Loading video...

Video Failed to Load

Go Home

mlx distributed uses mpi for machine to machine communication, and this is how you can handle uneven distribution of model layers. Thanks Awni Hannun for the help

31,265 views • 1 year ago •via X (Twitter)

11 Comments

abdou's profile picture
abdou1 year ago

@awnihannun v interesting, would love to know how it was ordering the machines before you specified it in the rank file. would be funny if it was just sorting the host names 😅

Alex Ziskind's profile picture
Alex Ziskind1 year ago

@awnihannun the mysteries of mpi.

Rainmaker's profile picture
Rainmaker2 years ago

Can Machine Learning beat the market? Check out this post on my free Substack where I share code and commentary for an XGBoost model and a Random Forest model that both deliver powerful performances.

Mason James.3.2-12b_q4_K_S's profile picture
Mason James.3.2-12b_q4_K_S1 year ago

@awnihannun Repo Prompt! ❤️ cc @pvncher

Art Friedman's profile picture
Art Friedman1 year ago

@awnihannun thanks for sharing! Great question for you, m3 ultra 512 vs multiple Mac mini m4 with the same budget, which option is better for LLM context?

Joe Burnett's profile picture
Joe Burnett1 year ago

@awnihannun Awesome!

Mid Season Group's profile picture
Mid Season Group1 year ago

@awnihannun @Apple engineering should be ashamed.

Leonard Sullivan's profile picture
Leonard Sullivan1 year ago

@awnihannun I love these mini racks.

Alex Ziskind's profile picture
Alex Ziskind1 year ago

@awnihannun yep It’s a DeskPi- @Mandy_Li888

moskstraumen's profile picture
moskstraumen1 year ago

@awnihannun Tell us more about the technical aspects, if you want!

CodeBreaker's profile picture
CodeBreaker1 year ago

@awnihannun love to see it

Related Videos