Lemon / stage2_7b /trainer_state.json
cheryyunl's picture
Upload folder using huggingface_hub
ed05174 verified
{
"best_global_step": 1000,
"best_metric": 2.343059778213501,
"best_model_checkpoint": "saves/qwen2.5_3d/full/sft_7b_stage2/checkpoint-1000",
"epoch": 2.9977671451355663,
"eval_steps": 1000,
"global_step": 1173,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.025518341307814992,
"grad_norm": 41.38909683442534,
"learning_rate": 1.1440677966101696e-06,
"loss": 13.0588,
"step": 10
},
{
"epoch": 0.051036682615629984,
"grad_norm": 24.45025201744234,
"learning_rate": 2.4152542372881355e-06,
"loss": 11.6024,
"step": 20
},
{
"epoch": 0.07655502392344497,
"grad_norm": 8.918549090671922,
"learning_rate": 3.686440677966102e-06,
"loss": 9.677,
"step": 30
},
{
"epoch": 0.10207336523125997,
"grad_norm": 5.740463305990162,
"learning_rate": 4.957627118644068e-06,
"loss": 7.5451,
"step": 40
},
{
"epoch": 0.12759170653907495,
"grad_norm": 4.261996908905627,
"learning_rate": 6.2288135593220344e-06,
"loss": 6.2001,
"step": 50
},
{
"epoch": 0.15311004784688995,
"grad_norm": 2.9963531250158075,
"learning_rate": 7.5e-06,
"loss": 5.4325,
"step": 60
},
{
"epoch": 0.17862838915470494,
"grad_norm": 2.452719493622236,
"learning_rate": 8.771186440677966e-06,
"loss": 4.8224,
"step": 70
},
{
"epoch": 0.20414673046251994,
"grad_norm": 2.296924769718695,
"learning_rate": 1.0042372881355933e-05,
"loss": 4.3397,
"step": 80
},
{
"epoch": 0.22966507177033493,
"grad_norm": 2.02789705030659,
"learning_rate": 1.1313559322033899e-05,
"loss": 4.0637,
"step": 90
},
{
"epoch": 0.2551834130781499,
"grad_norm": 2.0102538064720865,
"learning_rate": 1.2584745762711864e-05,
"loss": 3.8254,
"step": 100
},
{
"epoch": 0.2807017543859649,
"grad_norm": 1.9988271290927235,
"learning_rate": 1.385593220338983e-05,
"loss": 3.6752,
"step": 110
},
{
"epoch": 0.3062200956937799,
"grad_norm": 1.9671546379874392,
"learning_rate": 1.4999966747387387e-05,
"loss": 3.5183,
"step": 120
},
{
"epoch": 0.3317384370015949,
"grad_norm": 2.3495671945712764,
"learning_rate": 1.4995976790645575e-05,
"loss": 3.4113,
"step": 130
},
{
"epoch": 0.3572567783094099,
"grad_norm": 2.4845114503377284,
"learning_rate": 1.4985340365156019e-05,
"loss": 3.3294,
"step": 140
},
{
"epoch": 0.3827751196172249,
"grad_norm": 2.2685356609887966,
"learning_rate": 1.496806690193372e-05,
"loss": 3.2443,
"step": 150
},
{
"epoch": 0.4082934609250399,
"grad_norm": 1.8552414857417485,
"learning_rate": 1.494417171686568e-05,
"loss": 3.1948,
"step": 160
},
{
"epoch": 0.43381180223285487,
"grad_norm": 2.5986726068440635,
"learning_rate": 1.4913675997130732e-05,
"loss": 3.1538,
"step": 170
},
{
"epoch": 0.45933014354066987,
"grad_norm": 2.3542672335324646,
"learning_rate": 1.4876606782413468e-05,
"loss": 3.0716,
"step": 180
},
{
"epoch": 0.48484848484848486,
"grad_norm": 2.1550658582198516,
"learning_rate": 1.4832996940928936e-05,
"loss": 3.038,
"step": 190
},
{
"epoch": 0.5103668261562998,
"grad_norm": 2.1753949765417206,
"learning_rate": 1.4782885140279318e-05,
"loss": 2.9772,
"step": 200
},
{
"epoch": 0.5358851674641149,
"grad_norm": 1.9692819011886733,
"learning_rate": 1.4726315813168478e-05,
"loss": 2.9469,
"step": 210
},
{
"epoch": 0.5614035087719298,
"grad_norm": 1.833984536566349,
"learning_rate": 1.466333911800475e-05,
"loss": 2.9032,
"step": 220
},
{
"epoch": 0.5869218500797448,
"grad_norm": 2.021073458068275,
"learning_rate": 1.4594010894426907e-05,
"loss": 2.8903,
"step": 230
},
{
"epoch": 0.6124401913875598,
"grad_norm": 1.8342599659729268,
"learning_rate": 1.4518392613792747e-05,
"loss": 2.8585,
"step": 240
},
{
"epoch": 0.6379585326953748,
"grad_norm": 1.8015476338333145,
"learning_rate": 1.4436551324674196e-05,
"loss": 2.8197,
"step": 250
},
{
"epoch": 0.6634768740031898,
"grad_norm": 1.774909204385973,
"learning_rate": 1.4348559593407249e-05,
"loss": 2.7953,
"step": 260
},
{
"epoch": 0.6889952153110048,
"grad_norm": 1.913402588745503,
"learning_rate": 1.4254495439749472e-05,
"loss": 2.7817,
"step": 270
},
{
"epoch": 0.7145135566188198,
"grad_norm": 1.8231676178000666,
"learning_rate": 1.4154442267702106e-05,
"loss": 2.7489,
"step": 280
},
{
"epoch": 0.7400318979266348,
"grad_norm": 1.8647896227394318,
"learning_rate": 1.4048488791558127e-05,
"loss": 2.7492,
"step": 290
},
{
"epoch": 0.7655502392344498,
"grad_norm": 1.7710269150416975,
"learning_rate": 1.3936728957241805e-05,
"loss": 2.7359,
"step": 300
},
{
"epoch": 0.7910685805422647,
"grad_norm": 1.7971354575124974,
"learning_rate": 1.3819261859009536e-05,
"loss": 2.682,
"step": 310
},
{
"epoch": 0.8165869218500797,
"grad_norm": 1.8543518497527265,
"learning_rate": 1.3696191651585793e-05,
"loss": 2.6693,
"step": 320
},
{
"epoch": 0.8421052631578947,
"grad_norm": 1.671172953201345,
"learning_rate": 1.3567627457812107e-05,
"loss": 2.6371,
"step": 330
},
{
"epoch": 0.8676236044657097,
"grad_norm": 1.793308965007255,
"learning_rate": 1.3433683271890945e-05,
"loss": 2.6105,
"step": 340
},
{
"epoch": 0.8931419457735247,
"grad_norm": 1.702891636275614,
"learning_rate": 1.3294477858310318e-05,
"loss": 2.5997,
"step": 350
},
{
"epoch": 0.9186602870813397,
"grad_norm": 1.8004220505203588,
"learning_rate": 1.315013464653869e-05,
"loss": 2.6038,
"step": 360
},
{
"epoch": 0.9441786283891547,
"grad_norm": 1.6762412987987798,
"learning_rate": 1.300078162158359e-05,
"loss": 2.5724,
"step": 370
},
{
"epoch": 0.9696969696969697,
"grad_norm": 1.8651248798134779,
"learning_rate": 1.2846551210510956e-05,
"loss": 2.572,
"step": 380
},
{
"epoch": 0.9952153110047847,
"grad_norm": 1.8166685018367434,
"learning_rate": 1.2687580165025842e-05,
"loss": 2.5526,
"step": 390
},
{
"epoch": 1.0229665071770335,
"grad_norm": 1.9595305580742408,
"learning_rate": 1.2524009440218576e-05,
"loss": 2.6489,
"step": 400
},
{
"epoch": 1.0484848484848486,
"grad_norm": 1.9448735030077777,
"learning_rate": 1.2355984069583902e-05,
"loss": 2.3534,
"step": 410
},
{
"epoch": 1.0740031897926634,
"grad_norm": 1.9568519341241308,
"learning_rate": 1.218365303642392e-05,
"loss": 2.3209,
"step": 420
},
{
"epoch": 1.0995215311004785,
"grad_norm": 1.97257201705927,
"learning_rate": 1.200716914174885e-05,
"loss": 2.3114,
"step": 430
},
{
"epoch": 1.1250398724082935,
"grad_norm": 2.0078778706593234,
"learning_rate": 1.1826688868792723e-05,
"loss": 2.3196,
"step": 440
},
{
"epoch": 1.1505582137161086,
"grad_norm": 1.9231486716564377,
"learning_rate": 1.1642372244264167e-05,
"loss": 2.2972,
"step": 450
},
{
"epoch": 1.1760765550239234,
"grad_norm": 1.9501054668769524,
"learning_rate": 1.1454382696455303e-05,
"loss": 2.2779,
"step": 460
},
{
"epoch": 1.2015948963317384,
"grad_norm": 1.8856831931474254,
"learning_rate": 1.126288691033452e-05,
"loss": 2.2627,
"step": 470
},
{
"epoch": 1.2271132376395535,
"grad_norm": 2.0213311293812573,
"learning_rate": 1.106805467975168e-05,
"loss": 2.2969,
"step": 480
},
{
"epoch": 1.2526315789473683,
"grad_norm": 1.9724633188826075,
"learning_rate": 1.087005875688676e-05,
"loss": 2.2616,
"step": 490
},
{
"epoch": 1.2781499202551834,
"grad_norm": 2.02716350744511,
"learning_rate": 1.0669074699075414e-05,
"loss": 2.2511,
"step": 500
},
{
"epoch": 1.3036682615629984,
"grad_norm": 2.0652621663353803,
"learning_rate": 1.0465280713147303e-05,
"loss": 2.2503,
"step": 510
},
{
"epoch": 1.3291866028708135,
"grad_norm": 2.033130261702073,
"learning_rate": 1.0258857497415165e-05,
"loss": 2.2372,
"step": 520
},
{
"epoch": 1.3547049441786285,
"grad_norm": 2.056888858914891,
"learning_rate": 1.0049988081454802e-05,
"loss": 2.2551,
"step": 530
},
{
"epoch": 1.3802232854864434,
"grad_norm": 2.0022262070955525,
"learning_rate": 9.838857663817956e-06,
"loss": 2.2482,
"step": 540
},
{
"epoch": 1.4057416267942584,
"grad_norm": 2.078167028605348,
"learning_rate": 9.625653447822052e-06,
"loss": 2.2707,
"step": 550
},
{
"epoch": 1.4312599681020735,
"grad_norm": 2.126752097583203,
"learning_rate": 9.410564475562346e-06,
"loss": 2.2295,
"step": 560
},
{
"epoch": 1.4567783094098883,
"grad_norm": 1.9806408754738973,
"learning_rate": 9.193781460293695e-06,
"loss": 2.1887,
"step": 570
},
{
"epoch": 1.4822966507177033,
"grad_norm": 2.0059818507445715,
"learning_rate": 8.975496617330568e-06,
"loss": 2.212,
"step": 580
},
{
"epoch": 1.5078149920255184,
"grad_norm": 1.958138514087202,
"learning_rate": 8.755903493615193e-06,
"loss": 2.2421,
"step": 590
},
{
"epoch": 1.5333333333333332,
"grad_norm": 1.9444870040248599,
"learning_rate": 8.535196796105018e-06,
"loss": 2.2001,
"step": 600
},
{
"epoch": 1.5588516746411485,
"grad_norm": 2.0790121588452677,
"learning_rate": 8.3135722191316e-06,
"loss": 2.1919,
"step": 610
},
{
"epoch": 1.5843700159489633,
"grad_norm": 2.09803077018995,
"learning_rate": 8.09122627088403e-06,
"loss": 2.181,
"step": 620
},
{
"epoch": 1.6098883572567781,
"grad_norm": 2.103389635373312,
"learning_rate": 7.868356099170721e-06,
"loss": 2.2188,
"step": 630
},
{
"epoch": 1.6354066985645934,
"grad_norm": 2.2123492409708,
"learning_rate": 7.645159316614082e-06,
"loss": 2.2191,
"step": 640
},
{
"epoch": 1.6609250398724082,
"grad_norm": 2.023742495950975,
"learning_rate": 7.421833825433035e-06,
"loss": 2.1925,
"step": 650
},
{
"epoch": 1.6864433811802233,
"grad_norm": 2.0082728157794567,
"learning_rate": 7.198577641968766e-06,
"loss": 2.1633,
"step": 660
},
{
"epoch": 1.7119617224880384,
"grad_norm": 2.1289624844092665,
"learning_rate": 6.97558872110929e-06,
"loss": 2.1864,
"step": 670
},
{
"epoch": 1.7374800637958532,
"grad_norm": 2.0363999925653395,
"learning_rate": 6.753064780768488e-06,
"loss": 2.1915,
"step": 680
},
{
"epoch": 1.7629984051036682,
"grad_norm": 2.1325211739700323,
"learning_rate": 6.5312031265752946e-06,
"loss": 2.1748,
"step": 690
},
{
"epoch": 1.7885167464114833,
"grad_norm": 2.073383778430727,
"learning_rate": 6.310200476928403e-06,
"loss": 2.1473,
"step": 700
},
{
"epoch": 1.8140350877192981,
"grad_norm": 2.0597448767291557,
"learning_rate": 6.090252788571685e-06,
"loss": 2.1668,
"step": 710
},
{
"epoch": 1.8395534290271134,
"grad_norm": 2.0432434933668033,
"learning_rate": 5.87155508284494e-06,
"loss": 2.1442,
"step": 720
},
{
"epoch": 1.8650717703349282,
"grad_norm": 2.0612494703787,
"learning_rate": 5.654301272764029e-06,
"loss": 2.1346,
"step": 730
},
{
"epoch": 1.8905901116427433,
"grad_norm": 2.0948535247185496,
"learning_rate": 5.438683991083748e-06,
"loss": 2.1436,
"step": 740
},
{
"epoch": 1.9161084529505583,
"grad_norm": 2.085105020269672,
"learning_rate": 5.224894419495842e-06,
"loss": 2.1523,
"step": 750
},
{
"epoch": 1.9416267942583731,
"grad_norm": 1.990712509803804,
"learning_rate": 5.013122119113674e-06,
"loss": 2.1595,
"step": 760
},
{
"epoch": 1.9671451355661882,
"grad_norm": 2.0676031547838565,
"learning_rate": 4.803554862393779e-06,
"loss": 2.1328,
"step": 770
},
{
"epoch": 1.9926634768740032,
"grad_norm": 2.0958105821364095,
"learning_rate": 4.596378466643397e-06,
"loss": 2.1694,
"step": 780
},
{
"epoch": 2.0204146730462518,
"grad_norm": 2.3390863338233348,
"learning_rate": 4.391776629261555e-06,
"loss": 2.126,
"step": 790
},
{
"epoch": 2.045933014354067,
"grad_norm": 2.168148536310652,
"learning_rate": 4.1899307648598355e-06,
"loss": 1.8096,
"step": 800
},
{
"epoch": 2.071451355661882,
"grad_norm": 2.107040319291286,
"learning_rate": 3.991019844407206e-06,
"loss": 1.7884,
"step": 810
},
{
"epoch": 2.096969696969697,
"grad_norm": 2.3752722791144096,
"learning_rate": 3.7952202365415787e-06,
"loss": 1.802,
"step": 820
},
{
"epoch": 2.122488038277512,
"grad_norm": 2.2622817947741636,
"learning_rate": 3.602705551188764e-06,
"loss": 1.7952,
"step": 830
},
{
"epoch": 2.148006379585327,
"grad_norm": 2.2685137603537058,
"learning_rate": 3.4136464856275042e-06,
"loss": 1.7794,
"step": 840
},
{
"epoch": 2.173524720893142,
"grad_norm": 2.3902445863895094,
"learning_rate": 3.228210673137064e-06,
"loss": 1.788,
"step": 850
},
{
"epoch": 2.199043062200957,
"grad_norm": 2.3611950118719314,
"learning_rate": 3.0465625343615967e-06,
"loss": 1.8184,
"step": 860
},
{
"epoch": 2.2245614035087717,
"grad_norm": 2.3628012509214322,
"learning_rate": 2.8688631315230326e-06,
"loss": 1.7802,
"step": 870
},
{
"epoch": 2.250079744816587,
"grad_norm": 2.274090723203016,
"learning_rate": 2.6952700256118117e-06,
"loss": 1.7664,
"step": 880
},
{
"epoch": 2.275598086124402,
"grad_norm": 2.4526876937461646,
"learning_rate": 2.5259371366820507e-06,
"loss": 1.8002,
"step": 890
},
{
"epoch": 2.301116427432217,
"grad_norm": 2.3336921883837616,
"learning_rate": 2.3610146073750176e-06,
"loss": 1.7763,
"step": 900
},
{
"epoch": 2.326634768740032,
"grad_norm": 2.271646264284567,
"learning_rate": 2.200648669791939e-06,
"loss": 1.7612,
"step": 910
},
{
"epoch": 2.3521531100478468,
"grad_norm": 2.4913302090031952,
"learning_rate": 2.044981515834167e-06,
"loss": 1.7946,
"step": 920
},
{
"epoch": 2.377671451355662,
"grad_norm": 2.439411059369649,
"learning_rate": 1.8941511711256728e-06,
"loss": 1.7869,
"step": 930
},
{
"epoch": 2.403189792663477,
"grad_norm": 2.3140599984104138,
"learning_rate": 1.7482913726296778e-06,
"loss": 1.7664,
"step": 940
},
{
"epoch": 2.4287081339712917,
"grad_norm": 2.500843401322422,
"learning_rate": 1.607531450067886e-06,
"loss": 1.7834,
"step": 950
},
{
"epoch": 2.454226475279107,
"grad_norm": 2.4192381369051033,
"learning_rate": 1.4719962112475227e-06,
"loss": 1.7711,
"step": 960
},
{
"epoch": 2.479744816586922,
"grad_norm": 2.381315270199767,
"learning_rate": 1.3418058313978064e-06,
"loss": 1.7727,
"step": 970
},
{
"epoch": 2.5052631578947366,
"grad_norm": 2.4411828126850454,
"learning_rate": 1.2170757466140037e-06,
"loss": 1.7536,
"step": 980
},
{
"epoch": 2.530781499202552,
"grad_norm": 2.4714371413125504,
"learning_rate": 1.0979165515035402e-06,
"loss": 1.7918,
"step": 990
},
{
"epoch": 2.5562998405103667,
"grad_norm": 2.3600482331288557,
"learning_rate": 9.844339011249114e-07,
"loss": 1.7683,
"step": 1000
},
{
"epoch": 2.5562998405103667,
"eval_loss": 2.343059778213501,
"eval_runtime": 20.7074,
"eval_samples_per_second": 254.982,
"eval_steps_per_second": 7.968,
"step": 1000
},
{
"epoch": 2.581818181818182,
"grad_norm": 2.42847050890942,
"learning_rate": 8.767284173063628e-07,
"loss": 1.7602,
"step": 1010
},
{
"epoch": 2.607336523125997,
"grad_norm": 2.495286470584022,
"learning_rate": 7.748955994273793e-07,
"loss": 1.7623,
"step": 1020
},
{
"epoch": 2.6328548644338117,
"grad_norm": 2.380920410548109,
"learning_rate": 6.790257397421221e-07,
"loss": 1.7764,
"step": 1030
},
{
"epoch": 2.658373205741627,
"grad_norm": 2.4026078375726305,
"learning_rate": 5.892038433198463e-07,
"loss": 1.7552,
"step": 1040
},
{
"epoch": 2.6838915470494418,
"grad_norm": 2.4208779504436673,
"learning_rate": 5.055095526733442e-07,
"loss": 1.7624,
"step": 1050
},
{
"epoch": 2.709409888357257,
"grad_norm": 2.349970978991061,
"learning_rate": 4.280170771421957e-07,
"loss": 1.7642,
"step": 1060
},
{
"epoch": 2.734928229665072,
"grad_norm": 2.3526689883363248,
"learning_rate": 3.5679512709346773e-07,
"loss": 1.7577,
"step": 1070
},
{
"epoch": 2.7604465709728867,
"grad_norm": 2.450274730203489,
"learning_rate": 2.919068529981894e-07,
"loss": 1.7654,
"step": 1080
},
{
"epoch": 2.7859649122807015,
"grad_norm": 2.3651571766241246,
"learning_rate": 2.3340978943763164e-07,
"loss": 1.761,
"step": 1090
},
{
"epoch": 2.811483253588517,
"grad_norm": 2.404642620983574,
"learning_rate": 1.8135580408903674e-07,
"loss": 1.7544,
"step": 1100
},
{
"epoch": 2.8370015948963316,
"grad_norm": 2.473317448160841,
"learning_rate": 1.357910517360264e-07,
"loss": 1.7586,
"step": 1110
},
{
"epoch": 2.862519936204147,
"grad_norm": 2.325829716379361,
"learning_rate": 9.675593334447674e-08,
"loss": 1.7541,
"step": 1120
},
{
"epoch": 2.8880382775119617,
"grad_norm": 2.430438279086189,
"learning_rate": 6.428506024012904e-08,
"loss": 1.77,
"step": 1130
},
{
"epoch": 2.9135566188197766,
"grad_norm": 2.3903497741918414,
"learning_rate": 3.840722341971664e-08,
"loss": 1.7565,
"step": 1140
},
{
"epoch": 2.939074960127592,
"grad_norm": 2.3801298605080023,
"learning_rate": 1.914536802281208e-08,
"loss": 1.7433,
"step": 1150
},
{
"epoch": 2.9645933014354067,
"grad_norm": 2.364114165792993,
"learning_rate": 6.516572987022851e-09,
"loss": 1.7701,
"step": 1160
},
{
"epoch": 2.990111642743222,
"grad_norm": 2.2932824036528032,
"learning_rate": 5.320359045882239e-10,
"loss": 1.7531,
"step": 1170
},
{
"epoch": 2.9977671451355663,
"step": 1173,
"total_flos": 102874311622656.0,
"train_loss": 2.6594049156169453,
"train_runtime": 4231.9475,
"train_samples_per_second": 71.111,
"train_steps_per_second": 0.277
}
],
"logging_steps": 10,
"max_steps": 1173,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 1000,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 3000,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 102874311622656.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}