|
--- |
|
license: apache-2.0 |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: wav2vec2-base-cynthia-timit |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# wav2vec2-base-cynthia-timit |
|
|
|
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.4607 |
|
- Wer: 0.1903 |
|
|
|
## 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: 0.0001 |
|
- train_batch_size: 32 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 1000 |
|
- num_epochs: 30 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:| |
|
| 0.0689 | 3.45 | 500 | 0.3667 | 0.1970 | |
|
| 0.0758 | 6.9 | 1000 | 0.4734 | 0.2154 | |
|
| 0.0581 | 10.34 | 1500 | 0.4375 | 0.2179 | |
|
| 0.0468 | 13.79 | 2000 | 0.4866 | 0.2112 | |
|
| 0.0408 | 17.24 | 2500 | 0.5170 | 0.2043 | |
|
| 0.0365 | 20.69 | 3000 | 0.5125 | 0.1987 | |
|
| 0.0323 | 24.14 | 3500 | 0.4754 | 0.2008 | |
|
| 0.0297 | 27.59 | 4000 | 0.4607 | 0.1903 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.11.3 |
|
- Pytorch 1.10.0+cu102 |
|
- Datasets 1.14.0 |
|
- Tokenizers 0.10.3 |
|
|
