1b3a47b626
The model config, vision-check, and embedding-check endpoints now read the token from the X-Unsloth-HF-Token header (preferred) and fall back to the hf_token query param, matching /gguf-variants. This keeps older clients that still send the query param working after the backend updates, while new clients send the header and never put the token in the URL. |
||
|---|---|---|
| .. | ||
| data_recipe | ||
| .gitkeep | ||
| __init__.py | ||
| auth.py | ||
| chat_history.py | ||
| datasets.py | ||
| export.py | ||
| inference.py | ||
| llama.py | ||
| mcp_servers.py | ||
| models.py | ||
| preview.py | ||
| prompts.py | ||
| providers.py | ||
| rag.py | ||
| settings.py | ||
| training.py | ||
| training_history.py | ||
| training_vram.py | ||