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
num_epochs: 25
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
0.6317	1.0	68	0.4699
0.4599	2.0	136	0.3850
0.4199	3.0	204	0.3784
0.4039	4.0	272	0.3705
0.3937	5.0	340	0.3860
0.3865	6.0	408	0.3515
0.3779	7.0	476	0.3541
0.3808	8.0	544	0.3514
0.3694	9.0	612	0.3464
0.3654	10.0	680	0.3380
0.3598	11.0	748	0.3399
0.358	12.0	816	0.3404
0.352	13.0	884	0.3315
0.3518	14.0	952	0.3290
0.3488	15.0	1020	0.3353
0.3464	16.0	1088	0.3252
0.3432	17.0	1156	0.3209
0.3398	18.0	1224	0.3189
0.3373	19.0	1292	0.3196
0.335	20.0	1360	0.3228
0.3353	21.0	1428	0.3179
0.3322	22.0	1496	0.3145
0.329	23.0	1564	0.3124
0.3299	24.0	1632	0.3126
0.3285	25.0	1700	0.3115

kavinda123321/speecht5_kavinda_new1

speecht5_kavinda_new1