09132024 upstream fwd bwd optimized #74
Draft
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…u/hip_kernel_inc/
This reverts commit cb8cfc9. This is not due to missing files but broken #include paths
* Rename common header * Delete redundant files related to fwd pass * Add hip bwd kernel template generation * Remove old instantiation code, clean-up leftovers (except debug prints)
Summary: X-link: https://github.com/facebookresearch/FBGEMM/pull/244 - Fix logging in benchmarks (redo D62973297 bc diff cannot be unlinked) Reviewed By: basilwong, spcyppt Differential Revision: D62983422 fbshipit-source-id: 9551cf15df7c1c0281193758b25aaae177e34b67
avbokovoy
requested review from
amathews-amd,
amd-griguer,
liligwu and
xinyazhang
xinyazhang
reviewed
Oct 7, 2024
| kEmbeddingDim, | ||
| kWeighDecayMode | ||
| ) | ||
| }} |
There was a problem hiding this comment.
I like this. No indents look weird at first glance but much more readable than nested inner loop.
|
|
||
| {%- if weighted %} | ||
| vals[_j*kMaxVecsPerThread + i] = weights_slice; | ||
| //{%- if weighted %} |
There was a problem hiding this comment.
Consider removing them if they are not useful anymore.
If it's some upcoming feature, we can add them back later.
| {%- endif %} | ||
| {%- endmacro %} | ||
|
|
||
| {#-/* |
There was a problem hiding this comment.
There is one question. Should we keep the comment in generated source files?
(Just a concern, not a blocking factor)
fbgemm_gpu/codegen/training/forward/embedding_forward_split_kernel_template.cu
Outdated
Show resolved
Hide resolved
Definitely. I did a bit, but still more to come |
Closed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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Outdated suggestions cannot be applied.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.
Cleaned-up branch history for easier review. This branch contains merged bwd and fwd pass optimizations and targeting relatively up-to-date upstream. Current limitations:
BWD:
FWD: