metadata
library_name: transformers
language:
- vi
license: apache-2.0
base_model: facebook/wav2vec2-base-960h
tags:
- speech-to-text
- vietnamese
- uit-vimd
- generated_from_trainer
datasets:
- uit-vimd
metrics:
- wer
model-index:
- name: wav2vec2-base-960h_041109
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: UIT-ViMD
type: uit-vimd
metrics:
- name: Wer
type: wer
value: 0.999681224099458
wav2vec2-base-960h_041109
This model is a fine-tuned version of facebook/wav2vec2-base-960h on the UIT-ViMD dataset. It achieves the following results on the evaluation set:
- Loss: nan
- Wer: 0.9997
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
- gradient_accumulation_steps: 4
- 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_ratio: 0.1
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 13.6602 | 0.0639 | 30 | 13.3538 | 1.0 |
| 12.5275 | 0.1278 | 60 | 10.9900 | 0.9997 |
| 9.0758 | 0.1917 | 90 | 5.6756 | 0.9997 |
| 5.4378 | 0.2556 | 120 | 4.1876 | 0.9997 |
| 4.3313 | 0.3195 | 150 | 3.8115 | 0.9997 |
| 3.8956 | 0.3834 | 180 | 3.5275 | 0.9997 |
| 18.6702 | 0.4473 | 210 | nan | 0.9997 |
| 0.0 | 0.5112 | 240 | nan | 0.9997 |
| 0.0 | 0.5751 | 270 | nan | 0.9997 |
| 0.0 | 0.6390 | 300 | nan | 0.9997 |
| 0.0 | 0.7029 | 330 | nan | 0.9997 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.6.0+cu124
- Datasets 3.3.2
- Tokenizers 0.21.0