Closed daniel-thom closed 2 years ago
When running the SFO dataset I noticed that all long-running jobs were bunched together in the same batches and all short-running jobs were bunched together. This will mitigate that problem until we estimate the runtime of each job.
@wenbowangnrel FYI
That makes sense, for example, one large system will have many scenarios run together in one batch and all of them can be large jobs.
When running the SFO dataset I noticed that all long-running jobs were bunched together in the same batches and all short-running jobs were bunched together. This will mitigate that problem until we estimate the runtime of each job.