rocblas -> hipblas changes for ROCm #5401
Merged
Conversation
This file contains 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
Merged
rraminen
force-pushed
the
rocblas_to_hipblas_fix
branch
from
735ac6a
to
e74d5bf
Compare
rraminen
force-pushed
the
rocblas_to_hipblas_fix
branch
2 times, most recently
from
1309094
to
bef085e
Compare
rraminen
force-pushed
the
rocblas_to_hipblas_fix
branch
3 times, most recently
from
dda4bba
to
d288d36
Compare
rraminen
force-pushed
the
rocblas_to_hipblas_fix
branch
from
dfc8c6c
to
6d9fee9
Compare
|
@rraminen Formatting checks failed with trailing whitespace error: https://github.com/microsoft/DeepSpeed/actions/runs/9115455800/job/25064328013?pr=5401#step:5:60 Should be a straightforward one, can you please check? |
loadams
approved these changes
May 16, 2024
loadams
reviewed
May 16, 2024
loadams
approved these changes
May 17, 2024
|
Verified the extension builds in the following cases rocm/pytorch:rocm6.1_ubuntu20.04_py3.9_pytorch_1.13.1 |
github-merge-queue bot
pushed a commit
that referenced
this pull request
May 17, 2024
This PR enables building the below extensions for AMD GPUs with warp size 32. - transformer_inference - quantizer - random_ltd This PR works stand-alone for torch version <=2.0. For the latest versions, #5401 is required to be merged in addition to this PR. Unit test results (rocm/pytorch:rocm6.1_ubuntu20.04_py3.9_pytorch_2.1.2) on NAVI3x: **transformer_inference:** pytest --color=yes --durations=0 --verbose -s -m "inference_ops" -rF -n 4 unit/ops/transformer/inference Before this PR: ===== 674 failed, 622 skipped, 8 warnings, 1728 errors in 69.37s (0:01:09) ===== After this PR: ========== 476 failed, 1062 passed, 1486 skipped, 8 warnings in 9.31s ========== **quantizer:** pytest --color=yes --durations=0 --verbose -s -m "inference_ops" -rF -n 4 unit/ops/quantizer Before this PR: ==== 244 failed, 8 warnings in 30.53s ==== After this PR: ====== 186 failed, 58 passed, 8 warnings in 8.89s ====== I could not find random_ltd related unit tests to run. Fixes: #4753 #5474 ROCm#68 cc: @jithunnair-amd --------- Co-authored-by: rraminen@amd.com <rraminen> Co-authored-by: Logan Adams <114770087+loadams@users.noreply.github.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Fixes #4989
In addition to this PR, below changes are required to build below extensions successfully. Please note that not all unit tests for these extensions will pass with this PR. More details on the unit test results are below. These unit tests are skipped in CI anyway, so they will not break the CI.
transformer_inference
quantizer
random_ltd
[ROCm] Add cublasGemmAlgo_t -> hipblasGemmAlgo_t pytorch/pytorch#121030
Rocm warp size fix #5402
Unit test results (rocm/pytorch:rocm6.1_ubuntu20.04_py3.9_pytorch_2.1.2) on MI200:
transformer_inference:
pytest --color=yes --durations=0 --verbose -s -m "inference_ops" -rF -n 4 unit/ops/transformer/inference
Before this PR:
==== 674 failed, 622 skipped, 8 warnings, 1728 errors in 123.66s (0:02:03) =====
After this PR:
========== 555 failed, 983 passed, 1486 skipped, 8 warnings in 14.35s ==========
quantizer:
pytest --color=yes --durations=0 --verbose -s -m "inference_ops" -rF -n 4 unit/ops/quantizer
Before this PR:
==== 244 failed, 8 warnings in 48.02s ====
After this PR:
===== 187 failed, 57 passed, 8 warnings in 14.74s ====
I could not find random_ltd related unit tests to run.