metadata
library_name: transformers
license: cc-by-nc-4.0
base_model: nguyenvulebinh/wav2vec2-large-vi-vlsp2020
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: wp-v1
results: []
wp-v1
This model is a fine-tuned version of nguyenvulebinh/wav2vec2-large-vi-vlsp2020 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.7195
- Wer: 0.9789
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: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- total_train_batch_size: 32
- total_eval_batch_size: 32
- optimizer: Use 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: 40
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 2.7954 | 1.0 | 129 | 2.7902 | 0.9801 |
| 2.7317 | 2.0 | 258 | 2.7682 | 0.9888 |
| 2.7232 | 3.0 | 387 | 2.8586 | 0.9733 |
| 2.699 | 4.0 | 516 | 2.7598 | 0.9756 |
| 2.6634 | 5.0 | 645 | 2.7744 | 0.9833 |
| 2.666 | 6.0 | 774 | 2.7401 | 0.9746 |
| 2.6142 | 7.0 | 903 | 2.7946 | 1.0044 |
| 2.6361 | 8.0 | 1032 | 2.7195 | 0.9789 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.7.1+cu118
- Datasets 3.6.0
- Tokenizers 0.21.2