Back to all discussions

  •   over 1 year ago

Specific Model or Technology Stack

Is there any specific model we have to make or technologies to use?
Is there a specific way we need to present our work? like eg. in a web app etc..

  • 10 comments

  • Manager   •   over 1 year ago

    Hello Mairya and welcome!
    The main requirement for this hackathon is to create (or update) a Gen AI or an AI Retrieval Augmented Generation (RAG) application that makes use of the Red Hat OpenShift AI environment on AWS with Intel’s Xeon Processor AMX features.

    Access to the required Red Hat OpenShift AI environment mentioned above will be given out on October 21st (submission start date) to those accepted through the Environment request form found here-> https://redhat-intel.devpost.com/details/sandbox.

    For more details on what to submit and the main requirement, please review the *Requirements* section on the Overview page here-> https://redhat-intel.devpost.com/#challenge-requirements

    Please let me know if you have any questions. Best of luck!

    Thank you!

  •   •   over 1 year ago

    So if i made a chatbot that uses rag as well would that be fine?

  • Manager   •   over 1 year ago

    Yes, it can be any application as long as it uses the environment/tools provided.

  •   •   over 1 year ago

    Hi Michelle, when I click on the environment request form link, I'm getting a blank page.

  • Manager   •   over 1 year ago

    Uh Oh! It does take a moment for it to load as it is an embedded form. You can go directly to the form here as well: https://challengepost.wufoo.com/forms/mbcb8km1qlam7m/
    Best of luck!

  •   •   over 1 year ago

    @Michelle Will there be any AWS credits provide? AWS is a bit expensive these days.

  • Manager   •   over 1 year ago

    Hi Robert, Good question! Let me ask the team if that will be included in the official hackathon environment.

  • Manager   •   over 1 year ago

    Hi again, We provide a set amount of resources within the official hackathon environment. You may need to change your project scope or specs to fit within the provided environment.
    Thank you!

  •   •   over 1 year ago

    @MichellDesrochers This may end up being a key issue. I just checked out the Red Hat Openshift Slack and the first thing I saw is this message. Note I don't know BBarber at all, just reporting this:

    From: @Bbarber

    1:47 PM

    Hey, I'm here from the devpost hackathon and the OPEA linux foundation group, I just want to mention that your openshift environment is a pain in the @#$% to use, and the resource limits are so low, that I cant do any meaningful work on them. I was going to generate embeddings for a bunch of arxiv papers, but this seems like an unrealistic goal with the resources you provide.

    1:50
    also from an ease of use perspective, huggingface is much easier to use, and from a power-user perspective I would just rather spin up a fully fledged VM or container.

    1:51
    also it appears that spinning up a habana jupyter notebook, does not actually attach a gaudi gpu to the instance, and the tutorial that you provided for openvino is also impossible to run, because you have not enabled the openvino connector

  • Manager   •   over 1 year ago

    Hi Robert, Thank you for letting me know! We are not using their Slack for this hackathon so I haven't seen those messages but will let the team know.

    Please note that we have specific channels set up in the Devpost Discord community (https://discord.com/invite/HP4BhW3hnp). Once you are in you can checkin to this hackathon's specific channels in the check in channel here (https://discord.com/channels/861966823054639134/1283865621318205450/1283883446153777215)

    Thanks again!

Comments are closed.