mms-1b-all-nyagen-combined-ft-62
This model is a fine-tuned version of facebook/mms-1b-all on the NYAGEN - NYA dataset. It achieves the following results on the evaluation set:
- Loss: 0.3598
- Wer: 0.2645
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: 0.0003
- train_batch_size: 8
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 30.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 4.3846 | 2.5126 | 500 | 4.0447 | 0.9993 |
| 3.0421 | 5.0251 | 1000 | 3.0433 | 0.9932 |
| 0.2856 | 7.5377 | 1500 | 0.4490 | 0.3511 |
| 0.2413 | 10.0503 | 2000 | 0.3894 | 0.3138 |
| 0.2197 | 12.5628 | 2500 | 0.3717 | 0.2928 |
| 0.2063 | 15.0754 | 3000 | 0.3672 | 0.2824 |
| 0.1861 | 17.5879 | 3500 | 0.3638 | 0.2735 |
| 0.1805 | 20.1005 | 4000 | 0.3598 | 0.2643 |
| 0.1735 | 22.6131 | 4500 | 0.3618 | 0.2652 |
| 0.1741 | 25.1256 | 5000 | 0.3633 | 0.2608 |
| 0.1627 | 27.6382 | 5500 | 0.3660 | 0.2621 |
Framework versions
- Transformers 4.53.0
- Pytorch 2.10.0+cu128
- Datasets 4.5.0
- Tokenizers 0.21.4