Migrate fuse_chunk_reshape_concat_pass to PT2 #134026
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/134026
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (1 Unrelated Failure)As of commit 556a012 with merge base 2a73ba2 ( BROKEN TRUNK - The following job failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
This pull request was exported from Phabricator. Differential Revision: D60789747 |
|
This pull request was exported from Phabricator. Differential Revision: D60789747 |
oniononion36
force-pushed
the
export-D60789747
branch
from
cbd230b to
0339249
Compare
Summary: Pull Request resolved: pytorch#134026 This is part of the work of dper pass migration https://fburl.com/gdoc/wxwykxns This pass has ~2.4% perf impact for adfinder_reels_ctr_model Test Plan: Still in test Differential Revision: D60789747
Summary: Pull Request resolved: pytorch#134026 This is part of the work of dper pass migration https://fburl.com/gdoc/wxwykxns This pass has ~2.4% perf impact for adfinder_reels_ctr_model Test Plan: For unit tests: ``` %run ~/fbsource/fbcode/caffe2/test/inductor/fb/test_fuse_chunk_reshape_unsqueeze_concat_pass.py ``` For the pass itself, we can run ``` buck2 run mode/opt-split-dwarf mode/inplace -c fbcode.platform010_cuda_version=12 -c fbcode.nvcc_arch=a100 caffe2/torch/fb/model_transform/experimental/benchmark:mts_gpu_benchmark -- --model-path=manifold://ads_storage_fblearner/tree/user/facebook/fblearner/predictor/565872608/1881/gpu_lowering/input.predictor.disagg.gpu.merge --lower-backend=AOT_INDUCTOR --remove-passes= --disable-acc-tracer=True ``` and verify that the QPS is increased by ~2% after the diff is applied. Reviewed By: huxintong Differential Revision: D60789747
oniononion36
force-pushed
the
export-D60789747
branch
from
0339249 to
556a012
Compare
|
This pull request was exported from Phabricator. Differential Revision: D60789747 |
|
@pytorchbot merge -f 'Landed internally' (Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally) |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Summary: This is part of the work of dper pass migration https://fburl.com/gdoc/wxwykxns This pass has ~2.4% perf impact for adfinder_reels_ctr_model Test Plan: Still in test Differential Revision: D60789747 Pull Request resolved: pytorch#134026 Approved by: https://github.com/huxintong
4 participants
Summary:
This is part of the work of dper pass migration https://fburl.com/gdoc/wxwykxns
This pass has ~2.4% perf impact for adfinder_reels_ctr_model
Test Plan: Still in test
Differential Revision: D60789747
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @yf225 @chenyang78 @kadeng @muchulee8 @ColinPeppler @amjames @desertfire @chauhang