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
5.8208	0.1332	100	0.8827	0.8290
0.8914	0.2663	200	0.7362	0.6395
0.7704	0.3995	300	0.6462	0.5044
0.7695	0.5326	400	0.6256	0.4941
0.8153	0.6658	500	0.6210	0.4896
0.7625	0.7989	600	0.6180	0.4707
0.7062	0.9321	700	0.6003	0.4789
0.8157	1.0652	800	0.5861	0.4652
0.7057	1.1984	900	0.6160	0.4779
0.7667	1.3316	1000	0.5939	0.4770
0.6655	1.4647	1100	0.5949	0.4519
0.7098	1.5979	1200	0.5774	0.4535
0.6758	1.7310	1300	0.5673	0.4423
0.6529	1.8642	1400	0.5641	0.4336
0.6438	1.9973	1500	0.5889	0.4540
0.7389	2.1305	1600	0.5709	0.4368
0.6208	2.2636	1700	0.5773	0.4471
0.5822	2.3968	1800	0.5520	0.4379
0.6487	2.5300	1900	0.5496	0.4335
0.6113	2.6631	2000	0.5410	0.4367
0.6428	2.7963	2100	0.5401	0.4289
0.7135	2.9294	2200	0.5404	0.4224
0.6212	3.0626	2300	0.5464	0.4241
0.5827	3.1957	2400	0.5326	0.4225
0.5823	3.3289	2500	0.5285	0.4274
0.5544	3.4621	2600	0.5334	0.4346
0.6308	3.5952	2700	0.5204	0.4217
0.6724	3.7284	2800	0.5235	0.4226
0.5923	3.8615	2900	0.5275	0.4157
0.6517	3.9947	3000	0.5328	0.4212
0.6033	4.1278	3100	0.5208	0.4140
0.5838	4.2610	3200	0.5178	0.4084
0.5495	4.3941	3300	0.5242	0.4152
0.5191	4.5273	3400	0.5197	0.4144
0.5841	4.6605	3500	0.5161	0.4119
0.5908	4.7936	3600	0.5118	0.4063
0.6478	4.9268	3700	0.5121	0.4057

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

wav2vec2-large-mms-1b-aft-ukv