fix: disable gradio hot reload for CUDA-safe startup

Switch to LLM2Vec 3.1 pair to fix Space 401 on gated Llama 3.0

Sync text encoder 3.0 fix and smoke checks

Fix HF auth propagation for text encoder loads

Switch LLM2Vec to Meta-Llama-3.1-8B-Instruct (accessible with current token)

Revert early spaces import in text encoder server (fix gradio_auto_wrap crash)

Spaces: import spaces early in text-encoder server to avoid CUDA init warning thread crash

Use mirrored LLM2Vec base model repo and env overrides

Debug: print full traceback on text encoder init failure to expose real error

Add core kimodo package modules required by native demo