Changhai improve workload layer #157
Closed
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary
This PR contains commits belong to #151. New commits start from 39c83b0. This PR is tested with chakra updated to mlcommons/chakra#76. Not sure if original chakra works.Added attr checking before issuing nodesDo not supported because pending chakra PR for optional attrs not merged.Fixed bug of queue_levels when queues.size()=1Covered in other PRTest Plan
Using runs/example, assuming workload generated with new chakra(with #76, also after updated text converter in #78 to make sure the generated workloads is correct).
Positive case:
and has output
negative caseBecause there is not attr check thus no negative case for now.
Additional Notes
Should be merged after PR #151 to fast-forward a clean history.Fixed the bug: in previous example the communication will be routing to
skip_invalidbecausenode->is_cpu_op()==Trueandnode->runtime()==0, according toworkload.cc:130it will be invalid, which should not be correct.