learning_rate: 0.001
train_batch_size: 2
eval_batch_size: 8
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 5
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer
9.46	0.2392	100	0.9074	0.6628
0.6801	0.4785	200	0.8081	0.6265
0.6238	0.7177	300	0.7703	0.6106
0.6027	0.9569	400	0.7490	0.5896
0.5722	1.1962	500	0.7583	0.6051
0.5235	1.4354	600	0.7694	0.5841
0.5746	1.6746	700	0.7194	0.6011
0.5639	1.9139	800	0.7501	0.5704
0.5807	2.1531	900	0.7525	0.5730
0.4724	2.3923	1000	0.7501	0.5698
0.5164	2.6316	1100	0.7402	0.5948
0.5014	2.8708	1200	0.7531	0.5984
0.4917	3.1100	1300	0.7782	0.5500
0.5206	3.3493	1400	0.7267	0.5574
0.4475	3.5885	1500	0.7055	0.5501
0.4856	3.8278	1600	0.7066	0.5651
0.4432	4.0670	1700	0.7223	0.5625
0.4527	4.3062	1800	0.7014	0.5448
0.4599	4.5455	1900	0.7024	0.5460
0.4216	4.7847	2000	0.7062	0.5488

robertp408/wav2vec2-large-mms-1b-aft-tob

wav2vec2-large-mms-1b-aft-tob