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

add shuffling support (PR 6 of N) #9

Open
wants to merge 1 commit into
base: bkmartinjr/add-dataloader-api
Choose a base branch
from

Conversation

bkmartinjr
Copy link
Member

Add shuffling per the original CZI API - a random gather of small contiguous regions as a single I/O buffer, followed by a per-I/O batch shuffle.

The primary deviation from the CZI implementation:

  • add explicit epoch support, so that the shuffle order by epoch changes
  • fix RNG handling so that all worker processes agree on the random shuffling state (ie., ensure same shuffle across all partitions).

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

Successfully merging this pull request may close these issues.

1 participant