This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the common_voice_13_0 dataset. It achieves the following results on the evaluation set:

Loss: 1.8458
Wer: 0.8188

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: 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_steps: 500
num_epochs: 30
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
14.1341	2.5974	200	3.4817	1.0
3.468	5.1948	400	3.4658	1.0
3.4545	7.7922	600	3.5070	1.0
3.3847	10.3896	800	3.3519	1.0
2.5303	12.9870	1000	1.9302	1.0742
0.7082	15.5844	1200	1.6488	0.9012
0.3553	18.1818	1400	1.5767	0.8388
0.2409	20.7792	1600	1.6941	0.8354
0.1931	23.3766	1800	1.7294	0.8184
0.1659	25.9740	2000	1.8157	0.8304
0.1423	28.5714	2200	1.8458	0.8188

Framework versions

Transformers 4.48.3
Pytorch 2.5.1+cu124
Datasets 3.3.2
Tokenizers 0.21.0

qnho/wav2vec2-large-xlsr-vi

wav2vec2-large-xlsr-vi

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

qnho/
wav2vec2-large-xlsr-vi