|
---
|
|
license: apache-2.0
|
|
base_model: facebook/hubert-base-ls960
|
|
tags:
|
|
- generated_from_trainer
|
|
metrics:
|
|
- accuracy
|
|
- precision
|
|
- recall
|
|
- f1
|
|
model-index:
|
|
- name: hubert-classifier-aug-fold-4
|
|
results: []
|
|
---
|
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
should probably proofread and complete it, then remove this comment. -->
|
|
|
|
# hubert-classifier-aug-fold-4
|
|
|
|
This model is a fine-tuned version of [facebook/hubert-base-ls960](https://huggingface.co/facebook/hubert-base-ls960) on an unknown dataset.
|
|
It achieves the following results on the evaluation set:
|
|
- Loss: 0.6366
|
|
- Accuracy: 0.8369
|
|
- Precision: 0.8588
|
|
- Recall: 0.8369
|
|
- F1: 0.8354
|
|
- Binary: 0.8865
|
|
|
|
## 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: 0.0001
|
|
- train_batch_size: 32
|
|
- eval_batch_size: 32
|
|
- seed: 42
|
|
- gradient_accumulation_steps: 4
|
|
- total_train_batch_size: 128
|
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
|
- lr_scheduler_type: linear
|
|
- lr_scheduler_warmup_steps: 500
|
|
- num_epochs: 100
|
|
- mixed_precision_training: Native AMP
|
|
|
|
### Training results
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | Binary |
|
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|:------:|
|
|
| No log | 0.24 | 50 | 4.4216 | 0.0247 | 0.0070 | 0.0247 | 0.0085 | 0.1557 |
|
|
| No log | 0.48 | 100 | 4.3485 | 0.0652 | 0.0204 | 0.0652 | 0.0249 | 0.2965 |
|
|
| No log | 0.72 | 150 | 4.0225 | 0.0577 | 0.0118 | 0.0577 | 0.0151 | 0.3349 |
|
|
| No log | 0.96 | 200 | 3.7163 | 0.0780 | 0.0180 | 0.0780 | 0.0245 | 0.3495 |
|
|
| 4.2635 | 1.2 | 250 | 3.4497 | 0.1454 | 0.0633 | 0.1454 | 0.0699 | 0.3975 |
|
|
| 4.2635 | 1.44 | 300 | 3.2114 | 0.1912 | 0.0913 | 0.1912 | 0.0994 | 0.4306 |
|
|
| 4.2635 | 1.68 | 350 | 2.9633 | 0.2399 | 0.1697 | 0.2399 | 0.1533 | 0.4645 |
|
|
| 4.2635 | 1.92 | 400 | 2.7158 | 0.2684 | 0.2035 | 0.2684 | 0.1801 | 0.4852 |
|
|
| 3.2875 | 2.16 | 450 | 2.3488 | 0.4033 | 0.3287 | 0.4033 | 0.3141 | 0.5813 |
|
|
| 3.2875 | 2.4 | 500 | 2.2100 | 0.4070 | 0.3778 | 0.4070 | 0.3362 | 0.5768 |
|
|
| 3.2875 | 2.63 | 550 | 1.8435 | 0.5232 | 0.4362 | 0.5232 | 0.4500 | 0.6669 |
|
|
| 3.2875 | 2.87 | 600 | 1.6225 | 0.5847 | 0.5594 | 0.5847 | 0.5397 | 0.7085 |
|
|
| 2.2923 | 3.11 | 650 | 1.4552 | 0.6027 | 0.5886 | 0.6027 | 0.5546 | 0.7217 |
|
|
| 2.2923 | 3.35 | 700 | 1.2618 | 0.6529 | 0.6733 | 0.6529 | 0.6201 | 0.7570 |
|
|
| 2.2923 | 3.59 | 750 | 1.1393 | 0.6904 | 0.7135 | 0.6904 | 0.6672 | 0.7822 |
|
|
| 2.2923 | 3.83 | 800 | 1.1629 | 0.6904 | 0.7141 | 0.6904 | 0.6713 | 0.7825 |
|
|
| 1.6069 | 4.07 | 850 | 1.0495 | 0.7264 | 0.7485 | 0.7264 | 0.7120 | 0.8105 |
|
|
| 1.6069 | 4.31 | 900 | 0.9086 | 0.7571 | 0.7644 | 0.7571 | 0.7409 | 0.8304 |
|
|
| 1.6069 | 4.55 | 950 | 0.8281 | 0.7669 | 0.7855 | 0.7669 | 0.7580 | 0.8379 |
|
|
| 1.6069 | 4.79 | 1000 | 0.8820 | 0.7534 | 0.7784 | 0.7534 | 0.7412 | 0.8272 |
|
|
| 1.2646 | 5.03 | 1050 | 0.7262 | 0.7886 | 0.7982 | 0.7886 | 0.7817 | 0.8525 |
|
|
| 1.2646 | 5.27 | 1100 | 0.7475 | 0.7954 | 0.8113 | 0.7954 | 0.7923 | 0.8577 |
|
|
| 1.2646 | 5.51 | 1150 | 0.7535 | 0.7999 | 0.8152 | 0.7999 | 0.7960 | 0.8600 |
|
|
| 1.2646 | 5.75 | 1200 | 0.7443 | 0.8021 | 0.8159 | 0.8021 | 0.7997 | 0.8624 |
|
|
| 1.2646 | 5.99 | 1250 | 0.5991 | 0.8351 | 0.8471 | 0.8351 | 0.8301 | 0.8854 |
|
|
| 1.0169 | 6.23 | 1300 | 0.6740 | 0.8193 | 0.8387 | 0.8193 | 0.8142 | 0.8741 |
|
|
| 1.0169 | 6.47 | 1350 | 0.6129 | 0.8358 | 0.8535 | 0.8358 | 0.8326 | 0.8843 |
|
|
| 1.0169 | 6.71 | 1400 | 0.6051 | 0.8358 | 0.8486 | 0.8358 | 0.8324 | 0.8851 |
|
|
| 1.0169 | 6.95 | 1450 | 0.6603 | 0.8216 | 0.8388 | 0.8216 | 0.8176 | 0.8748 |
|
|
| 0.89 | 7.19 | 1500 | 0.6105 | 0.8388 | 0.8499 | 0.8388 | 0.8363 | 0.8871 |
|
|
| 0.89 | 7.43 | 1550 | 0.6328 | 0.8298 | 0.8433 | 0.8298 | 0.8265 | 0.8808 |
|
|
| 0.89 | 7.66 | 1600 | 0.7041 | 0.8246 | 0.8405 | 0.8246 | 0.8203 | 0.8765 |
|
|
| 0.89 | 7.9 | 1650 | 0.6693 | 0.8268 | 0.8453 | 0.8268 | 0.8257 | 0.8788 |
|
|
| 0.7796 | 8.14 | 1700 | 0.6774 | 0.8223 | 0.8399 | 0.8223 | 0.8198 | 0.8768 |
|
|
| 0.7796 | 8.38 | 1750 | 0.6875 | 0.8201 | 0.8334 | 0.8201 | 0.8181 | 0.8755 |
|
|
| 0.7796 | 8.62 | 1800 | 0.6783 | 0.8253 | 0.8409 | 0.8253 | 0.8225 | 0.8795 |
|
|
| 0.7796 | 8.86 | 1850 | 0.6815 | 0.8261 | 0.8431 | 0.8261 | 0.8241 | 0.8798 |
|
|
|
|
|
|
### Framework versions
|
|
|
|
- Transformers 4.38.2
|
|
- Pytorch 2.3.0
|
|
- Datasets 2.19.1
|
|
- Tokenizers 0.15.1
|
|
|
