https://github.com/qoyyuum/mychatbot
Ollama (on AMD GPU) + OpenWebUI locally and easily with docker-compose
https://github.com/qoyyuum/mychatbot
docker docker-compose ollama webui
Last synced: about 2 months ago
JSON representation
Ollama (on AMD GPU) + OpenWebUI locally and easily with docker-compose
- Host: GitHub
- URL: https://github.com/qoyyuum/mychatbot
- Owner: Qoyyuum
- Created: 2024-11-24T15:48:31.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-11-26T05:05:33.000Z (over 1 year ago)
- Last Synced: 2025-06-05T01:39:00.085Z (about 1 year ago)
- Topics: docker, docker-compose, ollama, webui
- Homepage:
- Size: 74.2 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Funding: .github/FUNDING.yml
Awesome Lists containing this project
README
# MyChatBot
This is a local chatbot that you can run on your own machine, download a model or two and start chatting.
## Setting Up
The only pre-requisite is to install Docker and Docker Compose :whale:
### For AMD GPU
The image of ollama relies and uses a different tag for AMD GPU specifically:
```shell
ollama/ollama:rocm
```
### For CPU only
Use the tag `:latest` if you are running without a GPU
```shell
ollama/ollama:latest
```
### For NVidia GPU
Need to add a new data parameter in the `docker-compose.yml` to get all GPUs.
```yaml
...
data:
- gpus=all
```
Run the following command to get it started:
```shell
docker compose up -d
```
Refer to the [official ollama docs](https://github.com/ollama/ollama/blob/52bbad12f96e84f7d62c5dfdd7dbba2b10b37344/docs/docker.md) for more info
Then navigate to http://localhost:8080 (feel free to change the port for the openwebui in `docker-compose.yml` if 8080 is taken)
## Experimented with CPU only
If you're curious about how it works with only a CPU, you can read the experience here: https://buymeacoffee.com/qoyyuum/how-i-made-ai-system-admin-assistant
TL;DR Get a GPU