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: 3.3264
- Wer: 1.0
- Cer: 0.9960
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: 5e-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: 500
- training_steps: 2000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 9.7192 | 1.09 | 200 | 10.1504 | 1.0 | 1.0000 |
| 3.6967 | 2.19 | 400 | 3.9101 | 1.0 | 1.0000 |
| 3.5978 | 3.28 | 600 | 3.8292 | 1.0 | 1.0000 |
| 3.5324 | 4.37 | 800 | 3.6538 | 1.0 | 1.0000 |
| 3.529 | 5.46 | 1000 | 3.5816 | 1.0 | 1.0000 |
| 3.5145 | 6.56 | 1200 | 3.5487 | 1.0 | 1.0000 |
| 3.4544 | 7.65 | 1400 | 3.4912 | 1.0 | 0.9999 |
| 3.3666 | 8.74 | 1600 | 3.3998 | 1.0 | 0.9974 |
| 3.2963 | 9.84 | 1800 | 3.3264 | 1.0 | 0.9960 |
| 3.2361 | 10.93 | 2000 | 3.2402 | 1.0001 | 0.9950 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu116
- Datasets 2.9.0
- Tokenizers 0.13.2