AutoGPT4All is a simple bash script that sets up and configures AutoGPT running with the GPT4All model on the LocalAI server. This setup allows you to run queries against an open-source licensed model without any limits, completely free and offline.
- Clone this repository.
- Run
./autogtp4all.sh
to start the server.- If needed, run
chmod +x autogtp4all.sh
to make the script executable.
- If needed, run
❗️ Please note this script has been primarily tested on MacOS with an M1 processor. It should work on Linux and Windows, but it has not been thoroughly tested on these platforms.
Specify a custom URL for the model download step. By default, the script will use https://gpt4all.io/models/ggml-gpt4all-l13b-snoozy.bin
.
Example:
./autogtp4all.sh --custom_model_url "https://example.com/path/to/model.bin"
Uninstall the projects from your local machine by deleting the LocalAI
and Auto-GPT
directories.
Example:
./autogtp4all.sh --uninstall
- The script checks if the directories exist before cloning the repositories.
- On macOS, the script installs
cmake
andgo
usingbrew
. - If the
--uninstall
argument is passed, the script stops executing after the uninstallation step.
After executing ./autogtp4all.sh
, the script configures everything needed to use AutoGPT in CLI mode.
To run the GPT4All model locally, use cd LocalAI && ./local-ai --models-path ./models/ --debug
. The model will be available at localhost:8080
.
Then, navigate to the Auto-GPT
folder and run ./run.sh
to start the CLI.
Please note, to simplify this script and setup, by default the downloaded model is renamed to gpt-3.5-turbo
which serves as a base model that auto gpt expects for FAST_LLM_MODEL
parameter.
Special thanks to everyone who starred the repository ❤️