cerebros-core-algorithm-alpha icon indicating copy to clipboard operation
cerebros-core-algorithm-alpha copied to clipboard

from-230-add-system-metrics-telemetry-to-mlflow-tracking

Open david-thrower opened this issue 2 months ago • 0 comments

On the Generative model proof of concept: Add system telemetry using MlFlow system metrics.

  • I want to collect metrics on the CPU / GPU and RAM pressure, etc.
  • This will be a crucial factor especially when multiple models are "optimal" from an ML metric or loss perspective.
  • This information will be useful for planning what hardware we need for what scale of trials and plan capabilities and model sizes.

From #240

david-thrower avatar Sep 20 '25 19:09 david-thrower