metadata
license: apache-2.0
base_model: facebook/wav2vec2-base
tags:
- generated_from_trainer
datasets:
- vivos
metrics:
- wer
model-index:
- name: working
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: vivos
type: vivos
config: default
split: None
args: default
metrics:
- name: Wer
type: wer
value: 0.1690242779130206
working
This model is a fine-tuned version of facebook/wav2vec2-base on the vivos dataset. It achieves the following results on the evaluation set:
- Loss: 0.2700
- Wer: 0.1690
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
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.2111 | 2.0 | 292 | 0.9069 | 0.5315 |
| 0.6195 | 4.0 | 584 | 0.4762 | 0.3436 |
| 0.4338 | 6.0 | 876 | 0.3627 | 0.2657 |
| 0.3471 | 8.0 | 1168 | 0.4132 | 0.2817 |
| 0.2897 | 10.0 | 1460 | 0.3098 | 0.2178 |
| 0.2471 | 12.0 | 1752 | 0.2761 | 0.1957 |
| 0.2184 | 14.0 | 2044 | 0.2694 | 0.1807 |
| 0.1927 | 16.0 | 2336 | 0.2705 | 0.1753 |
| 0.1754 | 18.0 | 2628 | 0.2692 | 0.1709 |
| 0.1622 | 20.0 | 2920 | 0.2700 | 0.1690 |
Framework versions
- Transformers 4.42.3
- Pytorch 2.1.2
- Datasets 2.20.0
- Tokenizers 0.19.1