sft-metamath-mistral-7b-vi-v1 / trainer_state.json
BaoLocTown's picture
Model save
a491beb
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.4480959833072509,
"eval_steps": 500,
"global_step": 1718,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 2.604166666666667e-07,
"loss": 0.7721,
"step": 1
},
{
"epoch": 0.01,
"learning_rate": 2.604166666666667e-06,
"loss": 0.7229,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 5.208333333333334e-06,
"loss": 0.6469,
"step": 20
},
{
"epoch": 0.02,
"learning_rate": 7.8125e-06,
"loss": 0.5906,
"step": 30
},
{
"epoch": 0.02,
"learning_rate": 1.0416666666666668e-05,
"loss": 0.5379,
"step": 40
},
{
"epoch": 0.03,
"learning_rate": 1.3020833333333334e-05,
"loss": 0.4474,
"step": 50
},
{
"epoch": 0.03,
"learning_rate": 1.5625e-05,
"loss": 0.4287,
"step": 60
},
{
"epoch": 0.04,
"learning_rate": 1.8229166666666668e-05,
"loss": 0.371,
"step": 70
},
{
"epoch": 0.04,
"learning_rate": 2.0833333333333336e-05,
"loss": 0.3793,
"step": 80
},
{
"epoch": 0.05,
"learning_rate": 2.34375e-05,
"loss": 0.3635,
"step": 90
},
{
"epoch": 0.05,
"learning_rate": 2.604166666666667e-05,
"loss": 0.3436,
"step": 100
},
{
"epoch": 0.06,
"learning_rate": 2.8645833333333333e-05,
"loss": 0.3837,
"step": 110
},
{
"epoch": 0.06,
"learning_rate": 3.125e-05,
"loss": 0.354,
"step": 120
},
{
"epoch": 0.07,
"learning_rate": 3.385416666666667e-05,
"loss": 0.3472,
"step": 130
},
{
"epoch": 0.07,
"learning_rate": 3.6458333333333336e-05,
"loss": 0.347,
"step": 140
},
{
"epoch": 0.08,
"learning_rate": 3.90625e-05,
"loss": 0.3444,
"step": 150
},
{
"epoch": 0.08,
"learning_rate": 4.166666666666667e-05,
"loss": 0.342,
"step": 160
},
{
"epoch": 0.09,
"learning_rate": 4.4270833333333337e-05,
"loss": 0.3392,
"step": 170
},
{
"epoch": 0.09,
"learning_rate": 4.6875e-05,
"loss": 0.3424,
"step": 180
},
{
"epoch": 0.1,
"learning_rate": 4.947916666666667e-05,
"loss": 0.3616,
"step": 190
},
{
"epoch": 0.1,
"learning_rate": 4.99994047380455e-05,
"loss": 0.3431,
"step": 200
},
{
"epoch": 0.11,
"learning_rate": 4.999698653493815e-05,
"loss": 0.3418,
"step": 210
},
{
"epoch": 0.11,
"learning_rate": 4.999270836660003e-05,
"loss": 0.3645,
"step": 220
},
{
"epoch": 0.12,
"learning_rate": 4.998657055135927e-05,
"loss": 0.3418,
"step": 230
},
{
"epoch": 0.13,
"learning_rate": 4.9978573545915854e-05,
"loss": 0.335,
"step": 240
},
{
"epoch": 0.13,
"learning_rate": 4.996871794530757e-05,
"loss": 0.3441,
"step": 250
},
{
"epoch": 0.14,
"learning_rate": 4.9957004482865796e-05,
"loss": 0.3463,
"step": 260
},
{
"epoch": 0.14,
"learning_rate": 4.994343403016093e-05,
"loss": 0.3284,
"step": 270
},
{
"epoch": 0.15,
"learning_rate": 4.992800759693746e-05,
"loss": 0.3232,
"step": 280
},
{
"epoch": 0.15,
"learning_rate": 4.9910726331038935e-05,
"loss": 0.3251,
"step": 290
},
{
"epoch": 0.16,
"learning_rate": 4.989159151832251e-05,
"loss": 0.3346,
"step": 300
},
{
"epoch": 0.16,
"learning_rate": 4.987060458256324e-05,
"loss": 0.3397,
"step": 310
},
{
"epoch": 0.17,
"learning_rate": 4.985013406298429e-05,
"loss": 0.3273,
"step": 320
},
{
"epoch": 0.17,
"learning_rate": 4.9825632509424134e-05,
"loss": 0.3227,
"step": 330
},
{
"epoch": 0.18,
"learning_rate": 4.979928374067127e-05,
"loss": 0.329,
"step": 340
},
{
"epoch": 0.18,
"learning_rate": 4.977108971727373e-05,
"loss": 0.3177,
"step": 350
},
{
"epoch": 0.19,
"learning_rate": 4.9741052537080565e-05,
"loss": 0.3428,
"step": 360
},
{
"epoch": 0.19,
"learning_rate": 4.970917443508558e-05,
"loss": 0.3077,
"step": 370
},
{
"epoch": 0.2,
"learning_rate": 4.967545778326114e-05,
"loss": 0.3248,
"step": 380
},
{
"epoch": 0.2,
"learning_rate": 4.963990509038167e-05,
"loss": 0.3073,
"step": 390
},
{
"epoch": 0.21,
"learning_rate": 4.9602519001836933e-05,
"loss": 0.3386,
"step": 400
},
{
"epoch": 0.21,
"learning_rate": 4.9563302299435246e-05,
"loss": 0.3272,
"step": 410
},
{
"epoch": 0.22,
"learning_rate": 4.952225790119644e-05,
"loss": 0.3355,
"step": 420
},
{
"epoch": 0.22,
"learning_rate": 4.947938886113482e-05,
"loss": 0.3166,
"step": 430
},
{
"epoch": 0.23,
"learning_rate": 4.943469836903181e-05,
"loss": 0.3289,
"step": 440
},
{
"epoch": 0.23,
"learning_rate": 4.938818975019871e-05,
"loss": 0.3369,
"step": 450
},
{
"epoch": 0.24,
"learning_rate": 4.933986646522921e-05,
"loss": 0.3161,
"step": 460
},
{
"epoch": 0.25,
"learning_rate": 4.928973210974192e-05,
"loss": 0.3296,
"step": 470
},
{
"epoch": 0.25,
"learning_rate": 4.9237790414112806e-05,
"loss": 0.2899,
"step": 480
},
{
"epoch": 0.26,
"learning_rate": 4.918404524319766e-05,
"loss": 0.3216,
"step": 490
},
{
"epoch": 0.26,
"learning_rate": 4.912850059604448e-05,
"loss": 0.3128,
"step": 500
},
{
"epoch": 0.26,
"eval_loss": 0.5092809200286865,
"eval_runtime": 107.6589,
"eval_samples_per_second": 7.914,
"eval_steps_per_second": 1.978,
"step": 500
},
{
"epoch": 0.27,
"learning_rate": 4.907116060559596e-05,
"loss": 0.3152,
"step": 510
},
{
"epoch": 0.27,
"learning_rate": 4.901202953838191e-05,
"loss": 0.3091,
"step": 520
},
{
"epoch": 0.28,
"learning_rate": 4.8951111794201845e-05,
"loss": 0.3238,
"step": 530
},
{
"epoch": 0.28,
"learning_rate": 4.8888411905797574e-05,
"loss": 0.3462,
"step": 540
},
{
"epoch": 0.29,
"learning_rate": 4.8823934538515946e-05,
"loss": 0.3253,
"step": 550
},
{
"epoch": 0.29,
"learning_rate": 4.875768448996173e-05,
"loss": 0.3321,
"step": 560
},
{
"epoch": 0.3,
"learning_rate": 4.868966668964057e-05,
"loss": 0.3204,
"step": 570
},
{
"epoch": 0.3,
"learning_rate": 4.8619886198592275e-05,
"loss": 0.3109,
"step": 580
},
{
"epoch": 0.31,
"learning_rate": 4.854834820901419e-05,
"loss": 0.3277,
"step": 590
},
{
"epoch": 0.31,
"learning_rate": 4.8475058043874875e-05,
"loss": 0.3081,
"step": 600
},
{
"epoch": 0.32,
"learning_rate": 4.840002115651802e-05,
"loss": 0.3314,
"step": 610
},
{
"epoch": 0.32,
"learning_rate": 4.832324313025669e-05,
"loss": 0.3164,
"step": 620
},
{
"epoch": 0.33,
"learning_rate": 4.824472967795788e-05,
"loss": 0.294,
"step": 630
},
{
"epoch": 0.33,
"learning_rate": 4.816448664161743e-05,
"loss": 0.3211,
"step": 640
},
{
"epoch": 0.34,
"learning_rate": 4.808251999192536e-05,
"loss": 0.3309,
"step": 650
},
{
"epoch": 0.34,
"learning_rate": 4.7998835827821536e-05,
"loss": 0.3374,
"step": 660
},
{
"epoch": 0.35,
"learning_rate": 4.7913440376041975e-05,
"loss": 0.3269,
"step": 670
},
{
"epoch": 0.35,
"learning_rate": 4.782633999065541e-05,
"loss": 0.3159,
"step": 680
},
{
"epoch": 0.36,
"learning_rate": 4.77375411525906e-05,
"loss": 0.3138,
"step": 690
},
{
"epoch": 0.37,
"learning_rate": 4.764705046915402e-05,
"loss": 0.3362,
"step": 700
},
{
"epoch": 0.37,
"learning_rate": 4.755487467353829e-05,
"loss": 0.3315,
"step": 710
},
{
"epoch": 0.38,
"learning_rate": 4.7461020624321104e-05,
"loss": 0.3344,
"step": 720
},
{
"epoch": 0.38,
"learning_rate": 4.7365495304955e-05,
"loss": 0.3148,
"step": 730
},
{
"epoch": 0.39,
"learning_rate": 4.7268305823247635e-05,
"loss": 0.3054,
"step": 740
},
{
"epoch": 0.39,
"learning_rate": 4.7169459410832986e-05,
"loss": 0.3213,
"step": 750
},
{
"epoch": 0.4,
"learning_rate": 4.706896342263319e-05,
"loss": 0.3165,
"step": 760
},
{
"epoch": 0.4,
"learning_rate": 4.6966825336311376e-05,
"loss": 0.3516,
"step": 770
},
{
"epoch": 0.41,
"learning_rate": 4.6863052751715175e-05,
"loss": 0.3083,
"step": 780
},
{
"epoch": 0.41,
"learning_rate": 4.67576533903113e-05,
"loss": 0.3044,
"step": 790
},
{
"epoch": 0.42,
"learning_rate": 4.665063509461097e-05,
"loss": 0.3207,
"step": 800
},
{
"epoch": 0.42,
"learning_rate": 4.65420058275864e-05,
"loss": 0.3172,
"step": 810
},
{
"epoch": 0.43,
"learning_rate": 4.643177367207827e-05,
"loss": 0.3177,
"step": 820
},
{
"epoch": 0.43,
"learning_rate": 4.6319946830194314e-05,
"loss": 0.3072,
"step": 830
},
{
"epoch": 0.44,
"learning_rate": 4.620653362269902e-05,
"loss": 0.3115,
"step": 840
},
{
"epoch": 0.44,
"learning_rate": 4.609154248839449e-05,
"loss": 0.2946,
"step": 850
},
{
"epoch": 1.0,
"learning_rate": 4.597498198349254e-05,
"loss": 0.2979,
"step": 860
},
{
"epoch": 1.01,
"learning_rate": 4.585686078097806e-05,
"loss": 0.2866,
"step": 870
},
{
"epoch": 1.01,
"learning_rate": 4.5737187669963675e-05,
"loss": 0.2918,
"step": 880
},
{
"epoch": 1.02,
"learning_rate": 4.561597155503574e-05,
"loss": 0.2801,
"step": 890
},
{
"epoch": 1.02,
"learning_rate": 4.549322145559181e-05,
"loss": 0.2839,
"step": 900
},
{
"epoch": 1.03,
"learning_rate": 4.536894650516952e-05,
"loss": 0.2815,
"step": 910
},
{
"epoch": 1.03,
"learning_rate": 4.524315595076695e-05,
"loss": 0.2825,
"step": 920
},
{
"epoch": 1.04,
"learning_rate": 4.511585915215462e-05,
"loss": 0.2788,
"step": 930
},
{
"epoch": 1.04,
"learning_rate": 4.498706558117903e-05,
"loss": 0.3051,
"step": 940
},
{
"epoch": 1.05,
"learning_rate": 4.485678482105789e-05,
"loss": 0.2595,
"step": 950
},
{
"epoch": 1.05,
"learning_rate": 4.4725026565667055e-05,
"loss": 0.2741,
"step": 960
},
{
"epoch": 1.06,
"learning_rate": 4.45918006188192e-05,
"loss": 0.2862,
"step": 970
},
{
"epoch": 1.06,
"learning_rate": 4.44571168935344e-05,
"loss": 0.2712,
"step": 980
},
{
"epoch": 1.07,
"learning_rate": 4.432098541130247e-05,
"loss": 0.2935,
"step": 990
},
{
"epoch": 1.07,
"learning_rate": 4.418341630133733e-05,
"loss": 0.2751,
"step": 1000
},
{
"epoch": 1.07,
"eval_loss": 0.4884406328201294,
"eval_runtime": 107.9712,
"eval_samples_per_second": 7.891,
"eval_steps_per_second": 1.973,
"step": 1000
},
{
"epoch": 1.08,
"learning_rate": 4.404441979982329e-05,
"loss": 0.2706,
"step": 1010
},
{
"epoch": 1.08,
"learning_rate": 4.3904006249153395e-05,
"loss": 0.2509,
"step": 1020
},
{
"epoch": 1.09,
"learning_rate": 4.376218609715992e-05,
"loss": 0.2774,
"step": 1030
},
{
"epoch": 1.09,
"learning_rate": 4.36189698963369e-05,
"loss": 0.2675,
"step": 1040
},
{
"epoch": 1.1,
"learning_rate": 4.3474368303055e-05,
"loss": 0.2946,
"step": 1050
},
{
"epoch": 1.1,
"learning_rate": 4.3328392076768597e-05,
"loss": 0.275,
"step": 1060
},
{
"epoch": 1.11,
"learning_rate": 4.318105207921515e-05,
"loss": 0.2635,
"step": 1070
},
{
"epoch": 1.12,
"learning_rate": 4.303235927360706e-05,
"loss": 0.2917,
"step": 1080
},
{
"epoch": 1.12,
"learning_rate": 4.288232472381589e-05,
"loss": 0.275,
"step": 1090
},
{
"epoch": 1.13,
"learning_rate": 4.273095959354914e-05,
"loss": 0.2551,
"step": 1100
},
{
"epoch": 1.13,
"learning_rate": 4.257827514551957e-05,
"loss": 0.2716,
"step": 1110
},
{
"epoch": 1.14,
"learning_rate": 4.2424282740607205e-05,
"loss": 0.2693,
"step": 1120
},
{
"epoch": 1.14,
"learning_rate": 4.2268993837013935e-05,
"loss": 0.2862,
"step": 1130
},
{
"epoch": 1.15,
"learning_rate": 4.2112419989411026e-05,
"loss": 0.2869,
"step": 1140
},
{
"epoch": 1.15,
"learning_rate": 4.195457284807927e-05,
"loss": 0.2698,
"step": 1150
},
{
"epoch": 1.16,
"learning_rate": 4.179546415804221e-05,
"loss": 0.275,
"step": 1160
},
{
"epoch": 1.16,
"learning_rate": 4.163510575819212e-05,
"loss": 0.2569,
"step": 1170
},
{
"epoch": 1.17,
"learning_rate": 4.1473509580409174e-05,
"loss": 0.249,
"step": 1180
},
{
"epoch": 1.17,
"learning_rate": 4.131068764867363e-05,
"loss": 0.283,
"step": 1190
},
{
"epoch": 1.18,
"learning_rate": 4.11466520781711e-05,
"loss": 0.2735,
"step": 1200
},
{
"epoch": 1.18,
"learning_rate": 4.0981415074391124e-05,
"loss": 0.2677,
"step": 1210
},
{
"epoch": 1.19,
"learning_rate": 4.081498893221899e-05,
"loss": 0.2928,
"step": 1220
},
{
"epoch": 1.19,
"learning_rate": 4.0647386035020884e-05,
"loss": 0.2819,
"step": 1230
},
{
"epoch": 1.2,
"learning_rate": 4.047861885372249e-05,
"loss": 0.2644,
"step": 1240
},
{
"epoch": 1.2,
"learning_rate": 4.0308699945881055e-05,
"loss": 0.2757,
"step": 1250
},
{
"epoch": 1.21,
"learning_rate": 4.013764195475101e-05,
"loss": 0.2782,
"step": 1260
},
{
"epoch": 1.21,
"learning_rate": 3.9965457608343194e-05,
"loss": 0.279,
"step": 1270
},
{
"epoch": 1.22,
"learning_rate": 3.979215971847783e-05,
"loss": 0.2726,
"step": 1280
},
{
"epoch": 1.22,
"learning_rate": 3.961776117983123e-05,
"loss": 0.2542,
"step": 1290
},
{
"epoch": 1.23,
"learning_rate": 3.944227496897629e-05,
"loss": 0.263,
"step": 1300
},
{
"epoch": 1.24,
"learning_rate": 3.9265714143416967e-05,
"loss": 0.2762,
"step": 1310
},
{
"epoch": 1.24,
"learning_rate": 3.9088091840616705e-05,
"loss": 0.2825,
"step": 1320
},
{
"epoch": 1.25,
"learning_rate": 3.890942127702089e-05,
"loss": 0.2862,
"step": 1330
},
{
"epoch": 1.25,
"learning_rate": 3.872971574707345e-05,
"loss": 0.2598,
"step": 1340
},
{
"epoch": 1.26,
"learning_rate": 3.8548988622227666e-05,
"loss": 0.2636,
"step": 1350
},
{
"epoch": 1.26,
"learning_rate": 3.836725334995123e-05,
"loss": 0.258,
"step": 1360
},
{
"epoch": 1.27,
"learning_rate": 3.81845234527256e-05,
"loss": 0.2562,
"step": 1370
},
{
"epoch": 1.27,
"learning_rate": 3.800081252703993e-05,
"loss": 0.2711,
"step": 1380
},
{
"epoch": 1.28,
"learning_rate": 3.781613424237926e-05,
"loss": 0.249,
"step": 1390
},
{
"epoch": 1.28,
"learning_rate": 3.763050234020752e-05,
"loss": 0.2386,
"step": 1400
},
{
"epoch": 1.29,
"learning_rate": 3.744393063294499e-05,
"loss": 0.2517,
"step": 1410
},
{
"epoch": 1.29,
"learning_rate": 3.725643300294056e-05,
"loss": 0.2762,
"step": 1420
},
{
"epoch": 1.3,
"learning_rate": 3.706802340143881e-05,
"loss": 0.2344,
"step": 1430
},
{
"epoch": 1.3,
"learning_rate": 3.68787158475419e-05,
"loss": 0.2555,
"step": 1440
},
{
"epoch": 1.31,
"learning_rate": 3.668852442716645e-05,
"loss": 0.2632,
"step": 1450
},
{
"epoch": 1.31,
"learning_rate": 3.649746329199545e-05,
"loss": 0.2422,
"step": 1460
},
{
"epoch": 1.32,
"learning_rate": 3.630554665842525e-05,
"loss": 0.2545,
"step": 1470
},
{
"epoch": 1.32,
"learning_rate": 3.611278880650779e-05,
"loss": 0.2518,
"step": 1480
},
{
"epoch": 1.33,
"learning_rate": 3.5919204078888004e-05,
"loss": 0.2467,
"step": 1490
},
{
"epoch": 1.33,
"learning_rate": 3.572480687973665e-05,
"loss": 0.2585,
"step": 1500
},
{
"epoch": 1.33,
"eval_loss": 0.49433010816574097,
"eval_runtime": 107.5959,
"eval_samples_per_second": 7.919,
"eval_steps_per_second": 1.98,
"step": 1500
},
{
"epoch": 1.34,
"learning_rate": 3.5529611673678543e-05,
"loss": 0.2635,
"step": 1510
},
{
"epoch": 1.34,
"learning_rate": 3.5333632984716226e-05,
"loss": 0.2351,
"step": 1520
},
{
"epoch": 1.35,
"learning_rate": 3.5136885395149345e-05,
"loss": 0.2663,
"step": 1530
},
{
"epoch": 1.36,
"learning_rate": 3.493938354448954e-05,
"loss": 0.2527,
"step": 1540
},
{
"epoch": 1.36,
"learning_rate": 3.474114212837123e-05,
"loss": 0.2453,
"step": 1550
},
{
"epoch": 1.37,
"learning_rate": 3.454217589745809e-05,
"loss": 0.2462,
"step": 1560
},
{
"epoch": 1.37,
"learning_rate": 3.43424996563455e-05,
"loss": 0.2612,
"step": 1570
},
{
"epoch": 1.38,
"learning_rate": 3.414212826245898e-05,
"loss": 0.2333,
"step": 1580
},
{
"epoch": 1.38,
"learning_rate": 3.394107662494872e-05,
"loss": 0.257,
"step": 1590
},
{
"epoch": 1.39,
"learning_rate": 3.3739359703580144e-05,
"loss": 0.249,
"step": 1600
},
{
"epoch": 1.39,
"learning_rate": 3.3536992507620854e-05,
"loss": 0.2542,
"step": 1610
},
{
"epoch": 1.4,
"learning_rate": 3.3333990094723826e-05,
"loss": 0.2789,
"step": 1620
},
{
"epoch": 1.4,
"learning_rate": 3.3130367569806965e-05,
"loss": 0.2743,
"step": 1630
},
{
"epoch": 1.41,
"learning_rate": 3.292614008392923e-05,
"loss": 0.2592,
"step": 1640
},
{
"epoch": 1.41,
"learning_rate": 3.272132283316324e-05,
"loss": 0.2371,
"step": 1650
},
{
"epoch": 1.42,
"learning_rate": 3.25159310574646e-05,
"loss": 0.2574,
"step": 1660
},
{
"epoch": 1.42,
"learning_rate": 3.230998003953793e-05,
"loss": 0.2278,
"step": 1670
},
{
"epoch": 1.43,
"learning_rate": 3.210348510369972e-05,
"loss": 0.2496,
"step": 1680
},
{
"epoch": 1.43,
"learning_rate": 3.1896461614738026e-05,
"loss": 0.2349,
"step": 1690
},
{
"epoch": 1.44,
"learning_rate": 3.1688924976769324e-05,
"loss": 0.2339,
"step": 1700
},
{
"epoch": 1.44,
"learning_rate": 3.1480890632092236e-05,
"loss": 0.249,
"step": 1710
},
{
"epoch": 1.45,
"step": 1718,
"total_flos": 3.091619088653353e+17,
"train_loss": 0.30544669220139453,
"train_runtime": 6733.7325,
"train_samples_per_second": 2.277,
"train_steps_per_second": 0.569
}
],
"logging_steps": 10,
"max_steps": 3834,
"num_train_epochs": 2,
"save_steps": 500,
"total_flos": 3.091619088653353e+17,
"trial_name": null,
"trial_params": null
}