HuBERT-base-F4-New
This model is a fine-tuned version of facebook/hubert-base-ls960 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.9168
- Accuracy: 0.8243
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: 3e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- 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: linear
- num_epochs: 9
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.1111 | 0.5714 | 400 | 0.9669 | 0.6179 |
| 0.3809 | 1.1429 | 800 | 0.8532 | 0.6907 |
| 0.675 | 1.7143 | 1200 | 0.7515 | 0.7286 |
| 0.9991 | 2.2857 | 1600 | 0.8572 | 0.7136 |
| 0.4101 | 2.8571 | 2000 | 0.6870 | 0.7800 |
| 0.2602 | 3.4286 | 2400 | 0.7185 | 0.7893 |
| 0.0872 | 4.0 | 2800 | 0.7470 | 0.7821 |
| 0.3991 | 4.5714 | 3200 | 0.6624 | 0.8107 |
| 0.1878 | 5.1429 | 3600 | 0.7700 | 0.8093 |
| 0.7543 | 5.7143 | 4000 | 0.8749 | 0.7950 |
| 0.5348 | 6.2857 | 4400 | 0.8467 | 0.8143 |
| 0.055 | 6.8571 | 4800 | 0.8527 | 0.8229 |
| 0.6014 | 7.4286 | 5200 | 0.9119 | 0.8150 |
| 0.4068 | 8.0 | 5600 | 0.8984 | 0.8250 |
| 0.0286 | 8.5714 | 6000 | 0.9168 | 0.8243 |
Framework versions
- Transformers 4.57.1
- Pytorch 2.8.0+cu126
- Datasets 4.0.0
- Tokenizers 0.22.1