Update README.md
Browse files
README.md
CHANGED
|
@@ -28,6 +28,9 @@ It achieves the following results on the evaluation set:
|
|
| 28 |
- Recall: 0.6937
|
| 29 |
- F1: 0.6917
|
| 30 |
|
|
|
|
|
|
|
|
|
|
| 31 |
## Model description
|
| 32 |
|
| 33 |
The code begins by loading a DistilBERT model and tokenizer configured for sequence classification with two possible labels. It then preprocesses the data: training and testing text sequences are tokenized using BERT, ensuring uniform length with padding and truncation to 256 tokens.
|
|
|
|
| 28 |
- Recall: 0.6937
|
| 29 |
- F1: 0.6917
|
| 30 |
|
| 31 |
+
## IMPORTANT NOTE
|
| 32 |
+
When using the model, please note that `LABEL_0` refers to non-violation and `LABEL_1` refers to a violation. Moreso, you can find the code for building the model within this repository. It is titled `code.ipynb`.
|
| 33 |
+
|
| 34 |
## Model description
|
| 35 |
|
| 36 |
The code begins by loading a DistilBERT model and tokenizer configured for sequence classification with two possible labels. It then preprocesses the data: training and testing text sequences are tokenized using BERT, ensuring uniform length with padding and truncation to 256 tokens.
|