twitter-data-distilbert-base-uncased-sentiment-finetuned-memes-v2
This model is a fine-tuned version of jayantapaul888/twitter-data-distilbert-base-uncased-sentiment-finetuned-memes on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.3345
- Accuracy: 0.6527
- Precision: 0.6528
- Recall: 0.6527
- F1: 0.6527
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| 0.859 | 1.0 | 663 | 0.7251 | 0.6635 | 0.6722 | 0.6635 | 0.6620 |
| 0.6619 | 2.0 | 1326 | 0.7307 | 0.6723 | 0.6737 | 0.6723 | 0.6720 |
| 0.5684 | 3.0 | 1989 | 0.8724 | 0.6629 | 0.6645 | 0.6629 | 0.6626 |
| 0.3338 | 4.0 | 2652 | 1.0549 | 0.6551 | 0.6571 | 0.6551 | 0.6548 |
| 0.2512 | 5.0 | 3315 | 1.2275 | 0.6554 | 0.6555 | 0.6554 | 0.6553 |
| 0.1945 | 6.0 | 3978 | 1.3345 | 0.6527 | 0.6528 | 0.6527 | 0.6527 |
Framework versions
- Transformers 4.23.1
- Pytorch 1.12.1+cu113
- Datasets 2.6.1
- Tokenizers 0.13.1