Skip to content

Improve the sharding and shuffling documentation - #6484

Merged
rostan-t merged 2 commits into
NVIDIA:mainfrom
rostan-t:shard-shuffling-doc
Sep 15, 2026
Merged

rostan-t merged 2 commits into
NVIDIA:mainfrom
rostan-t:shard-shuffling-doc

Conversation

@rostan-t

Copy link
Copy Markdown
Collaborator

Category:

Other (e.g. Documentation, Tests, Configuration)

Description:

The documentation didn't clearly state the differences between random_shuffle and shuffle_after_epoch, nor which readers allow combining them.

This PR adds a section about shuffling to clarify the related arguments in readers.

Additional information:

Affected modules and functionalities:

Sharding documentation becomes Sharding and Shuffling documentation.

Key points relevant for the review:

Tests:

  • Existing tests apply
  • New tests added
    • Python tests
    • GTests
    • Benchmark
    • Other
  • N/A

Checklist

Documentation

  • Existing documentation applies
  • Documentation updated
    • Docstring
    • Doxygen
    • RST
    • Jupyter
    • Other
  • N/A

DALI team only

Requirements

  • Implements new requirements
  • Affects existing requirements
  • N/A

REQ IDs: N/A

JIRA TASK: DALI-2763

@greptile-apps

greptile-apps Bot commented Sep 14, 2026 •

Copy link
Copy Markdown
Contributor

RetriggerConfidence Score: 5/5

The documentation-only PR appears safe to merge, with no outstanding correctness or repository-rule violations.

Summary

This PR expands the sharding guide into a combined sharding and shuffling guide and clarifies the relationships among random_shuffle, shuffle_after_epoch, shuffle_after_epoch_seed, and stick_to_shard.

  • Documents how epoch shuffling interacts with multi-pipeline sharding.
  • Describes reader-specific compatibility between shuffle modes.
  • Adds consistent pointers from affected reader argument documentation.
  • Clarifies that pipelines first compute a shared shuffled ordering and then partition it into shards.

Reviews (4) · Last reviewed commit: "Remove shuffle_after_epoch table and poi..."

Comment thread docs/advanced_topics_sharding.rst Resolved
@rostan-t

Copy link
Copy Markdown
Collaborator Author

!build

Comment thread dali/operators/reader/coco_reader_op.cc Resolved
@dali-automaton

Copy link
Copy Markdown
Collaborator

CI MESSAGE: [67804207]: BUILD STARTED

@dali-automaton

Copy link
Copy Markdown
Collaborator

CI MESSAGE: [67804207]: BUILD FAILED

keeps shard membership stable across epochs, so these options cannot be enabled together.

Whether ``random_shuffle`` can be combined with ``shuffle_after_epoch`` depends on the reader.
See the reader's ``shuffle_after_epoch`` argument documentation for supported combinations.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

👍

@dali-automaton

Copy link
Copy Markdown
Collaborator

CI MESSAGE: [67804207]: BUILD PASSED

Comment thread docs/advanced_topics_sharding.rst Resolved
@rostan-t
rostan-t merged commit 5eb891e into NVIDIA:main Sep 15, 2026
10 checks passed
@rostan-t
rostan-t deleted the shard-shuffling-doc branch September 15, 2026 16:35
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.

5 participants