Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
33.61k
stars
5.71k
forks
source link
[metrics] Report metrics to be used for debugging load balancing issues #14369
Open
stephanie-wang opened 3 years ago
Describe your feature request
Add some metrics for the raylet scheduler.