Create app.py

app.py

@@ -0,0 +1,41 @@
+from fastapi import FastAPI
+from pydantic import BaseModel
+from llama_cpp import Llama
+from huggingface_hub import hf_hub_download
+
+app = FastAPI()
+
+MODEL_REPO = "bartowski/Qwen2.5-Coder-1.5B-Instruct-GGUF"
+MODEL_FILE = "Qwen2.5-Coder-1.5B-Instruct-Q4_K_M.gguf"
+
+model_path = hf_hub_download(
+    repo_id=MODEL_REPO,
+    filename=MODEL_FILE
+)
+
+llm = Llama(
+    model_path=model_path,
+    n_ctx=2048,
+    n_threads=2
+)
+
+class ChatRequest(BaseModel):
+    message: str
+
+@app.get("/")
+def root():
+    return {"status": "AI engine running"}
+
+@app.post("/chat")
+def chat(req: ChatRequest):
+    output = llm(
+        f"<|user|>{req.message}<|assistant|>",
+        max_tokens=300,
+        stop=["<|end|>"]
+    )
+    return {"reply": output["choices"][0]["text"]}
+
+# ⭐ THIS PART WAS MISSING
+if __name__ == "__main__":
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=7860)