update app.py

app.py

@@ -19,25 +19,32 @@ For more information on `huggingface_hub` Inference API support, please check th
 # text_generator = pipeline("text-generation", model="microsoft/Phi-3-mini-4k-instruct", use_auth_token=hf_token, trust_remote_code=True)
 
 def authenticate_and_generate(message, history, system_message, max_tokens, temperature, top_p):
-    # Initialize the text-generation pipeline with the provided token
-    text_generator = pipeline("text-generation", model="microsoft/Phi-3-mini-4k-instruct", use_auth_token=hf_token, trust_remote_code=True)
-    
-    # Ensure that system_message is a string
-    system_message = str(system_message)
-    
-    # Construct the prompt with system message, history, and user input
-    history_str = "\n".join([f"User: {str(msg[0])}\nAssistant: {str(msg[1])}" for msg in history if isinstance(msg, (tuple, list)) and len(msg) == 2])
-    prompt = system_message + "\n" + history_str
-    prompt += f"\nUser: {message}\nAssistant:"
+    try:
+        # Initialize the text-generation pipeline with the provided token
+        text_generator = pipeline("text-generation", model="microsoft/Phi-3-mini-4k-instruct", use_auth_token=hf_token, trust_remote_code=True)
+        
+        if text_generator.tokenizer is None:
+            raise RuntimeError("Failed to load the tokenizer. Ensure the model and API token are correct.")
+        
+        # Ensure that system_message is a string
+        system_message = str(system_message)
+        
+        # Construct the prompt with system message, history, and user input
+        history_str = "\n".join([f"User: {str(msg[0])}\nAssistant: {str(msg[1])}" for msg in history if isinstance(msg, (tuple, list)) and len(msg) == 2])
+        prompt = system_message + "\n" + history_str
+        prompt += f"\nUser: {message}\nAssistant:"
 
-    # Generate a response using the model
-    response = text_generator(prompt, max_length=max_tokens, temperature=temperature, top_p=top_p, do_sample=True, truncation=True)
+        # Generate a response using the model
+        response = text_generator(prompt, max_length=max_tokens, temperature=temperature, top_p=top_p, do_sample=True, truncation=True)
 
-    # Extract the generated text from the response list
-    assistant_response = response[0]['generated_text']
-    # Optionally trim the assistant response if it includes the prompt again
-    assistant_response = assistant_response.split("Assistant:", 1)[-1].strip()
-    return assistant_response
+        # Extract the generated text from the response list
+        assistant_response = response[0]['generated_text']
+        # Optionally trim the assistant response if it includes the prompt again
+        assistant_response = assistant_response.split("Assistant:", 1)[-1].strip()
+        return assistant_response
+
+    except Exception as e:
+        return str(e)  # Return the error message for debugging
 
 """
 For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface