News for hackers. Stuff that matters.

A comment on DeepSpeed Chat: Easy, fast and affordable RLHF training of ChatGPT-like models.

Re:DeepSpeed Chat: Easy, fast and affordable RLHF training o... by scottydog51834 on Wednesday April 12, 2023 @07:48PM (#35548756)
This is a really cool step but as someone without the suggested GPU, it isn't easy or one click for me yet.

I am hoping that someone makes a very simple Jupyter notebook where I can enter my RLHF file and select a few other settings and just run (on AWS or Azure; willing to pay per fine-tuned model say $100-$500 for cloud credits + notebook access).

[ Parent | on HN ]
Re:DeepSpeed Chat: Easy, fast and affordable RLHF training o... by UncleEntity on Wednesday April 12, 2023 @08:52PM (#35549389)
TFA is fairly literally an advertisement for easy LLM training on azure GPUs.
[ Parent | on HN ]
Re:DeepSpeed Chat: Easy, fast and affordable RLHF training o... by scottydog51834 on Wednesday April 12, 2023 @09:18PM (#35549563)
What is TFA? Thanks.
[ Parent | on HN ]
Re:DeepSpeed Chat: Easy, fast and affordable RLHF training o... by DoreenMichele on Wednesday April 12, 2023 @09:29PM (#35549659)
The featured article.

Aka, the article under discussion.

(Though some people use other words for the F part.)

[ Parent | on HN ]
Re:DeepSpeed Chat: Easy, fast and affordable RLHF training o... by scottydog51834 on Wednesday April 12, 2023 @11:09PM (#35550425)
Thanks!
[ Parent | on HN ]
Re:DeepSpeed Chat: Easy, fast and affordable RLHF training o... by r3trohack3r on Thursday April 13, 2023 @10:31AM (#35556059)
Woah, I’ve never heard “featured” before! And suspect that many people mean that, I’ve just always taken the acronym as being crass and condescending.

Haven’t felt this way since I realized about 50% of my social network meant “In My Honest Opinion” and the other 50% meant “In My Humble Opinion”

[ Parent | on HN ]
Re:DeepSpeed Chat: Easy, fast and affordable RLHF training o... by kranner on Wednesday April 12, 2023 @09:28PM (#35549651)
"The Freaking Article", possibly
[ Parent | on HN ]
Re:DeepSpeed Chat: Easy, fast and affordable RLHF training o... by scottydog51834 on Wednesday April 12, 2023 @11:09PM (#35550426)
Thanks!
[ Parent | on HN ]
Re:DeepSpeed Chat: Easy, fast and affordable RLHF training o... by scottydog51834 on Wednesday April 12, 2023 @11:09PM (#35550423)
In case anyone else is curious, this yt video helps me understand how Azure ML works and I think with this understanding, combined with the Github read me, it will be doable. https://www.youtube.com/watch?v=yBVXR8G8Bg8
[ Parent | on HN ]