|
--- |
|
license: apache-2.0 |
|
base_model: facebook/wav2vec2-base-960h |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- wer |
|
model-index: |
|
- name: wav2vecvanilla_ctc_zero_infinity |
|
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. --> |
|
|
|
# wav2vecvanilla_ctc_zero_infinity |
|
|
|
This model is a fine-tuned version of [facebook/wav2vec2-base-960h](https://huggingface.co./facebook/wav2vec2-base-960h) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.8214 |
|
- Wer: 0.3168 |
|
|
|
## 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: 4 |
|
- 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: 500 |
|
- num_epochs: 7 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:| |
|
| 1.4684 | 0.43 | 100 | 1.0567 | 0.4018 | |
|
| 1.2572 | 0.85 | 200 | 0.9726 | 0.3706 | |
|
| 1.139 | 1.28 | 300 | 0.9748 | 0.3602 | |
|
| 1.0956 | 1.71 | 400 | 0.9989 | 0.3619 | |
|
| 1.0891 | 2.14 | 500 | 0.9133 | 0.3606 | |
|
| 1.063 | 2.56 | 600 | 0.9272 | 0.3548 | |
|
| 1.0339 | 2.99 | 700 | 1.0183 | 0.3444 | |
|
| 0.9709 | 3.42 | 800 | 0.8244 | 0.3488 | |
|
| 0.958 | 3.85 | 900 | 0.8335 | 0.3410 | |
|
| 0.8954 | 4.27 | 1000 | 0.8641 | 0.3336 | |
|
| 0.8735 | 4.7 | 1100 | 0.8671 | 0.3306 | |
|
| 0.8411 | 5.13 | 1200 | 0.8373 | 0.3281 | |
|
| 0.805 | 5.56 | 1300 | 0.8197 | 0.3198 | |
|
| 0.8452 | 5.98 | 1400 | 0.8343 | 0.3158 | |
|
| 0.8078 | 6.41 | 1500 | 0.8392 | 0.3165 | |
|
| 0.7946 | 6.84 | 1600 | 0.8214 | 0.3168 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.38.2 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
|
|
