Create app.py

app.py

@@ -0,0 +1,36 @@
+import gradio as gr
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch, os
+from fastapi import FastAPI, Request
+import uvicorn
+
+MODEL_ID = "Qwen/Qwen2.5-1.5B-Instruct"
+tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained(MODEL_ID, device_map="auto", torch_dtype=torch.float16)
+
+PROMPT_TEMPLATE = """You are a GAIA final-answer extractor.
+Extract only what follows "Final Answer:" from the text, or infer if missing.
+Text:
+{text}
+"""
+
+def extract_final_answer(raw):
+    prompt = PROMPT_TEMPLATE.format(text=raw)
+    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+    out = model.generate(**inputs, max_new_tokens=64, temperature=0.2)
+    return tokenizer.decode(out[0], skip_special_tokens=True).splitlines()[-1]
+
+app = FastAPI()
+
+@app.post("/api/predict")
+async def predict(request: Request):
+    data = await request.json()
+    text = data.get("data", [""])[0]
+    return {"data": [extract_final_answer(text)]}
+
+iface = gr.Interface(fn=extract_final_answer, inputs="text", outputs="text", title="Final Answer Agent")
+
+if __name__ == "__main__":
+    import threading
+    threading.Thread(target=lambda: iface.launch(server_name="0.0.0.0", server_port=7863)).start()
+    uvicorn.run(app, host="0.0.0.0", port=7863)