hubert-base-ls960-3kfoldfull-finetuned-bmd-V1-10-20250824_130250-LOSO-sec-out1
This model is a fine-tuned version of facebook/hubert-base-ls960 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.9744
- Accuracy: 0.4167
- F1: 0.3660
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: 2e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 1968
- gradient_accumulation_steps: 4
- 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: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|---|---|---|---|---|---|
| No log | 1.0 | 6 | 1.0949 | 0.3333 | 0.1667 |
| 1.0986 | 2.0 | 12 | 1.0883 | 0.4583 | 0.3657 |
| 1.0986 | 3.0 | 18 | 1.0816 | 0.4583 | 0.3913 |
| 1.0877 | 4.0 | 24 | 1.0726 | 0.4167 | 0.3620 |
| 1.0682 | 5.0 | 30 | 1.0597 | 0.4583 | 0.3913 |
| 1.0682 | 6.0 | 36 | 1.0307 | 0.5417 | 0.5097 |
| 1.0418 | 7.0 | 42 | 1.0012 | 0.5417 | 0.5507 |
| 1.0418 | 8.0 | 48 | 0.9725 | 0.5417 | 0.5442 |
| 1.0157 | 9.0 | 54 | 0.9712 | 0.4583 | 0.4237 |
| 0.9868 | 10.0 | 60 | 0.9744 | 0.4167 | 0.3660 |
Framework versions
- Transformers 4.55.2
- Pytorch 2.8.0+cu126
- Datasets 3.6.0
- Tokenizers 0.21.4