mst_hp2
This model is a fine-tuned version of Sjdan/mst_1 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2128
- Wer: 1.5888
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: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 7
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 6.1539 | 1.36 | 500 | 2.8040 | 1.5607 |
| 2.2802 | 2.72 | 1000 | 1.2387 | 2.3707 |
| 1.1976 | 4.09 | 1500 | 0.4206 | 1.8754 |
| 0.6861 | 5.45 | 2000 | 0.2622 | 1.6916 |
| 0.5078 | 6.81 | 2500 | 0.2128 | 1.5888 |
Framework versions
- Transformers 4.17.0
- Pytorch 1.13.1+cu116
- Datasets 1.18.3
- Tokenizers 0.13.2