Update app.py

app.py

@@ -1,24 +1,11 @@
 import gradio as gr
-import os
-import torch
-from transformers import AutoModelForCausalLM, AutoTokenizer
+from huggingface_hub import InferenceClient
 
 """
-Load model and tokenizer directly using transformers
+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
 """
-model_name = "PlantWisdom/Data_Management_Mistral"
+client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
-# Load tokenizer and model
-print("Loading tokenizer...")
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-
-print("Loading model...")
-model = AutoModelForCausalLM.from_pretrained(
-    model_name,
-    torch_dtype=torch.float16,
-    device_map="auto",
-    low_cpu_mem_usage=True
-)
 
 def respond(
     message,
@@ -28,40 +15,30 @@ def respond(
     temperature,
     top_p,
 ):
-    # Format the conversation in Mistral's instruction format
-    prompt = f"<s>[INST] {system_message} [/INST]\n\n"
-    
-    for user_msg, assistant_msg in history:
-        if user_msg:
-            prompt += f"[INST] {user_msg} [/INST]\n"
-        if assistant_msg:
-            prompt += f"{assistant_msg}\n\n"
-    
-    prompt += f"[INST] {message} [/INST]\n"
+    messages = [{"role": "system", "content": system_message}]
+
+    for val in history:
+        if val[0]:
+            messages.append({"role": "user", "content": val[0]})
+        if val[1]:
+            messages.append({"role": "assistant", "content": val[1]})
+
+    messages.append({"role": "user", "content": message})
 
-    # Print the prompt for debugging
-    print(f"Prompt: {prompt}")
+    response = ""
 
-    # Encode the prompt
-    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
-    
-    # Generate tokens
-    print("Generating response...")
-    
-    # Generate without streaming for simplicity
-    generated_ids = model.generate(
-        inputs.input_ids,
-        max_new_tokens=max_tokens,
-        do_sample=True,
+    for message in client.chat_completion(
+        messages,
+        max_tokens=max_tokens,
+        stream=True,
         temperature=temperature,
         top_p=top_p,
-        pad_token_id=tokenizer.eos_token_id,
-    )
-    
-    # Decode the response
-    full_response = tokenizer.decode(generated_ids[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)
-    print(f"Response generated: {full_response[:50]}...")
-    return full_response
+    ):
+        token = message.choices[0].delta.content
+
+        response += token
+        yield response
+
 
 """
 For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
@@ -69,20 +46,19 @@ For information on how to customize the ChatInterface, peruse the gradio docs: h
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Textbox(value="You are a Microsoft 365 data management assistant specialized in SharePoint and OneDrive. Answer questions concisely and accurately.", label="System message"),
-        gr.Slider(minimum=1, maximum=1024, value=256, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=1.0, value=0.7, step=0.1, label="Temperature"),
+        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
+        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
+        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
         gr.Slider(
             minimum=0.1,
             maximum=1.0,
-            value=0.9,
+            value=0.95,
             step=0.05,
             label="Top-p (nucleus sampling)",
         ),
     ],
-    title="Microsoft 365 Data Management Assistant",
-    description="Ask questions about SharePoint, OneDrive, and other Microsoft 365 data management topics."
 )
 
+
 if __name__ == "__main__":
     demo.launch()