-
Notifications
You must be signed in to change notification settings - Fork 43
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
article: Optimizing GPU Utilization in Containerized Environments #38
base: main
Are you sure you want to change the base?
Conversation
28abb6b
to
d4d430d
Compare
@nkkko Just opening it as a draft as I have worked on almost everything except a few and need to test this through daytona env and manually it is working. It is blocked by daytonaio/daytona#940 Once this issue is fixed, I will push my code to the repository of this project by testing it thoroughly and documenting the challenges encountered. I have been actively following up in Slack with @Tpuljak on this. |
articles/20240820_optimizing-gpu-utilization-in-containerized-environments.md
Show resolved
Hide resolved
@varshith257 this PR fails DCO checks. Equally please give an update. |
@mojafa Sure! Will update with the same |
@varshith257 any updates? |
@varshith257 your PR fails DCO checks |
6c61e3c
to
e96dd84
Compare
Signed-off-by: Vamshi Maskuri <[email protected]>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@varshith257 Please create a sample project that addresses the issue and use the article to guide us through how you made your project run.
Add screenshots and be very detailed.
Avoid AI generated content.
|
||
## Introduction | ||
|
||
With the rapid adoption of deep learning and large language models (LLMs), the need for efficient GPU utilization in containerized environments has become more pressing than ever. Containers provide a lightweight, portable solution for deploying applications, but when it comes to GPU-intensive tasks like LLM inference or fine-tuning, careful setup and configuration are crucial to achieving optimal performance. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@varshith257 this sounds really AI generated
|
||
# | ||
|
||
Let's adjust the draft to align it more closely with the suggested format: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
ChatGPT...😅
Please review this whole article again.
|
||
Example configuration: | ||
|
||
```plaintext |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
no need for plaintext, just add ascreenshot of this step
Install the NVIDIA Container Toolkit to enable GPU support in Docker: | ||
|
||
```bash | ||
sudo apt-get update && sudo apt-get install -y nvidia-container-toolkit |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm using MacOs, I don't think this will work locally or even in devcontainer.
Daytona manages the initialization and configuration of the workspace, displaying progress information in the terminal. Once completed, the workspace will be ready for use. Example output during workspace creation: | ||
|
||
```plaintext | ||
WORKSPACE | ✓ Request submitted |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
add as a screenshot please
|
||
### Create a Dockerfile for nanoGPT | ||
|
||
Next, create a Dockerfile tailored to run nanoGPT with GPU support. Here’s an example Dockerfile: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I suggest you create a project, push to github, do daytona create your-repo-url
, and share screenshots or steps to replicate yout project.
you can easily define the dependencies or packages devconatiner.json file.
you can check out this PR did there's step by step: #117 |
Pull Request Template
Closes #8
/claim #8
Writer's Checklist
Follow Writing Structure
code elements
where appropriate.Fact-Check
Assets
/assets
folder.YYYYMMDD_title_of_the_article_IMG_NAME_NO.png
.Interlinking
CTRL+F
to search for relevant keywords on:Glossary/Definitions
Review and Edit