AI-powered Jupyter Notebook
Thread is a Jupyter alternative that integrates an AI copilot into your Jupyter Notebook editing experience.
Best of all, Thread runs locally and can be used for free with Ollama or your own API key. To start:
pip install thread-dev
To start thread-dev, run the following
thread
![image](https://private-user-images.githubusercontent.com/18422723/338213122-93a16931-4236-4ab0-b5b5-673100af2ca0.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.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.sw_hxAbkuUYHWz3Td_c4CKmCwC8RGciUFI1fKDEDJKw)
ThreadDemo720.mp4
These are some of the features we are hoping to launch in the next few month. If you have any suggestions or would like to see a feature added, please don't hesitate to open an issue or reach out to us via email or discord.
- Add co-pilot style inline code suggestions
- Data warehouse + SQL support
- No code data exploration
- UI based chart creation
- Ability to collaborate on notebooks
- Publish notebooks as shareable webapps
- Add support for Jupyter Widgets
- Add file preview for all file types
Eventually we hope to integrate Thread into a cloud platform that can support collaboration features as well hosting of notebooks as web application. If this sounds interesting to you, we are looking for enterprise design partners to partner with and customize the solution for. If you're interested, please reach out to us via email or join our waitlist.
To run the repo in development mode, you need to run two terminal commands. One will run Jupyter Server, the other will run the NextJS front end.
To begin, run:
yarn install
Then in one terminal, run:
sh ./run_dev.sh
And in another, run:
yarn dev
Navigate to localhost:3000/thread
and you should see your local version of Thread running.
If you would like to develop with the AI features, navigate to the proxy
folder and run:
yarn install
Then:
yarn dev --port 5001
You can use Ollama for a fully offline AI experience. To begin, install and run thread using the commands above.
Once you have run thread, in the bottom left, select the Settings icon:
Next, select the Model Settings setting:
This is what you will see:
Navigate to Ollama and enter your model details:
Use Ctrl / Cmd + K and try running a query to see how it looks!
We initially got the idea when building Vizly a tool that lets non-technical users ask questions from their data. While Vizly is powerful at performing data transformations, as engineers, we often felt that natural language didn't give us enough freedom to edit the code that was generated or to explore the data further for ourselves. That is what gave us the inspiration to start Thread.