<!-- 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. -->
wav2vec2-9
This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 3.0821
- Wer: 1.0
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: 0.005
- 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: 400
- num_epochs: 30
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
3.2803 | 1.56 | 200 | 3.1231 | 1.0 |
2.8809 | 3.12 | 400 | 3.0366 | 1.0 |
2.8761 | 4.69 | 600 | 3.1217 | 1.0 |
2.8641 | 6.25 | 800 | 3.0584 | 1.0 |
2.866 | 7.81 | 1000 | 3.0318 | 1.0 |
2.865 | 9.38 | 1200 | 3.0789 | 1.0 |
2.8642 | 10.94 | 1400 | 3.0560 | 1.0 |
2.8617 | 12.5 | 1600 | 2.9985 | 1.0 |
2.8573 | 14.06 | 1800 | 3.1928 | 1.0 |
2.8609 | 15.62 | 2000 | 3.0782 | 1.0 |
2.8605 | 17.19 | 2200 | 3.1244 | 1.0 |
2.8638 | 18.75 | 2400 | 3.0417 | 1.0 |
2.8578 | 20.31 | 2600 | 3.1586 | 1.0 |
2.8579 | 21.88 | 2800 | 3.0409 | 1.0 |
2.8569 | 23.44 | 3000 | 3.0537 | 1.0 |
2.8574 | 25.0 | 3200 | 3.0105 | 1.0 |
2.8536 | 26.56 | 3400 | 3.0901 | 1.0 |
2.8571 | 28.12 | 3600 | 3.0904 | 1.0 |
2.8532 | 29.69 | 3800 | 3.0821 | 1.0 |
Framework versions
- Transformers 4.19.2
- Pytorch 1.11.0+cu113
- Datasets 2.2.2
- Tokenizers 0.12.1