mlcommons / mlcommons/storage

[Training 3.0] Add interlock file to data directories

Open
#390 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

enhancement Training
Dominant language
Python
Stars
203
Forks
67
Avg merge
20m
Merged PRs (30d)
8

Description

In order to avoid situations where the wrong data directory is used, or when an old data directory is used, we should write an "interlock" file into the data directory that indicates the following information:

  • The version of the MLPerf tools used to create the data
  • The type of data stored in the data directory

When a data directory is used, the MLPerf tool should check the interlock file to make sure that the versions match up, and that the right type of data is present. If not, an error message should be displayed indicating the problem.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by locating the MLPerf tools that create and consume data directories, then trace where the directory version and data type are known. Define the interlock file contents and validation points; done means mismatched versions or data types produce clear errors instead of using the directory.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
data-engineering
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
45/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.