learning_rate: 0.0001
train_batch_size: 4
eval_batch_size: 2
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 32
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: 100
training_steps: 1200
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
0.6698	0.4427	100	0.6503
0.6695	0.8854	200	0.6329
0.6585	1.3320	300	0.6239
0.653	1.7748	400	0.6197
0.6565	2.2214	500	0.6075
0.6455	2.6641	600	0.6101
0.6469	3.1107	700	0.5997
0.6295	3.5534	800	0.5985
0.622	3.9961	900	0.5946
0.6203	4.4427	1000	0.5928
0.6155	4.8854	1100	0.5902
0.607	5.3320	1200	0.5849

Kaizu07/speecht5_finetuned_Kaizu_bn_v1

speecht5_finetuned_Kaizu_bn_v1