Office hours #3 and #4:

https://www.loom.com/share/b57b995f79364da68da1d1826a766794

Links:

- https://gist.github.com/alexeygrigorev/ea13d4a70cd7afb1f3ab70aa53094e46
- https://github.com/DataTalksClub/datatalksclub.github.io

We've started module 5

Materials: https://github.com/DataTalksClub/llm-zoomcamp/tree/main/05-orchestration

We're still working on homework questions, so the forms for the previous homework and the workshop remain open

We finally created homework for the orchestration module:

https://github.com/DataTalksClub/llm-zoomcamp/blob/main/cohorts/2024/05-orchestration/homework.md

You have one week to complete it.

Have fun!

Next week we will start working on projects.

Here's more information: https://github.com/DataTalksClub/llm-zoomcamp/blob/main/project.md

Two things are important there:

- Evaluation criteria - this is how your project will be evaluated
- Potential ideas - in case you don't have any idea yet, you can use the list from the page for getting some inspiration

Have you already thought about your project? What is it going to be?

If you have problems coming up with an idea, ask in Slack!

While we're waiting for module 6 materials, we also started working on bonus materials - an end-to-end project example that may be useful for your projects

Here's the first video - https://www.loom.com/share/999cc828f62c4a5787d2e21dcf181eb6

It's optional, so you don't have to do it. There will be no homework for this extra module (apart from doing your projects)

Let us know what you think about it and if you have any questions that we should also cover there.

We won't have homework for module 6 and on Monday we start the project

https://github.com/DataTalksClub/llm-zoomcamp/blob/main/project.md

We almost finished the bonus module with an end-to-end project example

https://github.com/DataTalksClub/llm-zoomcamp/tree/main/07-project-example

It won't get the highest possible score, but hopefully it gives you a good idea of what to do

(We'll start uploading the videos to YouTube soon)

Also on Monday at 17:00 we will have office hours. You can ask your questions here

https://app.sli.do/event/bmZD6KbpuBqxPqjgQAsemF

See you soon!

A reminder that we have office hours today in ~90 minutes

We will talk about projects

Ask your questions here: https://app.sli.do/event/bmZD6KbpuBqxPqjgQAsemF

Stream: https://www.youtube.com/watch?v=pA9S1mTqAwU
Questions: https://app.sli.do/event/bmZD6KbpuBqxPqjgQAsemF

Starting in 4 minutes

As a follow-up to yesterday's stream, I recorded another video:

https://www.loom.com/share/d1361ff3db574d43ae86c2cde700cdb3

I talk about different ways to approach cases when your knowledge base is

- multiple articles
- one article or youtube video
- a book
- slide decks from your uni studies

https://lu.ma/zqzpbod1

Could be quite interesting for many of you

We've updated the end date of the competition: now it's September 30

Have fun!

The link: https://www.kaggle.com/competitions/llm-zoomcamp-2024-competition/overview

How is it going with your projects? Everyone is on track?

A talk about evaluating LLMs:

https://www.youtube.com/watch?v=ac6ZB5QEwGU

Watch now or later in replay

By the way, the deadline for the first project attempt was pushed to Monday end of day. Enjoy!

OpenAI has new reasoning models

https://openai.com/index/introducing-openai-o1-preview/

Perhaps they could be useful for the competition

The deadline for submitting the project for attempt 1 is over

We closed the form and assigned each one of you 3 projects to review. Don't forget that in order to receive a certificate you need to complete this peer review process.

You can find them here: https://courses.datatalks.club/llm-zoomcamp-2024/project/project1/eval

It's a great chance to learn from your peers!

Have fun and happy learning!

The deadline for submitting project attempt 2 is extended till October 7 EOD

Note that there will be no further extensions, so use the extra week wisely, and if you haven't started working on your project or haven't selected a dataset, do it now.

As always, if you have any problems, we will help you in slack.

Happy learning!

We added "bonus points" (up to 3 extra project points) that you can award to the projects you evaluate for something cool - when they went the extra mile in their project

For example:

- voice recognition
- CI/CD
- anything else you find interesting

Be sure to describe in feedback why you gave bonus points

If you already finished your evaluations, but you think the projects deserve extra points, you can update your submission

We've scored the projects from attempt #1! 56 of you have passed - congratulations!