update the UI

app.py

@@ -2,6 +2,7 @@ import urllib.request
 import fitz
 import re
 import numpy as np
+import tensorflow as tf
 import tensorflow_hub as hub
 import openai
 import gradio as gr
@@ -36,7 +37,7 @@ def pdf_to_text(path, start_page=1, end_page=None):
     return text_list
 
 
-def text_to_chunks(texts, word_length=150, start_page=1):
+def text_to_chunks(texts, word_length=200, start_page=1):
     text_toks = [t.split(' ') for t in texts]
     page_nums = []
     chunks = []
@@ -98,20 +99,17 @@ def load_recommender(path, start_page=1):
 
 
 ####################
-def generate_text(openAI_key,
-                  openAI_base,
-                  openAI_API_version,
-                  prompt,
-                  engine="chatgpt"):
-    openai.api_type = "azure"
+def generate_text(api_type, engine, openAI_key, openAI_base,
+                  openAI_API_version, temperature, prompt):
+    openai.api_type = api_type
     openai.api_base = openAI_base
     openai.api_version = openAI_API_version
     openai.api_key = openAI_key
-    completions = openai.ChatCompletion.create(engine="chatgpt",
-                                               max_tokens=1024,
+    completions = openai.ChatCompletion.create(engine=engine,
+                                               max_tokens=2056,
                                                n=1,
                                                stop=None,
-                                               temperature=1.0,
+                                               temperature=temperature,
                                                messages=[{
                                                    "role": "user",
                                                    "content": prompt
@@ -121,47 +119,30 @@ def generate_text(openAI_key,
     return message
 
 
-def generate_answer(question, openAI_key, openAI_base, openAI_API_version):
+def generate_answer(api_type, engine, openAI_key, openAI_base,
+                    openAI_API_version, temperature, user_prompt, question):
     topn_chunks = recommender(question)
-    #     print(len(topn_chunks))
-    #     print(*topn_chunks, sep="\n")
     prompt = ""
     prompt += 'search results:\n\n'
     for c in topn_chunks:
         prompt += c + '\n\n'
 
+    prompt += f"Instructions: {user_prompt}"\
+              f"\n\nQuery: {question}\nAnswer: "
 
-#         prompt += "Instructions: Compose a comprehensive reply to the query using the search results given. "\
-#               "Cite each reference using [ Page Number] notation (every result has this number at the beginning). "\
-#               "Citation should be done at the end of each sentence. "\
-#               "If the search results mention multiple subjects with the same name, create separate answers for each. "\
-#               "Make sure the answer is correct and don't output false content. "\
-#               "Only answer what is asked. The answer should be in details."\
-#               "\n\nQuery: {question}\nAnswer: "
-    prompt += "Instructions: Compose a comprehensive reply to the query using the search results given. "\
-              "Cite each reference using [ Page Number] notation (every result has this number at the beginning). "\
-              "Citation should be done at the end of each sentence. "\
-              "If the search results mention multiple subjects with the same name, create separate answers for each. "\
-              "Only include information found in the results and don't add any additional information."\
-              "Make sure the answer is correct and don't output false content. "\
-              "If the text does not relate to the query, simply state 'Found Nothing'. "\
-              "Ignore outlier search results which has nothing to do with the question."\
-              "Only answer what is asked. The answer should be short and concise."\
-              "\n\nQuery: {question}\nAnswer: "
-
-    prompt += f"Query: {question}\nAnswer:"
     print(prompt)
-    answer = generate_text(openAI_key, openAI_base, openAI_API_version, prompt,
-                           "chatgpt")
+    answer = generate_text(api_type, engine, openAI_key, openAI_base,
+                           openAI_API_version, temperature, prompt)
     return answer
 
 
-def question_answer(url, file, question, openAI_key, openAI_base,
-                    openAI_API_version):
+def question_answer(api_type, engine, openAI_key, openAI_base,
+                    openAI_API_version, url, file, temperature, user_prompt,
+                    question):
     if openAI_key.strip() == '':
         return '[ERROR]: Please enter you Open AI Key. Get your key here : https://platform.openai.com/account/api-keys'
     if url.strip() == '' and file == None:
-        return '[ERROR]: Both URL and PDF is empty. Provide atleast one.'
+        return '[ERROR]: Both URL and PDF is empty. Provide at least one.'
 
     if url.strip() != '' and file != None:
         return '[ERROR]: Both URL and PDF is provided. Please provide only one (eiter URL or PDF).'
@@ -181,71 +162,122 @@ def question_answer(url, file, question, openAI_key, openAI_base,
     if question.strip() == '':
         return '[ERROR]: Question field is empty'
 
-    return generate_answer(question, openAI_key, openAI_base,
-                           openAI_API_version)
+    return generate_answer(api_type, engine, openAI_key, openAI_base,
+                           openAI_API_version, temperature, user_prompt,
+                           question)
 
 
-def chatbot_respond(question, chat_history, url, file, openAI_key, openAI_base,
-            openAI_API_version):
+def chatbot_respond(api_type, engine, openAI_key, openAI_base,
+                    openAI_API_version, url, file, temperature, user_prompt,
+                    question, chatbot):
 
-    bot_message = question_answer(url, file, question, openAI_key, openAI_base,
-                                  openAI_API_version)
-    chat_history.append((question, bot_message))
-    return "", chat_history
+    bot_message = question_answer(api_type, engine, openAI_key, openAI_base,
+                                  openAI_API_version, url, file, temperature,
+                                  user_prompt, question)
+    chatbot.append((question, bot_message))
+    return "", chatbot
 
 recommender = SemanticSearch()
 
-title = 'PDF GPT Azure'
+title = 'HKU PDF GPT Azure'
 description = """ 
-PDF GPT allows you to chat with your PDF file using Universal Sentence Encoder and Open AI. 
-It gives hallucination free response than other tools as the embeddings are better than OpenAI. 
-The returned response can even cite the page number in square brackets([]) where the information is located,
-adding credibility to the responses and helping to locate pertinent information quickly.
+PDF GPT allows you to chat with your PDF file using Universal Sentence Encoder and Open AI. It gives hallucination free response than other tools as the embeddings are better than OpenAI. The returned response can even cite the page number in square brackets([]) where the information is located,adding credibility to the responses and helping to locate pertinent information quickly.
 """
 
+
+def save_settings():
+    return gr.Tabs.update(selected=1)
+
+
 with gr.Blocks() as demo:
 
     gr.Markdown(f'<center><h1>{title}</h1></center>')
     gr.Markdown(description)
 
-    with gr.Row():
-
-        with gr.Group():
-            gr.Markdown(
-                f'<p style="text-align:center">Get your Open AI API key <a href="https://platform.openai.com/account/api-keys">here</a></p>'
-            )
-            gr.Dropdown(label="API Type",
-                        choices=["azure", "OpenAI"],
-                        info="Azure or Open AI",
-                        value="azure"),
+    with gr.Tabs() as tabs:
+        with gr.TabItem("Setup", id=0):
+            with gr.Accordion("Detail Settings", open=False):
+                api_type = gr.Dropdown(label="API Type",
+                                       choices=["azure", "OpenAI"],
+                                       info="Azure or Open AI",
+                                       value="azure",
+                                       interactive=False)
+                openAI_base = gr.Textbox(label='api_base',
+                                         value="https://api.hku.hk",
+                                         interactive=True)
+                openAI_API_version = gr.Textbox(label='API version',
+                                                value="2023-03-15-preview",
+                                                interactive=True)
+                engine = gr.Dropdown(
+                    label="Engine",
+                    choices=["chatgpt", "chatgpt-4", "chatgpt-4-32k"],
+                    info="ChatGPT 3.5, ChatGPT 4 ,ChatGPT 4-32k",
+                    value="chatgpt",
+                    interactive=True)
             #####################
             ##
             ## REMEMBER to remove the key before public deploy
             ##
             #####################
             openAI_key = gr.Textbox(
-                label='Enter your Azure OpenAI API key here')
-            openAI_base = gr.Textbox(label='api_base',
-                                     value="https://api.hku.hk")
-            openAI_API_version = gr.Textbox(label='API version',
-                                            value="2023-03-15-preview")
+                label='Enter your HKU Azure OpenAI API key here',
+                value="561c52b8d3ec4733bab55ee9515e6deb")
+
             url = gr.Textbox(label='Enter PDF URL here')
             gr.Markdown("<center><h4>OR<h4></center>")
             file = gr.File(label='Upload your PDF/ Research Paper / Book here',
                            file_types=['.pdf'])
-        with gr.Group():
-            chatbot = gr.Chatbot()
+            btn = gr.Button(value="Save Settings")
+            btn.click(save_settings, None, tabs)
+        with gr.TabItem("Chat", id=1):
+            with gr.Accordion("Edit Customized Prompt", open=False):
+                user_prompt = gr.Textbox(
+                    lines=10,
+                    interactive=True,
+                    label="Prompt",
+                    value=
+                    """Compose a comprehensive reply to the query using the search results given. 
+Cite each reference using [p: Page Number] notation (every result has this number at the beginning). 
+Citation should be done at the end of each sentence. 
+If the search results mention multiple subjects with the same name, create separate answers for each. 
+Only include information found in the results and don't add any additional information. 
+Make sure the answer is correct and don't output false content. 
+If the text does not relate to the query, simply state 'Found Nothing'. 
+Ignore outlier search results which has nothing to do with the question.
+Only answer what is asked. The answer should be short and concise.""")
+            temperature = gr.Slider(
+                0.0,
+                1.0,
+                label="Temperature",
+                info="More focused 0.0 <---> 1.0 Highly creative",
+                value=0.5,
+                interactive=True)
+            chatbot = gr.Chatbot(interactive=True)
             question = gr.Textbox()
-            clear = gr.ClearButton([question, chatbot])
-
+            with gr.Row():
+                clear = gr.ClearButton([question, chatbot],
+                                       scale=1,
+                                       interactive=True)
+                submit_btn = gr.Button(value="Submit",
+                                       scale=2,
+                                       interactive=True)
+
+            submit_btn.click(
+                chatbot_respond,
+                inputs=[
+                    api_type, engine, openAI_key, openAI_base,
+                    openAI_API_version, url, file, temperature, user_prompt,
+                    question, chatbot
+                ],
+                outputs=[question, chatbot],
+            )
             question.submit(
                 chatbot_respond,
                 inputs=[
-                    question, chatbot, url, file, openAI_key, openAI_base,
-                    openAI_API_version
+                    api_type, engine, openAI_key, openAI_base,
+                    openAI_API_version, url, file, temperature, user_prompt,
+                    question, chatbot
                 ],
                 outputs=[question, chatbot],
             )
-
-
 demo.launch()