zeroGPU

app.py

@@ -1,11 +1,13 @@
 import gradio as gr
 from huggingface_hub import InferenceClient
+import spaces
 
 """
 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("csfufu/Revisual-R1-final")
+client = InferenceClient("HuggingFaceTB/SmolVLM-Instruct")
 
+@spaces.GPU
 def respond(
     message,
     history: list[tuple[str, str]],
@@ -16,7 +18,7 @@ def respond(
 ):
     messages = [{"role": "system", "content": system_message}]
 
-    messages.extend(chat_history)
+    messages.extend(history)
 
     messages.append({"role": "user", "content": message})