metadata
license: apache-2.0
base_model: openai/whisper-large
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: whisper-medium-r22-e
results: []
whisper-medium-r22-e
This model is a fine-tuned version of openai/whisper-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2567
- Wer: 32.4317
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: 5
- training_steps: 150
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 4.2827 | 0.02 | 10 | 2.4594 | 30.6262 |
| 1.5383 | 0.04 | 20 | 0.9529 | 36.0561 |
| 0.5967 | 0.05 | 30 | 0.4230 | 34.7607 |
| 0.3559 | 0.07 | 40 | 0.3960 | 33.9352 |
| 0.314 | 0.09 | 50 | 0.3285 | 32.7270 |
| 0.3339 | 0.11 | 60 | 0.3362 | 33.3244 |
| 0.3148 | 0.13 | 70 | 0.2927 | 31.6464 |
| 0.3128 | 0.14 | 80 | 0.2896 | 32.5458 |
| 0.3136 | 0.16 | 90 | 0.2828 | 32.8613 |
| 0.272 | 0.18 | 100 | 0.2818 | 33.9419 |
| 0.1936 | 0.2 | 110 | 0.2702 | 30.9148 |
| 0.2541 | 0.22 | 120 | 0.2644 | 31.8209 |
| 0.2957 | 0.23 | 130 | 0.2614 | 31.6531 |
| 0.2867 | 0.25 | 140 | 0.2574 | 31.6397 |
| 0.2085 | 0.27 | 150 | 0.2567 | 32.4317 |
Framework versions
- Transformers 4.35.0.dev0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1