learning_rate: 0.001
train_batch_size: 16
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: 25
num_epochs: 36
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
6.2634	1.1905	50	1.1763	0.9333	0.2775
1.2315	2.3810	100	0.8655	0.8603	0.2226
1.0631	3.5714	150	0.8182	0.8175	0.2164
0.9782	4.7619	200	0.7653	0.7873	0.1946
0.9386	5.9524	250	0.7318	0.8095	0.1954
0.8568	7.1429	300	0.7501	0.7603	0.1900
0.8183	8.3333	350	0.7080	0.7571	0.1868
0.7862	9.5238	400	0.6988	0.7540	0.1804
0.7581	10.7143	450	0.7104	0.7571	0.1780
0.7122	11.9048	500	0.6984	0.7921	0.1858
0.6844	13.0952	550	0.6938	0.7270	0.1754
0.6783	14.2857	600	0.6841	0.7206	0.1727
0.6277	15.4762	650	0.6793	0.7111	0.1648
0.5747	16.6667	700	0.6875	0.7206	0.1654
0.6067	17.8571	750	0.6637	0.7238	0.1703
0.5725	19.0476	800	0.6746	0.7063	0.1654

tadgeis/mms-1b-ckt-chuklang-only

mms-1b-ckt-chuklang-only