perf: switch to Q4_K_M model for 2x faster CPU inference

fix: increase context size (n_ctx) to 8192

chore: clean up debug middleware and logger from app.py

fix: disable RequestIdPlugin UUID validation for compatibility with Hugging Face proxy

debug: monkeypatch RouteErrorHandler to output exceptions

debug: add request logging middleware

fix: use h11 parser without proxy headers to resolve 400 Bad Request

chore: enable uvicorn debug logging

fix: revert to working uvicorn config with h11 and proxy headers

fix: restore default uvicorn run without custom proxy headers

fix: use h11 parser for uvicorn to prevent 400 bad request

fix: remove duplicate CORSMiddleware causing 400 Bad Request

fix: add proxy headers and trust all IPs for uvicorn

fix: adiciona CORS middleware para chamadas diretas do browser

Update app.py

Update app.py

Update app.py

Update app.py

Create app.py