auto-dl/k8s-deployment

Kubernetes configuration for deployment of Auto-DL

Introducing the "Futuristic" way of building Deep Learning Models

---

Auto-DL is just DL but without code. 🤩

Building deep learning models may take a lot of time. But not with AutoDL.
Automate your meta-work so that you can focus on building your models without the hassle of writing code.

• ✨ Use our interactive model builder and AI powered suggestion system to quickly whip out a working DL model. ✨

• ✨ Trained your model for multiple hours and found out that it's bugged? Not anymore. With AutoDL's AI ecosystem validate your model before you train it. ✨

• ✨ Host your model just as easily as you created it. ✨

Hacktoberfest Guidelines 📝

• If an issue is not "assigned," you can assume that no one is working on it (may NOT be true). So to be sure, you can ask the following:

  • Can I take this up?
  • Also, mention a rough deadline until which you plan to fix the issue
  • (Optionally) Potential fixes DO NOT CREATE a pull request FIRST if someone else is working on it then your efforts might be wasted. Claiming doable and unclaimed issues should be the priority

• Conventional commits in the PR title are a must

• For frontend, while creating a pull request attach a screenshot of the change

• Last but not least, one pull request to fix a typo is fine, but 5 pull requests to remove a stray whitespace is not (reference).

User Guidelines 📋

The docs are the perfect place for you to start getting used to Auto-DL.

Want to setup locally? 🏠

• Start with our getting started docs.

• The doc didn't help? Join our slack and ask your questions!

Just want to get a gist of "How to use the application"?

• You can directly jump to the user guide :)

Developer Guidelines 👩‍💻

The contributing guidelines hosts all the information about building Auto-DL from source, how to contribute code and documentation, who to contact about what, etc.

Get started with Auto-DL (For Developers) 🌐

Setup Auto-DL:

• Bare-metal way
• Using Docker

---

Features/Enhancements planned 💡

• Improve the UI and UX.

• Show model training related stats on the frontend.

• Visualization and data preprocessing steps.

• Model Explainability.

Contact 📧

You can find us on or on or join our community on Slack.

Have doubts? 💬

Read the FAQ or send us a mail at [email protected]

For more details, visit our website

---

Contributors 💯

Copyright and Licensing ©

GNU General Public License v3.0

k8s-deployment's Issues

Configure SSL for HTTPS ingress

Feature Request

Is your feature request related to an existing Issue or a new implementation?

No

Feature Description

As of now, all the communication between the services is using HTTP. It would be better to use HTTPS instead making it more secure. Create a configuration such that it supports multiple domains.

Possible Solution

Use a certificate issuer like letsencrypt. Create a new pod/cluster containing the letsencrypt service to automate the generation and renewal of SSL certificates.

Update or remove CodeQL & other workflows

Bug Report

Current Behavior

If codeql checks whether k8s secured, then great else remove the config
Same goes for other workflows & CIs

Make use of secrets and configmap for env variables

Feature Request

Is your feature request related to an existing Issue or a new implementation?

No

Feature Description

Instead of configuring the environment variables in multiple different configuration yamls use a configMap and secrets file to configure the required environment variables

Update linter to check YAML config & linting

Bug Report

Current Behavior

Python linting is checked right now

Expected behavior/code

YAML config checking needed