Update app.py

app.py

@@ -5,37 +5,38 @@ import threading
 import time
 import os
 
-# Your Hugging Face API token
+# Get Hugging Face API token from secrets
 API_TOKEN = os.getenv("token")
 
 if not API_TOKEN:
     print("ERROR: API token not found!")
 else:
     print("API token retrieved successfully.")
-"""
-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
-"""
+
+# Initialize inference client with Zephyr-7B
 client = InferenceClient("HuggingFaceH4/zephyr-7b-beta", token=API_TOKEN)
 
 
 def load_data():
+    """Load dataset from Hugging Face and store it in a dictionary."""
     dataset = load_dataset("accesscreate012/abhinav-academy-chatbot", split="train")
-    return {entry["instruction"]: entry["response"] for entry in dataset}
+    return {entry["instruction"].strip(): entry["response"].strip() for entry in dataset}
 
 
-# Global variable to store the dataset
+# Global dataset
 data = load_data()
 
 
 def auto_update():
+    """Automatically refresh the dataset every 24 hours."""
     global data
     while True:
-        time.sleep(86400)  # Wait for 24 hours before refreshing the dataset
-        data = load_data()  # Reload the dataset
+        time.sleep(86400)  # 24 hours
+        data = load_data()
         print("Dataset updated.")
 
 
-# Start the auto-update thread (runs in the background)
+# Start dataset auto-update in a separate thread
 threading.Thread(target=auto_update, daemon=True).start()
 
 
@@ -49,35 +50,47 @@ def respond(
 ):
     print("Received message:", message)
 
-    # Check if the question exists in the dataset
-    if message in data:
+    # Check if the message matches an entry in the dataset
+    if message.strip() in data:
         print("Found exact match in dataset.")
-        yield data[message]  # Return the exact response from the dataset
+        yield data[message.strip()]  # Return the exact response from the dataset
         return
 
-    print("No exact match found, using LLM.")
+    print("No exact match found, using Zephyr-7B.")
+
+    # Construct system message with dataset context
+    dataset_context = "\n".join([f"Q: {q}\nA: {a}" for q, a in data.items()])
+    full_system_message = (
+        f"{system_message}\n\n"
+        "Only use the following dataset for answers:\n"
+        f"{dataset_context}\n"
+        "If the exact answer is not found, infer based on the data.\n"
+        "Do NOT generate unrelated information.\n"
+        "Keep responses short and accurate."
+    )
 
-    messages = [{"role": "system", "content": system_message}]
+    # Construct conversation history
+    messages = [{"role": "system", "content": full_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]})
+    for user_input, bot_response in history:
+        if user_input:
+            messages.append({"role": "user", "content": user_input})
+        if bot_response:
+            messages.append({"role": "assistant", "content": bot_response})
 
     messages.append({"role": "user", "content": message})
 
     response = ""
 
     try:
-        for message in client.chat_completion(
+        for msg in client.chat_completion(
             messages,
             max_tokens=max_tokens,
             stream=True,
             temperature=temperature,
             top_p=top_p,
         ):
-            token = message.choices[0].delta.content
+            token = msg.choices[0].delta.content
             response += token
             yield response
     except Exception as e:
@@ -85,25 +98,17 @@ def respond(
         yield "An error occurred: " + str(e)
 
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# Gradio Chat UI
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        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.Textbox(value="You are a helpful and knowledgeable chatbot.", label="System message"),
+        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max 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=1.0, value=0.95, step=0.05, label="Top-p"),
     ],
 )
 
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()