You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

wav2vec2-xls-r-ewe-100-hours

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on an unknown dataset. It achieves the following results on the evaluation set:

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

learning_rate: 0.0003
train_batch_size: 32
eval_batch_size: 16
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 64
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_ratio: 0.1
num_epochs: 50

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
13.6452	1.7007	500	2.9602	1.0	1.0
3.0135	3.4014	1000	0.4361	0.3881	0.1111
1.0094	5.1020	1500	0.3739	0.3388	0.0976
0.8932	6.8027	2000	0.3530	0.3250	0.0933
0.8274	8.5034	2500	0.3506	0.3237	0.0933
0.7824	10.2041	3000	0.3480	0.3183	0.0921
0.7387	11.9048	3500	0.3651	0.3338	0.1028
0.6923	13.6054	4000	0.3551	0.3128	0.0918
0.6584	15.3061	4500	0.3511	0.3171	0.0916
0.6181	17.0068	5000	0.3572	0.3206	0.0924
0.5805	18.7075	5500	0.3604	0.3265	0.0958
0.5508	20.4082	6000	0.3734	0.3172	0.0929
0.5195	22.1088	6500	0.3913	0.3182	0.0932
0.4902	23.8095	7000	0.3914	0.3291	0.0945
0.4585	25.5102	7500	0.4277	0.3279	0.0934
0.4222	27.2109	8000	0.4336	0.3346	0.0964
0.3911	28.9116	8500	0.4419	0.3252	0.0938
0.3627	30.6122	9000	0.4687	0.3284	0.0951
0.3441	32.3129	9500	0.5102	0.3378	0.0962
0.3176	34.0136	10000	0.5415	0.3352	0.0950
0.2875	35.7143	10500	0.5497	0.3472	0.0978
0.2665	37.4150	11000	0.5724	0.3425	0.0975
0.2507	39.1156	11500	0.6267	0.3415	0.0968
0.2312	40.8163	12000	0.6423	0.3369	0.0960
0.2118	42.5170	12500	0.6688	0.3413	0.0964
0.2015	44.2177	13000	0.6806	0.3472	0.0970
0.1917	45.9184	13500	0.6875	0.3437	0.0970
0.1833	47.6190	14000	0.7128	0.3368	0.0951
0.1756	49.3197	14500	0.7355	0.3374	0.0950