albert-base-v2-finetuned-TRAC-DS
This model is a fine-tuned version of albert-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.8271
- Accuracy: 0.6315
- Precision: 0.6206
- Recall: 0.6201
- F1: 0.6147
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: 5.919508251872584e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 43
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| 1.0373 | 1.0 | 612 | 1.1241 | 0.3627 | 0.5914 | 0.3618 | 0.2414 |
| 1.0617 | 2.0 | 1224 | 1.1039 | 0.3350 | 0.2781 | 0.3354 | 0.1740 |
| 0.9791 | 3.0 | 1836 | 0.8365 | 0.5989 | 0.6192 | 0.5887 | 0.5883 |
| 0.798 | 3.99 | 2448 | 0.8271 | 0.6315 | 0.6206 | 0.6201 | 0.6147 |
Framework versions
- Transformers 4.21.3
- Pytorch 1.12.1+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1