Update app.py

app.py

@@ -1,65 +1,70 @@
 import gradio as gr
 import json
+import os
 from huggingface_hub import InferenceClient
+from groq import Groq
+from groq.types.chat.chat_completion_tool_param import ChatCompletionToolParam
 
 # Use the fine-tuned maritime legal model
 MODEL = "nomiChroma3.1"
+##client = Groq(api_key=os.environ["GROQ_API_KEY"])
 
-# Initialize the client with the correct model
 client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
-def respond(message, history: list[tuple[str, str]], system_message, max_tokens, temperature, top_p):
-    # Build the messages list for the model
+
+def respond(
+    message,
+    history: list[tuple[str, str]],
+    system_message,
+    max_tokens,
+    temperature,
+    top_p,
+):
     messages = [{"role": "system", "content": system_message}]
-    
-    # Add conversation history
-    for user_input, assistant_response in history:
-        if user_input:
-            messages.append({"role": "user", "content": user_input})
-        if assistant_response:
-            messages.append({"role": "assistant", "content": assistant_response})
 
-    # Add the latest user input
+    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})
 
     response = ""
-    
-    try:
-        # Streaming the chat completion response
-        for message in client.chat_completion(
-            messages,
-            max_tokens=max_tokens,
-            stream=True,
-            temperature=temperature,
-            top_p=top_p,
-        ):
-            try:
-                # Safely parse each streamed message item
-                payload = message.choices[0].delta.content
-                response += payload
-                yield response
-            except json.JSONDecodeError as e:
-                # Handle JSON parsing error
-                yield f"Error decoding JSON: {str(e)}"
 
-    except Exception as e:
-        yield f"Error: {str(e)}"
+    for message in client.chat_completion(
+        messages,
+        max_tokens=max_tokens,
+        stream=True,
+        temperature=temperature,
+        top_p=top_p,
+    ):
+        token = message.choices[0].delta.content
+
+        response += token
+        yield response
 
-# Gradio interface
+"""
+For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
+"""
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Textbox(
-            value="You are a maritime legal assistant with expertise strictly in Indian maritime law. Provide detailed legal advice and information based on Indian maritime legal principles and regulations.",
-            label="System message"
-        ),
+        gr.Textbox(value="You are a maritime legal assistant with expertise strictly in indian maritime law. Provide detailed legal advice and information based on indian maritime legal principles and regulations.", 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=0.95, step=0.05, label="Top-p (nucleus sampling)"),
+        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=0.95,
+            step=0.05,
+            label="Top-p (nucleus sampling)",
+        ),
     ],
     title="Maritime Legal Compliance",
-    description="This chatbot uses the fine-tuned model for providing legal advice on Indian maritime law.",
+    description="This chatbot uses the fine tune Llama 3.1 which has the capabilities of responding and helping in legal advices regarding maritime",
 )
 
+
 if __name__ == "__main__":
     demo.launch()