# Uploading to the Hugging Face Hub

This folder is a self-contained, upload-ready HF dataset repo: `README.md` (dataset card with
YAML front-matter), `TRAINING.md` (the Universal-Eve training guide, incl. NVIDIA/A100),
`data/*.parquet` (the four splits), `manifest.json` (provenance), a standalone loader, the
generation notebook, and a **snapshot of the dataset source** (`src/covcollab/` — the
build/verify/load closure **plus the `covcollab-eve-mtl` trainer** — + a trimmed
`pyproject.toml`) so the repo is a complete code + data bundle you can regenerate, verify, and
train on with no external checkout.

> If you are uploading from a fresh checkout, first materialize the two regenerable pieces:
> `covcollab-eve-controlled --out .` (builds `data/*.parquet`) and
> `covcollab-eve-controlled --bundle-source --out .` (copies the source snapshot in).
> Both are already present in this folder.

## One-time setup

```bash
uv pip install huggingface_hub          # or: pip install huggingface_hub
huggingface-cli login                   # paste a WRITE token from https://huggingface.co/settings/tokens
```

> Run the login yourself — never share the token here. In this session you can run it inline by
> typing:  `! huggingface-cli login`

## Upload

```bash
# from inside this folder:
huggingface-cli upload <your-username>/covcollab-eve-detection . . --repo-type dataset
```

This creates the dataset repo if needed and pushes every file; the large `*.parquet` are stored via
Git-LFS automatically (see `.gitattributes`). The `configs:` block in `README.md` maps the parquet
files to the `train` / `validation` / `test_iid` / `test_ood` splits, so the dataset viewer and
`load_dataset` work out of the box.

## Confirm

```python
from datasets import load_dataset
ds = load_dataset("<your-username>/covcollab-eve-detection")
print(ds)                               # train / validation / test_iid / test_ood
```

## Before you upload — checklist

- [ ] Confirm the **license** in `README.md` (currently `cc-by-4.0`) is what you want.
- [ ] (Optional) Regenerate + verify locally:
      `uv run --extra hf covcollab-eve-controlled --verify --out .`
- [ ] (Optional) Fill in the citation / your username in `README.md`.