You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The Adams2019 auto_scheduler uses beam search for searching, and a cost_model for evaluating schedules (or partial schedules) during search.
But cost model may wrongly rank schedules. I'm thus wondering how to use real hardware to evaluate schedules during the process of search? After reading the tutorial, it seems the using JIT mode we might achieve this. But I'm still not very clear how exactly this can be done, e.g., I cannot even find the input information in AutoSchedule.cpp.
Anyone knows how to do this? Is there any example might help?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
The Adams2019 auto_scheduler uses beam search for searching, and a cost_model for evaluating schedules (or partial schedules) during search.
But cost model may wrongly rank schedules. I'm thus wondering how to use real hardware to evaluate schedules during the process of search? After reading the tutorial, it seems the using JIT mode we might achieve this. But I'm still not very clear how exactly this can be done, e.g., I cannot even find the input information in AutoSchedule.cpp.
Anyone knows how to do this? Is there any example might help?
Thanks very much..
Beta Was this translation helpful? Give feedback.
All reactions