wav2vec2-base-CALLCENTER
This model is a fine-tuned version of Niccogrillo/wav2vec2-base-CALLCENTER on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4605
- Wer: 0.2139
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: 8
- 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: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
6.9258 | 0.41 | 500 | 2.6959 | 0.9996 |
0.7059 | 0.82 | 1000 | 0.6081 | 0.2841 |
0.1633 | 1.22 | 1500 | 0.5037 | 0.2302 |
0.2361 | 1.63 | 2000 | 0.4207 | 0.2206 |
0.2289 | 2.04 | 2500 | 0.4433 | 0.2184 |
0.1794 | 2.45 | 3000 | 0.4648 | 0.2172 |
0.1827 | 2.86 | 3500 | 0.4592 | 0.2151 |
0.1844 | 3.27 | 4000 | 0.4507 | 0.2143 |
0.1723 | 3.67 | 4500 | 0.4561 | 0.2143 |
0.1762 | 4.08 | 5000 | 0.4633 | 0.2143 |
0.1762 | 4.49 | 5500 | 0.4610 | 0.2140 |
0.1552 | 4.9 | 6000 | 0.4605 | 0.2139 |
Framework versions
- Transformers 4.30.0
- Pytorch 2.2.1+cu121
- Datasets 2.19.0
- Tokenizers 0.13.3
- Downloads last month
- 30