xlm-roberta-targin-final
This model is a fine-tuned version of xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.8172
- Accuracy: 0.6873
- Precision: 0.6494
- Recall: 0.6422
- F1: 0.6450
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 296 | 0.6065 | 0.6873 | 0.6537 | 0.5833 | 0.5748 |
| 0.597 | 2.0 | 592 | 0.5822 | 0.7015 | 0.6652 | 0.6279 | 0.6332 |
| 0.597 | 3.0 | 888 | 0.5704 | 0.7015 | 0.6654 | 0.6551 | 0.6589 |
| 0.5156 | 4.0 | 1184 | 0.6393 | 0.7044 | 0.6684 | 0.6552 | 0.6597 |
| 0.5156 | 5.0 | 1480 | 0.5924 | 0.7082 | 0.6752 | 0.6720 | 0.6735 |
| 0.4479 | 6.0 | 1776 | 0.7029 | 0.7006 | 0.6629 | 0.6351 | 0.6408 |
| 0.3783 | 7.0 | 2072 | 0.6963 | 0.7072 | 0.6715 | 0.6554 | 0.6606 |
| 0.3783 | 8.0 | 2368 | 0.7636 | 0.6987 | 0.6627 | 0.6549 | 0.6579 |
| 0.3253 | 9.0 | 2664 | 0.7804 | 0.6901 | 0.6549 | 0.6523 | 0.6535 |
| 0.3253 | 10.0 | 2960 | 0.8172 | 0.6873 | 0.6494 | 0.6422 | 0.6450 |
Framework versions
- Transformers 4.24.0.dev0
- Pytorch 1.11.0+cu102
- Datasets 2.6.1
- Tokenizers 0.13.1