Hugging Face's logo

Wonder-Griffin
/
TraXLMistralForCausalLM

Text Generation
Transformers
Safetensors
English
TraXLMistral
text-generation-inference
causal-LM
question-answering
conversational
Inference Endpoints
Model card Files Community
TraXLMistralForCausalLM / checkpoint-5740 /trainer_state.json
Wonder-Griffin's picture
Wonder-Griffin
Upload 68 files
40d58fc verified
raw
history blame contribute delete
94 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 5740,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008710801393728223,
"grad_norm": Infinity,
"learning_rate": 3.71949339942388e-07,
"loss": 23.9832,
"step": 10
},
{
"epoch": 0.017421602787456445,
"grad_norm": 759.6871337890625,
"learning_rate": 1.2088353548127608e-06,
"loss": 68.5664,
"step": 20
},
{
"epoch": 0.02613240418118467,
"grad_norm": 1186.6964111328125,
"learning_rate": 2.045721369683134e-06,
"loss": 113.461,
"step": 30
},
{
"epoch": 0.03484320557491289,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 40
},
{
"epoch": 0.04355400696864112,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 50
},
{
"epoch": 0.05226480836236934,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 60
},
{
"epoch": 0.06097560975609756,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 70
},
{
"epoch": 0.06968641114982578,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 80
},
{
"epoch": 0.078397212543554,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 90
},
{
"epoch": 0.08710801393728224,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 100
},
{
"epoch": 0.09581881533101046,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 110
},
{
"epoch": 0.10452961672473868,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 120
},
{
"epoch": 0.1132404181184669,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 130
},
{
"epoch": 0.12195121951219512,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 140
},
{
"epoch": 0.13066202090592335,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 150
},
{
"epoch": 0.13937282229965156,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 160
},
{
"epoch": 0.1480836236933798,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 170
},
{
"epoch": 0.156794425087108,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 180
},
{
"epoch": 0.16550522648083624,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 190
},
{
"epoch": 0.17421602787456447,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 200
},
{
"epoch": 0.18292682926829268,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 210
},
{
"epoch": 0.1916376306620209,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 220
},
{
"epoch": 0.20034843205574912,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 230
},
{
"epoch": 0.20905923344947736,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 240
},
{
"epoch": 0.21777003484320556,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 250
},
{
"epoch": 0.2264808362369338,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 260
},
{
"epoch": 0.23519163763066203,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 270
},
{
"epoch": 0.24390243902439024,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 280
},
{
"epoch": 0.25261324041811845,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 290
},
{
"epoch": 0.2613240418118467,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 300
},
{
"epoch": 0.2700348432055749,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 310
},
{
"epoch": 0.2787456445993031,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 320
},
{
"epoch": 0.2874564459930314,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 330
},
{
"epoch": 0.2961672473867596,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 340
},
{
"epoch": 0.3048780487804878,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 350
},
{
"epoch": 0.313588850174216,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 360
},
{
"epoch": 0.32229965156794427,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 370
},
{
"epoch": 0.3310104529616725,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 380
},
{
"epoch": 0.3397212543554007,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 390
},
{
"epoch": 0.34843205574912894,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 400
},
{
"epoch": 0.35714285714285715,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 410
},
{
"epoch": 0.36585365853658536,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 420
},
{
"epoch": 0.37456445993031356,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 430
},
{
"epoch": 0.3832752613240418,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 440
},
{
"epoch": 0.39198606271777003,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 450
},
{
"epoch": 0.40069686411149824,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 460
},
{
"epoch": 0.4094076655052265,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 470
},
{
"epoch": 0.4181184668989547,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 480
},
{
"epoch": 0.4268292682926829,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 490
},
{
"epoch": 0.4355400696864111,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 500
},
{
"epoch": 0.4442508710801394,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 510
},
{
"epoch": 0.4529616724738676,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 520
},
{
"epoch": 0.4616724738675958,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 530
},
{
"epoch": 0.47038327526132406,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 540
},
{
"epoch": 0.47909407665505227,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 550
},
{
"epoch": 0.4878048780487805,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 560
},
{
"epoch": 0.4965156794425087,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 570
},
{
"epoch": 0.5052264808362369,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 580
},
{
"epoch": 0.5139372822299652,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 590
},
{
"epoch": 0.5226480836236934,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 600
},
{
"epoch": 0.5313588850174216,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 610
},
{
"epoch": 0.5400696864111498,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 620
},
{
"epoch": 0.5487804878048781,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 630
},
{
"epoch": 0.5574912891986062,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 640
},
{
"epoch": 0.5662020905923345,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 650
},
{
"epoch": 0.5749128919860628,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 660
},
{
"epoch": 0.5836236933797909,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 670
},
{
"epoch": 0.5923344947735192,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 680
},
{
"epoch": 0.6010452961672473,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 690
},
{
"epoch": 0.6097560975609756,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 700
},
{
"epoch": 0.6184668989547039,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 710
},
{
"epoch": 0.627177700348432,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 720
},
{
"epoch": 0.6358885017421603,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 730
},
{
"epoch": 0.6445993031358885,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 740
},
{
"epoch": 0.6533101045296167,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 750
},
{
"epoch": 0.662020905923345,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 760
},
{
"epoch": 0.6707317073170732,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 770
},
{
"epoch": 0.6794425087108014,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 780
},
{
"epoch": 0.6881533101045296,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 790
},
{
"epoch": 0.6968641114982579,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 800
},
{
"epoch": 0.705574912891986,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 810
},
{
"epoch": 0.7142857142857143,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 820
},
{
"epoch": 0.7229965156794426,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 830
},
{
"epoch": 0.7317073170731707,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 840
},
{
"epoch": 0.740418118466899,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 850
},
{
"epoch": 0.7491289198606271,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 860
},
{
"epoch": 0.7578397212543554,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 870
},
{
"epoch": 0.7665505226480837,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 880
},
{
"epoch": 0.7752613240418118,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 890
},
{
"epoch": 0.7839721254355401,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 900
},
{
"epoch": 0.7926829268292683,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 910
},
{
"epoch": 0.8013937282229965,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 920
},
{
"epoch": 0.8101045296167247,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 930
},
{
"epoch": 0.818815331010453,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 940
},
{
"epoch": 0.8275261324041812,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 950
},
{
"epoch": 0.8362369337979094,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 960
},
{
"epoch": 0.8449477351916377,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 970
},
{
"epoch": 0.8536585365853658,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 980
},
{
"epoch": 0.8623693379790941,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 990
},
{
"epoch": 0.8710801393728222,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1000
},
{
"epoch": 0.8797909407665505,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1010
},
{
"epoch": 0.8885017421602788,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1020
},
{
"epoch": 0.8972125435540069,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1030
},
{
"epoch": 0.9059233449477352,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1040
},
{
"epoch": 0.9146341463414634,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1050
},
{
"epoch": 0.9233449477351916,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1060
},
{
"epoch": 0.9320557491289199,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1070
},
{
"epoch": 0.9407665505226481,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1080
},
{
"epoch": 0.9494773519163763,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1090
},
{
"epoch": 0.9581881533101045,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1100
},
{
"epoch": 0.9668989547038328,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1110
},
{
"epoch": 0.975609756097561,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1120
},
{
"epoch": 0.9843205574912892,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1130
},
{
"epoch": 0.9930313588850174,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1140
},
{
"epoch": 1.0017421602787457,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1150
},
{
"epoch": 1.0104529616724738,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1160
},
{
"epoch": 1.019163763066202,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1170
},
{
"epoch": 1.0278745644599303,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1180
},
{
"epoch": 1.0365853658536586,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1190
},
{
"epoch": 1.0452961672473868,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1200
},
{
"epoch": 1.054006968641115,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1210
},
{
"epoch": 1.0627177700348431,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1220
},
{
"epoch": 1.0714285714285714,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1230
},
{
"epoch": 1.0801393728222997,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1240
},
{
"epoch": 1.088850174216028,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1250
},
{
"epoch": 1.0975609756097562,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1260
},
{
"epoch": 1.1062717770034842,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1270
},
{
"epoch": 1.1149825783972125,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1280
},
{
"epoch": 1.1236933797909407,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1290
},
{
"epoch": 1.132404181184669,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1300
},
{
"epoch": 1.1411149825783973,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1310
},
{
"epoch": 1.1498257839721253,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1320
},
{
"epoch": 1.1585365853658536,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1330
},
{
"epoch": 1.1672473867595818,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1340
},
{
"epoch": 1.17595818815331,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1350
},
{
"epoch": 1.1846689895470384,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1360
},
{
"epoch": 1.1933797909407666,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1370
},
{
"epoch": 1.202090592334495,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1380
},
{
"epoch": 1.210801393728223,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1390
},
{
"epoch": 1.2195121951219512,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1400
},
{
"epoch": 1.2282229965156795,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1410
},
{
"epoch": 1.2369337979094077,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1420
},
{
"epoch": 1.245644599303136,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1430
},
{
"epoch": 1.254355400696864,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1440
},
{
"epoch": 1.2630662020905923,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1450
},
{
"epoch": 1.2717770034843205,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1460
},
{
"epoch": 1.2804878048780488,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1470
},
{
"epoch": 1.289198606271777,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1480
},
{
"epoch": 1.297909407665505,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1490
},
{
"epoch": 1.3066202090592334,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1500
},
{
"epoch": 1.3153310104529616,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1510
},
{
"epoch": 1.32404181184669,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1520
},
{
"epoch": 1.3327526132404182,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1530
},
{
"epoch": 1.3414634146341464,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1540
},
{
"epoch": 1.3501742160278747,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1550
},
{
"epoch": 1.3588850174216027,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1560
},
{
"epoch": 1.367595818815331,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1570
},
{
"epoch": 1.3763066202090593,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1580
},
{
"epoch": 1.3850174216027875,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1590
},
{
"epoch": 1.3937282229965158,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1600
},
{
"epoch": 1.4024390243902438,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1610
},
{
"epoch": 1.411149825783972,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1620
},
{
"epoch": 1.4198606271777003,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1630
},
{
"epoch": 1.4285714285714286,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1640
},
{
"epoch": 1.4372822299651569,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1650
},
{
"epoch": 1.445993031358885,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1660
},
{
"epoch": 1.4547038327526132,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1670
},
{
"epoch": 1.4634146341463414,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1680
},
{
"epoch": 1.4721254355400697,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1690
},
{
"epoch": 1.480836236933798,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1700
},
{
"epoch": 1.489547038327526,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1710
},
{
"epoch": 1.4982578397212545,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1720
},
{
"epoch": 1.5069686411149825,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1730
},
{
"epoch": 1.5156794425087108,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1740
},
{
"epoch": 1.524390243902439,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1750
},
{
"epoch": 1.533101045296167,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1760
},
{
"epoch": 1.5418118466898956,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1770
},
{
"epoch": 1.5505226480836236,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1780
},
{
"epoch": 1.5592334494773519,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1790
},
{
"epoch": 1.5679442508710801,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1800
},
{
"epoch": 1.5766550522648084,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1810
},
{
"epoch": 1.5853658536585367,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1820
},
{
"epoch": 1.5940766550522647,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1830
},
{
"epoch": 1.6027874564459932,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1840
},
{
"epoch": 1.6114982578397212,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1850
},
{
"epoch": 1.6202090592334495,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1860
},
{
"epoch": 1.6289198606271778,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1870
},
{
"epoch": 1.6376306620209058,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1880
},
{
"epoch": 1.6463414634146343,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1890
},
{
"epoch": 1.6550522648083623,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1900
},
{
"epoch": 1.6637630662020906,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1910
},
{
"epoch": 1.6724738675958188,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1920
},
{
"epoch": 1.6811846689895469,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1930
},
{
"epoch": 1.6898954703832754,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1940
},
{
"epoch": 1.6986062717770034,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1950
},
{
"epoch": 1.7073170731707317,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1960
},
{
"epoch": 1.71602787456446,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1970
},
{
"epoch": 1.7247386759581882,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1980
},
{
"epoch": 1.7334494773519165,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 1990
},
{
"epoch": 1.7421602787456445,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2000
},
{
"epoch": 1.750871080139373,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2010
},
{
"epoch": 1.759581881533101,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2020
},
{
"epoch": 1.7682926829268293,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2030
},
{
"epoch": 1.7770034843205575,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2040
},
{
"epoch": 1.7857142857142856,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2050
},
{
"epoch": 1.794425087108014,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2060
},
{
"epoch": 1.8031358885017421,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2070
},
{
"epoch": 1.8118466898954704,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2080
},
{
"epoch": 1.8205574912891986,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2090
},
{
"epoch": 1.8292682926829267,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2100
},
{
"epoch": 1.8379790940766552,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2110
},
{
"epoch": 1.8466898954703832,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2120
},
{
"epoch": 1.8554006968641115,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2130
},
{
"epoch": 1.8641114982578397,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2140
},
{
"epoch": 1.872822299651568,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2150
},
{
"epoch": 1.8815331010452963,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2160
},
{
"epoch": 1.8902439024390243,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2170
},
{
"epoch": 1.8989547038327528,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2180
},
{
"epoch": 1.9076655052264808,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2190
},
{
"epoch": 1.916376306620209,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2200
},
{
"epoch": 1.9250871080139373,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2210
},
{
"epoch": 1.9337979094076654,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2220
},
{
"epoch": 1.9425087108013939,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2230
},
{
"epoch": 1.951219512195122,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2240
},
{
"epoch": 1.9599303135888502,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2250
},
{
"epoch": 1.9686411149825784,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2260
},
{
"epoch": 1.9773519163763065,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2270
},
{
"epoch": 1.986062717770035,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2280
},
{
"epoch": 1.994773519163763,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2290
},
{
"epoch": 2.0034843205574915,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2300
},
{
"epoch": 2.0121951219512195,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2310
},
{
"epoch": 2.0209059233449476,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2320
},
{
"epoch": 2.029616724738676,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2330
},
{
"epoch": 2.038327526132404,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2340
},
{
"epoch": 2.0470383275261326,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2350
},
{
"epoch": 2.0557491289198606,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2360
},
{
"epoch": 2.0644599303135887,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2370
},
{
"epoch": 2.073170731707317,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2380
},
{
"epoch": 2.081881533101045,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2390
},
{
"epoch": 2.0905923344947737,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2400
},
{
"epoch": 2.0993031358885017,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2410
},
{
"epoch": 2.10801393728223,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2420
},
{
"epoch": 2.1167247386759582,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2430
},
{
"epoch": 2.1254355400696863,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2440
},
{
"epoch": 2.1341463414634148,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2450
},
{
"epoch": 2.142857142857143,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2460
},
{
"epoch": 2.1515679442508713,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2470
},
{
"epoch": 2.1602787456445993,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2480
},
{
"epoch": 2.1689895470383274,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2490
},
{
"epoch": 2.177700348432056,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2500
},
{
"epoch": 2.186411149825784,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2510
},
{
"epoch": 2.1951219512195124,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2520
},
{
"epoch": 2.2038327526132404,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2530
},
{
"epoch": 2.2125435540069684,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2540
},
{
"epoch": 2.221254355400697,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2550
},
{
"epoch": 2.229965156794425,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2560
},
{
"epoch": 2.2386759581881535,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2570
},
{
"epoch": 2.2473867595818815,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2580
},
{
"epoch": 2.2560975609756095,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2590
},
{
"epoch": 2.264808362369338,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2600
},
{
"epoch": 2.273519163763066,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2610
},
{
"epoch": 2.2822299651567945,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2620
},
{
"epoch": 2.2909407665505226,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2630
},
{
"epoch": 2.2996515679442506,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2640
},
{
"epoch": 2.308362369337979,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2650
},
{
"epoch": 2.317073170731707,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2660
},
{
"epoch": 2.3257839721254356,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2670
},
{
"epoch": 2.3344947735191637,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2680
},
{
"epoch": 2.343205574912892,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2690
},
{
"epoch": 2.35191637630662,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2700
},
{
"epoch": 2.3606271777003482,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2710
},
{
"epoch": 2.3693379790940767,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2720
},
{
"epoch": 2.3780487804878048,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2730
},
{
"epoch": 2.3867595818815333,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2740
},
{
"epoch": 2.3954703832752613,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2750
},
{
"epoch": 2.40418118466899,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2760
},
{
"epoch": 2.412891986062718,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2770
},
{
"epoch": 2.421602787456446,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2780
},
{
"epoch": 2.4303135888501743,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2790
},
{
"epoch": 2.4390243902439024,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2800
},
{
"epoch": 2.447735191637631,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2810
},
{
"epoch": 2.456445993031359,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2820
},
{
"epoch": 2.465156794425087,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2830
},
{
"epoch": 2.4738675958188154,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2840
},
{
"epoch": 2.4825783972125435,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2850
},
{
"epoch": 2.491289198606272,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2860
},
{
"epoch": 2.5,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2870
},
{
"epoch": 2.508710801393728,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2880
},
{
"epoch": 2.5174216027874565,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2890
},
{
"epoch": 2.5261324041811846,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2900
},
{
"epoch": 2.534843205574913,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2910
},
{
"epoch": 2.543554006968641,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2920
},
{
"epoch": 2.552264808362369,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2930
},
{
"epoch": 2.5609756097560976,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2940
},
{
"epoch": 2.5696864111498257,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2950
},
{
"epoch": 2.578397212543554,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2960
},
{
"epoch": 2.587108013937282,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2970
},
{
"epoch": 2.59581881533101,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2980
},
{
"epoch": 2.6045296167247387,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 2990
},
{
"epoch": 2.6132404181184667,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3000
},
{
"epoch": 2.6219512195121952,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3010
},
{
"epoch": 2.6306620209059233,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3020
},
{
"epoch": 2.6393728222996513,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3030
},
{
"epoch": 2.64808362369338,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3040
},
{
"epoch": 2.6567944250871083,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3050
},
{
"epoch": 2.6655052264808363,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3060
},
{
"epoch": 2.6742160278745644,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3070
},
{
"epoch": 2.682926829268293,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3080
},
{
"epoch": 2.691637630662021,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3090
},
{
"epoch": 2.7003484320557494,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3100
},
{
"epoch": 2.7090592334494774,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3110
},
{
"epoch": 2.7177700348432055,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3120
},
{
"epoch": 2.726480836236934,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3130
},
{
"epoch": 2.735191637630662,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3140
},
{
"epoch": 2.7439024390243905,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3150
},
{
"epoch": 2.7526132404181185,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3160
},
{
"epoch": 2.7613240418118465,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3170
},
{
"epoch": 2.770034843205575,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3180
},
{
"epoch": 2.778745644599303,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3190
},
{
"epoch": 2.7874564459930316,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3200
},
{
"epoch": 2.7961672473867596,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3210
},
{
"epoch": 2.8048780487804876,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3220
},
{
"epoch": 2.813588850174216,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3230
},
{
"epoch": 2.822299651567944,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3240
},
{
"epoch": 2.8310104529616726,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3250
},
{
"epoch": 2.8397212543554007,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3260
},
{
"epoch": 2.8484320557491287,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3270
},
{
"epoch": 2.857142857142857,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3280
},
{
"epoch": 2.8658536585365852,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3290
},
{
"epoch": 2.8745644599303137,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3300
},
{
"epoch": 2.8832752613240418,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3310
},
{
"epoch": 2.89198606271777,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3320
},
{
"epoch": 2.9006968641114983,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3330
},
{
"epoch": 2.9094076655052263,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3340
},
{
"epoch": 2.918118466898955,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3350
},
{
"epoch": 2.926829268292683,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3360
},
{
"epoch": 2.935540069686411,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3370
},
{
"epoch": 2.9442508710801394,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3380
},
{
"epoch": 2.952961672473868,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3390
},
{
"epoch": 2.961672473867596,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3400
},
{
"epoch": 2.970383275261324,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3410
},
{
"epoch": 2.979094076655052,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3420
},
{
"epoch": 2.9878048780487805,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3430
},
{
"epoch": 2.996515679442509,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3440
},
{
"epoch": 3.005226480836237,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3450
},
{
"epoch": 3.013937282229965,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3460
},
{
"epoch": 3.0226480836236935,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3470
},
{
"epoch": 3.0313588850174216,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3480
},
{
"epoch": 3.0400696864111496,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3490
},
{
"epoch": 3.048780487804878,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3500
},
{
"epoch": 3.057491289198606,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3510
},
{
"epoch": 3.0662020905923346,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3520
},
{
"epoch": 3.0749128919860627,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3530
},
{
"epoch": 3.083623693379791,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3540
},
{
"epoch": 3.092334494773519,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3550
},
{
"epoch": 3.1010452961672472,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3560
},
{
"epoch": 3.1097560975609757,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3570
},
{
"epoch": 3.1184668989547037,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3580
},
{
"epoch": 3.1271777003484322,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3590
},
{
"epoch": 3.1358885017421603,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3600
},
{
"epoch": 3.1445993031358883,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3610
},
{
"epoch": 3.153310104529617,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3620
},
{
"epoch": 3.162020905923345,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3630
},
{
"epoch": 3.1707317073170733,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3640
},
{
"epoch": 3.1794425087108014,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3650
},
{
"epoch": 3.1881533101045294,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3660
},
{
"epoch": 3.196864111498258,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3670
},
{
"epoch": 3.205574912891986,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3680
},
{
"epoch": 3.2142857142857144,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3690
},
{
"epoch": 3.2229965156794425,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3700
},
{
"epoch": 3.231707317073171,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3710
},
{
"epoch": 3.240418118466899,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3720
},
{
"epoch": 3.249128919860627,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3730
},
{
"epoch": 3.2578397212543555,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3740
},
{
"epoch": 3.2665505226480835,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3750
},
{
"epoch": 3.275261324041812,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3760
},
{
"epoch": 3.28397212543554,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3770
},
{
"epoch": 3.292682926829268,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3780
},
{
"epoch": 3.3013937282229966,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3790
},
{
"epoch": 3.3101045296167246,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3800
},
{
"epoch": 3.318815331010453,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3810
},
{
"epoch": 3.327526132404181,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3820
},
{
"epoch": 3.3362369337979096,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3830
},
{
"epoch": 3.3449477351916377,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3840
},
{
"epoch": 3.3536585365853657,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3850
},
{
"epoch": 3.362369337979094,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3860
},
{
"epoch": 3.3710801393728222,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3870
},
{
"epoch": 3.3797909407665507,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3880
},
{
"epoch": 3.3885017421602788,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3890
},
{
"epoch": 3.397212543554007,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3900
},
{
"epoch": 3.4059233449477353,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3910
},
{
"epoch": 3.4146341463414633,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3920
},
{
"epoch": 3.423344947735192,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3930
},
{
"epoch": 3.43205574912892,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3940
},
{
"epoch": 3.440766550522648,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3950
},
{
"epoch": 3.4494773519163764,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3960
},
{
"epoch": 3.4581881533101044,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3970
},
{
"epoch": 3.466898954703833,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3980
},
{
"epoch": 3.475609756097561,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 3990
},
{
"epoch": 3.484320557491289,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4000
},
{
"epoch": 3.4930313588850175,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4010
},
{
"epoch": 3.5017421602787455,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4020
},
{
"epoch": 3.510452961672474,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4030
},
{
"epoch": 3.519163763066202,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4040
},
{
"epoch": 3.52787456445993,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4050
},
{
"epoch": 3.5365853658536586,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4060
},
{
"epoch": 3.5452961672473866,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4070
},
{
"epoch": 3.554006968641115,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4080
},
{
"epoch": 3.562717770034843,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4090
},
{
"epoch": 3.571428571428571,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4100
},
{
"epoch": 3.5801393728222997,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4110
},
{
"epoch": 3.588850174216028,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4120
},
{
"epoch": 3.597560975609756,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4130
},
{
"epoch": 3.6062717770034842,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4140
},
{
"epoch": 3.6149825783972127,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4150
},
{
"epoch": 3.6236933797909407,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4160
},
{
"epoch": 3.6324041811846692,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4170
},
{
"epoch": 3.6411149825783973,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4180
},
{
"epoch": 3.6498257839721253,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4190
},
{
"epoch": 3.658536585365854,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4200
},
{
"epoch": 3.667247386759582,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4210
},
{
"epoch": 3.6759581881533103,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4220
},
{
"epoch": 3.6846689895470384,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4230
},
{
"epoch": 3.6933797909407664,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4240
},
{
"epoch": 3.702090592334495,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4250
},
{
"epoch": 3.710801393728223,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4260
},
{
"epoch": 3.7195121951219514,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4270
},
{
"epoch": 3.7282229965156795,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4280
},
{
"epoch": 3.7369337979094075,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4290
},
{
"epoch": 3.745644599303136,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4300
},
{
"epoch": 3.754355400696864,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4310
},
{
"epoch": 3.7630662020905925,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4320
},
{
"epoch": 3.7717770034843205,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4330
},
{
"epoch": 3.7804878048780486,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4340
},
{
"epoch": 3.789198606271777,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4350
},
{
"epoch": 3.797909407665505,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4360
},
{
"epoch": 3.8066202090592336,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4370
},
{
"epoch": 3.8153310104529616,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4380
},
{
"epoch": 3.8240418118466897,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4390
},
{
"epoch": 3.832752613240418,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4400
},
{
"epoch": 3.841463414634146,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4410
},
{
"epoch": 3.8501742160278747,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4420
},
{
"epoch": 3.8588850174216027,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4430
},
{
"epoch": 3.8675958188153308,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4440
},
{
"epoch": 3.8763066202090593,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4450
},
{
"epoch": 3.8850174216027873,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4460
},
{
"epoch": 3.8937282229965158,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4470
},
{
"epoch": 3.902439024390244,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4480
},
{
"epoch": 3.911149825783972,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4490
},
{
"epoch": 3.9198606271777003,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4500
},
{
"epoch": 3.928571428571429,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4510
},
{
"epoch": 3.937282229965157,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4520
},
{
"epoch": 3.945993031358885,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4530
},
{
"epoch": 3.9547038327526134,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4540
},
{
"epoch": 3.9634146341463414,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4550
},
{
"epoch": 3.97212543554007,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4560
},
{
"epoch": 3.980836236933798,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4570
},
{
"epoch": 3.989547038327526,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4580
},
{
"epoch": 3.9982578397212545,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4590
},
{
"epoch": 4.006968641114983,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4600
},
{
"epoch": 4.015679442508711,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4610
},
{
"epoch": 4.024390243902439,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4620
},
{
"epoch": 4.033101045296167,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4630
},
{
"epoch": 4.041811846689895,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4640
},
{
"epoch": 4.050522648083624,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4650
},
{
"epoch": 4.059233449477352,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4660
},
{
"epoch": 4.06794425087108,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4670
},
{
"epoch": 4.076655052264808,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4680
},
{
"epoch": 4.085365853658536,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4690
},
{
"epoch": 4.094076655052265,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4700
},
{
"epoch": 4.102787456445993,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4710
},
{
"epoch": 4.111498257839721,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4720
},
{
"epoch": 4.120209059233449,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4730
},
{
"epoch": 4.128919860627177,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4740
},
{
"epoch": 4.137630662020906,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4750
},
{
"epoch": 4.146341463414634,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4760
},
{
"epoch": 4.155052264808362,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4770
},
{
"epoch": 4.16376306620209,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4780
},
{
"epoch": 4.172473867595818,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4790
},
{
"epoch": 4.181184668989547,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4800
},
{
"epoch": 4.189895470383275,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4810
},
{
"epoch": 4.198606271777003,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4820
},
{
"epoch": 4.2073170731707314,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4830
},
{
"epoch": 4.21602787456446,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4840
},
{
"epoch": 4.224738675958188,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4850
},
{
"epoch": 4.2334494773519165,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4860
},
{
"epoch": 4.2421602787456445,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4870
},
{
"epoch": 4.2508710801393725,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4880
},
{
"epoch": 4.2595818815331015,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4890
},
{
"epoch": 4.2682926829268295,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4900
},
{
"epoch": 4.2770034843205575,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4910
},
{
"epoch": 4.285714285714286,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4920
},
{
"epoch": 4.294425087108014,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4930
},
{
"epoch": 4.303135888501743,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4940
},
{
"epoch": 4.311846689895471,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4950
},
{
"epoch": 4.320557491289199,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4960
},
{
"epoch": 4.329268292682927,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4970
},
{
"epoch": 4.337979094076655,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4980
},
{
"epoch": 4.346689895470384,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 4990
},
{
"epoch": 4.355400696864112,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5000
},
{
"epoch": 4.36411149825784,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5010
},
{
"epoch": 4.372822299651568,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5020
},
{
"epoch": 4.381533101045296,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5030
},
{
"epoch": 4.390243902439025,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5040
},
{
"epoch": 4.398954703832753,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5050
},
{
"epoch": 4.407665505226481,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5060
},
{
"epoch": 4.416376306620209,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5070
},
{
"epoch": 4.425087108013937,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5080
},
{
"epoch": 4.433797909407666,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5090
},
{
"epoch": 4.442508710801394,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5100
},
{
"epoch": 4.451219512195122,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5110
},
{
"epoch": 4.45993031358885,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5120
},
{
"epoch": 4.468641114982578,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5130
},
{
"epoch": 4.477351916376307,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5140
},
{
"epoch": 4.486062717770035,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5150
},
{
"epoch": 4.494773519163763,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5160
},
{
"epoch": 4.503484320557491,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5170
},
{
"epoch": 4.512195121951219,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5180
},
{
"epoch": 4.520905923344948,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5190
},
{
"epoch": 4.529616724738676,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5200
},
{
"epoch": 4.538327526132404,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5210
},
{
"epoch": 4.547038327526132,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5220
},
{
"epoch": 4.55574912891986,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5230
},
{
"epoch": 4.564459930313589,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5240
},
{
"epoch": 4.573170731707317,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5250
},
{
"epoch": 4.581881533101045,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5260
},
{
"epoch": 4.590592334494773,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5270
},
{
"epoch": 4.599303135888501,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5280
},
{
"epoch": 4.60801393728223,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5290
},
{
"epoch": 4.616724738675958,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5300
},
{
"epoch": 4.625435540069686,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5310
},
{
"epoch": 4.634146341463414,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5320
},
{
"epoch": 4.642857142857143,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5330
},
{
"epoch": 4.651567944250871,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5340
},
{
"epoch": 4.660278745644599,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5350
},
{
"epoch": 4.668989547038327,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5360
},
{
"epoch": 4.677700348432055,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5370
},
{
"epoch": 4.686411149825784,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5380
},
{
"epoch": 4.695121951219512,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5390
},
{
"epoch": 4.70383275261324,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5400
},
{
"epoch": 4.7125435540069684,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5410
},
{
"epoch": 4.7212543554006965,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5420
},
{
"epoch": 4.729965156794425,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5430
},
{
"epoch": 4.7386759581881535,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5440
},
{
"epoch": 4.7473867595818815,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5450
},
{
"epoch": 4.7560975609756095,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5460
},
{
"epoch": 4.7648083623693385,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5470
},
{
"epoch": 4.7735191637630665,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5480
},
{
"epoch": 4.7822299651567945,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5490
},
{
"epoch": 4.790940766550523,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5500
},
{
"epoch": 4.799651567944251,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5510
},
{
"epoch": 4.80836236933798,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5520
},
{
"epoch": 4.817073170731708,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5530
},
{
"epoch": 4.825783972125436,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5540
},
{
"epoch": 4.834494773519164,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5550
},
{
"epoch": 4.843205574912892,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5560
},
{
"epoch": 4.851916376306621,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5570
},
{
"epoch": 4.860627177700349,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5580
},
{
"epoch": 4.869337979094077,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5590
},
{
"epoch": 4.878048780487805,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5600
},
{
"epoch": 4.886759581881533,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5610
},
{
"epoch": 4.895470383275262,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5620
},
{
"epoch": 4.90418118466899,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5630
},
{
"epoch": 4.912891986062718,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5640
},
{
"epoch": 4.921602787456446,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5650
},
{
"epoch": 4.930313588850174,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5660
},
{
"epoch": 4.939024390243903,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5670
},
{
"epoch": 4.947735191637631,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5680
},
{
"epoch": 4.956445993031359,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5690
},
{
"epoch": 4.965156794425087,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5700
},
{
"epoch": 4.973867595818815,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5710
},
{
"epoch": 4.982578397212544,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5720
},
{
"epoch": 4.991289198606272,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5730
},
{
"epoch": 5.0,
"grad_norm": NaN,
"learning_rate": 2.045721369683134e-06,
"loss": 0.0,
"step": 5740
}
],
"logging_steps": 10,
"max_steps": 5740,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"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": 2828953243944960.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}