hello

app.py

@@ -1,8 +1,7 @@
 import gradio as gr
-from transformers import pipeline
+from huggingface_hub import InferenceClient
 
-# Load the model
-pipe = pipeline("text-generation", model="KoboldAI/fairseq-dense-13B-Shinen")
+client = InferenceClient("KoboldAI/fairseq-dense-13B-Shinen")
 
 def respond(
     message,
@@ -19,13 +18,13 @@ def respond(
     prompt += f"Human: {message}\nAI:"
 
     # Generate response
-    response = pipe(
+    response = client.text_generation(
         prompt,
-        max_length=len(prompt.split()) + max_tokens,
+        max_new_tokens=max_tokens,
         temperature=temperature,
         top_p=top_p,
         do_sample=True,
-    )[0]['generated_text']
+    )
 
     # Extract only the AI's response
     ai_response = response.split("AI:")[-1].strip()