Update app.py

app.py

@@ -19,7 +19,6 @@ from utils import PROMPT, INPUT_PROMPT, extract_after_final_answer
 
 # Initialize the HF Inference Client
 # It will automatically pick up the HF_TOKEN environment variable if it's set
-client = InferenceClient(timeout=60)
 
 # Define the models we want to use from the Hugging Face Hub
 LLM_MODEL = "Qwen/Qwen2.5-7B-Instruct"
@@ -121,6 +120,9 @@ def play_turn(human_response, loc, player, cur, custom_token, request: gr.Reques
         }
     ]
     try:
+        client = InferenceClient(
+            api_key=chosen_token
+        )
         response = client.chat_completion(
             model=LLM_MODEL,
             messages=messages,