Update app.py

app.py

@@ -1,14 +1,20 @@
-import gradio as gr
+from fastapi import FastAPI
+from pydantic import BaseModel
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import torch
 
+app = FastAPI()
+
+# Chargement du modèle
 model = AutoModelForCausalLM.from_pretrained("deepseek-ai/deepseek-llm-7b-chat", device_map="auto", torch_dtype=torch.float16)
 tokenizer = AutoTokenizer.from_pretrained("deepseek-ai/deepseek-llm-7b-chat")
 
-def chat(prompt):
-    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
-    outputs = model.generate(**inputs, max_new_tokens=256, do_sample=True)
-    return tokenizer.decode(outputs[0], skip_special_tokens=True)
+class Prompt(BaseModel):
+    prompt: str
 
-demo = gr.Interface(fn=chat, inputs="text", outputs="text")
-demo.launch()
+@app.post("/predict")
+async def predict(prompt: Prompt):
+    inputs = tokenizer(prompt.prompt, return_tensors="pt").to(model.device)
+    outputs = model.generate(**inputs, max_new_tokens=256, do_sample=True)
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return {"response": response}