File size: 886 Bytes
c296d62 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 | # Local checkpoints (not in Git)
Trained weights live here so clones stay small. After cloning, install the published bundle:
```bash
cd polyguard-rl
python scripts/install_hf_active_bundle.py
```
That creates **`active/`** with:
| Path | Contents |
|------|----------|
| `active/active_model_manifest.json` | Which artifact to load (GRPO vs merged vs SFT) |
| `active/grpo_adapter/` | PEFT GRPO adapter (+ tokenizer files) |
| `active/merged/` | Full merged Qwen 0.5B weights (~1 GB) |
| `active/sft_adapter/` | SFT LoRA fallback |
A Hub cache copy may also appear under `.hf_bundles/` (safe to delete after a successful install).
Enable in `.env`: `POLYGUARD_ENABLE_ACTIVE_MODEL=true` and `POLYGUARD_HF_MODEL=Qwen/Qwen2.5-0.5B-Instruct` (base for the adapter path).
**If this folder looks empty in the editor:** run the install command above; then confirm with `ls active/`.
|