merrymercy commented on a change in pull request #6671:
URL: https://github.com/apache/incubator-tvm/pull/6671#discussion_r511903881
##########
File path: python/tvm/auto_scheduler/measure.py
##########
@@ -950,26 +1017,30 @@ def rpc_runner_run(
res : List[MeasureResult]
The measure results of these MeasureInputs.
"""
- global GLOBAL_RUN_ARGUMENTS
- GLOBAL_RUN_ARGUMENTS = (
- inputs,
- build_results,
- key,
- host,
- port,
- priority,
- timeout,
- number,
- repeat,
- min_repeat_ms,
- cooldown_interval,
- enable_cpu_cache_flush,
- verbose,
- )
-
assert len(inputs) == len(build_results), "Measure input size should be
equal to build results"
- pool = NoDaemonPool(n_parallel)
- tuple_res = pool.map(rpc_run_worker, range(len(build_results)))
+ # This pool is not doing computationally intensive work, so we can use
threads
Review comment:
Did you benchmark the speed of ProcessingPool vs. ThreadPool?
For the comment, is this pool "not doing computational intensive work" or
"not doing computational intensive work in python"?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
