mms-1b-all-nyagen-combined-train-drodat-gdro-1.24e-5-dat-0.0272-sd-0.00016-42
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: 36.2341
- Cer: 0.0829
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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- 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 | Cer |
|---|---|---|---|---|
| 395.4834 | 1.2566 | 500 | 390.8813 | 0.8561 |
| 271.0132 | 2.5132 | 1000 | 285.4378 | 0.9147 |
| 28.1943 | 3.7698 | 1500 | 47.4892 | 0.0986 |
| 22.0784 | 5.0252 | 2000 | 43.0185 | 0.0917 |
| 20.2081 | 6.2818 | 2500 | 40.8579 | 0.0877 |
| 17.4232 | 7.5384 | 3000 | 41.3706 | 0.0850 |
| 18.7554 | 8.7950 | 3500 | 38.6634 | 0.0852 |
| 16.1057 | 10.0503 | 4000 | 38.3902 | 0.0830 |
| 17.6642 | 11.3069 | 4500 | 36.2347 | 0.0829 |
| 16.2232 | 12.5635 | 5000 | 39.2986 | 0.0809 |
| 13.0751 | 13.8201 | 5500 | 39.5855 | 0.0828 |
| 15.8405 | 15.0755 | 6000 | 39.6599 | 0.0811 |
Framework versions
- Transformers 4.53.0
- Pytorch 2.10.0+cu128
- Datasets 4.5.0
- Tokenizers 0.21.4