<!-- 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. -->
asr-tokenizor
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: 0.4323
- Wer: 0.2814
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
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
3.4838 | 3.45 | 500 | 1.1430 | 0.8192 |
0.5107 | 6.9 | 1000 | 0.4258 | 0.3776 |
0.198 | 10.34 | 1500 | 0.3897 | 0.3188 |
0.1173 | 13.79 | 2000 | 0.3641 | 0.2982 |
0.0824 | 17.24 | 2500 | 0.4068 | 0.2956 |
0.0614 | 20.69 | 3000 | 0.4007 | 0.2886 |
0.0503 | 24.14 | 3500 | 0.4141 | 0.2837 |
0.0396 | 27.59 | 4000 | 0.4323 | 0.2814 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.0.1+cu118
- Datasets 1.18.3
- Tokenizers 0.13.3