Whisper Large MN - Ankhbayasgalan Davaadorj
This model is a fine-tuned version of openai/whisper-large-v3 on the Common Voice 16.1 & FLEURS dataset. It achieves the following results on the evaluation set:
- Loss: 0.3245
- Wer: 37.0497
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: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 40
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.4691 | 0.3 | 100 | 0.5472 | 57.2191 |
| 0.3191 | 0.6 | 200 | 0.4417 | 49.0237 |
| 0.2677 | 0.9 | 300 | 0.3791 | 43.3530 |
| 0.1486 | 1.2 | 400 | 0.3560 | 40.1188 |
| 0.1387 | 1.5 | 500 | 0.3430 | 37.8912 |
| 0.1396 | 1.8 | 600 | 0.3245 | 37.0497 |
Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.2.0+cu121
- Datasets 2.17.0
- Tokenizers 0.15.2
- Downloads last month
- 0