fix: drop provider='hf-inference' (auto-route works for 72B)

fix: bump sdk_version to 5.49.1 to match requirements

fix: gradio>=5.49.1 + huggingface_hub<1.0 (HfFolder compat)

fix: upgrade huggingface_hub + explicit hf-inference provider

fix: switch to Qwen2.5-72B-Instruct (7B not available on free inference API)

fix: LLM timeout + no-token fallback

fix: pin gradio 5.0.0 + hf_hub 0.25.2

fix: pin python 3.11, gradio 5.x

Step 5: RAG with Qwen2.5

Upgrade Gradio to 5.x for huggingface_hub 1.x compat

Update

Initial deploy: FAQ bot with bge-small-zh

initial commit