WMACv2.2
This model is a fine-tuned version of 7ocho/WMACv2.1 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.2941
- Wer: 55.5454
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: 8e-06
- train_batch_size: 8
- eval_batch_size: 24
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 300
- training_steps: 5000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.6864 | 1.4753 | 1000 | 0.3015 | 56.8943 |
| 0.5709 | 2.9506 | 2000 | 0.2881 | 56.1667 |
| 0.4183 | 4.4251 | 3000 | 0.2890 | 55.8880 |
| 0.4151 | 5.9004 | 4000 | 0.2913 | 55.5706 |
| 0.3905 | 7.3749 | 5000 | 0.2941 | 55.5454 |
Framework versions
- Transformers 5.2.0
- Pytorch 2.10.0+cu128
- Datasets 3.6.0
- Tokenizers 0.22.2
- Downloads last month
- 93