metadata
license: apache-2.0
base_model: openai/whisper-medium
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: whisper-medium-ft-10000_2
results: []
whisper-medium-ft-10000_2
This model is a fine-tuned version of openai/whisper-medium on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.2654
- Wer: 10.1641
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.3583 | 1.0 | 625 | 0.2361 | 32.4521 |
| 0.0928 | 2.0 | 1250 | 0.2281 | 10.0273 |
| 0.0368 | 3.0 | 1875 | 0.2280 | 10.8478 |
| 0.0123 | 4.0 | 2500 | 0.2604 | 10.6655 |
| 0.0039 | 5.0 | 3125 | 0.2571 | 10.3008 |
| 0.0011 | 6.0 | 3750 | 0.2654 | 10.1641 |
Framework versions
- Transformers 4.33.1
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.13.3