algorithmic-efficiency icon indicating copy to clipboard operation
algorithmic-efficiency copied to clipboard

verified publisher icon mlcommons
→

Metadata

MLCommons Algorithmic Efficiency is a benchmark and competition measuring neural network training speedups due to algorithmic improvements in both training algorithms and models.

Results 62 algorithmic-efficiency issues
Sort by recently updated
recently updated
newest added

submission_runner_test.py is not included in CI tests

The submission_runner_test.py test appears that is not currently included in the CI test suite.

Naeemkh avatar Naeemkh

🧪 Test

Recalibrate eval time intervals

Some workloads e.g. Criteo1TB spent 75% of the total runtime time on evaluations. We should re-analyze and potentially reconfigure the intervals.

priyakasimbeg avatar priyakasimbeg

← Metadata

278
Stars
53
Forks
Watchers

Owner

verified publisher icon mlcommons

Metadata

MLCommons Algorithmic Efficiency is a benchmark and competition measuring neural network training speedups due to algorithmic improvements in both training algorithms and models.

Back