README.md

metadata

title: ChatGPT Plugins In Gradio
emoji: 💻
colorFrom: green
colorTo: gray
sdk: gradio
sdk_version: 3.35.2
app_file: app.py
pinned: true
license: mit
duplicated_from: ysharma/ChatGPT-Plugins-in-Gradio

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Steps to add new Plugins to your Gradio ChatGPT Chatbot

1. Acquire the API Endpoint

   • You need an API which you can query, and for this example let's consider using a text-to-speech demo hosted on Huggingface Spaces.
   • API Endpoint: https://gradio-neon-tts-plugin-coqui.hf.space/

2. Create a Function to Query the API

   • You can access any Gradio demo as an API via the Gradio Python Client.

   from gradio.client import Client
   
   def texttospeech(input_text):
       client = Client("https://gradio-neon-tts-plugin-coqui.hf.space/")
       result = client.predict(
           input_text,	# str  in 'Input' Textbox component
           "en",	    # str  in 'Language' Radio component
           api_name="/predict"
       )
       return result
   

3. Describe the Function to GPT-3.5

   • You need to describe your function to GPT3.5/4. This function definition will get passed to gpt and will suck up your token. GPT may or may not use this function based on user inputs later on.
   • You can either use the Gradio demo for converting any given function to the required JSON format for GPT-3.5.
     • Demo: Function to JSON
     • Or, you can create the dictionary object on your own. Note that, the correct format is super important here.
   • MAke sure to name your JSON object description as <function_name>_func.

   texttospeech_func = {
       "name": "texttospeech",
       "description": "generate speech from the given input text",
       "parameters": {
           "type": "object",
           "properties": {
               "input_text": {
                   "type": "string",
                   "description": "text that will be used to generate speech"
               }
           },
           "required": [
               "input_text"
           ]
       }
   }
   

4. Add Function and JSON Object Details

   • Add the function definition and description to the gpt_function_definitions.py file (simply copy and paste).
   • dict_plugin_functions is a dictionary of all available plugins. Add your plugin information to this dictionary in the required format.

   'texttospeech_func': {
       'dict': texttospeech_func,
       'func': texttospeech
   }
   

5. Update the Chatbot Layout

   • Go to the Blocks Chatbot layout and add a new checkbox for your plugin as:

   texttospeech = gr.Checkbox(label="📝🗣️Text-To-Speech", value=False)
   

   • Add the new checkbox component to your submit and click events for your chatbot and to the predict function accordingly.
   • And also to the plugins list in predict

     plugins = [music_gen, stable_diff, image_cap, top_news, texttospeech]
     

Thats it! you are have added your own brand new CHATGPT Plugin for yourself. Go PLAY!!