Create app.py

app.py

@@ -0,0 +1,21 @@
+from fastapi import FastAPI
+from llama_cpp import Llama
+import uvicorn
+
+app = FastAPI()
+
+# This loads the model we will download in the Dockerfile
+llm = Llama(model_path="./model.gguf", n_ctx=2048)
+
+@app.get("/")
+def home():
+    return {"status": "PIBIT NANO GPT is Online"}
+
+@app.post("/ask")
+async def ask_ai(data: dict):
+    prompt = data.get("prompt", "")
+    output = llm(f"User: {prompt}\nAI:", max_tokens=150, stop=["User:"], echo=False)
+    return {"response": output["choices"][0]["text"]}
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=7860)