
|
|
},
|
|
{
|
|
"entropy": 0.8289615735411644,
|
|
"epoch": 0.48845867460908415,
|
|
"grad_norm": 2.828125,
|
|
"learning_rate": 1.8752234219087538e-05,
|
|
"loss": 0.8527,
|
|
"mean_token_accuracy": 0.7677234709262848,
|
|
"num_tokens": 1201655.0,
|
|
"step": 82
|
|
},
|
|
{
|
|
"entropy": 0.8504510894417763,
|
|
"epoch": 0.49441548771407295,
|
|
"grad_norm": 2.953125,
|
|
"learning_rate": 1.8721912643966055e-05,
|
|
"loss": 0.8618,
|
|
"mean_token_accuracy": 0.7615446001291275,
|
|
"num_tokens": 1215941.0,
|
|
"step": 83
|
|
},
|
|
{
|
|
"entropy": 0.8274638205766678,
|
|
"epoch": 0.5003723008190618,
|
|
"grad_norm": 2.734375,
|
|
"learning_rate": 1.8691252186456465e-05,
|
|
"loss": 0.8268,
|
|
"mean_token_accuracy": 0.7684095054864883,
|
|
"num_tokens": 1230066.0,
|
|
"step": 84
|
|
},
|
|
{
|
|
"entropy": 0.8172023519873619,
|
|
"epoch": 0.5063291139240507,
|
|
"grad_norm": 2.703125,
|
|
"learning_rate": 1.866025403784439e-05,
|
|
"loss": 0.8084,
|
|
"mean_token_accuracy": 0.7703977599740028,
|
|
"num_tokens": 1245411.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 0.8355653882026672,
|
|
"epoch": 0.5122859270290394,
|
|
"grad_norm": 2.671875,
|
|
"learning_rate": 1.862891940253613e-05,
|
|
"loss": 0.8434,
|
|
"mean_token_accuracy": 0.7626368254423141,
|
|
"num_tokens": 1260081.0,
|
|
"step": 86
|
|
},
|
|
{
|
|
"entropy": 0.780564196407795,
|
|
"epoch": 0.5182427401340283,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 1.8597249498011906e-05,
|
|
"loss": 0.7753,
|
|
"mean_token_accuracy": 0.7802636325359344,
|
|
"num_tokens": 1274169.0,
|
|
"step": 87
|
|
},
|
|
{
|
|
"entropy": 0.7850174382328987,
|
|
"epoch": 0.5241995532390171,
|
|
"grad_norm": 4.03125,
|
|
"learning_rate": 1.8565245554778516e-05,
|
|
"loss": 0.7824,
|
|
"mean_token_accuracy": 0.7741290479898453,
|
|
"num_tokens": 1288810.0,
|
|
"step": 88
|
|
},
|
|
{
|
|
"entropy": 0.8584944009780884,
|
|
"epoch": 0.530156366344006,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 1.8532908816321557e-05,
|
|
"loss": 0.8295,
|
|
"mean_token_accuracy": 0.7657060846686363,
|
|
"num_tokens": 1303352.0,
|
|
"step": 89
|
|
},
|
|
{
|
|
"entropy": 0.8400231376290321,
|
|
"epoch": 0.5361131794489948,
|
|
"grad_norm": 2.734375,
|
|
"learning_rate": 1.8500240539057093e-05,
|
|
"loss": 0.8352,
|
|
"mean_token_accuracy": 0.7626395598053932,
|
|
"num_tokens": 1317948.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 0.8977006673812866,
|
|
"epoch": 0.5420699925539836,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 1.8467241992282842e-05,
|
|
"loss": 0.8941,
|
|
"mean_token_accuracy": 0.7465017959475517,
|
|
"num_tokens": 1332383.0,
|
|
"step": 91
|
|
},
|
|
{
|
|
"entropy": 0.7989586219191551,
|
|
"epoch": 0.5480268056589724,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 1.843391445812886e-05,
|
|
"loss": 0.793,
|
|
"mean_token_accuracy": 0.7722924873232841,
|
|
"num_tokens": 1346797.0,
|
|
"step": 92
|
|
},
|
|
{
|
|
"entropy": 0.7695709094405174,
|
|
"epoch": 0.5539836187639613,
|
|
"grad_norm": 2.71875,
|
|
"learning_rate": 1.8400259231507716e-05,
|
|
"loss": 0.7725,
|
|
"mean_token_accuracy": 0.779527023434639,
|
|
"num_tokens": 1360935.0,
|
|
"step": 93
|
|
},
|
|
{
|
|
"entropy": 0.789020448923111,
|
|
"epoch": 0.5599404318689502,
|
|
"grad_norm": 2.546875,
|
|
"learning_rate": 1.83662776200642e-05,
|
|
"loss": 0.7813,
|
|
"mean_token_accuracy": 0.7760330885648727,
|
|
"num_tokens": 1375392.0,
|
|
"step": 94
|
|
},
|
|
{
|
|
"entropy": 0.771543525159359,
|
|
"epoch": 0.5658972449739389,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 1.833197094412449e-05,
|
|
"loss": 0.7697,
|
|
"mean_token_accuracy": 0.7768829986453056,
|
|
"num_tokens": 1389642.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 0.7857540473341942,
|
|
"epoch": 0.5718540580789278,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 1.8297340536644877e-05,
|
|
"loss": 0.8023,
|
|
"mean_token_accuracy": 0.7720776870846748,
|
|
"num_tokens": 1404072.0,
|
|
"step": 96
|
|
},
|
|
{
|
|
"entropy": 0.7771449312567711,
|
|
"epoch": 0.5778108711839166,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 1.826238774315995e-05,
|
|
"loss": 0.7691,
|
|
"mean_token_accuracy": 0.7801476120948792,
|
|
"num_tokens": 1418644.0,
|
|
"step": 97
|
|
},
|
|
{
|
|
"entropy": 0.7695603221654892,
|
|
"epoch": 0.5837676842889055,
|
|
"grad_norm": 2.78125,
|
|
"learning_rate": 1.8227113921730336e-05,
|
|
"loss": 0.7678,
|
|
"mean_token_accuracy": 0.7779613733291626,
|
|
"num_tokens": 1432848.0,
|
|
"step": 98
|
|
},
|
|
{
|
|
"entropy": 0.7877998277544975,
|
|
"epoch": 0.5897244973938943,
|
|
"grad_norm": 3.171875,
|
|
"learning_rate": 1.819152044288992e-05,
|
|
"loss": 0.8406,
|
|
"mean_token_accuracy": 0.7601675689220428,
|
|
"num_tokens": 1446026.0,
|
|
"step": 99
|
|
},
|
|
{
|
|
"entropy": 0.7991646230220795,
|
|
"epoch": 0.5956813104988831,
|
|
"grad_norm": 2.890625,
|
|
"learning_rate": 1.8155608689592604e-05,
|
|
"loss": 0.8248,
|
|
"mean_token_accuracy": 0.7615427002310753,
|
|
"num_tokens": 1461039.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 0.7964242175221443,
|
|
"epoch": 0.6016381236038719,
|
|
"grad_norm": 2.671875,
|
|
"learning_rate": 1.811938005715857e-05,
|
|
"loss": 0.771,
|
|
"mean_token_accuracy": 0.7777570560574532,
|
|
"num_tokens": 1475526.0,
|
|
"step": 101
|
|
},
|
|
{
|
|
"entropy": 0.8020969182252884,
|
|
"epoch": 0.6075949367088608,
|
|
"grad_norm": 2.65625,
|
|
"learning_rate": 1.8082835953220055e-05,
|
|
"loss": 0.8012,
|
|
"mean_token_accuracy": 0.7739049047231674,
|
|
"num_tokens": 1489823.0,
|
|
"step": 102
|
|
},
|
|
{
|
|
"entropy": 0.8560249656438828,
|
|
"epoch": 0.6135517498138496,
|
|
"grad_norm": 2.953125,
|
|
"learning_rate": 1.8045977797666685e-05,
|
|
"loss": 0.853,
|
|
"mean_token_accuracy": 0.760221041738987,
|
|
"num_tokens": 1504383.0,
|
|
"step": 103
|
|
},
|
|
{
|
|
"entropy": 0.8747018277645111,
|
|
"epoch": 0.6195085629188384,
|
|
"grad_norm": 2.96875,
|
|
"learning_rate": 1.8008807022590283e-05,
|
|
"loss": 0.8782,
|
|
"mean_token_accuracy": 0.7500127330422401,
|
|
"num_tokens": 1517806.0,
|
|
"step": 104
|
|
},
|
|
{
|
|
"entropy": 0.8000111356377602,
|
|
"epoch": 0.6254653760238272,
|
|
"grad_norm": 2.734375,
|
|
"learning_rate": 1.7971325072229227e-05,
|
|
"loss": 0.8354,
|
|
"mean_token_accuracy": 0.768502801656723,
|
|
"num_tokens": 1532497.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 0.8312654495239258,
|
|
"epoch": 0.6314221891288161,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 1.7933533402912354e-05,
|
|
"loss": 0.8225,
|
|
"mean_token_accuracy": 0.7695776075124741,
|
|
"num_tokens": 1547634.0,
|
|
"step": 106
|
|
},
|
|
{
|
|
"entropy": 0.8169409111142159,
|
|
"epoch": 0.637379002233805,
|
|
"grad_norm": 2.5625,
|
|
"learning_rate": 1.7895433483002356e-05,
|
|
"loss": 0.8334,
|
|
"mean_token_accuracy": 0.7691005021333694,
|
|
"num_tokens": 1562026.0,
|
|
"step": 107
|
|
},
|
|
{
|
|
"entropy": 0.8326558768749237,
|
|
"epoch": 0.6433358153387937,
|
|
"grad_norm": 3.078125,
|
|
"learning_rate": 1.785702679283874e-05,
|
|
"loss": 0.8656,
|
|
"mean_token_accuracy": 0.7613467499613762,
|
|
"num_tokens": 1575785.0,
|
|
"step": 108
|
|
},
|
|
{
|
|
"entropy": 0.8096423745155334,
|
|
"epoch": 0.6492926284437825,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 1.78183148246803e-05,
|
|
"loss": 0.7975,
|
|
"mean_token_accuracy": 0.7697085440158844,
|
|
"num_tokens": 1591294.0,
|
|
"step": 109
|
|
},
|
|
{
|
|
"entropy": 0.8186523467302322,
|
|
"epoch": 0.6552494415487714,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 1.777929908264715e-05,
|
|
"loss": 0.8136,
|
|
"mean_token_accuracy": 0.7723612412810326,
|
|
"num_tokens": 1604958.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 0.836366131901741,
|
|
"epoch": 0.6612062546537603,
|
|
"grad_norm": 2.8125,
|
|
"learning_rate": 1.7739981082662275e-05,
|
|
"loss": 0.8341,
|
|
"mean_token_accuracy": 0.7610246166586876,
|
|
"num_tokens": 1620204.0,
|
|
"step": 111
|
|
},
|
|
{
|
|
"entropy": 0.7696946710348129,
|
|
"epoch": 0.6671630677587491,
|
|
"grad_norm": 2.578125,
|
|
"learning_rate": 1.7700362352392632e-05,
|
|
"loss": 0.7748,
|
|
"mean_token_accuracy": 0.7808224633336067,
|
|
"num_tokens": 1635257.0,
|
|
"step": 112
|
|
},
|
|
{
|
|
"entropy": 0.8294946998357773,
|
|
"epoch": 0.6731198808637379,
|
|
"grad_norm": 2.96875,
|
|
"learning_rate": 1.766044443118978e-05,
|
|
"loss": 0.8364,
|
|
"mean_token_accuracy": 0.764847993850708,
|
|
"num_tokens": 1649668.0,
|
|
"step": 113
|
|
},
|
|
{
|
|
"entropy": 0.8253164365887642,
|
|
"epoch": 0.6790766939687267,
|
|
"grad_norm": 2.765625,
|
|
"learning_rate": 1.762022887003011e-05,
|
|
"loss": 0.815,
|
|
"mean_token_accuracy": 0.7684815153479576,
|
|
"num_tokens": 1663868.0,
|
|
"step": 114
|
|
},
|
|
{
|
|
"entropy": 0.7470197826623917,
|
|
"epoch": 0.6850335070737156,
|
|
"grad_norm": 2.765625,
|
|
"learning_rate": 1.757971723145453e-05,
|
|
"loss": 0.741,
|
|
"mean_token_accuracy": 0.7875789180397987,
|
|
"num_tokens": 1678332.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 0.8128603473305702,
|
|
"epoch": 0.6909903201787044,
|
|
"grad_norm": 2.640625,
|
|
"learning_rate": 1.75389110895078e-05,
|
|
"loss": 0.8328,
|
|
"mean_token_accuracy": 0.7683836743235588,
|
|
"num_tokens": 1692798.0,
|
|
"step": 116
|
|
},
|
|
{
|
|
"entropy": 0.778105765581131,
|
|
"epoch": 0.6969471332836932,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 1.7497812029677344e-05,
|
|
"loss": 0.7876,
|
|
"mean_token_accuracy": 0.7749749645590782,
|
|
"num_tokens": 1706403.0,
|
|
"step": 117
|
|
},
|
|
{
|
|
"entropy": 0.7804563790559769,
|
|
"epoch": 0.702903946388682,
|
|
"grad_norm": 3.265625,
|
|
"learning_rate": 1.7456421648831658e-05,
|
|
"loss": 0.7908,
|
|
"mean_token_accuracy": 0.7710901945829391,
|
|
"num_tokens": 1720637.0,
|
|
"step": 118
|
|
},
|
|
{
|
|
"entropy": 0.8247266188263893,
|
|
"epoch": 0.7088607594936709,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 1.741474155515827e-05,
|
|
"loss": 0.8222,
|
|
"mean_token_accuracy": 0.7634738311171532,
|
|
"num_tokens": 1735839.0,
|
|
"step": 119
|
|
},
|
|
{
|
|
"entropy": 0.793066531419754,
|
|
"epoch": 0.7148175725986597,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 1.737277336810124e-05,
|
|
"loss": 0.7645,
|
|
"mean_token_accuracy": 0.782652385532856,
|
|
"num_tokens": 1750093.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 0.7985671162605286,
|
|
"epoch": 0.7207743857036486,
|
|
"grad_norm": 2.703125,
|
|
"learning_rate": 1.7330518718298263e-05,
|
|
"loss": 0.8126,
|
|
"mean_token_accuracy": 0.7698739618062973,
|
|
"num_tokens": 1765040.0,
|
|
"step": 121
|
|
},
|
|
{
|
|
"entropy": 0.815500408411026,
|
|
"epoch": 0.7267311988086373,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 1.7287979247517285e-05,
|
|
"loss": 0.8181,
|
|
"mean_token_accuracy": 0.7727770283818245,
|
|
"num_tokens": 1779610.0,
|
|
"step": 122
|
|
},
|
|
{
|
|
"entropy": 0.7722458392381668,
|
|
"epoch": 0.7326880119136262,
|
|
"grad_norm": 2.671875,
|
|
"learning_rate": 1.7245156608592727e-05,
|
|
"loss": 0.7501,
|
|
"mean_token_accuracy": 0.7853315845131874,
|
|
"num_tokens": 1794351.0,
|
|
"step": 123
|
|
},
|
|
{
|
|
"entropy": 0.8114070221781731,
|
|
"epoch": 0.7386448250186151,
|
|
"grad_norm": 2.71875,
|
|
"learning_rate": 1.7202052465361268e-05,
|
|
"loss": 0.8195,
|
|
"mean_token_accuracy": 0.7718819826841354,
|
|
"num_tokens": 1809510.0,
|
|
"step": 124
|
|
},
|
|
{
|
|
"entropy": 0.8247914463281631,
|
|
"epoch": 0.7446016381236039,
|
|
"grad_norm": 2.875,
|
|
"learning_rate": 1.7158668492597186e-05,
|
|
"loss": 0.8497,
|
|
"mean_token_accuracy": 0.7609348967671394,
|
|
"num_tokens": 1823089.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 0.8314331695437431,
|
|
"epoch": 0.7505584512285927,
|
|
"grad_norm": 2.828125,
|
|
"learning_rate": 1.7115006375947304e-05,
|
|
"loss": 0.8345,
|
|
"mean_token_accuracy": 0.7624246254563332,
|
|
"num_tokens": 1837660.0,
|
|
"step": 126
|
|
},
|
|
{
|
|
"entropy": 0.8372238278388977,
|
|
"epoch": 0.7565152643335815,
|
|
"grad_norm": 2.765625,
|
|
"learning_rate": 1.7071067811865477e-05,
|
|
"loss": 0.8322,
|
|
"mean_token_accuracy": 0.7600690796971321,
|
|
"num_tokens": 1851869.0,
|
|
"step": 127
|
|
},
|
|
{
|
|
"entropy": 0.7920485585927963,
|
|
"epoch": 0.7624720774385704,
|
|
"grad_norm": 2.71875,
|
|
"learning_rate": 1.7026854507546694e-05,
|
|
"loss": 0.7779,
|
|
"mean_token_accuracy": 0.7758874222636223,
|
|
"num_tokens": 1866454.0,
|
|
"step": 128
|
|
},
|
|
{
|
|
"entropy": 0.8146230578422546,
|
|
"epoch": 0.7684288905435592,
|
|
"grad_norm": 2.703125,
|
|
"learning_rate": 1.698236818086073e-05,
|
|
"loss": 0.8168,
|
|
"mean_token_accuracy": 0.763184204697609,
|
|
"num_tokens": 1880761.0,
|
|
"step": 129
|
|
},
|
|
{
|
|
"entropy": 0.8262524753808975,
|
|
"epoch": 0.774385703648548,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 1.693761056028542e-05,
|
|
"loss": 0.8409,
|
|
"mean_token_accuracy": 0.7651299685239792,
|
|
"num_tokens": 1895812.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 0.7881096228957176,
|
|
"epoch": 0.7803425167535368,
|
|
"grad_norm": 2.546875,
|
|
"learning_rate": 1.689258338483947e-05,
|
|
"loss": 0.764,
|
|
"mean_token_accuracy": 0.7817611619830132,
|
|
"num_tokens": 1910875.0,
|
|
"step": 131
|
|
},
|
|
{
|
|
"entropy": 0.7774896547198296,
|
|
"epoch": 0.7862993298585257,
|
|
"grad_norm": 2.53125,
|
|
"learning_rate": 1.6847288404014937e-05,
|
|
"loss": 0.7729,
|
|
"mean_token_accuracy": 0.7737534716725349,
|
|
"num_tokens": 1925210.0,
|
|
"step": 132
|
|
},
|
|
{
|
|
"entropy": 0.8020075485110283,
|
|
"epoch": 0.7922561429635145,
|
|
"grad_norm": 2.609375,
|
|
"learning_rate": 1.6801727377709195e-05,
|
|
"loss": 0.7858,
|
|
"mean_token_accuracy": 0.7749705463647842,
|
|
"num_tokens": 1939706.0,
|
|
"step": 133
|
|
},
|
|
{
|
|
"entropy": 0.7764975354075432,
|
|
"epoch": 0.7982129560685034,
|
|
"grad_norm": 2.765625,
|
|
"learning_rate": 1.6755902076156606e-05,
|
|
"loss": 0.7727,
|
|
"mean_token_accuracy": 0.7722419053316116,
|
|
"num_tokens": 1953951.0,
|
|
"step": 134
|
|
},
|
|
{
|
|
"entropy": 0.7186224311590195,
|
|
"epoch": 0.8041697691734921,
|
|
"grad_norm": 2.65625,
|
|
"learning_rate": 1.67098142798597e-05,
|
|
"loss": 0.7469,
|
|
"mean_token_accuracy": 0.7871624454855919,
|
|
"num_tokens": 1968478.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 0.7475928589701653,
|
|
"epoch": 0.810126582278481,
|
|
"grad_norm": 2.703125,
|
|
"learning_rate": 1.6663465779520042e-05,
|
|
"loss": 0.7718,
|
|
"mean_token_accuracy": 0.784703366458416,
|
|
"num_tokens": 1982972.0,
|
|
"step": 136
|
|
},
|
|
{
|
|
"entropy": 0.772842176258564,
|
|
"epoch": 0.8160833953834699,
|
|
"grad_norm": 2.65625,
|
|
"learning_rate": 1.6616858375968596e-05,
|
|
"loss": 0.7802,
|
|
"mean_token_accuracy": 0.7753547430038452,
|
|
"num_tokens": 1997617.0,
|
|
"step": 137
|
|
},
|
|
{
|
|
"entropy": 0.775102287530899,
|
|
"epoch": 0.8220402084884587,
|
|
"grad_norm": 2.8125,
|
|
"learning_rate": 1.6569993880095807e-05,
|
|
"loss": 0.8019,
|
|
"mean_token_accuracy": 0.7741298601031303,
|
|
"num_tokens": 2011907.0,
|
|
"step": 138
|
|
},
|
|
{
|
|
"entropy": 0.7909455075860023,
|
|
"epoch": 0.8279970215934475,
|
|
"grad_norm": 2.75,
|
|
"learning_rate": 1.6522874112781213e-05,
|
|
"loss": 0.8084,
|
|
"mean_token_accuracy": 0.7722871527075768,
|
|
"num_tokens": 2026748.0,
|
|
"step": 139
|
|
},
|
|
{
|
|
"entropy": 0.7726485431194305,
|
|
"epoch": 0.8339538346984363,
|
|
"grad_norm": 2.609375,
|
|
"learning_rate": 1.6475500904822707e-05,
|
|
"loss": 0.7771,
|
|
"mean_token_accuracy": 0.7788979560136795,
|
|
"num_tokens": 2041028.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 0.7483975887298584,
|
|
"epoch": 0.8399106478034252,
|
|
"grad_norm": 2.53125,
|
|
"learning_rate": 1.6427876096865394e-05,
|
|
"loss": 0.7464,
|
|
"mean_token_accuracy": 0.7817212343215942,
|
|
"num_tokens": 2056459.0,
|
|
"step": 141
|
|
},
|
|
{
|
|
"entropy": 0.8063411265611649,
|
|
"epoch": 0.845867460908414,
|
|
"grad_norm": 2.640625,
|
|
"learning_rate": 1.6380001539330088e-05,
|
|
"loss": 0.8058,
|
|
"mean_token_accuracy": 0.768462710082531,
|
|
"num_tokens": 2071572.0,
|
|
"step": 142
|
|
},
|
|
{
|
|
"entropy": 0.81895712018013,
|
|
"epoch": 0.8518242740134029,
|
|
"grad_norm": 2.640625,
|
|
"learning_rate": 1.6331879092341402e-05,
|
|
"loss": 0.823,
|
|
"mean_token_accuracy": 0.7620123773813248,
|
|
"num_tokens": 2086270.0,
|
|
"step": 143
|
|
},
|
|
{
|
|
"entropy": 0.7827034220099449,
|
|
"epoch": 0.8577810871183916,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 1.6283510625655474e-05,
|
|
"loss": 0.7777,
|
|
"mean_token_accuracy": 0.7804948911070824,
|
|
"num_tokens": 2100982.0,
|
|
"step": 144
|
|
},
|
|
{
|
|
"entropy": 0.7757274061441422,
|
|
"epoch": 0.8637379002233805,
|
|
"grad_norm": 2.40625,
|
|
"learning_rate": 1.6234898018587336e-05,
|
|
"loss": 0.7534,
|
|
"mean_token_accuracy": 0.7780917063355446,
|
|
"num_tokens": 2115803.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 0.7186077982187271,
|
|
"epoch": 0.8696947133283693,
|
|
"grad_norm": 2.421875,
|
|
"learning_rate": 1.6186043159937884e-05,
|
|
"loss": 0.7088,
|
|
"mean_token_accuracy": 0.7938716411590576,
|
|
"num_tokens": 2130971.0,
|
|
"step": 146
|
|
},
|
|
{
|
|
"entropy": 0.7525220066308975,
|
|
"epoch": 0.8756515264333582,
|
|
"grad_norm": 2.5,
|
|
"learning_rate": 1.6136947947920477e-05,
|
|
"loss": 0.7376,
|
|
"mean_token_accuracy": 0.7867058664560318,
|
|
"num_tokens": 2145552.0,
|
|
"step": 147
|
|
},
|
|
{
|
|
"entropy": 0.8051852136850357,
|
|
"epoch": 0.8816083395383469,
|
|
"grad_norm": 2.734375,
|
|
"learning_rate": 1.608761429008721e-05,
|
|
"loss": 0.8408,
|
|
"mean_token_accuracy": 0.7625209614634514,
|
|
"num_tokens": 2160069.0,
|
|
"step": 148
|
|
},
|
|
{
|
|
"entropy": 0.7129626348614693,
|
|
"epoch": 0.8875651526433358,
|
|
"grad_norm": 2.65625,
|
|
"learning_rate": 1.6038044103254775e-05,
|
|
"loss": 0.7091,
|
|
"mean_token_accuracy": 0.7939872592687607,
|
|
"num_tokens": 2175381.0,
|
|
"step": 149
|
|
},
|
|
{
|
|
"entropy": 0.7308422699570656,
|
|
"epoch": 0.8935219657483247,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 1.5988239313430004e-05,
|
|
"loss": 0.7568,
|
|
"mean_token_accuracy": 0.7810374870896339,
|
|
"num_tokens": 2190275.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 0.7620030343532562,
|
|
"epoch": 0.8994787788533135,
|
|
"grad_norm": 2.671875,
|
|
"learning_rate": 1.5938201855735017e-05,
|
|
"loss": 0.7797,
|
|
"mean_token_accuracy": 0.7755021527409554,
|
|
"num_tokens": 2204775.0,
|
|
"step": 151
|
|
},
|
|
{
|
|
"entropy": 0.8035119771957397,
|
|
"epoch": 0.9054355919583023,
|
|
"grad_norm": 2.734375,
|
|
"learning_rate": 1.5887933674332048e-05,
|
|
"loss": 0.8385,
|
|
"mean_token_accuracy": 0.7616443932056427,
|
|
"num_tokens": 2219016.0,
|
|
"step": 152
|
|
},
|
|
{
|
|
"entropy": 0.7269817739725113,
|
|
"epoch": 0.9113924050632911,
|
|
"grad_norm": 2.46875,
|
|
"learning_rate": 1.5837436722347902e-05,
|
|
"loss": 0.7148,
|
|
"mean_token_accuracy": 0.7955159544944763,
|
|
"num_tokens": 2233377.0,
|
|
"step": 153
|
|
},
|
|
{
|
|
"entropy": 0.7402635291218758,
|
|
"epoch": 0.91734921816828,
|
|
"grad_norm": 2.5625,
|
|
"learning_rate": 1.578671296179806e-05,
|
|
"loss": 0.7271,
|
|
"mean_token_accuracy": 0.7927682846784592,
|
|
"num_tokens": 2248523.0,
|
|
"step": 154
|
|
},
|
|
{
|
|
"entropy": 0.7693969383835793,
|
|
"epoch": 0.9233060312732688,
|
|
"grad_norm": 2.578125,
|
|
"learning_rate": 1.573576436351046e-05,
|
|
"loss": 0.7685,
|
|
"mean_token_accuracy": 0.7787287458777428,
|
|
"num_tokens": 2263288.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 0.8041390031576157,
|
|
"epoch": 0.9292628443782577,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 1.5684592907048925e-05,
|
|
"loss": 0.7886,
|
|
"mean_token_accuracy": 0.7773109897971153,
|
|
"num_tokens": 2278085.0,
|
|
"step": 156
|
|
},
|
|
{
|
|
"entropy": 0.7588407248258591,
|
|
"epoch": 0.9352196574832464,
|
|
"grad_norm": 2.5625,
|
|
"learning_rate": 1.563320058063622e-05,
|
|
"loss": 0.721,
|
|
"mean_token_accuracy": 0.7877694144845009,
|
|
"num_tokens": 2292673.0,
|
|
"step": 157
|
|
},
|
|
{
|
|
"entropy": 0.7770284041762352,
|
|
"epoch": 0.9411764705882353,
|
|
"grad_norm": 2.359375,
|
|
"learning_rate": 1.5581589381076843e-05,
|
|
"loss": 0.746,
|
|
"mean_token_accuracy": 0.7794747352600098,
|
|
"num_tokens": 2307646.0,
|
|
"step": 158
|
|
},
|
|
{
|
|
"entropy": 0.7631111741065979,
|
|
"epoch": 0.9471332836932241,
|
|
"grad_norm": 2.578125,
|
|
"learning_rate": 1.5529761313679396e-05,
|
|
"loss": 0.7196,
|
|
"mean_token_accuracy": 0.7826567068696022,
|
|
"num_tokens": 2321559.0,
|
|
"step": 159
|
|
},
|
|
{
|
|
"entropy": 0.7937072217464447,
|
|
"epoch": 0.953090096798213,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 1.5477718392178716e-05,
|
|
"loss": 0.8045,
|
|
"mean_token_accuracy": 0.771477721631527,
|
|
"num_tokens": 2336203.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 0.8479283452033997,
|
|
"epoch": 0.9590469099032017,
|
|
"grad_norm": 2.890625,
|
|
"learning_rate": 1.5425462638657597e-05,
|
|
"loss": 0.9102,
|
|
"mean_token_accuracy": 0.750169575214386,
|
|
"num_tokens": 2351344.0,
|
|
"step": 161
|
|
},
|
|
{
|
|
"entropy": 0.7240238338708878,
|
|
"epoch": 0.9650037230081906,
|
|
"grad_norm": 2.46875,
|
|
"learning_rate": 1.5372996083468242e-05,
|
|
"loss": 0.717,
|
|
"mean_token_accuracy": 0.7932362481951714,
|
|
"num_tokens": 2366108.0,
|
|
"step": 162
|
|
},
|
|
{
|
|
"entropy": 0.7222452163696289,
|
|
"epoch": 0.9709605361131795,
|
|
"grad_norm": 2.609375,
|
|
"learning_rate": 1.5320320765153367e-05,
|
|
"loss": 0.7294,
|
|
"mean_token_accuracy": 0.791483536362648,
|
|
"num_tokens": 2381068.0,
|
|
"step": 163
|
|
},
|
|
{
|
|
"entropy": 0.7367577031254768,
|
|
"epoch": 0.9769173492181683,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 1.526743873036701e-05,
|
|
"loss": 0.7407,
|
|
"mean_token_accuracy": 0.784386046230793,
|
|
"num_tokens": 2395295.0,
|
|
"step": 164
|
|
},
|
|
{
|
|
"entropy": 0.7757211104035378,
|
|
"epoch": 0.9828741623231572,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 1.5214352033794981e-05,
|
|
"loss": 0.8013,
|
|
"mean_token_accuracy": 0.7718920782208443,
|
|
"num_tokens": 2410456.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 0.7407322824001312,
|
|
"epoch": 0.9888309754281459,
|
|
"grad_norm": 2.828125,
|
|
"learning_rate": 1.5161062738075068e-05,
|
|
"loss": 0.7465,
|
|
"mean_token_accuracy": 0.7847936823964119,
|
|
"num_tokens": 2424439.0,
|
|
"step": 166
|
|
},
|
|
{
|
|
"entropy": 0.7530843243002892,
|
|
"epoch": 0.9947877885331348,
|
|
"grad_norm": 2.8125,
|
|
"learning_rate": 1.5107572913716859e-05,
|
|
"loss": 0.782,
|
|
"mean_token_accuracy": 0.7779848203063011,
|
|
"num_tokens": 2438585.0,
|
|
"step": 167
|
|
},
|
|
{
|
|
"entropy": 0.8105373467717852,
|
|
"epoch": 1.0,
|
|
"grad_norm": 2.84375,
|
|
"learning_rate": 1.505388463902131e-05,
|
|
"loss": 0.8184,
|
|
"mean_token_accuracy": 0.7624573196683612,
|
|
"num_tokens": 2452101.0,
|
|
"step": 168
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_entropy": 0.7692393018802007,
|
|
"eval_loss": 0.7709271311759949,
|
|
"eval_mean_token_accuracy": 0.7770779418945313,
|
|
"eval_model_preparation_time": 0.0048,
|
|
"eval_num_tokens": 2452101.0,
|
|
"eval_runtime": 26.6985,
|
|
"eval_samples_per_second": 5.618,
|
|
"eval_steps_per_second": 5.618,
|
|
"step": 168
|
|
}
|
|
],
|
|
"logging_steps": 1,
|
|
"max_steps": 504,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 3,
|
|
"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": 1.1041699975881523e+17,
|
|
"train_batch_size": 1,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|