sft-mistral-7b-v01-v1 / trainer_state.json
BaoLocTown's picture
Model save
30bce3d
raw
history blame
21.9 kB
{
"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": 0.7574,
"step": 1
},
{
"epoch": 0.01,
"learning_rate": 2.604166666666667e-06,
"loss": 0.7145,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 5.208333333333334e-06,
"loss": 0.6428,
"step": 20
},
{
"epoch": 0.02,
"learning_rate": 7.8125e-06,
"loss": 0.5893,
"step": 30
},
{
"epoch": 0.02,
"learning_rate": 1.0416666666666668e-05,
"loss": 0.537,
"step": 40
},
{
"epoch": 0.03,
"learning_rate": 1.3020833333333334e-05,
"loss": 0.4432,
"step": 50
},
{
"epoch": 0.03,
"learning_rate": 1.5625e-05,
"loss": 0.4239,
"step": 60
},
{
"epoch": 0.04,
"learning_rate": 1.8229166666666668e-05,
"loss": 0.3672,
"step": 70
},
{
"epoch": 0.04,
"learning_rate": 2.0833333333333336e-05,
"loss": 0.3798,
"step": 80
},
{
"epoch": 0.05,
"learning_rate": 2.34375e-05,
"loss": 0.3666,
"step": 90
},
{
"epoch": 0.05,
"learning_rate": 2.604166666666667e-05,
"loss": 0.341,
"step": 100
},
{
"epoch": 0.06,
"learning_rate": 2.8645833333333333e-05,
"loss": 0.3855,
"step": 110
},
{
"epoch": 0.06,
"learning_rate": 3.125e-05,
"loss": 0.3517,
"step": 120
},
{
"epoch": 0.07,
"learning_rate": 3.385416666666667e-05,
"loss": 0.3482,
"step": 130
},
{
"epoch": 0.07,
"learning_rate": 3.6458333333333336e-05,
"loss": 0.3465,
"step": 140
},
{
"epoch": 0.08,
"learning_rate": 3.90625e-05,
"loss": 0.3455,
"step": 150
},
{
"epoch": 0.08,
"learning_rate": 4.166666666666667e-05,
"loss": 0.3418,
"step": 160
},
{
"epoch": 0.09,
"learning_rate": 4.4270833333333337e-05,
"loss": 0.338,
"step": 170
},
{
"epoch": 0.09,
"learning_rate": 4.6875e-05,
"loss": 0.3419,
"step": 180
},
{
"epoch": 0.1,
"learning_rate": 4.947916666666667e-05,
"loss": 0.3593,
"step": 190
},
{
"epoch": 0.1,
"learning_rate": 4.99994047380455e-05,
"loss": 0.3421,
"step": 200
},
{
"epoch": 0.11,
"learning_rate": 4.999698653493815e-05,
"loss": 0.3433,
"step": 210
},
{
"epoch": 0.11,
"learning_rate": 4.999270836660003e-05,
"loss": 0.3644,
"step": 220
},
{
"epoch": 0.12,
"learning_rate": 4.998657055135927e-05,
"loss": 0.3415,
"step": 230
},
{
"epoch": 0.13,
"learning_rate": 4.9978573545915854e-05,
"loss": 0.337,
"step": 240
},
{
"epoch": 0.13,
"learning_rate": 4.996871794530757e-05,
"loss": 0.3445,
"step": 250
},
{
"epoch": 0.14,
"learning_rate": 4.9957004482865796e-05,
"loss": 0.3434,
"step": 260
},
{
"epoch": 0.14,
"learning_rate": 4.994343403016093e-05,
"loss": 0.3308,
"step": 270
},
{
"epoch": 0.15,
"learning_rate": 4.992800759693746e-05,
"loss": 0.3223,
"step": 280
},
{
"epoch": 0.15,
"learning_rate": 4.9910726331038935e-05,
"loss": 0.3283,
"step": 290
},
{
"epoch": 0.16,
"learning_rate": 4.989159151832251e-05,
"loss": 0.3344,
"step": 300
},
{
"epoch": 0.16,
"learning_rate": 4.987060458256324e-05,
"loss": 0.3395,
"step": 310
},
{
"epoch": 0.17,
"learning_rate": 4.98477670853482e-05,
"loss": 0.3255,
"step": 320
},
{
"epoch": 0.17,
"learning_rate": 4.982308072596025e-05,
"loss": 0.3075,
"step": 330
},
{
"epoch": 0.18,
"learning_rate": 4.979654734125161e-05,
"loss": 0.3294,
"step": 340
},
{
"epoch": 0.18,
"learning_rate": 4.976816890550717e-05,
"loss": 0.3196,
"step": 350
},
{
"epoch": 0.19,
"learning_rate": 4.9737947530297606e-05,
"loss": 0.3436,
"step": 360
},
{
"epoch": 0.19,
"learning_rate": 4.9705885464322266e-05,
"loss": 0.3114,
"step": 370
},
{
"epoch": 0.2,
"learning_rate": 4.967198509324183e-05,
"loss": 0.3257,
"step": 380
},
{
"epoch": 0.2,
"learning_rate": 4.9636248939500805e-05,
"loss": 0.3062,
"step": 390
},
{
"epoch": 0.21,
"learning_rate": 4.959867966213986e-05,
"loss": 0.3381,
"step": 400
},
{
"epoch": 0.21,
"learning_rate": 4.955928005659792e-05,
"loss": 0.3234,
"step": 410
},
{
"epoch": 0.22,
"learning_rate": 4.951805305450422e-05,
"loss": 0.3345,
"step": 420
},
{
"epoch": 0.22,
"learning_rate": 4.947500172346016e-05,
"loss": 0.317,
"step": 430
},
{
"epoch": 0.23,
"learning_rate": 4.9430129266811e-05,
"loss": 0.3306,
"step": 440
},
{
"epoch": 0.23,
"learning_rate": 4.9383439023407585e-05,
"loss": 0.3393,
"step": 450
},
{
"epoch": 0.24,
"learning_rate": 4.933493446735784e-05,
"loss": 0.3143,
"step": 460
},
{
"epoch": 0.25,
"learning_rate": 4.928461920776832e-05,
"loss": 0.3322,
"step": 470
},
{
"epoch": 0.25,
"learning_rate": 4.923249698847564e-05,
"loss": 0.2923,
"step": 480
},
{
"epoch": 0.26,
"learning_rate": 4.91785716877679e-05,
"loss": 0.3197,
"step": 490
},
{
"epoch": 0.26,
"learning_rate": 4.912284731809614e-05,
"loss": 0.313,
"step": 500
},
{
"epoch": 0.26,
"eval_loss": 0.5063315629959106,
"eval_runtime": 106.6235,
"eval_samples_per_second": 7.991,
"eval_steps_per_second": 1.998,
"step": 500
},
{
"epoch": 0.27,
"learning_rate": 4.906532802577575e-05,
"loss": 0.3151,
"step": 510
},
{
"epoch": 0.27,
"learning_rate": 4.9006018090677974e-05,
"loss": 0.3121,
"step": 520
},
{
"epoch": 0.28,
"learning_rate": 4.8944921925911444e-05,
"loss": 0.3254,
"step": 530
},
{
"epoch": 0.28,
"learning_rate": 4.888204407749382e-05,
"loss": 0.3487,
"step": 540
},
{
"epoch": 0.29,
"learning_rate": 4.881738922401353e-05,
"loss": 0.3261,
"step": 550
},
{
"epoch": 0.29,
"learning_rate": 4.8750962176281635e-05,
"loss": 0.334,
"step": 560
},
{
"epoch": 0.3,
"learning_rate": 4.868276787697389e-05,
"loss": 0.3191,
"step": 570
},
{
"epoch": 0.3,
"learning_rate": 4.861281140026296e-05,
"loss": 0.3118,
"step": 580
},
{
"epoch": 0.31,
"learning_rate": 4.854109795144084e-05,
"loss": 0.3249,
"step": 590
},
{
"epoch": 0.31,
"learning_rate": 4.8467632866531596e-05,
"loss": 0.3095,
"step": 600
},
{
"epoch": 0.32,
"learning_rate": 4.8392421611894275e-05,
"loss": 0.3334,
"step": 610
},
{
"epoch": 0.32,
"learning_rate": 4.83154697838162e-05,
"loss": 0.3193,
"step": 620
},
{
"epoch": 0.33,
"learning_rate": 4.8236783108096514e-05,
"loss": 0.297,
"step": 630
},
{
"epoch": 0.33,
"learning_rate": 4.815636743962022e-05,
"loss": 0.3232,
"step": 640
},
{
"epoch": 0.34,
"learning_rate": 4.8074228761922444e-05,
"loss": 0.3348,
"step": 650
},
{
"epoch": 0.34,
"learning_rate": 4.799037318674329e-05,
"loss": 0.3401,
"step": 660
},
{
"epoch": 0.35,
"learning_rate": 4.7904806953573e-05,
"loss": 0.3299,
"step": 670
},
{
"epoch": 0.35,
"learning_rate": 4.781753642918778e-05,
"loss": 0.3152,
"step": 680
},
{
"epoch": 0.36,
"learning_rate": 4.772856810717599e-05,
"loss": 0.3104,
"step": 690
},
{
"epoch": 0.37,
"learning_rate": 4.763790860745502e-05,
"loss": 0.3331,
"step": 700
},
{
"epoch": 0.37,
"learning_rate": 4.7545564675778677e-05,
"loss": 0.332,
"step": 710
},
{
"epoch": 0.38,
"learning_rate": 4.745154318323528e-05,
"loss": 0.3318,
"step": 720
},
{
"epoch": 0.38,
"learning_rate": 4.7355851125736394e-05,
"loss": 0.3166,
"step": 730
},
{
"epoch": 0.39,
"learning_rate": 4.725849562349627e-05,
"loss": 0.3055,
"step": 740
},
{
"epoch": 0.39,
"learning_rate": 4.715948392050206e-05,
"loss": 0.3213,
"step": 750
},
{
"epoch": 0.4,
"learning_rate": 4.7058823383974794e-05,
"loss": 0.315,
"step": 760
},
{
"epoch": 0.4,
"learning_rate": 4.6956521503821215e-05,
"loss": 0.3525,
"step": 770
},
{
"epoch": 0.41,
"learning_rate": 4.685258589207648e-05,
"loss": 0.3089,
"step": 780
},
{
"epoch": 0.41,
"learning_rate": 4.6747024282337735e-05,
"loss": 0.3047,
"step": 790
},
{
"epoch": 0.42,
"learning_rate": 4.6639844529188734e-05,
"loss": 0.3192,
"step": 800
},
{
"epoch": 0.42,
"learning_rate": 4.653105460761533e-05,
"loss": 0.319,
"step": 810
},
{
"epoch": 0.43,
"learning_rate": 4.642066261241212e-05,
"loss": 0.3179,
"step": 820
},
{
"epoch": 0.43,
"learning_rate": 4.630867675758013e-05,
"loss": 0.3061,
"step": 830
},
{
"epoch": 0.44,
"learning_rate": 4.61951053757156e-05,
"loss": 0.3123,
"step": 840
},
{
"epoch": 0.44,
"learning_rate": 4.6079956917390024e-05,
"loss": 0.2952,
"step": 850
},
{
"epoch": 1.0,
"learning_rate": 4.5963239950521306e-05,
"loss": 0.2996,
"step": 860
},
{
"epoch": 1.01,
"learning_rate": 4.5844963159736303e-05,
"loss": 0.285,
"step": 870
},
{
"epoch": 1.01,
"learning_rate": 4.572513534572457e-05,
"loss": 0.2925,
"step": 880
},
{
"epoch": 1.02,
"learning_rate": 4.560376542458354e-05,
"loss": 0.2836,
"step": 890
},
{
"epoch": 1.02,
"learning_rate": 4.5480862427155134e-05,
"loss": 0.2824,
"step": 900
},
{
"epoch": 1.03,
"learning_rate": 4.535643549835373e-05,
"loss": 0.2814,
"step": 910
},
{
"epoch": 1.03,
"learning_rate": 4.5230493896485774e-05,
"loss": 0.2838,
"step": 920
},
{
"epoch": 1.04,
"learning_rate": 4.5103046992560855e-05,
"loss": 0.2802,
"step": 930
},
{
"epoch": 1.04,
"learning_rate": 4.4974104269594444e-05,
"loss": 0.3026,
"step": 940
},
{
"epoch": 1.05,
"learning_rate": 4.484367532190228e-05,
"loss": 0.2613,
"step": 950
},
{
"epoch": 1.05,
"learning_rate": 4.47117698543865e-05,
"loss": 0.276,
"step": 960
},
{
"epoch": 1.06,
"learning_rate": 4.457839768181349e-05,
"loss": 0.2865,
"step": 970
},
{
"epoch": 1.06,
"learning_rate": 4.444356872808362e-05,
"loss": 0.2705,
"step": 980
},
{
"epoch": 1.07,
"learning_rate": 4.43072930254928e-05,
"loss": 0.2942,
"step": 990
},
{
"epoch": 1.07,
"learning_rate": 4.4169580713986037e-05,
"loss": 0.279,
"step": 1000
},
{
"epoch": 1.07,
"eval_loss": 0.48916903138160706,
"eval_runtime": 106.511,
"eval_samples_per_second": 7.999,
"eval_steps_per_second": 2.0,
"step": 1000
},
{
"epoch": 1.08,
"learning_rate": 4.4030442040402915e-05,
"loss": 0.2714,
"step": 1010
},
{
"epoch": 1.08,
"learning_rate": 4.388988735771518e-05,
"loss": 0.2521,
"step": 1020
},
{
"epoch": 1.09,
"learning_rate": 4.374792712425637e-05,
"loss": 0.2797,
"step": 1030
},
{
"epoch": 1.09,
"learning_rate": 4.360457190294366e-05,
"loss": 0.2666,
"step": 1040
},
{
"epoch": 1.1,
"learning_rate": 4.345983236049189e-05,
"loss": 0.2916,
"step": 1050
},
{
"epoch": 1.1,
"learning_rate": 4.33137192666199e-05,
"loss": 0.2758,
"step": 1060
},
{
"epoch": 1.11,
"learning_rate": 4.3166243493249136e-05,
"loss": 0.2632,
"step": 1070
},
{
"epoch": 1.12,
"learning_rate": 4.301741601369475e-05,
"loss": 0.2913,
"step": 1080
},
{
"epoch": 1.12,
"learning_rate": 4.286724790184906e-05,
"loss": 0.2753,
"step": 1090
},
{
"epoch": 1.13,
"learning_rate": 4.27157503313576e-05,
"loss": 0.2565,
"step": 1100
},
{
"epoch": 1.13,
"learning_rate": 4.256293457478769e-05,
"loss": 0.2745,
"step": 1110
},
{
"epoch": 1.14,
"learning_rate": 4.24088120027897e-05,
"loss": 0.2663,
"step": 1120
},
{
"epoch": 1.14,
"learning_rate": 4.2253394083250946e-05,
"loss": 0.2836,
"step": 1130
},
{
"epoch": 1.15,
"learning_rate": 4.209669238044245e-05,
"loss": 0.2894,
"step": 1140
},
{
"epoch": 1.15,
"learning_rate": 4.19387185541584e-05,
"loss": 0.2705,
"step": 1150
},
{
"epoch": 1.16,
"learning_rate": 4.1779484358848644e-05,
"loss": 0.275,
"step": 1160
},
{
"epoch": 1.16,
"learning_rate": 4.161900164274403e-05,
"loss": 0.2566,
"step": 1170
},
{
"epoch": 1.17,
"learning_rate": 4.145728234697479e-05,
"loss": 0.2475,
"step": 1180
},
{
"epoch": 1.17,
"learning_rate": 4.12943385046821e-05,
"loss": 0.2853,
"step": 1190
},
{
"epoch": 1.18,
"learning_rate": 4.113018224012262e-05,
"loss": 0.2749,
"step": 1200
},
{
"epoch": 1.18,
"learning_rate": 4.0964825767766465e-05,
"loss": 0.2693,
"step": 1210
},
{
"epoch": 1.19,
"learning_rate": 4.079828139138827e-05,
"loss": 0.2938,
"step": 1220
},
{
"epoch": 1.19,
"learning_rate": 4.063056150315177e-05,
"loss": 0.2793,
"step": 1230
},
{
"epoch": 1.2,
"learning_rate": 4.046167858268766e-05,
"loss": 0.2647,
"step": 1240
},
{
"epoch": 1.2,
"learning_rate": 4.0291645196165026e-05,
"loss": 0.2764,
"step": 1250
},
{
"epoch": 1.21,
"learning_rate": 4.012047399535642e-05,
"loss": 0.2802,
"step": 1260
},
{
"epoch": 1.21,
"learning_rate": 3.9948177716696324e-05,
"loss": 0.278,
"step": 1270
},
{
"epoch": 1.22,
"learning_rate": 3.977476918033357e-05,
"loss": 0.275,
"step": 1280
},
{
"epoch": 1.22,
"learning_rate": 3.960026128917741e-05,
"loss": 0.2548,
"step": 1290
},
{
"epoch": 1.23,
"learning_rate": 3.942466702793738e-05,
"loss": 0.2629,
"step": 1300
},
{
"epoch": 1.24,
"learning_rate": 3.924799946215723e-05,
"loss": 0.2776,
"step": 1310
},
{
"epoch": 1.24,
"learning_rate": 3.9070271737242656e-05,
"loss": 0.2868,
"step": 1320
},
{
"epoch": 1.25,
"learning_rate": 3.889149707748327e-05,
"loss": 0.2874,
"step": 1330
},
{
"epoch": 1.25,
"learning_rate": 3.8711688785068555e-05,
"loss": 0.2624,
"step": 1340
},
{
"epoch": 1.26,
"learning_rate": 3.853086023909808e-05,
"loss": 0.2674,
"step": 1350
},
{
"epoch": 1.26,
"learning_rate": 3.8349024894586014e-05,
"loss": 0.2613,
"step": 1360
},
{
"epoch": 1.27,
"learning_rate": 3.816619628146e-05,
"loss": 0.2559,
"step": 1370
},
{
"epoch": 1.27,
"learning_rate": 3.798238800355436e-05,
"loss": 0.2713,
"step": 1380
},
{
"epoch": 1.28,
"learning_rate": 3.77976137375979e-05,
"loss": 0.2493,
"step": 1390
},
{
"epoch": 1.28,
"learning_rate": 3.761188723219628e-05,
"loss": 0.2425,
"step": 1400
},
{
"epoch": 1.29,
"learning_rate": 3.742522230680896e-05,
"loss": 0.2541,
"step": 1410
},
{
"epoch": 1.29,
"learning_rate": 3.7237632850720995e-05,
"loss": 0.278,
"step": 1420
},
{
"epoch": 1.3,
"learning_rate": 3.70491328220095e-05,
"loss": 0.2362,
"step": 1430
},
{
"epoch": 1.3,
"learning_rate": 3.685973624650511e-05,
"loss": 0.255,
"step": 1440
},
{
"epoch": 1.31,
"learning_rate": 3.6669457216748326e-05,
"loss": 0.2668,
"step": 1450
},
{
"epoch": 1.31,
"learning_rate": 3.647830989094095e-05,
"loss": 0.2416,
"step": 1460
},
{
"epoch": 1.32,
"learning_rate": 3.6286308491892565e-05,
"loss": 0.2492,
"step": 1470
},
{
"epoch": 1.32,
"learning_rate": 3.609346730596229e-05,
"loss": 0.2546,
"step": 1480
},
{
"epoch": 1.33,
"learning_rate": 3.589980068199575e-05,
"loss": 0.2502,
"step": 1490
},
{
"epoch": 1.33,
"learning_rate": 3.570532303025742e-05,
"loss": 0.2584,
"step": 1500
},
{
"epoch": 1.33,
"eval_loss": 0.4950372576713562,
"eval_runtime": 106.6983,
"eval_samples_per_second": 7.985,
"eval_steps_per_second": 1.996,
"step": 1500
},
{
"epoch": 1.34,
"learning_rate": 3.551004882135839e-05,
"loss": 0.2638,
"step": 1510
},
{
"epoch": 1.34,
"learning_rate": 3.531399258517962e-05,
"loss": 0.2383,
"step": 1520
},
{
"epoch": 1.35,
"learning_rate": 3.511716890979084e-05,
"loss": 0.2677,
"step": 1530
},
{
"epoch": 1.36,
"learning_rate": 3.4919592440365085e-05,
"loss": 0.255,
"step": 1540
},
{
"epoch": 1.36,
"learning_rate": 3.472127787808893e-05,
"loss": 0.2474,
"step": 1550
},
{
"epoch": 1.37,
"learning_rate": 3.452223997906871e-05,
"loss": 0.2474,
"step": 1560
},
{
"epoch": 1.37,
"learning_rate": 3.432249355323244e-05,
"loss": 0.2632,
"step": 1570
},
{
"epoch": 1.38,
"learning_rate": 3.412205346322794e-05,
"loss": 0.2338,
"step": 1580
},
{
"epoch": 1.38,
"learning_rate": 3.392093462331688e-05,
"loss": 0.2553,
"step": 1590
},
{
"epoch": 1.39,
"learning_rate": 3.371915199826506e-05,
"loss": 0.2519,
"step": 1600
},
{
"epoch": 1.39,
"learning_rate": 3.3516720602228945e-05,
"loss": 0.2549,
"step": 1610
},
{
"epoch": 1.4,
"learning_rate": 3.331365549763848e-05,
"loss": 0.2802,
"step": 1620
},
{
"epoch": 1.4,
"learning_rate": 3.310997179407631e-05,
"loss": 0.2734,
"step": 1630
},
{
"epoch": 1.41,
"learning_rate": 3.2905684647153536e-05,
"loss": 0.2609,
"step": 1640
},
{
"epoch": 1.41,
"learning_rate": 3.270080925738202e-05,
"loss": 0.2383,
"step": 1650
},
{
"epoch": 1.42,
"learning_rate": 3.2495360869043354e-05,
"loss": 0.2576,
"step": 1660
},
{
"epoch": 1.42,
"learning_rate": 3.2289354769054556e-05,
"loss": 0.2305,
"step": 1670
},
{
"epoch": 1.43,
"learning_rate": 3.2082806285830645e-05,
"loss": 0.2508,
"step": 1680
},
{
"epoch": 1.43,
"learning_rate": 3.187573078814402e-05,
"loss": 0.2367,
"step": 1690
},
{
"epoch": 1.44,
"learning_rate": 3.166814368398098e-05,
"loss": 0.2335,
"step": 1700
},
{
"epoch": 1.44,
"learning_rate": 3.146006041939522e-05,
"loss": 0.2523,
"step": 1710
},
{
"epoch": 1.45,
"step": 1718,
"total_flos": 3.091617473007452e+17,
"train_loss": 0.305820442848905,
"train_runtime": 6698.0985,
"train_samples_per_second": 2.289,
"train_steps_per_second": 0.572
}
],
"logging_steps": 10,
"max_steps": 3834,
"num_train_epochs": 2,
"save_steps": 500,
"total_flos": 3.091617473007452e+17,
"trial_name": null,
"trial_params": null
}