Explanation for pai_job_duration_estimate_100K.csv?

Hello,

Where exactly does the data in `clusterdata/cluster-trace-gpu-v2020/simulator/traces/pai/pai_job_duration_estimate_100K.csv` come from?

I'm assuming it's a subset of the `clusterdata/cluster-trace-gpu-v2020/data/` files, but it's unclear to me exactly how exactly it was generated.

Is it possible to create a complete trace like `pai_job_duration_estimate_100K.csv` to use with the 2020 simulator what has been posted to this repo?

Thanks for your help!