Upload step03_chatbot.py with huggingface_hub

step03_chatbot.py

@@ -8,6 +8,9 @@ import os
 import json
 import numpy as np
 import gradio as gr
+
+# Désactiver le warning tokenizers sur ZeroGPU
+os.environ["TOKENIZERS_PARALLELISM"] = "false"
 from gradio import ChatMessage
 from typing import List, Dict, Optional, Tuple
 import time
@@ -255,13 +258,15 @@ class Qwen3Reranker:
         if self.is_mps:
             self.device = torch.device("mps")
             self.model = self.model.to(self.device)
-        elif self.is_cuda:
+        elif self.is_cuda and not os.getenv("SPACE_ID"):
+            # Utiliser CUDA seulement si pas sur ZeroGPU Spaces
             if hasattr(self.model, 'device'):
                 self.device = next(self.model.parameters()).device
             else:
                 self.device = torch.device("cuda")
                 self.model = self.model.to(self.device)
         else:
+            # Forcer CPU sur ZeroGPU pour éviter l'erreur CUDA init
             self.device = torch.device("cpu")
             self.model = self.model.to(self.device)