Fix 410 error with token + text_generation

app.py

@@ -2,7 +2,7 @@ import gradio as gr
 from huggingface_hub import InferenceClient
 import os
 
-# Your secret token
+# Get your secret token for new endpoint
 token = os.getenv("HF_TOKEN")
 
 # Uncensored model
@@ -11,14 +11,14 @@ model_id = "HuggingFaceH4/zephyr-7b-beta"
 def chat(message, history):
     try:
         client = InferenceClient(token=token)
-        messages = [{"role": "user", "content": message}]
-        output = client.chat_completion(
-            messages,
+        response = client.text_generation(
+            message,
             model=model_id,
-            max_tokens=200,
-            temperature=0.8
+            max_new_tokens=200,
+            temperature=0.8,
+            do_sample=True,
         )
-        return output.choices[0].message.content
+        return response
     except Exception as e:
         return f"Sorry, glitch: {str(e)}. Try again!"