Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

add run on sco #555

Closed
wants to merge 69 commits into from
Closed
Changes from 15 commits
Commits
Show all changes
69 commits
Select commit Hold shift + click to select a range
71d4e8f
run on sco
wugeshui Dec 19, 2023
7a5efb4
run on sco
wugeshui Dec 19, 2023
8e1ce79
run on sco
wugeshui Dec 19, 2023
d9b8af3
run on sco
wugeshui Dec 19, 2023
7fb7bdb
Update main.yml
wugeshui Dec 20, 2023
c905efc
Update main.yml
wugeshui Dec 20, 2023
8bc9d7e
Merge branch 'DeepLink-org:main' into wgs/run_on_sco
wugeshui Dec 20, 2023
142ebbc
change v100 to sco
wugeshui Dec 20, 2023
9cb7853
add run on sco
wugeshui Dec 20, 2023
f2f0ae1
Update main.yml
wugeshui Dec 20, 2023
629d1a8
update runs-on-sco.yml
wugeshui Dec 20, 2023
f59fd3b
Update runs-on-sco.yml
wugeshui Dec 20, 2023
40e3d17
update source-main
wugeshui Dec 20, 2023
35ca3f8
Update runs-on-sco.yml
wugeshui Dec 20, 2023
a35efae
Update runs-on-sco.yml
wugeshui Dec 20, 2023
683fcad
Update path
wugeshui Dec 20, 2023
443883a
add sco
wugeshui Dec 20, 2023
21d2699
modify one iter
wugeshui Dec 20, 2023
e8f2526
update diopi path
wugeshui Dec 20, 2023
be33fd1
update sco
wugeshui Dec 20, 2023
d547472
add sco
wugeshui Dec 21, 2023
61a1043
add sco
wugeshui Dec 21, 2023
2fdc34a
update sco
wugeshui Dec 21, 2023
5672886
add sco
wugeshui Dec 21, 2023
6b5f41c
add sco
wugeshui Dec 21, 2023
d3531f8
run on sco
wugeshui Dec 21, 2023
1a8e52b
run on sco
wugeshui Dec 22, 2023
204edaa
run on sco
wugeshui Dec 22, 2023
a4a8f74
test model
wugeshui Dec 27, 2023
5dfcede
rm cancel for test
wugeshui Dec 27, 2023
1b799d4
echo env
wugeshui Dec 29, 2023
fe5b2d2
add sco ci
wugeshui Jan 2, 2024
3f1d12a
add sco ci
wugeshui Jan 2, 2024
1663327
add sco ci
wugeshui Jan 2, 2024
7f67964
add sco ci
wugeshui Jan 2, 2024
fa0aa76
add sco ci
wugeshui Jan 2, 2024
397f583
add sco ci
wugeshui Jan 2, 2024
c6d9506
add sco ci
wugeshui Jan 2, 2024
d6ff2bf
add sco ci
wugeshui Jan 2, 2024
4e7bfc4
add sco ci
wugeshui Jan 2, 2024
e369264
rm bash
wugeshui Jan 3, 2024
8d71c10
add bash
wugeshui Jan 3, 2024
30a2981
add other sco ci
wugeshui Jan 8, 2024
316034f
update env path
wugeshui Jan 8, 2024
d615103
update ENV_PATH
wugeshui Jan 8, 2024
169f114
add sco ci
wugeshui Jan 9, 2024
8f366f5
Merge branch 'DeepLink-org:main' into wgs/run_on_sco
wugeshui Jan 15, 2024
e0442f8
run on sco
wugeshui Jan 15, 2024
3fba918
run on sco
wugeshui Jan 15, 2024
762072e
run on sco
wugeshui Jan 15, 2024
f06e6fd
run on sco
wugeshui Jan 15, 2024
e826e08
Update ci_nv_tidy.sh
wugeshui Jan 15, 2024
f052a2f
run on sco
wugeshui Jan 16, 2024
e82dbc7
run on sco
wugeshui Jan 16, 2024
04472cb
run on sco
wugeshui Jan 16, 2024
1066562
run on sco
wugeshui Jan 16, 2024
9b140e1
add sco ci
wugeshui Jan 16, 2024
ba9d1b6
add sco ci
wugeshui Jan 16, 2024
56e83fa
run on sco
wugeshui Jan 17, 2024
0361e4f
run on sco
wugeshui Jan 17, 2024
37a2730
add bashrc
wugeshui Jan 17, 2024
b02bcc7
add bashrc
wugeshui Jan 17, 2024
8351804
add proxy
wugeshui Jan 17, 2024
06f82f6
add sco ci
wugeshui Jan 17, 2024
0f81b64
Merge branch 'main' into wgs/run_on_sco
wugeshui Jan 17, 2024
eb261bf
add sco ci
wugeshui Jan 17, 2024
ef0fb0e
add sco ci
wugeshui Jan 18, 2024
47ee3b1
Merge branch 'DeepLink-org:main' into wgs/run_on_sco
wugeshui Jan 22, 2024
548b29f
add sco ci
wugeshui Jan 23, 2024
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
126 changes: 126 additions & 0 deletions .github/workflows/runs-on-sco.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,126 @@
name: runs on sco

on:
workflow_dispatch:
push:
branches:
- main
pull_request:
paths-ignore:
- "**.md"
- ".github/ISSUE_TEMPLATE/**"
- ".git*"
- "CODE_OF_CONDUCT**"

concurrency:
group: sco-${{ github.head_ref || github.ref }}
cancel-in-progress: true

env:
DEEPLINK_PATH: '/mnt/cache/share/deeplinkci/github/${{ github.repository }}'
ALL_COVERAGE: ${{ (contains( github.ref, 'main') || startsWith(github.ref, 'refs/heads/v') || startsWith(github.ref, 'refs/heads/dev')) && 'ON' || 'OFF' }}
REQUIRE_COVERAGE: ${{ vars.REQUIRE_COVERAGE != '' && vars.REQUIRE_COVERAGE || '40' }}

jobs:
checkout_code:
name: checkout code
runs-on: tps-sco-ci
steps:
- name: Checkout Code
uses: DeepLink-org/deeplink.framework/.github/actions/checkout-code@main
- name: add mmlab_pack
run: |
set -e
cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER} && cp -R source source-main
cd source/dipu && bash /home/autolink/rsync/sourcecode/update_code.sh
rsync -a /home/autolink/rsync/sourcecode/mmlab_pack . && cd mmlab_pack
bash ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/source/dipu/scripts/ci/ci_one_iter.sh clone
cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/source-main/dipu/third_party && rm -rf DIOPI && git clone https://github.com/DeepLink-org/DIOPI.git

Build-Cuda:
name: Build-dipu-cuda
needs: checkout_code
runs-on: tps-sco-ci
steps:
- name: Build cuda
run: |
set -e
cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER} && rm -rf ${GITHUB_JOB} && cp -R source ${GITHUB_JOB}
srun --job-name=${GITHUB_JOB} bash -c "export USE_COVERAGE=ON && cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/${GITHUB_JOB}/dipu \
&& source /mnt/cache/share/deeplinkci/github/dipu_env \
&& rsync -a /mnt/lustre/share_data/PAT/datasets/huggingface mmlab_pack/ \
&& bash scripts/ci/nv/ci_nv_script.sh build_dipu" || ( cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/ && rm -rf ${GITHUB_JOB} && exit 1 )

Test-Cuda:
name: Test-dipu-cuda
needs: [Build-Cuda]
runs-on: tps-sco-ci
steps:
- name: Run-test
run: |
set -e
srun --job-name=${GITHUB_JOB} bash -c "export USE_COVERAGE=ON && cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/Build-Cuda/dipu \
&& source /mnt/cache/share/deeplinkci/github/dipu_env \
&& bash tests/run_nv_tests.sh"

Test-One-Iter_Cuda:
name: Test-one-iter-cuda
needs: [Build-Cuda]
runs-on: tps-sco-ci
steps:
- name: build some env
run: |
set -e
srun --job-name=${GITHUB_JOB} bash -c "cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/Build-Cuda/dipu/mmlab_pack \
&& source /mnt/cache/share/deeplinkci/github/dipu_env \
&& bash ../scripts/ci/ci_one_iter.sh build_cuda"
- name: run-one-iter-for-tradition
run: |
set -e
srun --job-name=${GITHUB_JOB} bash -c "cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/Build-Cuda/dipu/mmlab_pack \
&& source /mnt/cache/share/deeplinkci/github/dipu_env \
&& rm -rf one_iter_data \
&& python ../scripts/ci/ci_run_one_iter.py cuda ${GITHUB_RUN_NUMBER}_${GITHUB_JOB} gpu:${GPU_REQUESTS} \"${CUDA_PARTATION}\" && rm -rf one_iter_data || (rm -rf one_iter_data && exit 1)"
- name: run-one-iter-for-llm
run: |
set -e
srun --job-name=${GITHUB_JOB} bash -c "cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/Build-Cuda/dipu/mmlab_pack \
&& source /mnt/cache/share/deeplinkci/github/dipu_env \
&& rm -rf one_iter_data \
&& python ../scripts/ci/ci_run_one_iter.py cuda ${GITHUB_RUN_NUMBER}_${GITHUB_JOB} gpu:${GPU_REQUESTS} \"${CUDA_PARTATION}\" "llm" && rm -rf one_iter_data || (rm -rf one_iter_data && exit 1)"
- name: Perform cleanup one iter data
if: always()
run: |
set -e
cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/Build-Cuda/dipu/mmlab_pack \
rm -rf one_iter_data
touch one_iter_data #用于占位,防止创建新的one_iter_data文件夹
- name: Check for failure
if: ${{ failure() }}
run: exit 1

Build-Cuda-Latest-Target:
name: Build-dipu-cuda-latest-target
needs: [checkout_code]
runs-on: tps-sco-ci
steps:
- name: Build dipu diopi-latest-target
run: |
set -e
cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER} && rm -rf ${GITHUB_JOB} && cp -R source-main ${GITHUB_JOB}
srun --job-name=${GITHUB_JOB} bash -c "export USE_COVERAGE=ON && cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/${GITHUB_JOB}/dipu \
&& source /mnt/cache/share/deeplinkci/github/dipu_env \
&& bash scripts/ci/nv/ci_nv_script.sh build_dipu" || ( cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/ && rm -rf ${GITHUB_JOB} && exit 1 )

Test-Cuda-Latest-Target:
name: Test-dipu-cuda-latest-target
needs: [Build-Cuda-Latest-Target]
runs-on: tps-sco-ci
steps:
- name: Run-test
run: |
set -e
srun --job-name=${GITHUB_JOB} bash -c "cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/Build-Cuda-Latest-Target/dipu \
&& source /mnt/cache/share/deeplinkci/github/dipu_env \
&& bash tests/run_nv_tests.sh" && cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/ && rm -rf Build-Cuda-Latest-Target \
|| ( cd ${DEEPLINK_PATH}/${GITHUB_RUN_NUMBER}/ && rm -rf ${GITHUB_JOB} && exit 1 )