WhiteOwlEd

hace 2 añosdiscuss

With OpenAI and other LLMs, web development is accelerating. For example, I put together a AI call center demo ( https://www.youtube.com/watch?v=Vv7mI_qRrhE ) by using Open AI o1-preview. There I could take a lot of different files on typescript and backend server stuff written in python. I would add logs into the mix to make one massive prompt, and then I would let the AI work on reasoning in the cases where I needed to accelerate the writing of additional code.

WhiteOwlEd

hace 2 añosdiscuss

Building on this, Human preference optimization (such as Direct Preference Optimization or Kahneman Tversky Optimization) could be used to help in refining models to create better data.

I wrote about this more recently in the context of using LLMs to improve data pipelines. That blog post is at: https://www.linkedin.com/posts/ralphbrooks_bigdata-dataengin...

WhiteOwlEd

hace 2 añosdiscuss

The author is actually doing something that will help with the job search, and that is reaching out by methods other than the resume.

Having a blog on the front page (coupled with the value that he can bring) should at least give him a few warm leads for job opportunities.

WhiteOwlEd

hace 3 añosdiscuss

If you are using no-code solutions, increasing an "idea" in a dataset will make that idea more likely to appear.

If you are fine-tuning your own LLM, there are other ways to get your idea to appear. In the literature this is sometimes called RLHF or preference optimization, and here are a few approaches:

Direct Preference Optimization

This uses Elo-scores to learn pairwise preferences. Elo is used in chess and basketball to rank individuals who compete in pairs.

@argilla_io on X.com has been doing some work in evaluating DPO.

Here is a decent thread on this: https://x.com/argilla_io/status/1745057571696693689?s=20

Identity Preference Optimization

IPO is research from Google DeepMind. It removes the reliance of Elo scores to address overfitting issues in DPO.

Paper: https://x.com/kylemarieb/status/1728281581306233036?s=20

Kahneman-Tversky Optimization

KTO is an approach that uses mono preference data. For example, it asks if a response is "good or not." This is helpful for a lot of real word situations (e.g. "Is the restaurant well liked?").

Here is a brief discussion on it:

https://x.com/ralphbrooks/status/1744840033872330938?s=20

Here is more on KTO:

* Paper: https://github.com/ContextualAI/HALOs/blob/main/assets/repor...

* Code: https://github.com/ContextualAI/HALOs

WhiteOwlEd

hace 3 añosdiscuss

Data scientist here who spent a couple of years working with Unreal (to produce high end data visualizations). Here are my thoughts

> Blueprints suck! Not really. Think of Blueprints like python. Its good for routing and keeping track of things at a high level. Think of C++ as handling things at a lower level.

> I heard you need to start with blueprints. Not really. After going through the basic tutorial that Unreal Sensei has on YouTube (https://youtu.be/gQmiqmxJMtA?si=TqBiiIe12M5hiCda) , it is better to do a mix of blueprints and C++ if you have any programming background.

> I don't know what to use for the IDE. I used Rider for Unreal Engine and it has good integration into Unreal Engine.

> So when do you use C++? When I was doing data vis of census data, I needed a way to load in 10,000 data points into memory. The "out of the box" tools for Unreal didn't support this, so custom C++ was the way to go.

> But really, if I just want to get started with Unreal and want official tutorials, where do I go?

After going through Unreal Sensei, I looked at at https://dev.epicgames.com/community/unreal-engine/getting-st..., there are a ton of tutorials there for game developers.

Also, a year ago I put together an online course that looked at how to ramp up on Unreal Engine. The course ("Data Visualization in the Metaverse") is ideal if you already have a programming background. I put the course out on YouTube for free (https://www.youtube.com/playlist?list=PLKH3Xg62luIgPaB4fiFuT...) and happy to answer any questions about it.

comments