Update README.md
Browse files
README.md
CHANGED
|
@@ -85,17 +85,6 @@ Accuracy (overall correct predictions), Macro F1-score (average F1 across classe
|
|
| 85 |
- Macro F1-Score: 66.35%
|
| 86 |
|
| 87 |
Detailed per-class metrics and confusion matrix available in the repository.
|
| 88 |
-
#### Summary
|
| 89 |
-
The model shows robust performance on unseen Persian tweet data, with good generalization across emotions.
|
| 90 |
-
|
| 91 |
-
|
| 92 |
-
## Technical Specifications
|
| 93 |
-
### Model Architecture and Objective
|
| 94 |
-
BERT base model with a sequence classification head for multi-class emotion prediction.
|
| 95 |
-
#### Hardware
|
| 96 |
-
T4.
|
| 97 |
-
#### Software
|
| 98 |
-
Hugging Face Transformers library, Python.
|
| 99 |
|
| 100 |
|
| 101 |
**BibTeX:**
|
|
|
|
| 85 |
- Macro F1-Score: 66.35%
|
| 86 |
|
| 87 |
Detailed per-class metrics and confusion matrix available in the repository.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 88 |
|
| 89 |
|
| 90 |
**BibTeX:**
|