metadata
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: wav2vec2-large-asr-th
results: []
wav2vec2-large-asr-th
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 5.1807
- Wer: 1.0
- Cer: 1.0000
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 200
- training_steps: 2000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 17.808 | 1.09 | 200 | 25.2775 | 1.0 | 1.0000 |
| 6.2017 | 2.19 | 400 | 7.3944 | 1.0 | 1.0000 |
| 4.858 | 3.28 | 600 | 5.1807 | 1.0 | 1.0000 |
| 3.9172 | 4.37 | 800 | 4.4228 | 1.0 | 1.0000 |
| 3.7413 | 5.46 | 1000 | 4.1509 | 1.0 | 1.0000 |
| 3.7496 | 6.56 | 1200 | 4.0403 | 1.0 | 1.0000 |
| 3.6742 | 7.65 | 1400 | 4.0565 | 1.0 | 1.0000 |
| 3.7246 | 8.74 | 1600 | 3.9468 | 1.0 | 1.0000 |
| 3.6329 | 9.84 | 1800 | 3.9271 | 1.0 | 1.0000 |
| 3.6562 | 10.93 | 2000 | 3.9091 | 1.0 | 1.0000 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu116
- Datasets 2.9.0
- Tokenizers 0.13.2