Feature Extraction
PEFT
embeddings
text-classification
distillation
knowledge-distillation
generated-code
status:scaffold
Instructions to use Kodep/jev-topic-head with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use Kodep/jev-topic-head with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
|
Download data/README.md from Kodep/jev-topic-head: direct link, hf CLI and curl.
- Browser
- Download file 710 Bytes
-
https://huggingface.co/Kodep/jev-topic-head/resolve/main/data/README.md
- Command line
-
hf download hf://Kodep/jev-topic-head/data/README.md
-
curl -L -o README.md https://huggingface.co/Kodep/jev-topic-head/resolve/main/data/README.md
710 Bytes
| # data/ — pointer (data is NOT stored here) | |
| ⚠️ The raw item text (school-email-derived), Jev's saved answers, and the human-labeled ("gold") set are kept **private**, in a separate dataset repo: | |
| 🔒 **https://huggingface.co/datasets/Kodep/jev-topic-data** | |
| **Why:** code and weights are public, but the raw text is family/school content. Decision = "2b": public code, gated data. This folder in the public repo is only a pointer. | |
| ## To run the code | |
| Pull the data into a local `data/` directory (collaborators only — ask Kodep for access): | |
| ``` | |
| hf download Kodep/jev-topic-data --repo-type dataset --local-dir data | |
| ``` | |
| Schema and expected files are documented in the private dataset repo's README. | |