Alfred is a LangChain and GPT-3.5-turbo powered personal assistant for your PC!
Image generated with Blue Willow AI on Discord
- Better ALFRED, better name!
- Added chat window GUI using tkinter for quiet interaction with the bot (Can also be launched with chat.py)
- Updated most variables to use settings.ini for easy update
- Added a Settings window so that user can update settings from the GUI
- Added a Keys window off of settings to update API keys
- Added hotkey to main.py to launch chat window
- Moved from ConversationBufferMemory to TokeBufferMemory to clear out memory when it hits 1000 tokens
- Introduced a probably pretty hacky solution for the Wikipedia tool to stop constantly hitting the token limit
- Updated the "Pinecone tool" so that user can specific their own tool name and description depending on what they use pinecone for
- Added additional error handling to clear the current memory when you get an exception regarding the token limit so that you don't need to restart
- Added a script for testing voices for pyttsx3
- Added a notebook to help with ingesting into Pinecone (I have not tested all of the methods)
- Clone the repo, cd into the root, create an environment based on Python 3.10 and activate it:
git clone https://github.com/masrad/ALFRED.git
cd ALFRED
1.1 If you have multiple versions of Python installed, specify the location of 3.10:
C:\PATHTOYOUR\Python310\python.exe -m venv venv
.\venv\Scripts\activate
1.2 If you only have 3.10 (verified with python --version) then you don't need a path:
python -m venv venv
.\venv\Scripts\activate
- If you want to use GPU with pytorch for Bark (recommended if you have a GPU), get your download string here:
https://pytorch.org/get-started/locally/
Since you're probably using pip on Windows, this should work for you, double check if you're unsure:
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117
2.1 Install ffmpeg, as whisper requires this to work:
# on Ubuntu or Debian
sudo apt update && sudo apt install ffmpeg
# on Arch Linux
sudo pacman -S ffmpeg
# on MacOS using Homebrew (https://brew.sh/)
brew install ffmpeg
# on Windows using Chocolatey (https://chocolatey.org/)
choco install ffmpeg
# on Windows using Scoop (https://scoop.sh/)
scoop install ffmpeg
# on Windows using the more manual way
https://www.geeksforgeeks.org/how-to-install-ffmpeg-on-windows/2.2 Next, install the required packages:
pip install -r requirements.txt
- Create a .env file or rename the example in the root and update the information:
OPENAI_API_KEY=your-key-here
GOOGLE_API_KEY=your-key-here
GOOGLE_CSE_ID=your-key-here
WOLFRAM_ALPHA_APPID=your-key-here
OPENWEATHERMAP_API_KEY=your-key-here
ZAPIER_NLA_API_KEY=your-key-here
PINE_API_KEY=your-key-here
IMPORTANT: If you are not using a certain tool, you still need SOMETHING in the environment variable, otherwise it will error. Just leave your-key-here if you're not using a tool and ensure it's disabled in settings.ini
- If you'd like, you can update the settings.ini before launching, most things are turned off by default
- Run the main script:
python main.py
-
Speak the Wake Word "Alfred"
-
Once you hear a chime, Speak your query or command.
-
The assistant will respond with synthesized speech.
-
To end the conversation, say "Thank you" or a similar phrase.
- Run the main script:
python main.py
-
Press the hotkey (default is ctrl+shift+9)
-
A chat window will popup allowing you to interact with the bot over text
-
You can access settings from this GUI as well
Note: You can also launch only the chat window portion, and not even use the voice assistant by running the chat alone:
python chat.py
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.
If you would like to use any of the tools I have included, you will need to follow a few instructions
- Create a Zapier Dev account
- Go to https://zapier.com/l/natural-language-actions and request API access (it's pretty instant)
- Begin making Actions in Zapier, choose "Let AI Guess" for most fields, and the AI will determine what should go in that field
- Copy your key into the .env file at ZAPIER_NLA_API_KEY
- You're done. You can use complex chains such as "Summarize the last email I got about the TPS reports and add it to my todo list so that I will remember to review them tomorrow. Also ask Greg in Teams what he thinks about the TPS reports and add a funny gif to the message too."
- Get Google API key
- Go to https://console.cloud.google.com/apis/dashboard while signed in
- Create a New Project
- Name the project and assign a location (if applicable)
- Select "Credentials" on the side bar, and click CREATE CREDENTIALS at the top of the page
- Click Create an API Key and copy the key that it creates to your .env file under GOOGLE_API_KEY
- Click on "Enabled APIs & Services" on the sidebar, and click ENABLE APIS AND SERVICES at the top of the page
- Search for, and click "Custom Search API", click ENABLE
- Create a custom search engine
- From the Google Custom Search homepage ( http://www.google.com/cse/ ), click Create a Custom Search Engine.
- Type a name and description for your search engine.
- Under Define your search engine, in the Sites to Search box, enter at least one valid URL (For now, just put www.anyurl.com to get past this screen. More on this later ).
- Select the CSE edition you want and accept the Terms of Service, then click Next. Select the layout option you want, and then click Next.
- Click any of the links under the Next steps section to navigate to your Control panel.
- In the left-hand menu, under Control Panel, click Basics.
- In the Search Preferences section, select Search the entire web but emphasize included sites.
- Click Save Changes.
- In the left-hand menu, under Control Panel, click Sites.
- Delete the site you entered during the initial setup process.
- Enter the Search engine ID you get from the Basics page into the .env file at GOOGLE_CSE_ID
- Create a developer account and get your APP ID from here: https://developer.wolframalpha.com/
- Enter the Application ID from Wolfram Alpha into the .env file at WOLFRAM_ALPHA_APPID
- Create a developer account and sign up for an API key here: https://openweathermap.org/api/
- Follow the prompts to make an API, as long as you aren't asking too often a free acount is fine
- Enter the API into the .env at OPENWEATHERMAP_API_KEY
- Added Pinecone as a tool for getting answers based on the latest LangChain docs site
- Added a notebook so that I can more easily vectorize stuff
- Created tools folder for the voice checker and the pinecone notebook
- Natural language integration with Zapier, enabling Alfred to access, use and manage over 5000 applications
- Speech recognition using the speech_recognition and whisper libraries
- Speech synthesis using pyttsx3 or Bark
- Uses LangChain to use tools to answer questions with more current data
- Activate with a spoken wake word "Alfred" - user editable
- Terminates the conversation when the bot says "You're welcome" or a similar phrase
- Configurable tools for LangChain, so if you don't have an API for a tool or don't want the tool, you can disable it
- Added ability to switch between pyttsx3 and a new project called Bark for voice synthesis
- Removed hotkey so that I don't have to uncleanly thread both listeners
- Added intro.wav so that you know when it's done loading and is waiting for wake word (it was generated with Bark, that should give you an idea of the voice quality)
- Made Barks history_prompt (also known as the voice) changable in one spot
- Exceptions now loop back to wake word detection instead of crashing
This project is licensed with GNU General Public License Version 3, see LICENSE.md for more information
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent