mim-asr-interviews-full-small-no-augmented
This model is a fine-tuned version of openai/whisper-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.8421
- Wer: 75.7006
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: 32
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- training_steps: 10000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0003 | 52.63 | 1000 | 0.6102 | 55.2783 |
| 0.0001 | 105.26 | 2000 | 0.6581 | 76.8522 |
| 0.0 | 157.89 | 3000 | 0.6936 | 62.1113 |
| 0.0 | 210.53 | 4000 | 0.7215 | 64.2994 |
| 0.0 | 263.16 | 5000 | 0.7532 | 65.4127 |
| 0.0 | 315.79 | 6000 | 0.7775 | 65.1440 |
| 0.0 | 368.42 | 7000 | 0.8028 | 65.0288 |
| 0.0 | 421.05 | 8000 | 0.8207 | 69.8656 |
| 0.0 | 473.68 | 9000 | 0.8364 | 70.2495 |
| 0.0 | 526.32 | 10000 | 0.8421 | 75.7006 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu117
- Datasets 2.11.0
- Tokenizers 0.13.3