metadata
license: apache-2.0
library_name: peft
tags:
- generated_from_trainer
base_model: openai/whisper-small
datasets:
- librispeech
model-index:
- name: Whisper Small English 1h
results: []
Whisper Small English 1h
This model is a fine-tuned version of openai/whisper-small on the librispeech dataset. It achieves the following results on the evaluation set:
- Loss: 1.9444
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-06
- train_batch_size: 8
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 0.9744 | 19 | 2.1407 |
| 1.3439 | 2.0 | 39 | 2.1284 |
| 1.3611 | 2.9744 | 58 | 2.1101 |
| 1.3218 | 4.0 | 78 | 2.0900 |
| 1.3218 | 4.9744 | 97 | 2.0717 |
| 1.3312 | 6.0 | 117 | 2.0546 |
| 1.3108 | 6.9744 | 136 | 2.0389 |
| 1.2979 | 8.0 | 156 | 2.0235 |
| 1.2662 | 8.9744 | 175 | 2.0105 |
| 1.2662 | 10.0 | 195 | 1.9979 |
| 1.2706 | 10.9744 | 214 | 1.9876 |
| 1.2673 | 12.0 | 234 | 1.9781 |
| 1.2302 | 12.9744 | 253 | 1.9695 |
| 1.2302 | 14.0 | 273 | 1.9620 |
| 1.2478 | 14.9744 | 292 | 1.9564 |
| 1.2536 | 16.0 | 312 | 1.9516 |
| 1.2125 | 16.9744 | 331 | 1.9480 |
| 1.2406 | 18.0 | 351 | 1.9457 |
| 1.2406 | 18.9744 | 370 | 1.9445 |
| 1.218 | 19.4872 | 380 | 1.9444 |
Framework versions
- PEFT 0.10.0
- Transformers 4.40.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1