A free-to-use, locally running, privacy-aware chatbot. No GPU or internet required (internet only required for initial installation, can be closed after).
Introduction
GPT4All is an ecosystem to train and deploy powerful and customized large language models that run locally on consumer grade CPUs.
It can answer to anything, write creative stories, and is capable of answering your questions based on the documents that you provided. No internet connection is required, so you don't have to share your information with any external sources, nor do you need to preprocess the data to ensure it doesn't contain private information.
Requirements
Access to a Workspace with the role of Privileged Member or Accountable.
Upload the installer to your Workspace and follow the installation instructions below.
Installation instructions
1. Submit a ticket for full internet access via support.mydre.org. Please specify the reason (e.g. installation GPT4All), time period and Workspace name (e.g. dws-xxx-YYY). If you're not the Accountable, then please also include the approval of the Accountable. Be aware that opening up full internet comes with (security) risks. Incomplete requests are not processed.
2. GPT4All itself is not that large, however the models can take up quite some space, 3GB - 8GB. By default, it installs on the C: drive which also hosts the Operating System (OS). Clogging the C: drive might leave your VM unable to be used. Therefore, it might also be good to submit a ticket for an extra disk via support.mydre.org. Please specify the desired size of the disk and to which VM it needs to be attached. If you're not the Accountable, then please also include the approval of the Accountable. Be aware that extra disks will generate extra costs. Incomplete requests are not processed.
3. After confirmation by the Support Team,
initialize the extra disk within the VM according to the instructions provided by the Support Team.
4. GPT4All also requires quite some working memory. It might be good to resize the VM to a 4 cores, 16 GB RAM (Standard_B4ms) machine.
Be aware that resizing VMs will generate extra costs. See
Resize virtual machines.
5. Within the VM, go to the magnifying glass on the bottom left and search for 'Change proxy settings'.
6. Under 'Manual proxy setup' put the slider for 'Use a proxy server' on Off.
7. On the extra disk, create a folder called 'GPT4All'.
8. Install GPT4All on the extra disk and download a model on the extra disk. A recommended model is the groovy model.
9. Go to C:\Users\[yourusername]\AppData\Local\nomic.ai and rename the 'GPT4All' folder to something like 'GPT4All Models'.
10. Copy the 'GPT4All Models' folder to where you have installed GPT4All on the extra disk.
11. Move the groovy model to the 'bin' folder.
12. Request the closing of full internet access.
13. Under 'Manual proxy setup' put the slider for 'Use a proxy server' on On.
14. Enjoy GPT4all in your Workspace.
Related Articles
Host ALL your AI locally
Build an AI server in your myDRE Workspace and run all of AI models locally with a beautiful chat interface! It's customizable and super fast, much faster than anything else I've used. And again, it's local, it's private. You control it! Isn't that ...
Extra data disk request & initialization - Windows & Linux
Introduction The Z-drive (fileshare) is rather slow compared to a managed disk such as the C-drive. Although expansion of the C-drive is possible, it is good to keep in mind that the Operating System (OS) is also installed on the C-drive. It is not a ...
AI, LLMs, GPTs and myDRE
When using AI, LLMs, and GPTs, especially when using external services, consult your CISO, DPO, and Legal. At the very least go over all the fine prints! AI, LLMs, and GPTs will most likely, or even already are, part of processing data in general. ...
Perplexity.ai
When using AI, LLMs, and GPTs, especially when using external services, consult your CISO, DPO, and Legal. At the very least go over all the fine prints! More information on https://www.perplexity.ai/ To be used as an external service. If a ...