Evenly distribute complete chunks/batches across partial dataset splits

[laggui]

Checklist

• Confirmed that cargo run-checks command has been executed.
• Made sure the book is up to date with changes in this PR.

Related Issues/PRs

#2373
#3448

The current implementation prioritizes the number of workers, so the dataset is first split into num_workers partial subsets. Then each partial dataset is used to retrieve a batch. This can yield additional partial batches and doesn't match most users' expected behavior.

Changes

• Added PartialDataset::split_chunks which evenly distributes complete chunks/batches across multiple partial datasets.
• Changed multi-thread dataloader behavior to use split_chunks when BatchStrategy defines the expected batch size (default with fixed batch size)

Testing

Added unit tests

[codecov]

Codecov Report

❌ Patch coverage is 98.78049% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 62.52%. Comparing base (689aa77) to head (d94a956).
⚠️ Report is 4 commits behind head on main.

Files with missing lines	Patch %	Lines
...rates/burn-core/src/data/dataloader/multithread.rs	97.05%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3476      +/-   ##
==========================================
+ Coverage   62.49%   62.52%   +0.03%     
==========================================
  Files        1016     1016              
  Lines      113736   113816      +80     
==========================================
+ Hits        71080    71165      +85     
+ Misses      42656    42651       -5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.