Files
d1_science_gpt_0.3k/trainer_state.json
ModelHub XC ab771acdea 初始化项目,由ModelHub XC社区提供模型
Model: mlfoundations-dev/d1_science_gpt_0.3k
Source: Original Platform
2026-06-03 05:09:13 +08:00

953 lines
21 KiB
JSON

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 13.0,
"eval_steps": 500,
"global_step": 130,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.1,
"grad_norm": 6.733106782508708,
"learning_rate": 7.692307692307694e-07,
"loss": 1.0626,
"step": 1
},
{
"epoch": 0.2,
"grad_norm": 6.149285803125056,
"learning_rate": 1.5384615384615387e-06,
"loss": 1.0435,
"step": 2
},
{
"epoch": 0.3,
"grad_norm": 6.152320728736608,
"learning_rate": 2.307692307692308e-06,
"loss": 1.0193,
"step": 3
},
{
"epoch": 0.4,
"grad_norm": 6.517614170207162,
"learning_rate": 3.0769230769230774e-06,
"loss": 1.02,
"step": 4
},
{
"epoch": 0.5,
"grad_norm": 5.305198033715196,
"learning_rate": 3.846153846153847e-06,
"loss": 1.0348,
"step": 5
},
{
"epoch": 0.6,
"grad_norm": 3.8592003927171907,
"learning_rate": 4.615384615384616e-06,
"loss": 0.954,
"step": 6
},
{
"epoch": 0.7,
"grad_norm": 2.800854752532226,
"learning_rate": 5.384615384615385e-06,
"loss": 0.9604,
"step": 7
},
{
"epoch": 0.8,
"grad_norm": 3.5023896682586497,
"learning_rate": 6.153846153846155e-06,
"loss": 0.8856,
"step": 8
},
{
"epoch": 0.9,
"grad_norm": 4.1684273728378685,
"learning_rate": 6.923076923076923e-06,
"loss": 0.8593,
"step": 9
},
{
"epoch": 1.0,
"grad_norm": 4.8032937097662,
"learning_rate": 7.692307692307694e-06,
"loss": 0.8837,
"step": 10
},
{
"epoch": 1.1,
"grad_norm": 4.082576868292361,
"learning_rate": 8.461538461538462e-06,
"loss": 0.8455,
"step": 11
},
{
"epoch": 1.2,
"grad_norm": 3.907890503246759,
"learning_rate": 9.230769230769232e-06,
"loss": 0.7795,
"step": 12
},
{
"epoch": 1.3,
"grad_norm": 3.5914196704208172,
"learning_rate": 1e-05,
"loss": 0.8492,
"step": 13
},
{
"epoch": 1.4,
"grad_norm": 2.942570514389922,
"learning_rate": 9.998197638354428e-06,
"loss": 0.8569,
"step": 14
},
{
"epoch": 1.5,
"grad_norm": 2.2999462111794804,
"learning_rate": 9.992791852820709e-06,
"loss": 0.771,
"step": 15
},
{
"epoch": 1.6,
"grad_norm": 2.6948422206665636,
"learning_rate": 9.983786540671052e-06,
"loss": 0.7751,
"step": 16
},
{
"epoch": 1.7,
"grad_norm": 2.681033931109484,
"learning_rate": 9.971188194237141e-06,
"loss": 0.7374,
"step": 17
},
{
"epoch": 1.8,
"grad_norm": 2.742683686105219,
"learning_rate": 9.955005896229543e-06,
"loss": 0.7825,
"step": 18
},
{
"epoch": 1.9,
"grad_norm": 2.071739361645055,
"learning_rate": 9.935251313189564e-06,
"loss": 0.7741,
"step": 19
},
{
"epoch": 2.0,
"grad_norm": 1.8767253510635002,
"learning_rate": 9.911938687078324e-06,
"loss": 0.7026,
"step": 20
},
{
"epoch": 2.1,
"grad_norm": 1.7974374045731076,
"learning_rate": 9.885084825009085e-06,
"loss": 0.6487,
"step": 21
},
{
"epoch": 2.2,
"grad_norm": 1.7364077012351282,
"learning_rate": 9.854709087130261e-06,
"loss": 0.6709,
"step": 22
},
{
"epoch": 2.3,
"grad_norm": 1.3747907551026266,
"learning_rate": 9.820833372667813e-06,
"loss": 0.6251,
"step": 23
},
{
"epoch": 2.4,
"grad_norm": 1.2086788755323006,
"learning_rate": 9.783482104137127e-06,
"loss": 0.5849,
"step": 24
},
{
"epoch": 2.5,
"grad_norm": 1.2212416207318195,
"learning_rate": 9.742682209735727e-06,
"loss": 0.6776,
"step": 25
},
{
"epoch": 2.6,
"grad_norm": 1.3324486836498435,
"learning_rate": 9.698463103929542e-06,
"loss": 0.6056,
"step": 26
},
{
"epoch": 2.7,
"grad_norm": 1.2812363796358743,
"learning_rate": 9.650856666246693e-06,
"loss": 0.6272,
"step": 27
},
{
"epoch": 2.8,
"grad_norm": 1.3121250186867432,
"learning_rate": 9.599897218294122e-06,
"loss": 0.6117,
"step": 28
},
{
"epoch": 2.9,
"grad_norm": 1.1416700701522662,
"learning_rate": 9.54562149901362e-06,
"loss": 0.6137,
"step": 29
},
{
"epoch": 3.0,
"grad_norm": 0.8638570875229467,
"learning_rate": 9.488068638195072e-06,
"loss": 0.5514,
"step": 30
},
{
"epoch": 3.1,
"grad_norm": 1.1680224672840247,
"learning_rate": 9.427280128266049e-06,
"loss": 0.4937,
"step": 31
},
{
"epoch": 3.2,
"grad_norm": 0.9985846334069853,
"learning_rate": 9.363299794378072e-06,
"loss": 0.5146,
"step": 32
},
{
"epoch": 3.3,
"grad_norm": 0.993554703522936,
"learning_rate": 9.296173762811084e-06,
"loss": 0.4297,
"step": 33
},
{
"epoch": 3.4,
"grad_norm": 1.0979981540776833,
"learning_rate": 9.225950427718974e-06,
"loss": 0.4366,
"step": 34
},
{
"epoch": 3.5,
"grad_norm": 0.9947660773449873,
"learning_rate": 9.152680416240059e-06,
"loss": 0.4638,
"step": 35
},
{
"epoch": 3.6,
"grad_norm": 1.22170267986729,
"learning_rate": 9.076416551997721e-06,
"loss": 0.4385,
"step": 36
},
{
"epoch": 3.7,
"grad_norm": 1.2669424656892199,
"learning_rate": 8.997213817017508e-06,
"loss": 0.4564,
"step": 37
},
{
"epoch": 3.8,
"grad_norm": 1.1385305262473597,
"learning_rate": 8.915129312088112e-06,
"loss": 0.4638,
"step": 38
},
{
"epoch": 3.9,
"grad_norm": 0.85752544440622,
"learning_rate": 8.83022221559489e-06,
"loss": 0.4355,
"step": 39
},
{
"epoch": 4.0,
"grad_norm": 0.9350416528577298,
"learning_rate": 8.742553740855507e-06,
"loss": 0.4131,
"step": 40
},
{
"epoch": 4.1,
"grad_norm": 1.0044872736721613,
"learning_rate": 8.652187091988516e-06,
"loss": 0.3543,
"step": 41
},
{
"epoch": 4.2,
"grad_norm": 0.9158552416245116,
"learning_rate": 8.559187418346703e-06,
"loss": 0.3948,
"step": 42
},
{
"epoch": 4.3,
"grad_norm": 0.9500005615720003,
"learning_rate": 8.463621767547998e-06,
"loss": 0.3216,
"step": 43
},
{
"epoch": 4.4,
"grad_norm": 1.1593134176446123,
"learning_rate": 8.36555903713785e-06,
"loss": 0.3338,
"step": 44
},
{
"epoch": 4.5,
"grad_norm": 1.036997277407171,
"learning_rate": 8.265069924917925e-06,
"loss": 0.3533,
"step": 45
},
{
"epoch": 4.6,
"grad_norm": 0.9081359993684659,
"learning_rate": 8.162226877976886e-06,
"loss": 0.3303,
"step": 46
},
{
"epoch": 4.7,
"grad_norm": 1.0357743695490003,
"learning_rate": 8.057104040460062e-06,
"loss": 0.3076,
"step": 47
},
{
"epoch": 4.8,
"grad_norm": 1.003725556135553,
"learning_rate": 7.949777200115617e-06,
"loss": 0.2905,
"step": 48
},
{
"epoch": 4.9,
"grad_norm": 0.8306334207263184,
"learning_rate": 7.84032373365578e-06,
"loss": 0.3255,
"step": 49
},
{
"epoch": 5.0,
"grad_norm": 0.9674484368296132,
"learning_rate": 7.728822550972523e-06,
"loss": 0.2633,
"step": 50
},
{
"epoch": 5.1,
"grad_norm": 0.8974490960860493,
"learning_rate": 7.615354038247889e-06,
"loss": 0.2282,
"step": 51
},
{
"epoch": 5.2,
"grad_norm": 0.8365328600849284,
"learning_rate": 7.500000000000001e-06,
"loss": 0.2057,
"step": 52
},
{
"epoch": 5.3,
"grad_norm": 1.0261442314495612,
"learning_rate": 7.382843600106539e-06,
"loss": 0.2344,
"step": 53
},
{
"epoch": 5.4,
"grad_norm": 1.0045141958077353,
"learning_rate": 7.263969301848188e-06,
"loss": 0.2055,
"step": 54
},
{
"epoch": 5.5,
"grad_norm": 0.791279076653458,
"learning_rate": 7.143462807015271e-06,
"loss": 0.2188,
"step": 55
},
{
"epoch": 5.6,
"grad_norm": 0.8526401510959523,
"learning_rate": 7.021410994121525e-06,
"loss": 0.2022,
"step": 56
},
{
"epoch": 5.7,
"grad_norm": 0.8913810238298835,
"learning_rate": 6.897901855769483e-06,
"loss": 0.2313,
"step": 57
},
{
"epoch": 5.8,
"grad_norm": 0.9650702324016643,
"learning_rate": 6.773024435212678e-06,
"loss": 0.1914,
"step": 58
},
{
"epoch": 5.9,
"grad_norm": 0.7872208893856325,
"learning_rate": 6.646868762160399e-06,
"loss": 0.2626,
"step": 59
},
{
"epoch": 6.0,
"grad_norm": 0.9217811368898755,
"learning_rate": 6.519525787871235e-06,
"loss": 0.2043,
"step": 60
},
{
"epoch": 6.1,
"grad_norm": 0.6948999282360269,
"learning_rate": 6.391087319582264e-06,
"loss": 0.1684,
"step": 61
},
{
"epoch": 6.2,
"grad_norm": 0.8075198651452326,
"learning_rate": 6.261645954321109e-06,
"loss": 0.1655,
"step": 62
},
{
"epoch": 6.3,
"grad_norm": 0.8528235582977988,
"learning_rate": 6.131295012148613e-06,
"loss": 0.1035,
"step": 63
},
{
"epoch": 6.4,
"grad_norm": 0.8642283154141255,
"learning_rate": 6.000128468880223e-06,
"loss": 0.1257,
"step": 64
},
{
"epoch": 6.5,
"grad_norm": 0.8562595202564899,
"learning_rate": 5.8682408883346535e-06,
"loss": 0.1591,
"step": 65
},
{
"epoch": 6.6,
"grad_norm": 0.7667412263031629,
"learning_rate": 5.735727354158581e-06,
"loss": 0.1463,
"step": 66
},
{
"epoch": 6.7,
"grad_norm": 0.758115793445322,
"learning_rate": 5.6026834012766155e-06,
"loss": 0.157,
"step": 67
},
{
"epoch": 6.8,
"grad_norm": 0.7474467056534099,
"learning_rate": 5.469204947015897e-06,
"loss": 0.1524,
"step": 68
},
{
"epoch": 6.9,
"grad_norm": 0.7093244081518412,
"learning_rate": 5.335388221955012e-06,
"loss": 0.1188,
"step": 69
},
{
"epoch": 7.0,
"grad_norm": 0.6945048553836589,
"learning_rate": 5.201329700547077e-06,
"loss": 0.1194,
"step": 70
},
{
"epoch": 7.1,
"grad_norm": 0.632963012825732,
"learning_rate": 5.067126031566988e-06,
"loss": 0.1055,
"step": 71
},
{
"epoch": 7.2,
"grad_norm": 0.535098574774597,
"learning_rate": 4.932873968433014e-06,
"loss": 0.0691,
"step": 72
},
{
"epoch": 7.3,
"grad_norm": 0.6563557436695596,
"learning_rate": 4.798670299452926e-06,
"loss": 0.1127,
"step": 73
},
{
"epoch": 7.4,
"grad_norm": 0.6600862624801698,
"learning_rate": 4.664611778044988e-06,
"loss": 0.1064,
"step": 74
},
{
"epoch": 7.5,
"grad_norm": 0.6077118042914474,
"learning_rate": 4.530795052984104e-06,
"loss": 0.0528,
"step": 75
},
{
"epoch": 7.6,
"grad_norm": 0.7605226736684684,
"learning_rate": 4.397316598723385e-06,
"loss": 0.1128,
"step": 76
},
{
"epoch": 7.7,
"grad_norm": 0.618480612781212,
"learning_rate": 4.264272645841419e-06,
"loss": 0.0804,
"step": 77
},
{
"epoch": 7.8,
"grad_norm": 0.6591734340885593,
"learning_rate": 4.131759111665349e-06,
"loss": 0.1166,
"step": 78
},
{
"epoch": 7.9,
"grad_norm": 0.706476292562821,
"learning_rate": 3.999871531119779e-06,
"loss": 0.0954,
"step": 79
},
{
"epoch": 8.0,
"grad_norm": 0.6511060587514005,
"learning_rate": 3.86870498785139e-06,
"loss": 0.0996,
"step": 80
},
{
"epoch": 8.1,
"grad_norm": 0.5877512031587164,
"learning_rate": 3.7383540456788915e-06,
"loss": 0.0643,
"step": 81
},
{
"epoch": 8.2,
"grad_norm": 0.5458773628267196,
"learning_rate": 3.6089126804177373e-06,
"loss": 0.0923,
"step": 82
},
{
"epoch": 8.3,
"grad_norm": 0.5072513687549686,
"learning_rate": 3.480474212128766e-06,
"loss": 0.0444,
"step": 83
},
{
"epoch": 8.4,
"grad_norm": 0.6566938015455519,
"learning_rate": 3.3531312378396026e-06,
"loss": 0.0742,
"step": 84
},
{
"epoch": 8.5,
"grad_norm": 0.5801962223341335,
"learning_rate": 3.226975564787322e-06,
"loss": 0.0588,
"step": 85
},
{
"epoch": 8.6,
"grad_norm": 0.6460146970130528,
"learning_rate": 3.1020981442305187e-06,
"loss": 0.0505,
"step": 86
},
{
"epoch": 8.7,
"grad_norm": 0.5597931785794817,
"learning_rate": 2.978589005878476e-06,
"loss": 0.0595,
"step": 87
},
{
"epoch": 8.8,
"grad_norm": 0.736772650278511,
"learning_rate": 2.8565371929847286e-06,
"loss": 0.0828,
"step": 88
},
{
"epoch": 8.9,
"grad_norm": 0.6093005863730493,
"learning_rate": 2.736030698151815e-06,
"loss": 0.0546,
"step": 89
},
{
"epoch": 9.0,
"grad_norm": 0.5542138970933609,
"learning_rate": 2.6171563998934605e-06,
"loss": 0.0663,
"step": 90
},
{
"epoch": 9.1,
"grad_norm": 0.5234184327196851,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.0448,
"step": 91
},
{
"epoch": 9.2,
"grad_norm": 0.5305920522906186,
"learning_rate": 2.384645961752113e-06,
"loss": 0.0459,
"step": 92
},
{
"epoch": 9.3,
"grad_norm": 0.4197505273523392,
"learning_rate": 2.2711774490274767e-06,
"loss": 0.0434,
"step": 93
},
{
"epoch": 9.4,
"grad_norm": 0.3934591059055063,
"learning_rate": 2.159676266344222e-06,
"loss": 0.0515,
"step": 94
},
{
"epoch": 9.5,
"grad_norm": 0.43218959356908915,
"learning_rate": 2.050222799884387e-06,
"loss": 0.0382,
"step": 95
},
{
"epoch": 9.6,
"grad_norm": 0.4622104651280975,
"learning_rate": 1.942895959539939e-06,
"loss": 0.0463,
"step": 96
},
{
"epoch": 9.7,
"grad_norm": 0.5535724362191721,
"learning_rate": 1.8377731220231144e-06,
"loss": 0.0447,
"step": 97
},
{
"epoch": 9.8,
"grad_norm": 0.4640259982574777,
"learning_rate": 1.7349300750820758e-06,
"loss": 0.0517,
"step": 98
},
{
"epoch": 9.9,
"grad_norm": 0.5266934894357415,
"learning_rate": 1.6344409628621482e-06,
"loss": 0.0585,
"step": 99
},
{
"epoch": 10.0,
"grad_norm": 0.4676281163285699,
"learning_rate": 1.5363782324520033e-06,
"loss": 0.0242,
"step": 100
},
{
"epoch": 10.1,
"grad_norm": 0.35967933721052214,
"learning_rate": 1.4408125816532981e-06,
"loss": 0.0357,
"step": 101
},
{
"epoch": 10.2,
"grad_norm": 0.27447636992783964,
"learning_rate": 1.347812908011485e-06,
"loss": 0.0214,
"step": 102
},
{
"epoch": 10.3,
"grad_norm": 0.3014381976097687,
"learning_rate": 1.257446259144494e-06,
"loss": 0.0267,
"step": 103
},
{
"epoch": 10.4,
"grad_norm": 0.38294713899905675,
"learning_rate": 1.1697777844051105e-06,
"loss": 0.0665,
"step": 104
},
{
"epoch": 10.5,
"grad_norm": 0.3684690274136007,
"learning_rate": 1.0848706879118893e-06,
"loss": 0.0258,
"step": 105
},
{
"epoch": 10.6,
"grad_norm": 0.31478908256541077,
"learning_rate": 1.0027861829824953e-06,
"loss": 0.0171,
"step": 106
},
{
"epoch": 10.7,
"grad_norm": 0.34994824968823507,
"learning_rate": 9.235834480022788e-07,
"loss": 0.0422,
"step": 107
},
{
"epoch": 10.8,
"grad_norm": 0.4015273337656147,
"learning_rate": 8.473195837599419e-07,
"loss": 0.0474,
"step": 108
},
{
"epoch": 10.9,
"grad_norm": 0.3512467702811788,
"learning_rate": 7.740495722810271e-07,
"loss": 0.0387,
"step": 109
},
{
"epoch": 11.0,
"grad_norm": 0.2819446830081912,
"learning_rate": 7.03826237188916e-07,
"loss": 0.0265,
"step": 110
},
{
"epoch": 11.1,
"grad_norm": 0.26544732968956203,
"learning_rate": 6.367002056219285e-07,
"loss": 0.0206,
"step": 111
},
{
"epoch": 11.2,
"grad_norm": 0.28547087305675023,
"learning_rate": 5.727198717339511e-07,
"loss": 0.0346,
"step": 112
},
{
"epoch": 11.3,
"grad_norm": 0.29019462565393106,
"learning_rate": 5.119313618049309e-07,
"loss": 0.0263,
"step": 113
},
{
"epoch": 11.4,
"grad_norm": 0.2587759144580898,
"learning_rate": 4.54378500986381e-07,
"loss": 0.0262,
"step": 114
},
{
"epoch": 11.5,
"grad_norm": 0.2964554848018806,
"learning_rate": 4.001027817058789e-07,
"loss": 0.045,
"step": 115
},
{
"epoch": 11.6,
"grad_norm": 0.286044305327069,
"learning_rate": 3.49143333753309e-07,
"loss": 0.0283,
"step": 116
},
{
"epoch": 11.7,
"grad_norm": 0.2675893029381175,
"learning_rate": 3.015368960704584e-07,
"loss": 0.0185,
"step": 117
},
{
"epoch": 11.8,
"grad_norm": 0.3300518406278238,
"learning_rate": 2.573177902642726e-07,
"loss": 0.0422,
"step": 118
},
{
"epoch": 11.9,
"grad_norm": 0.32684186571188295,
"learning_rate": 2.1651789586287442e-07,
"loss": 0.0444,
"step": 119
},
{
"epoch": 12.0,
"grad_norm": 0.23225636748688117,
"learning_rate": 1.7916662733218848e-07,
"loss": 0.0162,
"step": 120
},
{
"epoch": 12.1,
"grad_norm": 0.18555282856615898,
"learning_rate": 1.4529091286973994e-07,
"loss": 0.0111,
"step": 121
},
{
"epoch": 12.2,
"grad_norm": 0.2272995406967833,
"learning_rate": 1.1491517499091498e-07,
"loss": 0.0211,
"step": 122
},
{
"epoch": 12.3,
"grad_norm": 0.3170863977174301,
"learning_rate": 8.80613129216762e-08,
"loss": 0.0489,
"step": 123
},
{
"epoch": 12.4,
"grad_norm": 0.2869870238374455,
"learning_rate": 6.474868681043578e-08,
"loss": 0.0338,
"step": 124
},
{
"epoch": 12.5,
"grad_norm": 0.19605392462041799,
"learning_rate": 4.499410377045765e-08,
"loss": 0.0136,
"step": 125
},
{
"epoch": 12.6,
"grad_norm": 0.24074331934174073,
"learning_rate": 2.8811805762860578e-08,
"loss": 0.0236,
"step": 126
},
{
"epoch": 12.7,
"grad_norm": 0.2555618190181957,
"learning_rate": 1.6213459328950355e-08,
"loss": 0.0334,
"step": 127
},
{
"epoch": 12.8,
"grad_norm": 0.3138210130616997,
"learning_rate": 7.2081471792911914e-09,
"loss": 0.0412,
"step": 128
},
{
"epoch": 12.9,
"grad_norm": 0.26393331869796904,
"learning_rate": 1.8023616455731253e-09,
"loss": 0.0253,
"step": 129
},
{
"epoch": 13.0,
"grad_norm": 0.24454986517667737,
"learning_rate": 0.0,
"loss": 0.0263,
"step": 130
},
{
"epoch": 13.0,
"step": 130,
"total_flos": 3.2858174247665664e+16,
"train_loss": 0.2939408891905959,
"train_runtime": 1300.2962,
"train_samples_per_second": 3.159,
"train_steps_per_second": 0.1
}
],
"logging_steps": 1,
"max_steps": 130,
"num_input_tokens_seen": 0,
"num_train_epochs": 13,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.2858174247665664e+16,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}