Skip to content

Commit a09987a

Browse files
jackalcooperoneflow-ci-bot
jackalcooper
and
oneflow-ci-bot
authored
CI uses cu118 (#10359)
Co-authored-by: oneflow-ci-bot <ci-bot@oneflow.org>
1 parent a27f657 commit a09987a

File tree

8 files changed

+45
-37
lines changed

8 files changed

+45
-37
lines changed

‎.github/workflows/canary.yml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -54,7 +54,7 @@ jobs:
5454
- name: Checkout Oneflow-Inc/oneflow
5555
if: ${{ github.event.inputs.oneflow-ref == '' }}
5656
uses: actions/checkout@v2
57-
- uses: Oneflow-Inc/get-oneflow@refactor-versions-wheels
57+
- uses: Oneflow-Inc/get-oneflow@ci-test-with-cu118
5858
name: Build manylinux
5959
id: build-cuda
6060
with:

‎.github/workflows/on_merge.yml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -15,6 +15,6 @@ jobs:
1515
if: github.event.pull_request.merged == true
1616
runs-on: ubuntu-latest
1717
steps:
18-
- uses: Oneflow-Inc/get-oneflow/update-benchmark-history@refactor-versions-wheels
18+
- uses: Oneflow-Inc/get-oneflow/update-benchmark-history@ci-test-with-cu118
1919
name: Update benchmark history
2020
timeout-minutes: 10

‎.github/workflows/release.yml

Lines changed: 4 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -70,7 +70,7 @@ jobs:
7070
ref: ${{ inputs.branch }}
7171
repository: ${{ secrets.ONEFLOW_PRIV_ORG }}/oneflow
7272
token: ${{ secrets.ONEFLOW_PRIV_GH_TOKEN }}
73-
- uses: Oneflow-Inc/get-oneflow/cache-complete/matrix/build@refactor-versions-wheels
73+
- uses: Oneflow-Inc/get-oneflow/cache-complete/matrix/build@ci-test-with-cu118
7474
name: Find build cache
7575
id: find-cache
7676
timeout-minutes: 5
@@ -149,7 +149,7 @@ jobs:
149149
if: ${{ inputs.is_priv }}
150150
run: |
151151
env
152-
- uses: Oneflow-Inc/get-oneflow@refactor-versions-wheels
152+
- uses: Oneflow-Inc/get-oneflow@ci-test-with-cu118
153153
name: Build ${{ matrix.entry }}
154154
if: ${{ matrix.entry =='cu118' || startsWith(matrix.entry, 'cu12') }}
155155
with:
@@ -175,7 +175,7 @@ jobs:
175175
3.10
176176
3.9
177177
3.8
178-
- uses: Oneflow-Inc/get-oneflow@refactor-versions-wheels
178+
- uses: Oneflow-Inc/get-oneflow@ci-test-with-cu118
179179
name: Build ${{ matrix.entry }}
180180
if: ${{ startsWith(matrix.entry, 'cu') && matrix.entry !='cu118' && !startsWith(matrix.entry, 'cu12') }}
181181
with:
@@ -201,7 +201,7 @@ jobs:
201201
3.10
202202
3.9
203203
3.8
204-
- uses: Oneflow-Inc/get-oneflow@refactor-versions-wheels
204+
- uses: Oneflow-Inc/get-oneflow@ci-test-with-cu118
205205
name: Build ${{ matrix.entry }}
206206
if: ${{ matrix.entry =='cpu' }}
207207
with:

‎.github/workflows/simple.yml

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -244,7 +244,7 @@ jobs:
244244
repository: Oneflow-Inc/conda-env
245245
ref: 30a7f00eb48ee9009d85a848e720823e5054c66b
246246
path: conda-env
247-
- uses: Oneflow-Inc/get-oneflow@refactor-versions-wheels
247+
- uses: Oneflow-Inc/get-oneflow@ci-test-with-cu118
248248
name: Build with gcc7
249249
if: ${{ matrix.build-type == 'gcc7'}}
250250
with:
@@ -253,7 +253,7 @@ jobs:
253253
oneflow-build-env: conda
254254
conda-env-file: conda-env/dev/gcc7/environment-v2.yml
255255
conda-env-name: oneflow-dev-gcc7-v2
256-
- uses: Oneflow-Inc/get-oneflow@refactor-versions-wheels
256+
- uses: Oneflow-Inc/get-oneflow@ci-test-with-cu118
257257
name: Build with clang10
258258
if: ${{ matrix.build-type == 'clang10'}}
259259
with:

‎.github/workflows/test.yml

Lines changed: 24 additions & 24 deletions
Original file line numberDiff line numberDiff line change
@@ -176,7 +176,7 @@ jobs:
176176
with:
177177
ref: ${{ github.event.pull_request.head.sha }}
178178
repository: ${{github.event.pull_request.head.repo.full_name}}
179-
- uses: Oneflow-Inc/get-oneflow/cache-complete/matrix/build@refactor-versions-wheels
179+
- uses: Oneflow-Inc/get-oneflow/cache-complete/matrix/build@ci-test-with-cu118
180180
name: find cache
181181
id: find-cache
182182
timeout-minutes: 5
@@ -188,7 +188,7 @@ jobs:
188188
builder
189189
oneflow-src: ${{ env.ONEFLOW_SRC }}
190190
entries: |
191-
cu116
191+
cu118
192192
cpu
193193
cpu-asan-ubsan
194194
cpu-tsan
@@ -223,7 +223,7 @@ jobs:
223223
with:
224224
ref: ${{ github.event.pull_request.head.sha }}
225225
repository: ${{github.event.pull_request.head.repo.full_name}}
226-
- uses: Oneflow-Inc/get-oneflow/cache-complete@refactor-versions-wheels
226+
- uses: Oneflow-Inc/get-oneflow/cache-complete@ci-test-with-cu118
227227
name: Save cache if successful
228228
id: save-cache
229229
timeout-minutes: 5
@@ -237,7 +237,7 @@ jobs:
237237
run: |
238238
echo "::error file=test.yml,line=204,col=10::steps.save-cache.outputs.cache-hit != matrix.cache-hit"
239239
exit 1
240-
- uses: Oneflow-Inc/get-oneflow@refactor-versions-wheels
240+
- uses: Oneflow-Inc/get-oneflow@ci-test-with-cu118
241241
name: Build manylinux ${{ matrix.entry }}
242242
id: build-cpu
243243
if: ${{ matrix.entry =='cpu' && !matrix.cache-hit }}
@@ -259,7 +259,7 @@ jobs:
259259
python-versions: |
260260
3.7
261261
3.8
262-
- uses: Oneflow-Inc/get-oneflow@refactor-versions-wheels
262+
- uses: Oneflow-Inc/get-oneflow@ci-test-with-cu118
263263
name: Build manylinux ${{ matrix.entry }}
264264
id: build-cpu-sanitizers
265265
if: ${{ (matrix.entry == 'cpu-asan-ubsan' || matrix.entry == 'cpu-tsan') && !matrix.cache-hit && false }}
@@ -280,10 +280,10 @@ jobs:
280280
clean-ccache: ${{ contains(github.event.pull_request.labels.*.name, 'need-clean-ccache') }}
281281
python-versions: |
282282
3.8
283-
- uses: Oneflow-Inc/get-oneflow@refactor-versions-wheels
283+
- uses: Oneflow-Inc/get-oneflow@ci-test-with-cu118
284284
name: Build manylinux ${{ matrix.entry }}
285285
id: build-cuda
286-
if: ${{ matrix.entry =='cu116' && !matrix.cache-hit }}
286+
if: ${{ matrix.entry =='cu118' && !matrix.cache-hit }}
287287
with:
288288
cmake-init-cache: ${{ env.ONEFLOW_SRC }}/cmake/caches/ci/cuda.cmake
289289
build-script: ${{ env.ONEFLOW_SRC }}/ci/manylinux/build-gcc9.sh
@@ -292,15 +292,15 @@ jobs:
292292
wheelhouse-dir: ${{ env.WHEELHOUSE_DIR }}
293293
clear-wheelhouse-dir: true
294294
self-hosted: ${{ contains(matrix.runs-on, 'self-hosted') }}
295-
cuda-version: "11.6"
295+
cuda-version: "11.8"
296296
manylinux-cache-dir: ${{ env.MANYLINUX_CACHE_DIR }}
297297
docker-run-use-system-http-proxy: false
298298
docker-run-use-lld: false
299299
retry-failed-build: true
300300
clean-ccache: ${{ contains(github.event.pull_request.labels.*.name, 'need-clean-ccache') }}
301301
python-versions: |
302302
3.7
303-
- uses: Oneflow-Inc/get-oneflow@refactor-versions-wheels
303+
- uses: Oneflow-Inc/get-oneflow@ci-test-with-cu118
304304
name: Build ${{ matrix.entry }}
305305
if: ${{ matrix.entry == 'llvm15' && !matrix.cache-hit }}
306306
with:
@@ -339,7 +339,7 @@ jobs:
339339
})
340340
- name: Upload packed liboneflow
341341
if: ${{ !fromJson(matrix.cache-hit) && matrix.entry != 'llvm15' && matrix.entry != 'cpu-asan-ubsan' && matrix.entry != 'cpu-tsan' }}
342-
uses: Oneflow-Inc/get-oneflow/digest/upload@refactor-versions-wheels
342+
uses: Oneflow-Inc/get-oneflow/digest/upload@ci-test-with-cu118
343343
timeout-minutes: 10
344344
with:
345345
digest: ${{ steps.save-cache.outputs.build-digest }}
@@ -350,7 +350,7 @@ jobs:
350350
dst-dir: cpack
351351
- name: Upload whl
352352
if: ${{ !fromJson(matrix.cache-hit) && matrix.entry != 'llvm15' && matrix.entry != 'cpu-asan-ubsan' && matrix.entry != 'cpu-tsan' }}
353-
uses: Oneflow-Inc/get-oneflow/digest/upload@refactor-versions-wheels
353+
uses: Oneflow-Inc/get-oneflow/digest/upload@ci-test-with-cu118
354354
timeout-minutes: 10
355355
with:
356356
digest: ${{ steps.save-cache.outputs.build-digest }}
@@ -375,7 +375,7 @@ jobs:
375375
with:
376376
ref: ${{ github.event.pull_request.head.sha }}
377377
repository: ${{github.event.pull_request.head.repo.full_name}}
378-
- uses: Oneflow-Inc/get-oneflow/cache-complete/matrix/test@refactor-versions-wheels
378+
- uses: Oneflow-Inc/get-oneflow/cache-complete/matrix/test@ci-test-with-cu118
379379
name: find cache
380380
id: find-cache
381381
timeout-minutes: 5
@@ -406,7 +406,7 @@ jobs:
406406
with:
407407
ref: ${{ github.event.pull_request.head.sha }}
408408
repository: ${{github.event.pull_request.head.repo.full_name}}
409-
- uses: Oneflow-Inc/get-oneflow/cache-complete/matrix/test@refactor-versions-wheels
409+
- uses: Oneflow-Inc/get-oneflow/cache-complete/matrix/test@ci-test-with-cu118
410410
name: find cache
411411
id: find-cache
412412
timeout-minutes: 5
@@ -488,7 +488,7 @@ jobs:
488488
if: ${{ contains(matrix.runs-on, 'self-hosted') }}
489489
run: |
490490
docker rm -f ${{ env.TEST_CONTAINER_NAME }} || true
491-
- uses: Oneflow-Inc/get-oneflow/cache-complete@refactor-versions-wheels
491+
- uses: Oneflow-Inc/get-oneflow/cache-complete@ci-test-with-cu118
492492
name: Save cache if successful
493493
id: save-cache
494494
timeout-minutes: 5
@@ -504,7 +504,7 @@ jobs:
504504
exit 1
505505
- name: Download wheel and packed liboneflow
506506
if: ${{ !fromJson(matrix.cache-hit) && contains(matrix.runs-on, 'self-hosted') }}
507-
uses: Oneflow-Inc/get-oneflow/digest/download@refactor-versions-wheels
507+
uses: Oneflow-Inc/get-oneflow/digest/download@ci-test-with-cu118
508508
id: download-digest
509509
timeout-minutes: 10
510510
with:
@@ -514,7 +514,7 @@ jobs:
514514
ssh-tank-path: ${{ env.SSH_TANK_PATH }}
515515
- name: Get primary node
516516
if: ${{ !fromJson(matrix.cache-hit) && contains(matrix.runs-on, 'self-hosted') }}
517-
uses: Oneflow-Inc/get-oneflow/master-address@refactor-versions-wheels
517+
uses: Oneflow-Inc/get-oneflow/master-address@ci-test-with-cu118
518518
id: get-primary-node
519519
with:
520520
rank: ${{ matrix.rank }}
@@ -650,7 +650,7 @@ jobs:
650650
TEST_CONTAINER_NAME: "pr-${{ github.event.pull_request.number }}-run-id-${{ github.run_id }}-${{ matrix.entry }}-test"
651651
TEST_MANYLINUX_CONTAINER_NAME: "pr-${{ github.event.pull_request.number }}-run-id-${{ github.run_id }}-${{ matrix.entry }}-test-manylinux"
652652
TEST_WITH_TF_IMG_TAG: registry.cn-beijing.aliyuncs.com/oneflow/test-with-tf-2.3.0:2f831e9354298a11447578e869d983959feb046f
653-
TEST_MANYLINUX_IMG_TAG: registry.cn-beijing.aliyuncs.com/oneflow/manylinux2014_x86_64_cuda11.6:328e477069c80035adb3cd4db9632997e6284edd
653+
TEST_MANYLINUX_IMG_TAG: registry.cn-beijing.aliyuncs.com/oneflow/manylinux2014_x86_64_cuda11.8:6455f9b8154333333e6285fde3747aaac4a92929
654654
METRICS_DIR: metrics
655655
steps:
656656
- name: Set proxy
@@ -718,7 +718,7 @@ jobs:
718718
if: ${{ contains(matrix.runs-on, 'self-hosted') }}
719719
run: |
720720
docker rm -f ${{ env.TEST_MANYLINUX_CONTAINER_NAME }} || true
721-
- uses: Oneflow-Inc/get-oneflow/cache-complete@refactor-versions-wheels
721+
- uses: Oneflow-Inc/get-oneflow/cache-complete@ci-test-with-cu118
722722
name: Save cache if successful
723723
id: save-cache
724724
timeout-minutes: 5
@@ -734,7 +734,7 @@ jobs:
734734
exit 1
735735
- name: Download wheel and packed liboneflow
736736
if: ${{ !fromJson(matrix.cache-hit) && contains(matrix.runs-on, 'self-hosted') }}
737-
uses: Oneflow-Inc/get-oneflow/digest/download@refactor-versions-wheels
737+
uses: Oneflow-Inc/get-oneflow/digest/download@ci-test-with-cu118
738738
id: download-digest
739739
timeout-minutes: 10
740740
with:
@@ -744,7 +744,7 @@ jobs:
744744
ssh-tank-path: ${{ env.SSH_TANK_PATH }}
745745
- name: Download ASAN and UBSAN wheel and packed liboneflow
746746
if: ${{ !fromJson(matrix.cache-hit) && contains(matrix.runs-on, 'self-hosted') && matrix.device == 'cpu' && false }}
747-
uses: Oneflow-Inc/get-oneflow/digest/download@refactor-versions-wheels
747+
uses: Oneflow-Inc/get-oneflow/digest/download@ci-test-with-cu118
748748
id: asan-ubsan-download-digest
749749
timeout-minutes: 10
750750
with:
@@ -754,7 +754,7 @@ jobs:
754754
ssh-tank-path: ${{ env.SSH_TANK_PATH }}
755755
- name: Download TSAN wheel and packed liboneflow
756756
if: ${{ !fromJson(matrix.cache-hit) && contains(matrix.runs-on, 'self-hosted') && matrix.device == 'cpu' && false }}
757-
uses: Oneflow-Inc/get-oneflow/digest/download@refactor-versions-wheels
757+
uses: Oneflow-Inc/get-oneflow/digest/download@ci-test-with-cu118
758758
id: tsan-download-digest
759759
timeout-minutes: 10
760760
with:
@@ -902,7 +902,7 @@ jobs:
902902
run: |
903903
ls ${ONEFLOW_WHEEL_PATH}
904904
docker exec ${TEST_CONTAINER_NAME} python3 -m pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple
905-
docker exec ${TEST_CONTAINER_NAME} python3 -m pip install --find-links=${ONEFLOW_WHEEL_PATH} oneflow
905+
docker exec ${TEST_CONTAINER_NAME} python3 -m pip install -U --find-links=${ONEFLOW_WHEEL_PATH} oneflow
906906
- name: Install downstream libs
907907
if: ${{ !fromJson(matrix.cache-hit) && contains(matrix.runs-on, 'self-hosted') }}
908908
run: |
@@ -1080,7 +1080,7 @@ jobs:
10801080
- name: Benchmark Test
10811081
timeout-minutes: 100
10821082
if: ${{ !fromJson(matrix.cache-hit) && matrix.test-type == 'benchmark' && matrix.device == 'cuda' }}
1083-
uses: Oneflow-Inc/get-oneflow/pytest-benchmark@refactor-versions-wheels
1083+
uses: Oneflow-Inc/get-oneflow/pytest-benchmark@ci-test-with-cu118
10841084
with:
10851085
collect-path: ${{ env.FLOW_VISION_SRC }}/benchmark
10861086
container-name: ${{ env.TEST_CONTAINER_NAME }}
@@ -1141,7 +1141,7 @@ jobs:
11411141
ref: ${{ github.event.pull_request.head.sha }}
11421142
repository: ${{github.event.pull_request.head.repo.full_name}}
11431143
fetch-depth: 0
1144-
- uses: Oneflow-Inc/get-oneflow/cache-complete@refactor-versions-wheels
1144+
- uses: Oneflow-Inc/get-oneflow/cache-complete@ci-test-with-cu118
11451145
name: Save cache if successful
11461146
id: save-cache
11471147
timeout-minutes: 5

‎cmake/caches/ci/cuda.cmake

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -16,3 +16,6 @@ set(WITH_MLIR ON CACHE BOOL "")
1616
set(BUILD_CPP_API ON CACHE BOOL "")
1717
set(CUDA_NVCC_THREADS_NUMBER 8 CACHE STRING "")
1818
set(BUILD_FOR_CI ON CACHE BOOL "")
19+
set(CMAKE_CXX_FLAGS
20+
"-Wno-unused-but-set-parameter -Wno-unused-variable -Wno-class-memaccess -Wno-cast-function-type -Wno-comment -Wno-reorder"
21+
CACHE STRING "")

‎cmake/third_party/flash_attention.cmake

Lines changed: 5 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -7,8 +7,9 @@ find_package(Git QUIET REQUIRED)
77

88
set(FLASH_ATTENTION_PROJECT flash_attention)
99

10-
set(FLASH_ATTENTION_URL https://github.com/Oneflow-Inc/flash-attention-v2.git)
11-
set(FLASH_ATTENTION_TAG eed2e82b880e06237af3e50ceac4cf6728b15645)
10+
set(FLASH_ATTENTION_URL
11+
https://oneflow-static.oss-cn-beijing.aliyuncs.com/third_party_mirror/flash-attention-v2-eed2e82b880e06237af3e50ceac4cf6728b15645.zip
12+
)
1213

1314
set(FLASH_ATTENTION_INSTALL_DIR ${THIRD_PARTY_DIR}/flash_attention)
1415
set(FLASH_ATTENTION_INCLUDE_DIR ${FLASH_ATTENTION_INSTALL_DIR}/include CACHE PATH "" FORCE)
@@ -19,8 +20,8 @@ if(THIRD_PARTY)
1920
ExternalProject_Add(
2021
${FLASH_ATTENTION_PROJECT}
2122
PREFIX flash_attention
22-
GIT_REPOSITORY ${FLASH_ATTENTION_URL}
23-
GIT_TAG ${FLASH_ATTENTION_TAG}
23+
URL ${FLASH_ATTENTION_URL}
24+
URL_HASH MD5=63192a05973f614aff594a8bd11813ce
2425
UPDATE_COMMAND ""
2526
BUILD_BYPRODUCTS ${FLASH_ATTENTION_LIBRARIES}
2627
CMAKE_ARGS -DCMAKE_BUILD_TYPE:STRING=${CMAKE_BUILD_TYPE}

‎python/oneflow/test/modules/test_normal.py

Lines changed: 5 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -62,7 +62,11 @@ def helper(self, device, dtype, ptype, t_transform, std_transform):
6262
t_transform(q[99:100]).std().item(), std_transform(1), atol=0.3, rtol=0
6363
)
6464
)
65-
self.assertTrue(flow.allclose(t_transform(q[0:1]).clone(), t_transform(q_row1)))
65+
self.assertTrue(
66+
flow.allclose(
67+
t_transform(q[0:1]).clone(), t_transform(q_row1), atol=0.3, rtol=0.3,
68+
)
69+
)
6670

6771
mean = flow.empty(100, 100, dtype=dtype, device=device)
6872
mean[:50].fill_(ptype(0))

0 commit comments

Comments
 (0)