learning_rate: 3e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 64
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 400
num_epochs: 20
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Wer
0.7748	1.0	96	0.3779	0.0074	0.1654
0.2836	2.0	192	0.4260	0.0074	0.1973
0.2354	3.0	288	0.3930	0.0074	0.1524
0.2097	4.0	384	0.4422	0.0074	0.2027
0.1942	5.0	480	0.4359	0.0074	0.2456
0.1625	6.0	576	0.4169	0.0074	0.1910
0.1318	7.0	672	0.4465	0.0074	0.1603
0.1106	8.0	768	0.4704	0.0074	0.1860
0.0946	9.0	864	0.4792	0.0074	0.1685
0.0801	10.0	960	0.4609	0.0074	0.1617
0.0711	11.0	1056	0.4906	0.0074	0.1621
0.0569	12.0	1152	0.5165	0.0074	0.1690
0.0471	13.0	1248	0.5087	0.0074	0.1492
0.0374	14.0	1344	0.5130	0.0074	0.1348
0.0337	15.0	1440	0.5367	0.0074	0.1394
0.0256	16.0	1536	0.5285	0.0074	0.1350
0.0241	17.0	1632	0.5178	0.0074	0.1370
0.0143	18.0	1728	0.5489	0.0074	0.1283
0.0114	19.0	1824	0.5605	0.0074	0.1349
0.0066	19.7927	1900	0.5599	0.0074	0.1343

dhaval0108/ap-gQPvCZE1ClYjiHJYVrmkTQ

ap-gQPvCZE1ClYjiHJYVrmkTQ