learning_rate: 1e-05
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: 500
training_steps: 4000
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
0.704	1.2207	200	0.6241
0.6022	2.4414	400	0.5469
0.58	3.6621	600	0.5264
0.5628	4.8828	800	0.5140
0.5355	6.0981	1000	0.5057
0.5414	7.3188	1200	0.5020
0.5374	8.5395	1400	0.4974
0.5327	9.7602	1600	0.4939
0.5329	10.9808	1800	0.4931
0.5258	12.1962	2000	0.4914
0.5263	13.4169	2200	0.4897
0.5232	14.6375	2400	0.4890
0.5199	15.8582	2600	0.4878
0.5067	17.0736	2800	0.4879
0.5215	18.2943	3000	0.4860
0.5137	19.5149	3200	0.4861
0.5146	20.7356	3400	0.4864
0.5158	21.9563	3600	0.4844
0.518	23.1716	3800	0.4849
0.5213	24.3923	4000	0.4847

mvsamsonov/speecht5_finetuned_voxpopuli_it

speecht5_finetuned_voxpopuli_it