Commit History
-
Update README.md
Verified signature91ad21d -
Upload README.md with huggingface_hub
Verified signaturedb7370e -
Update README.md
Verified signature841c708 -
Model save
Verified signature05a8b69 -
Training in progress, step 75
Verified signature6789740
-
Update README.md
Verified signature1865799
-
Update README.md
Verified signature9e837ef -
Upload train_theme.jsonl
Verified signature01e1e3a
-
Update README.md
Verified signature28c57e8 -
Upload theme_training_150.json
Verified signature650facf -
Update README.md
Verified signature01e8424 -
Update README.md
Verified signature82b4e1c
-
Upload theme_response.json
Verified signature37447ae -
Create README.md
Verified signature# Theme Model (Multi-Label Classifier) This model performs **multi-label theme detection** for community knowledge submissions. It classifies short narratives or summaries into one or more of the following themes: - **Mentorship** - **Entrepreneurship** - **Startup Success** ## 🧩 Base Model Fine-tuned from [`distilbert-base-uncased`](https://huggingface.co/distilbert-base-uncased). ## 📚 Training Data Trained on a synthetic dataset (`theme_response.json`) consisting of community stories with labeled themes such as: - “Mentorship strengthens community-based skill development.” - “Entrepreneurship drives local economic growth and sustainability.” - “Support programs help boost startup success rates.” ## ⚙️ Usage ```python from transformers import pipeline classifier = pipeline("text-classification", model="4nkh/theme_model", tokenizer="4nkh/theme_model", top_k=None) text = "Our incubator helped women founders scale their startups." preds = classifier(text) print(preds)d9d84f8 -
Upload train_theme_model.py
Verified signature36b732d -
initial commit
Verified signature7558b0f