squeezebert-uncased-News_About_Gold
This model is a fine-tuned version of squeezebert/squeezebert-uncased. It achieves the following results on the evaluation set:
- Loss: 0.2643
- Accuracy: 0.9167
- F1
- Weighted: 0.9166
- Micro: 0.9167
- Macro: 0.8749
- Recall
- Weighted: 0.9167
- Micro: 0.9167
- Macro: 0.8684
- Precision
- Weighted: 0.9168
- Micro: 0.9167
- Macro: 0.8822
Model description
For more information on how it was created, check out the following link: https://github.com/DunnBC22/NLP_Projects/blob/main/Sentiment%20Analysis/Sentiment%20Analysis%20of%20Commodity%20News%20-%20Gold%20(Transformer%20Comparison)/News%20About%20Gold%20-%20Sentiment%20Analysis%20-%20SqueezeBERT%20with%20W%26B.ipynb
This project is part of a comparison of seven (7) transformers. Here is the README page for the comparison: https://github.com/DunnBC22/NLP_Projects/tree/main/Sentiment%20Analysis/Sentiment%20Analysis%20of%20Commodity%20News%20-%20Gold%20(Transformer%20Comparison)
Intended uses & limitations
This model is intended to demonstrate my ability to solve a complex problem using technology.
Training and evaluation data
Dataset Source: https://www.kaggle.com/datasets/ankurzing/sentiment-analysis-in-commodity-market-gold
Input Word Length:
/Images/Input%20Word%20Length.png)
Class Distribution:
/Images/Class%20Distribution.png)
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Weighted F1 | Micro F1 | Macro F1 | Weighted Recall | Micro Recall | Macro Recall | Weighted Precision | Micro Precision | Macro Precision |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.8756 | 1.0 | 133 | 0.4529 | 0.8699 | 0.8557 | 0.8699 | 0.6560 | 0.8699 | 0.8699 | 0.6727 | 0.8437 | 0.8699 | 0.6414 |
| 0.4097 | 2.0 | 266 | 0.3196 | 0.9026 | 0.8982 | 0.9026 | 0.7826 | 0.9026 | 0.9026 | 0.7635 | 0.9059 | 0.9026 | 0.8743 |
| 0.3147 | 3.0 | 399 | 0.2824 | 0.9115 | 0.9111 | 0.9115 | 0.8470 | 0.9115 | 0.9115 | 0.8319 | 0.9138 | 0.9115 | 0.8751 |
| 0.2685 | 4.0 | 532 | 0.2649 | 0.9186 | 0.9187 | 0.9186 | 0.8681 | 0.9186 | 0.9186 | 0.8602 | 0.9203 | 0.9186 | 0.8797 |
| 0.2479 | 5.0 | 665 | 0.2643 | 0.9167 | 0.9166 | 0.9167 | 0.8749 | 0.9167 | 0.9167 | 0.8684 | 0.9168 | 0.9167 | 0.8822 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0
- Datasets 2.11.0
- Tokenizers 0.13.3
