Open
Conversation
WangJialei-A
requested review from
dchigarev,
kurapov-peter,
leshikus and
zhczhong
dchigarev
reviewed
Oct 14, 2024
| cmake -B build -G Ninja -DLLVM_DIR=${LLVM_INST_PATH}/lib/cmake/llvm -DMLIR_DIR=${LLVM_INST_PATH}/lib/cmake/mlir -DENABLE_GRAPH_COMPILER=ON -DENABLE_INTEL_GPU=ON -DENABLE_TESTS=ON | ||
| cmake --build build --target all | ||
|
|
||
| - name: Benchmark |
Contributor
There was a problem hiding this comment.
let's also add this step that runs sanity tests for the gpu integration in OV:
OV_MLIR_MODE=GC_GPU ./bin/intel64/Release/ov_gpu_func_tests --gtest_filter=MLIRExecution.*
71a4de7 to
02640c7
Compare
dchigarev
reviewed
Oct 15, 2024
| run: | | ||
| pip install openvino torch | ||
| for param in 'linear[512,512,512]' 'linear[1024,1024,1024]' 'linear[2048,2048,2048]', 'linear[4096,4096,4096]' 'linear[8192,8192,8192]' 'linear[4096,512,4096]'; do | ||
| python3 ./tools/mlir_bench/ov_model_gen.py -l=$param -t f16 -n test.xml |
Contributor
There was a problem hiding this comment.
I think the root-cause for the error we see in CI is this bug (#360)
loc(fused<{name = "aten::linear/Add", type = "Add"}>["aten::linear/Add"]): error: operand #0 does not dominate this use
The GPU pipeline is currently not working for models generated by ov_model_gen.py; only simple matmul modules like this are functioning (but I'm already working to fix this)
WangJialei-A
force-pushed
the
wangjial/ov_test
branch
from
02640c7 to
58ffe07
Compare
WangJialei-A
force-pushed
the
wangjial/ov_test
branch
from
58ffe07 to
6f16887
Compare
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR depends on a develop repo and branch 'dchigarev/openvino/gc-gpu' now
So it is not ready for merge.
But it is ready for review.
Support single matmul test with openvino on GPU.
It will be automatically trigger remotely by nightly test in gc-perf repository.
@kurapov-peter Please comment the PR if you need run more test case or collect more metrics.