Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

clarification in sharding chapter #35

Open
eliben opened this issue Feb 20, 2025 · 0 comments
Open

clarification in sharding chapter #35

eliben opened this issue Feb 20, 2025 · 0 comments

Comments

@eliben
Copy link

eliben commented Feb 20, 2025

I may be misunderstanding something, but the Sharding chapter has this paragraph:

https://github.com/jax-ml/scaling-book/blob/main/sharding.md?plain=1#L188

The two sharding options seem to be identical:

  • $A[I_X, J] \cdot B[J, K] \to C[I_X, K]$ in the main flow of the text
  • $A[I_X, J] \cdot B[J, K] \to C[I_X, K]$ in parenthesis

Yet the text proceeds with "These two choices have different communication costs...", so it's not clear what the two different choices are. Is the second supposed to be C[I, K] on the rhs, or am I missing something obvious?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant