README.md

---
title: Monu Portfolio Chat
emoji: 💬
colorFrom: blue
colorTo: purple
sdk: docker
app_port: 7860
pinned: false
---

# Monu Kumar Portfolio Chat API

FastAPI endpoint serving a fine-tuned Gemma 2 2B-it model (GGUF, Q4_K_M) via llama.cpp.

- `GET /` — health check
- `POST /chat` — body: `{"message": "...", "history": [{"role": "user", "content": "..."}]}` → `{"reply": "..."}`

Set the `MODEL_REPO` and `MODEL_FILE` variables in Space Settings if you rename the model repo.