Sandeep Chowdhary
commited on
Update model card
Browse files
README.md
CHANGED
|
@@ -23,15 +23,6 @@ This model classifies content related to plant-based diets, sustainable food sys
|
|
| 23 |
- Labels: 7
|
| 24 |
- Training Data: Reddit posts from climate subreddits (2010-2023)
|
| 25 |
|
| 26 |
-
## Training
|
| 27 |
-
|
| 28 |
-
Trained on GPT-labeled Reddit data:
|
| 29 |
-
1. Data collection from climate subreddits
|
| 30 |
-
2. Regex filtering for sector-specific content
|
| 31 |
-
3. GPT labeling for multilabel classification
|
| 32 |
-
4. 80/10/10 train/validation/test split
|
| 33 |
-
5. Fine-tuning with threshold optimization
|
| 34 |
-
|
| 35 |
## Labels
|
| 36 |
|
| 37 |
The model predicts 7 labels simultaneously:
|
|
@@ -234,6 +225,15 @@ for label, score in zip(label_names, predictions[0]):
|
|
| 234 |
```
|
| 235 |
|
| 236 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 237 |
## Citation
|
| 238 |
|
| 239 |
If you use this model in your research, please cite:
|
|
|
|
| 23 |
- Labels: 7
|
| 24 |
- Training Data: Reddit posts from climate subreddits (2010-2023)
|
| 25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
## Labels
|
| 27 |
|
| 28 |
The model predicts 7 labels simultaneously:
|
|
|
|
| 225 |
```
|
| 226 |
|
| 227 |
|
| 228 |
+
## Training
|
| 229 |
+
|
| 230 |
+
Trained on GPT-labeled Reddit data:
|
| 231 |
+
1. Data collection from climate subreddits
|
| 232 |
+
2. Regex filtering for sector-specific content
|
| 233 |
+
3. GPT labeling for multilabel classification
|
| 234 |
+
4. 80/10/10 train/validation/test split
|
| 235 |
+
5. Fine-tuning with threshold optimization
|
| 236 |
+
|
| 237 |
## Citation
|
| 238 |
|
| 239 |
If you use this model in your research, please cite:
|