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

Loss: 0.3514
Wer: 0.4838

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0003
train_batch_size: 16
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
5.512	1.32	400	3.2207	1.0
3.1562	2.65	800	3.0166	1.0
1.5211	3.97	1200	0.7134	0.8275
0.6724	5.3	1600	0.4713	0.6402
0.4693	6.62	2000	0.3904	0.5668
0.3693	7.95	2400	0.3609	0.5121
0.3004	9.27	2800	0.3514	0.4838

Framework versions

Transformers 4.11.3
Pytorch 1.10.0+cu111
Datasets 1.14.0
Tokenizers 0.10.3

emre/wav2vec2-xls-r-300m-Russian-small

wav2vec2-xls-r-300m-Russian-small

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

emre/
wav2vec2-xls-r-300m-Russian-small