Hugging Face's logo

ahmadmwali
/
afriteva_large_trial2

PEFT
TensorBoard
Safetensors
Generated from Trainer
Model card Files
xet
 Metrics Community
afriteva_large_trial2 / trainer_state.json
ahmadmwali's picture
ahmadmwali
End of training
ab0e80d verified
raw
history blame contribute delete
99.4 kB
Invalid JSON: Unexpected token 'I', ..."_metric": Infinity, "... is not valid JSON
{
 "best_global_step": null,
 "best_metric": Infinity,
 "best_model_checkpoint": null,
 "epoch": 3.0,
 "eval_steps": 500,
 "global_step": 37500,
 "is_hyper_param_search": false,
 "is_local_process_zero": true,
 "is_world_process_zero": true,
 "log_history": [
 {
 "epoch": 0.004,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 50
 },
 {
 "epoch": 0.008,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 100
 },
 {
 "epoch": 0.012,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 150
 },
 {
 "epoch": 0.016,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 200
 },
 {
 "epoch": 0.02,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 250
 },
 {
 "epoch": 0.024,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 300
 },
 {
 "epoch": 0.028,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 350
 },
 {
 "epoch": 0.032,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 400
 },
 {
 "epoch": 0.036,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 450
 },
 {
 "epoch": 0.04,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 500
 },
 {
 "epoch": 0.044,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 550
 },
 {
 "epoch": 0.048,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 600
 },
 {
 "epoch": 0.052,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 650
 },
 {
 "epoch": 0.056,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 700
 },
 {
 "epoch": 0.06,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 750
 },
 {
 "epoch": 0.064,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 800
 },
 {
 "epoch": 0.068,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 850
 },
 {
 "epoch": 0.072,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 900
 },
 {
 "epoch": 0.076,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 950
 },
 {
 "epoch": 0.08,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1000
 },
 {
 "epoch": 0.084,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1050
 },
 {
 "epoch": 0.088,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1100
 },
 {
 "epoch": 0.092,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1150
 },
 {
 "epoch": 0.096,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1200
 },
 {
 "epoch": 0.1,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1250
 },
 {
 "epoch": 0.104,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1300
 },
 {
 "epoch": 0.108,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1350
 },
 {
 "epoch": 0.112,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1400
 },
 {
 "epoch": 0.116,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1450
 },
 {
 "epoch": 0.12,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1500
 },
 {
 "epoch": 0.124,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1550
 },
 {
 "epoch": 0.128,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1600
 },
 {
 "epoch": 0.132,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1650
 },
 {
 "epoch": 0.136,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1700
 },
 {
 "epoch": 0.14,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1750
 },
 {
 "epoch": 0.144,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1800
 },
 {
 "epoch": 0.148,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1850
 },
 {
 "epoch": 0.152,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1900
 },
 {
 "epoch": 0.156,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 1950
 },
 {
 "epoch": 0.16,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2000
 },
 {
 "epoch": 0.164,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2050
 },
 {
 "epoch": 0.168,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2100
 },
 {
 "epoch": 0.172,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2150
 },
 {
 "epoch": 0.176,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2200
 },
 {
 "epoch": 0.18,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2250
 },
 {
 "epoch": 0.184,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2300
 },
 {
 "epoch": 0.188,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2350
 },
 {
 "epoch": 0.192,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2400
 },
 {
 "epoch": 0.196,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2450
 },
 {
 "epoch": 0.2,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2500
 },
 {
 "epoch": 0.204,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2550
 },
 {
 "epoch": 0.208,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2600
 },
 {
 "epoch": 0.212,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2650
 },
 {
 "epoch": 0.216,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2700
 },
 {
 "epoch": 0.22,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2750
 },
 {
 "epoch": 0.224,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2800
 },
 {
 "epoch": 0.228,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2850
 },
 {
 "epoch": 0.232,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2900
 },
 {
 "epoch": 0.236,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 2950
 },
 {
 "epoch": 0.24,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3000
 },
 {
 "epoch": 0.244,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3050
 },
 {
 "epoch": 0.248,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3100
 },
 {
 "epoch": 0.252,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3150
 },
 {
 "epoch": 0.256,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3200
 },
 {
 "epoch": 0.26,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3250
 },
 {
 "epoch": 0.264,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3300
 },
 {
 "epoch": 0.268,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3350
 },
 {
 "epoch": 0.272,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3400
 },
 {
 "epoch": 0.276,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3450
 },
 {
 "epoch": 0.28,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3500
 },
 {
 "epoch": 0.284,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3550
 },
 {
 "epoch": 0.288,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3600
 },
 {
 "epoch": 0.292,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3650
 },
 {
 "epoch": 0.296,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3700
 },
 {
 "epoch": 0.3,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3750
 },
 {
 "epoch": 0.304,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3800
 },
 {
 "epoch": 0.308,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3850
 },
 {
 "epoch": 0.312,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3900
 },
 {
 "epoch": 0.316,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 3950
 },
 {
 "epoch": 0.32,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4000
 },
 {
 "epoch": 0.324,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4050
 },
 {
 "epoch": 0.328,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4100
 },
 {
 "epoch": 0.332,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4150
 },
 {
 "epoch": 0.336,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4200
 },
 {
 "epoch": 0.34,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4250
 },
 {
 "epoch": 0.344,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4300
 },
 {
 "epoch": 0.348,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4350
 },
 {
 "epoch": 0.352,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4400
 },
 {
 "epoch": 0.356,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4450
 },
 {
 "epoch": 0.36,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4500
 },
 {
 "epoch": 0.364,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4550
 },
 {
 "epoch": 0.368,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4600
 },
 {
 "epoch": 0.372,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4650
 },
 {
 "epoch": 0.376,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4700
 },
 {
 "epoch": 0.38,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4750
 },
 {
 "epoch": 0.384,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4800
 },
 {
 "epoch": 0.388,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4850
 },
 {
 "epoch": 0.392,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4900
 },
 {
 "epoch": 0.396,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 4950
 },
 {
 "epoch": 0.4,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5000
 },
 {
 "epoch": 0.404,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5050
 },
 {
 "epoch": 0.408,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5100
 },
 {
 "epoch": 0.412,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5150
 },
 {
 "epoch": 0.416,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5200
 },
 {
 "epoch": 0.42,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5250
 },
 {
 "epoch": 0.424,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5300
 },
 {
 "epoch": 0.428,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5350
 },
 {
 "epoch": 0.432,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5400
 },
 {
 "epoch": 0.436,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5450
 },
 {
 "epoch": 0.44,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5500
 },
 {
 "epoch": 0.444,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5550
 },
 {
 "epoch": 0.448,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5600
 },
 {
 "epoch": 0.452,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5650
 },
 {
 "epoch": 0.456,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5700
 },
 {
 "epoch": 0.46,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5750
 },
 {
 "epoch": 0.464,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5800
 },
 {
 "epoch": 0.468,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5850
 },
 {
 "epoch": 0.472,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5900
 },
 {
 "epoch": 0.476,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 5950
 },
 {
 "epoch": 0.48,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6000
 },
 {
 "epoch": 0.484,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6050
 },
 {
 "epoch": 0.488,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6100
 },
 {
 "epoch": 0.492,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6150
 },
 {
 "epoch": 0.496,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6200
 },
 {
 "epoch": 0.5,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6250
 },
 {
 "epoch": 0.504,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6300
 },
 {
 "epoch": 0.508,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6350
 },
 {
 "epoch": 0.512,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6400
 },
 {
 "epoch": 0.516,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6450
 },
 {
 "epoch": 0.52,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6500
 },
 {
 "epoch": 0.524,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6550
 },
 {
 "epoch": 0.528,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6600
 },
 {
 "epoch": 0.532,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6650
 },
 {
 "epoch": 0.536,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6700
 },
 {
 "epoch": 0.54,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6750
 },
 {
 "epoch": 0.544,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6800
 },
 {
 "epoch": 0.548,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6850
 },
 {
 "epoch": 0.552,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6900
 },
 {
 "epoch": 0.556,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 6950
 },
 {
 "epoch": 0.56,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7000
 },
 {
 "epoch": 0.564,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7050
 },
 {
 "epoch": 0.568,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7100
 },
 {
 "epoch": 0.572,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7150
 },
 {
 "epoch": 0.576,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7200
 },
 {
 "epoch": 0.58,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7250
 },
 {
 "epoch": 0.584,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7300
 },
 {
 "epoch": 0.588,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7350
 },
 {
 "epoch": 0.592,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7400
 },
 {
 "epoch": 0.596,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7450
 },
 {
 "epoch": 0.6,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7500
 },
 {
 "epoch": 0.604,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7550
 },
 {
 "epoch": 0.608,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7600
 },
 {
 "epoch": 0.612,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7650
 },
 {
 "epoch": 0.616,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7700
 },
 {
 "epoch": 0.62,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7750
 },
 {
 "epoch": 0.624,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7800
 },
 {
 "epoch": 0.628,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7850
 },
 {
 "epoch": 0.632,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7900
 },
 {
 "epoch": 0.636,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 7950
 },
 {
 "epoch": 0.64,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8000
 },
 {
 "epoch": 0.644,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8050
 },
 {
 "epoch": 0.648,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8100
 },
 {
 "epoch": 0.652,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8150
 },
 {
 "epoch": 0.656,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8200
 },
 {
 "epoch": 0.66,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8250
 },
 {
 "epoch": 0.664,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8300
 },
 {
 "epoch": 0.668,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8350
 },
 {
 "epoch": 0.672,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8400
 },
 {
 "epoch": 0.676,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8450
 },
 {
 "epoch": 0.68,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8500
 },
 {
 "epoch": 0.684,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8550
 },
 {
 "epoch": 0.688,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8600
 },
 {
 "epoch": 0.692,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8650
 },
 {
 "epoch": 0.696,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8700
 },
 {
 "epoch": 0.7,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8750
 },
 {
 "epoch": 0.704,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8800
 },
 {
 "epoch": 0.708,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8850
 },
 {
 "epoch": 0.712,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8900
 },
 {
 "epoch": 0.716,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 8950
 },
 {
 "epoch": 0.72,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9000
 },
 {
 "epoch": 0.724,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9050
 },
 {
 "epoch": 0.728,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9100
 },
 {
 "epoch": 0.732,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9150
 },
 {
 "epoch": 0.736,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9200
 },
 {
 "epoch": 0.74,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9250
 },
 {
 "epoch": 0.744,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9300
 },
 {
 "epoch": 0.748,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9350
 },
 {
 "epoch": 0.752,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9400
 },
 {
 "epoch": 0.756,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9450
 },
 {
 "epoch": 0.76,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9500
 },
 {
 "epoch": 0.764,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9550
 },
 {
 "epoch": 0.768,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9600
 },
 {
 "epoch": 0.772,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9650
 },
 {
 "epoch": 0.776,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9700
 },
 {
 "epoch": 0.78,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9750
 },
 {
 "epoch": 0.784,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9800
 },
 {
 "epoch": 0.788,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9850
 },
 {
 "epoch": 0.792,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9900
 },
 {
 "epoch": 0.796,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 9950
 },
 {
 "epoch": 0.8,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10000
 },
 {
 "epoch": 0.804,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10050
 },
 {
 "epoch": 0.808,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10100
 },
 {
 "epoch": 0.812,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10150
 },
 {
 "epoch": 0.816,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10200
 },
 {
 "epoch": 0.82,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10250
 },
 {
 "epoch": 0.824,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10300
 },
 {
 "epoch": 0.828,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10350
 },
 {
 "epoch": 0.832,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10400
 },
 {
 "epoch": 0.836,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10450
 },
 {
 "epoch": 0.84,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10500
 },
 {
 "epoch": 0.844,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10550
 },
 {
 "epoch": 0.848,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10600
 },
 {
 "epoch": 0.852,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10650
 },
 {
 "epoch": 0.856,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10700
 },
 {
 "epoch": 0.86,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10750
 },
 {
 "epoch": 0.864,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10800
 },
 {
 "epoch": 0.868,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10850
 },
 {
 "epoch": 0.872,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10900
 },
 {
 "epoch": 0.876,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 10950
 },
 {
 "epoch": 0.88,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11000
 },
 {
 "epoch": 0.884,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11050
 },
 {
 "epoch": 0.888,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11100
 },
 {
 "epoch": 0.892,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11150
 },
 {
 "epoch": 0.896,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11200
 },
 {
 "epoch": 0.9,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11250
 },
 {
 "epoch": 0.904,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11300
 },
 {
 "epoch": 0.908,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11350
 },
 {
 "epoch": 0.912,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11400
 },
 {
 "epoch": 0.916,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11450
 },
 {
 "epoch": 0.92,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11500
 },
 {
 "epoch": 0.924,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11550
 },
 {
 "epoch": 0.928,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11600
 },
 {
 "epoch": 0.932,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11650
 },
 {
 "epoch": 0.936,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11700
 },
 {
 "epoch": 0.94,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11750
 },
 {
 "epoch": 0.944,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11800
 },
 {
 "epoch": 0.948,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11850
 },
 {
 "epoch": 0.952,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11900
 },
 {
 "epoch": 0.956,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 11950
 },
 {
 "epoch": 0.96,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12000
 },
 {
 "epoch": 0.964,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12050
 },
 {
 "epoch": 0.968,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12100
 },
 {
 "epoch": 0.972,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12150
 },
 {
 "epoch": 0.976,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12200
 },
 {
 "epoch": 0.98,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12250
 },
 {
 "epoch": 0.984,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12300
 },
 {
 "epoch": 0.988,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12350
 },
 {
 "epoch": 0.992,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12400
 },
 {
 "epoch": 0.996,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12450
 },
 {
 "epoch": 1.0,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12500
 },
 {
 "epoch": 1.0,
 "eval_bleu": 0.12212197044710801,
 "eval_cer": 0.8160053321787701,
 "eval_f1": 0.35202235691264105,
 "eval_loss": NaN,
 "eval_meteor": 0.2636940499037806,
 "eval_runtime": 655.0147,
 "eval_samples_per_second": 7.633,
 "eval_steps_per_second": 0.24,
 "eval_wer": 0.9984162999487562,
 "step": 12500
 },
 {
 "epoch": 1.004,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12550
 },
 {
 "epoch": 1.008,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12600
 },
 {
 "epoch": 1.012,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12650
 },
 {
 "epoch": 1.016,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12700
 },
 {
 "epoch": 1.02,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12750
 },
 {
 "epoch": 1.024,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12800
 },
 {
 "epoch": 1.028,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12850
 },
 {
 "epoch": 1.032,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12900
 },
 {
 "epoch": 1.036,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 12950
 },
 {
 "epoch": 1.04,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13000
 },
 {
 "epoch": 1.044,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13050
 },
 {
 "epoch": 1.048,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13100
 },
 {
 "epoch": 1.052,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13150
 },
 {
 "epoch": 1.056,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13200
 },
 {
 "epoch": 1.06,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13250
 },
 {
 "epoch": 1.064,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13300
 },
 {
 "epoch": 1.068,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13350
 },
 {
 "epoch": 1.072,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13400
 },
 {
 "epoch": 1.076,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13450
 },
 {
 "epoch": 1.08,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13500
 },
 {
 "epoch": 1.084,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13550
 },
 {
 "epoch": 1.088,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13600
 },
 {
 "epoch": 1.092,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13650
 },
 {
 "epoch": 1.096,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13700
 },
 {
 "epoch": 1.1,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13750
 },
 {
 "epoch": 1.104,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13800
 },
 {
 "epoch": 1.108,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13850
 },
 {
 "epoch": 1.112,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13900
 },
 {
 "epoch": 1.116,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 13950
 },
 {
 "epoch": 1.12,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14000
 },
 {
 "epoch": 1.124,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14050
 },
 {
 "epoch": 1.1280000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14100
 },
 {
 "epoch": 1.1320000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14150
 },
 {
 "epoch": 1.1360000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14200
 },
 {
 "epoch": 1.1400000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14250
 },
 {
 "epoch": 1.144,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14300
 },
 {
 "epoch": 1.148,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14350
 },
 {
 "epoch": 1.152,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14400
 },
 {
 "epoch": 1.156,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14450
 },
 {
 "epoch": 1.16,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14500
 },
 {
 "epoch": 1.164,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14550
 },
 {
 "epoch": 1.168,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14600
 },
 {
 "epoch": 1.172,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14650
 },
 {
 "epoch": 1.176,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14700
 },
 {
 "epoch": 1.18,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14750
 },
 {
 "epoch": 1.184,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14800
 },
 {
 "epoch": 1.188,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14850
 },
 {
 "epoch": 1.192,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14900
 },
 {
 "epoch": 1.196,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 14950
 },
 {
 "epoch": 1.2,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15000
 },
 {
 "epoch": 1.204,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15050
 },
 {
 "epoch": 1.208,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15100
 },
 {
 "epoch": 1.212,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15150
 },
 {
 "epoch": 1.216,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15200
 },
 {
 "epoch": 1.22,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15250
 },
 {
 "epoch": 1.224,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15300
 },
 {
 "epoch": 1.228,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15350
 },
 {
 "epoch": 1.232,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15400
 },
 {
 "epoch": 1.236,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15450
 },
 {
 "epoch": 1.24,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15500
 },
 {
 "epoch": 1.244,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15550
 },
 {
 "epoch": 1.248,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15600
 },
 {
 "epoch": 1.252,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15650
 },
 {
 "epoch": 1.256,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15700
 },
 {
 "epoch": 1.26,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15750
 },
 {
 "epoch": 1.264,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15800
 },
 {
 "epoch": 1.268,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15850
 },
 {
 "epoch": 1.272,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15900
 },
 {
 "epoch": 1.276,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 15950
 },
 {
 "epoch": 1.28,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16000
 },
 {
 "epoch": 1.284,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16050
 },
 {
 "epoch": 1.288,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16100
 },
 {
 "epoch": 1.292,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16150
 },
 {
 "epoch": 1.296,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16200
 },
 {
 "epoch": 1.3,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16250
 },
 {
 "epoch": 1.304,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16300
 },
 {
 "epoch": 1.308,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16350
 },
 {
 "epoch": 1.312,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16400
 },
 {
 "epoch": 1.316,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16450
 },
 {
 "epoch": 1.32,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16500
 },
 {
 "epoch": 1.324,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16550
 },
 {
 "epoch": 1.328,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16600
 },
 {
 "epoch": 1.332,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16650
 },
 {
 "epoch": 1.336,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16700
 },
 {
 "epoch": 1.34,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16750
 },
 {
 "epoch": 1.3439999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16800
 },
 {
 "epoch": 1.3479999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16850
 },
 {
 "epoch": 1.3519999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16900
 },
 {
 "epoch": 1.3559999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 16950
 },
 {
 "epoch": 1.3599999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17000
 },
 {
 "epoch": 1.3639999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17050
 },
 {
 "epoch": 1.3679999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17100
 },
 {
 "epoch": 1.3719999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17150
 },
 {
 "epoch": 1.376,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17200
 },
 {
 "epoch": 1.38,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17250
 },
 {
 "epoch": 1.384,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17300
 },
 {
 "epoch": 1.388,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17350
 },
 {
 "epoch": 1.392,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17400
 },
 {
 "epoch": 1.396,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17450
 },
 {
 "epoch": 1.4,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17500
 },
 {
 "epoch": 1.404,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17550
 },
 {
 "epoch": 1.408,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17600
 },
 {
 "epoch": 1.412,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17650
 },
 {
 "epoch": 1.416,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17700
 },
 {
 "epoch": 1.42,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17750
 },
 {
 "epoch": 1.424,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17800
 },
 {
 "epoch": 1.428,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17850
 },
 {
 "epoch": 1.432,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17900
 },
 {
 "epoch": 1.436,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 17950
 },
 {
 "epoch": 1.44,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18000
 },
 {
 "epoch": 1.444,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18050
 },
 {
 "epoch": 1.448,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18100
 },
 {
 "epoch": 1.452,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18150
 },
 {
 "epoch": 1.456,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18200
 },
 {
 "epoch": 1.46,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18250
 },
 {
 "epoch": 1.464,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18300
 },
 {
 "epoch": 1.468,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18350
 },
 {
 "epoch": 1.472,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18400
 },
 {
 "epoch": 1.476,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18450
 },
 {
 "epoch": 1.48,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18500
 },
 {
 "epoch": 1.484,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18550
 },
 {
 "epoch": 1.488,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18600
 },
 {
 "epoch": 1.492,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18650
 },
 {
 "epoch": 1.496,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18700
 },
 {
 "epoch": 1.5,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18750
 },
 {
 "epoch": 1.504,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18800
 },
 {
 "epoch": 1.508,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18850
 },
 {
 "epoch": 1.512,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18900
 },
 {
 "epoch": 1.516,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 18950
 },
 {
 "epoch": 1.52,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19000
 },
 {
 "epoch": 1.524,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19050
 },
 {
 "epoch": 1.528,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19100
 },
 {
 "epoch": 1.532,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19150
 },
 {
 "epoch": 1.536,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19200
 },
 {
 "epoch": 1.54,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19250
 },
 {
 "epoch": 1.544,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19300
 },
 {
 "epoch": 1.548,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19350
 },
 {
 "epoch": 1.552,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19400
 },
 {
 "epoch": 1.556,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19450
 },
 {
 "epoch": 1.56,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19500
 },
 {
 "epoch": 1.564,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19550
 },
 {
 "epoch": 1.568,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19600
 },
 {
 "epoch": 1.572,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19650
 },
 {
 "epoch": 1.576,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19700
 },
 {
 "epoch": 1.58,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19750
 },
 {
 "epoch": 1.584,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19800
 },
 {
 "epoch": 1.588,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19850
 },
 {
 "epoch": 1.592,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19900
 },
 {
 "epoch": 1.596,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 19950
 },
 {
 "epoch": 1.6,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20000
 },
 {
 "epoch": 1.604,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20050
 },
 {
 "epoch": 1.608,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20100
 },
 {
 "epoch": 1.612,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20150
 },
 {
 "epoch": 1.616,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20200
 },
 {
 "epoch": 1.62,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20250
 },
 {
 "epoch": 1.624,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20300
 },
 {
 "epoch": 1.6280000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20350
 },
 {
 "epoch": 1.6320000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20400
 },
 {
 "epoch": 1.6360000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20450
 },
 {
 "epoch": 1.6400000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20500
 },
 {
 "epoch": 1.6440000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20550
 },
 {
 "epoch": 1.6480000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20600
 },
 {
 "epoch": 1.6520000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20650
 },
 {
 "epoch": 1.6560000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20700
 },
 {
 "epoch": 1.6600000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20750
 },
 {
 "epoch": 1.6640000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20800
 },
 {
 "epoch": 1.6680000000000001,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20850
 },
 {
 "epoch": 1.6720000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20900
 },
 {
 "epoch": 1.6760000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 20950
 },
 {
 "epoch": 1.6800000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21000
 },
 {
 "epoch": 1.6840000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21050
 },
 {
 "epoch": 1.688,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21100
 },
 {
 "epoch": 1.692,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21150
 },
 {
 "epoch": 1.696,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21200
 },
 {
 "epoch": 1.7,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21250
 },
 {
 "epoch": 1.704,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21300
 },
 {
 "epoch": 1.708,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21350
 },
 {
 "epoch": 1.712,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21400
 },
 {
 "epoch": 1.716,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21450
 },
 {
 "epoch": 1.72,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21500
 },
 {
 "epoch": 1.724,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21550
 },
 {
 "epoch": 1.728,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21600
 },
 {
 "epoch": 1.732,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21650
 },
 {
 "epoch": 1.736,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21700
 },
 {
 "epoch": 1.74,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21750
 },
 {
 "epoch": 1.744,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21800
 },
 {
 "epoch": 1.748,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21850
 },
 {
 "epoch": 1.752,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21900
 },
 {
 "epoch": 1.756,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 21950
 },
 {
 "epoch": 1.76,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22000
 },
 {
 "epoch": 1.764,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22050
 },
 {
 "epoch": 1.768,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22100
 },
 {
 "epoch": 1.772,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22150
 },
 {
 "epoch": 1.776,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22200
 },
 {
 "epoch": 1.78,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22250
 },
 {
 "epoch": 1.784,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22300
 },
 {
 "epoch": 1.788,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22350
 },
 {
 "epoch": 1.792,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22400
 },
 {
 "epoch": 1.796,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22450
 },
 {
 "epoch": 1.8,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22500
 },
 {
 "epoch": 1.804,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22550
 },
 {
 "epoch": 1.808,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22600
 },
 {
 "epoch": 1.812,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22650
 },
 {
 "epoch": 1.8159999999999998,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22700
 },
 {
 "epoch": 1.8199999999999998,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22750
 },
 {
 "epoch": 1.8239999999999998,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22800
 },
 {
 "epoch": 1.8279999999999998,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22850
 },
 {
 "epoch": 1.8319999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22900
 },
 {
 "epoch": 1.8359999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 22950
 },
 {
 "epoch": 1.8399999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23000
 },
 {
 "epoch": 1.8439999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23050
 },
 {
 "epoch": 1.8479999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23100
 },
 {
 "epoch": 1.8519999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23150
 },
 {
 "epoch": 1.8559999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23200
 },
 {
 "epoch": 1.8599999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23250
 },
 {
 "epoch": 1.8639999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23300
 },
 {
 "epoch": 1.8679999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23350
 },
 {
 "epoch": 1.8719999999999999,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23400
 },
 {
 "epoch": 1.876,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23450
 },
 {
 "epoch": 1.88,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23500
 },
 {
 "epoch": 1.884,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23550
 },
 {
 "epoch": 1.888,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23600
 },
 {
 "epoch": 1.892,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23650
 },
 {
 "epoch": 1.896,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23700
 },
 {
 "epoch": 1.9,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23750
 },
 {
 "epoch": 1.904,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23800
 },
 {
 "epoch": 1.908,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23850
 },
 {
 "epoch": 1.912,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23900
 },
 {
 "epoch": 1.916,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 23950
 },
 {
 "epoch": 1.92,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24000
 },
 {
 "epoch": 1.924,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24050
 },
 {
 "epoch": 1.928,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24100
 },
 {
 "epoch": 1.932,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24150
 },
 {
 "epoch": 1.936,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24200
 },
 {
 "epoch": 1.94,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24250
 },
 {
 "epoch": 1.944,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24300
 },
 {
 "epoch": 1.948,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24350
 },
 {
 "epoch": 1.952,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24400
 },
 {
 "epoch": 1.956,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24450
 },
 {
 "epoch": 1.96,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24500
 },
 {
 "epoch": 1.964,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24550
 },
 {
 "epoch": 1.968,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24600
 },
 {
 "epoch": 1.972,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24650
 },
 {
 "epoch": 1.976,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24700
 },
 {
 "epoch": 1.98,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24750
 },
 {
 "epoch": 1.984,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24800
 },
 {
 "epoch": 1.988,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24850
 },
 {
 "epoch": 1.992,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24900
 },
 {
 "epoch": 1.996,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 24950
 },
 {
 "epoch": 2.0,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25000
 },
 {
 "epoch": 2.0,
 "eval_bleu": 0.12212197044710801,
 "eval_cer": 0.8160053321787701,
 "eval_f1": 0.35202235691264105,
 "eval_loss": NaN,
 "eval_meteor": 0.2636940499037806,
 "eval_runtime": 659.7772,
 "eval_samples_per_second": 7.578,
 "eval_steps_per_second": 0.238,
 "eval_wer": 0.9984162999487562,
 "step": 25000
 },
 {
 "epoch": 2.004,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25050
 },
 {
 "epoch": 2.008,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25100
 },
 {
 "epoch": 2.012,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25150
 },
 {
 "epoch": 2.016,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25200
 },
 {
 "epoch": 2.02,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25250
 },
 {
 "epoch": 2.024,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25300
 },
 {
 "epoch": 2.028,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25350
 },
 {
 "epoch": 2.032,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25400
 },
 {
 "epoch": 2.036,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25450
 },
 {
 "epoch": 2.04,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25500
 },
 {
 "epoch": 2.044,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25550
 },
 {
 "epoch": 2.048,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25600
 },
 {
 "epoch": 2.052,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25650
 },
 {
 "epoch": 2.056,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25700
 },
 {
 "epoch": 2.06,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25750
 },
 {
 "epoch": 2.064,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25800
 },
 {
 "epoch": 2.068,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25850
 },
 {
 "epoch": 2.072,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25900
 },
 {
 "epoch": 2.076,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 25950
 },
 {
 "epoch": 2.08,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26000
 },
 {
 "epoch": 2.084,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26050
 },
 {
 "epoch": 2.088,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26100
 },
 {
 "epoch": 2.092,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26150
 },
 {
 "epoch": 2.096,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26200
 },
 {
 "epoch": 2.1,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26250
 },
 {
 "epoch": 2.104,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26300
 },
 {
 "epoch": 2.108,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26350
 },
 {
 "epoch": 2.112,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26400
 },
 {
 "epoch": 2.116,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26450
 },
 {
 "epoch": 2.12,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26500
 },
 {
 "epoch": 2.124,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26550
 },
 {
 "epoch": 2.128,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26600
 },
 {
 "epoch": 2.132,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26650
 },
 {
 "epoch": 2.136,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26700
 },
 {
 "epoch": 2.14,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26750
 },
 {
 "epoch": 2.144,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26800
 },
 {
 "epoch": 2.148,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26850
 },
 {
 "epoch": 2.152,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26900
 },
 {
 "epoch": 2.156,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 26950
 },
 {
 "epoch": 2.16,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27000
 },
 {
 "epoch": 2.164,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27050
 },
 {
 "epoch": 2.168,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27100
 },
 {
 "epoch": 2.172,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27150
 },
 {
 "epoch": 2.176,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27200
 },
 {
 "epoch": 2.18,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27250
 },
 {
 "epoch": 2.184,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27300
 },
 {
 "epoch": 2.188,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27350
 },
 {
 "epoch": 2.192,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27400
 },
 {
 "epoch": 2.196,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27450
 },
 {
 "epoch": 2.2,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27500
 },
 {
 "epoch": 2.204,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27550
 },
 {
 "epoch": 2.208,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27600
 },
 {
 "epoch": 2.212,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27650
 },
 {
 "epoch": 2.216,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27700
 },
 {
 "epoch": 2.22,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27750
 },
 {
 "epoch": 2.224,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27800
 },
 {
 "epoch": 2.228,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27850
 },
 {
 "epoch": 2.232,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27900
 },
 {
 "epoch": 2.2359999999999998,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 27950
 },
 {
 "epoch": 2.24,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28000
 },
 {
 "epoch": 2.2439999999999998,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28050
 },
 {
 "epoch": 2.248,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28100
 },
 {
 "epoch": 2.252,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28150
 },
 {
 "epoch": 2.2560000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28200
 },
 {
 "epoch": 2.26,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28250
 },
 {
 "epoch": 2.2640000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28300
 },
 {
 "epoch": 2.268,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28350
 },
 {
 "epoch": 2.2720000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28400
 },
 {
 "epoch": 2.276,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28450
 },
 {
 "epoch": 2.2800000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28500
 },
 {
 "epoch": 2.284,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28550
 },
 {
 "epoch": 2.288,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28600
 },
 {
 "epoch": 2.292,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28650
 },
 {
 "epoch": 2.296,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28700
 },
 {
 "epoch": 2.3,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28750
 },
 {
 "epoch": 2.304,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28800
 },
 {
 "epoch": 2.308,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28850
 },
 {
 "epoch": 2.312,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28900
 },
 {
 "epoch": 2.316,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 28950
 },
 {
 "epoch": 2.32,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29000
 },
 {
 "epoch": 2.324,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29050
 },
 {
 "epoch": 2.328,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29100
 },
 {
 "epoch": 2.332,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29150
 },
 {
 "epoch": 2.336,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29200
 },
 {
 "epoch": 2.34,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29250
 },
 {
 "epoch": 2.344,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29300
 },
 {
 "epoch": 2.348,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29350
 },
 {
 "epoch": 2.352,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29400
 },
 {
 "epoch": 2.356,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29450
 },
 {
 "epoch": 2.36,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29500
 },
 {
 "epoch": 2.364,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29550
 },
 {
 "epoch": 2.368,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29600
 },
 {
 "epoch": 2.372,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29650
 },
 {
 "epoch": 2.376,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29700
 },
 {
 "epoch": 2.38,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29750
 },
 {
 "epoch": 2.384,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29800
 },
 {
 "epoch": 2.388,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29850
 },
 {
 "epoch": 2.392,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29900
 },
 {
 "epoch": 2.396,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 29950
 },
 {
 "epoch": 2.4,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30000
 },
 {
 "epoch": 2.404,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30050
 },
 {
 "epoch": 2.408,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30100
 },
 {
 "epoch": 2.412,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30150
 },
 {
 "epoch": 2.416,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30200
 },
 {
 "epoch": 2.42,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30250
 },
 {
 "epoch": 2.424,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30300
 },
 {
 "epoch": 2.428,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30350
 },
 {
 "epoch": 2.432,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30400
 },
 {
 "epoch": 2.436,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30450
 },
 {
 "epoch": 2.44,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30500
 },
 {
 "epoch": 2.444,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30550
 },
 {
 "epoch": 2.448,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30600
 },
 {
 "epoch": 2.452,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30650
 },
 {
 "epoch": 2.456,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30700
 },
 {
 "epoch": 2.46,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30750
 },
 {
 "epoch": 2.464,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30800
 },
 {
 "epoch": 2.468,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30850
 },
 {
 "epoch": 2.472,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30900
 },
 {
 "epoch": 2.476,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 30950
 },
 {
 "epoch": 2.48,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31000
 },
 {
 "epoch": 2.484,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31050
 },
 {
 "epoch": 2.488,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31100
 },
 {
 "epoch": 2.492,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31150
 },
 {
 "epoch": 2.496,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31200
 },
 {
 "epoch": 2.5,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31250
 },
 {
 "epoch": 2.504,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31300
 },
 {
 "epoch": 2.508,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31350
 },
 {
 "epoch": 2.512,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31400
 },
 {
 "epoch": 2.516,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31450
 },
 {
 "epoch": 2.52,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31500
 },
 {
 "epoch": 2.524,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31550
 },
 {
 "epoch": 2.528,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31600
 },
 {
 "epoch": 2.532,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31650
 },
 {
 "epoch": 2.536,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31700
 },
 {
 "epoch": 2.54,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31750
 },
 {
 "epoch": 2.544,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31800
 },
 {
 "epoch": 2.548,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31850
 },
 {
 "epoch": 2.552,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31900
 },
 {
 "epoch": 2.556,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 31950
 },
 {
 "epoch": 2.56,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32000
 },
 {
 "epoch": 2.564,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32050
 },
 {
 "epoch": 2.568,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32100
 },
 {
 "epoch": 2.572,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32150
 },
 {
 "epoch": 2.576,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32200
 },
 {
 "epoch": 2.58,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32250
 },
 {
 "epoch": 2.584,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32300
 },
 {
 "epoch": 2.588,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32350
 },
 {
 "epoch": 2.592,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32400
 },
 {
 "epoch": 2.596,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32450
 },
 {
 "epoch": 2.6,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32500
 },
 {
 "epoch": 2.604,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32550
 },
 {
 "epoch": 2.608,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32600
 },
 {
 "epoch": 2.612,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32650
 },
 {
 "epoch": 2.616,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32700
 },
 {
 "epoch": 2.62,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32750
 },
 {
 "epoch": 2.624,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32800
 },
 {
 "epoch": 2.628,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32850
 },
 {
 "epoch": 2.632,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32900
 },
 {
 "epoch": 2.636,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 32950
 },
 {
 "epoch": 2.64,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33000
 },
 {
 "epoch": 2.644,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33050
 },
 {
 "epoch": 2.648,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33100
 },
 {
 "epoch": 2.652,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33150
 },
 {
 "epoch": 2.656,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33200
 },
 {
 "epoch": 2.66,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33250
 },
 {
 "epoch": 2.664,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33300
 },
 {
 "epoch": 2.668,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33350
 },
 {
 "epoch": 2.672,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33400
 },
 {
 "epoch": 2.676,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33450
 },
 {
 "epoch": 2.68,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33500
 },
 {
 "epoch": 2.684,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33550
 },
 {
 "epoch": 2.6879999999999997,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33600
 },
 {
 "epoch": 2.692,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33650
 },
 {
 "epoch": 2.6959999999999997,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33700
 },
 {
 "epoch": 2.7,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33750
 },
 {
 "epoch": 2.7039999999999997,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33800
 },
 {
 "epoch": 2.708,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33850
 },
 {
 "epoch": 2.7119999999999997,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33900
 },
 {
 "epoch": 2.716,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 33950
 },
 {
 "epoch": 2.7199999999999998,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34000
 },
 {
 "epoch": 2.724,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34050
 },
 {
 "epoch": 2.7279999999999998,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34100
 },
 {
 "epoch": 2.732,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34150
 },
 {
 "epoch": 2.7359999999999998,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34200
 },
 {
 "epoch": 2.74,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34250
 },
 {
 "epoch": 2.7439999999999998,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34300
 },
 {
 "epoch": 2.748,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34350
 },
 {
 "epoch": 2.752,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34400
 },
 {
 "epoch": 2.7560000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34450
 },
 {
 "epoch": 2.76,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34500
 },
 {
 "epoch": 2.7640000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34550
 },
 {
 "epoch": 2.768,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34600
 },
 {
 "epoch": 2.7720000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34650
 },
 {
 "epoch": 2.776,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34700
 },
 {
 "epoch": 2.7800000000000002,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34750
 },
 {
 "epoch": 2.784,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34800
 },
 {
 "epoch": 2.7880000000000003,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34850
 },
 {
 "epoch": 2.792,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34900
 },
 {
 "epoch": 2.7960000000000003,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 34950
 },
 {
 "epoch": 2.8,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35000
 },
 {
 "epoch": 2.8040000000000003,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35050
 },
 {
 "epoch": 2.808,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35100
 },
 {
 "epoch": 2.8120000000000003,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35150
 },
 {
 "epoch": 2.816,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35200
 },
 {
 "epoch": 2.82,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35250
 },
 {
 "epoch": 2.824,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35300
 },
 {
 "epoch": 2.828,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35350
 },
 {
 "epoch": 2.832,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35400
 },
 {
 "epoch": 2.836,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35450
 },
 {
 "epoch": 2.84,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35500
 },
 {
 "epoch": 2.844,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35550
 },
 {
 "epoch": 2.848,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35600
 },
 {
 "epoch": 2.852,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35650
 },
 {
 "epoch": 2.856,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35700
 },
 {
 "epoch": 2.86,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35750
 },
 {
 "epoch": 2.864,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35800
 },
 {
 "epoch": 2.868,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35850
 },
 {
 "epoch": 2.872,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35900
 },
 {
 "epoch": 2.876,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 35950
 },
 {
 "epoch": 2.88,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36000
 },
 {
 "epoch": 2.884,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36050
 },
 {
 "epoch": 2.888,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36100
 },
 {
 "epoch": 2.892,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36150
 },
 {
 "epoch": 2.896,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36200
 },
 {
 "epoch": 2.9,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36250
 },
 {
 "epoch": 2.904,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36300
 },
 {
 "epoch": 2.908,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36350
 },
 {
 "epoch": 2.912,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36400
 },
 {
 "epoch": 2.916,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36450
 },
 {
 "epoch": 2.92,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36500
 },
 {
 "epoch": 2.924,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36550
 },
 {
 "epoch": 2.928,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36600
 },
 {
 "epoch": 2.932,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36650
 },
 {
 "epoch": 2.936,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36700
 },
 {
 "epoch": 2.94,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36750
 },
 {
 "epoch": 2.944,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36800
 },
 {
 "epoch": 2.948,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36850
 },
 {
 "epoch": 2.952,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36900
 },
 {
 "epoch": 2.956,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 36950
 },
 {
 "epoch": 2.96,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37000
 },
 {
 "epoch": 2.964,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37050
 },
 {
 "epoch": 2.968,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37100
 },
 {
 "epoch": 2.972,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37150
 },
 {
 "epoch": 2.976,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37200
 },
 {
 "epoch": 2.98,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37250
 },
 {
 "epoch": 2.984,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37300
 },
 {
 "epoch": 2.988,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37350
 },
 {
 "epoch": 2.992,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37400
 },
 {
 "epoch": 2.996,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37450
 },
 {
 "epoch": 3.0,
 "grad_norm": NaN,
 "learning_rate": 0.0005,
 "loss": 0.0,
 "step": 37500
 },
 {
 "epoch": 3.0,
 "eval_bleu": 0.12212197044710801,
 "eval_cer": 0.8160053321787701,
 "eval_f1": 0.35202235691264105,
 "eval_loss": NaN,
 "eval_meteor": 0.2636940499037806,
 "eval_runtime": 666.2051,
 "eval_samples_per_second": 7.505,
 "eval_steps_per_second": 0.236,
 "eval_wer": 0.9984162999487562,
 "step": 37500
 },
 {
 "epoch": 3.0,
 "step": 37500,
 "total_flos": 1.614293713378345e+18,
 "train_loss": 0.0,
 "train_runtime": 18366.5428,
 "train_samples_per_second": 65.336,
 "train_steps_per_second": 2.042
 }
 ],
 "logging_steps": 50,
 "max_steps": 37500,
 "num_input_tokens_seen": 0,
 "num_train_epochs": 3,
 "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": 1.614293713378345e+18,
 "train_batch_size": 32,
 "trial_name": null,
 "trial_params": null
}