Update app.py

app.py

@@ -1,65 +1,64 @@
-from fastapi import FastAPI, Request, HTTPException
-from fastapi.responses import StreamingResponse, HTMLResponse
-from fastapi.templating import Jinja2Templates
-import torch
-import asyncio
-from transformers import AutoTokenizer, AutoModelForCausalLM
+import gradio as gr
+from huggingface_hub import InferenceClient
 
-app = FastAPI()
-templates = Jinja2Templates(directory="templates")
+"""
+For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
+"""
+client = InferenceClient("microsoft/Phi-3.5-mini-instruct")
 
-# Configuration du modèle (optimisé pour 2000 tokens)
-model_name = "microsoft/Phi-3.5-mini-instruct"
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForCausalLM.from_pretrained(
-    model_name,
-    torch_dtype=torch.float16,
-    device_map="auto",
-    low_cpu_mem_usage=True  # Critique pour les longs contextes
-)
 
-async def generate_stream(prompt: str, max_tokens: int = 2000):
-    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
-    generated_count = 0
-    
-    while generated_count < max_tokens:
-        outputs = model.generate(
-            **inputs,
-            max_new_tokens=1,
-            do_sample=True,
-            temperature=0.7,
-            top_p=0.9
-        )
-        new_token = tokenizer.decode(outputs[0][-1], skip_special_tokens=True)
-        yield f"data: {new_token}\n\n"
-        generated_count += 1
-        
-        # Optimisation mémoire
-        if generated_count % 50 == 0:
-            await asyncio.sleep(0.01)  # Réduit la pression sur le GPU
-            torch.cuda.empty_cache()  # Nettoyage mémoire
-        
-        inputs = {"input_ids": outputs}
+def respond(
+    message,
+    history: list[tuple[str, str]],
+    system_message,
+    max_tokens,
+    temperature,
+    top_p,
+):
+    messages = [{"role": "system", "content": system_message}]
+
+    for val in history:
+        if val[0]:
+            messages.append({"role": "user", "content": val[0]})
+        if val[1]:
+            messages.append({"role": "assistant", "content": val[1]})
+
+    messages.append({"role": "user", "content": message})
+
+    response = ""
 
-@app.get("/", response_class=HTMLResponse)
-async def chat_page(request: Request):
-    return templates.TemplateResponse("stream.html", {"request": request})
+    for message in client.chat_completion(
+        messages,
+        max_tokens=max_tokens,
+        stream=True,
+        temperature=temperature,
+        top_p=top_p,
+    ):
+        token = message.choices[0].delta.content
+
+        response += token
+        yield response
+
+
+"""
+For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface  
+"""
+demo = gr.ChatInterface(
+    respond,
+    additional_inputs=[
+        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
+        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
+        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
+        gr.Slider(
+            minimum=0.1,
+            maximum=1.0,
+            value=0.95,
+            step=0.05,
+            label="Top-p (nucleus sampling)",
+        ),
+    ],
+)
 
-@app.get("/stream")
-async def stream_response(prompt: str):
-    try:
-        return StreamingResponse(
-            generate_stream(prompt),
-            media_type="text/event-stream",
-            headers={
-                "Cache-Control": "no-cache",
-                "Connection": "keep-alive",
-                "X-Accel-Buffering": "no"  # Critique pour les streams longs
-            }
-        )
-    except Exception as e:
-        raise HTTPException(status_code=500, detail=str(e))
 
 if __name__ == "__main__":
-    import uvicorn
-    uvicorn.run(app, host="0.0.0.0", port=7860)
+    demo.launch()