Readme refactor & Add Appache License #7
Conversation
|
|
||
| Deploy AI models to Kubernetes using LocalAI, vLLM, DeepSpeed-MII, NVIDIA Triton and custom Docker images. |
There was a problem hiding this comment.
I'd like to keep a one line description which specifies what the operator does with the minimum number of words. We could maybe add "easily" at the start?
There was a problem hiding this comment.
Can you suggest exact sentence you would like to see?
There was a problem hiding this comment.
That sentence with "Easily" on the front would be fine IMO
| - K8s cluster | ||
| - Helm | ||
| - Ingress controller (e.g. Traefik) | ||
| - Nvidia GPU Operator |
There was a problem hiding this comment.
The last three are recommendations. You can install without Helm, you don't need ingress if you are serving the model internally and you can use CPU for embeddings.
docs/deployment.md
Outdated
|
|
||
| ## Installation | ||
| After setting up K8s cluster, install [Nvidia GPU Operator](https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/latest/getting-started.html) and [Traefik](https://doc.traefik.io/traefik/getting-started/install-traefik/#use-the-helm-chart) as an ingress controller. | ||
| Note that Nvidia GPU Operator is required for GPU support and Traefik is required for handling ingress traffic. |
There was a problem hiding this comment.
These are recommendations, but it is good to point out this is what we have tested it with
There was a problem hiding this comment.
Also here if you can precise how exactly you would like this to look like?
There was a problem hiding this comment.
For instance add "optionally", before install and change "Traefik is required" to "Traefik is on option"
| "temperature": 0.1, | ||
| "max_tokens": 50 | ||
| }' | ||
| ``` |
docs/contributing.md
Outdated
|
|
||
| ### New Features | ||
|
|
||
| - For new features, we ask that you **open an issue first** to discuss your ideas with the project maintainers. |
There was a problem hiding this comment.
I don't mind if people make a PR first. I think the main thing is that they understand that without buy-in it's possible that we won't review the code, but just dismiss it on principle.
This refactors Readme file and adds Apache license file.
@tiero @richiejp please review.