mirror of
https://github.com/microsoft/mscclpp.git
synced 2026-05-11 17:00:22 +00:00
Align interfaces of put/get/putPackets/getPackets (#185)
This commit is contained in:
Changho Hwang
37
.github/workflows/integration-test-backup.yml
vendored
37
.github/workflows/integration-test-backup.yml
vendored
@@ -5,17 +5,20 @@ on: workflow_dispatch
|
||||
jobs:
|
||||
IntegrationTest:
|
||||
runs-on: self-hosted
|
||||
defaults:
|
||||
run:
|
||||
shell: bash
|
||||
strategy:
|
||||
matrix:
|
||||
container-image: [ghcr.io/microsoft/mscclpp/mscclpp:base-cuda11.8, ghcr.io/microsoft/mscclpp/mscclpp:base-cuda12.1]
|
||||
cuda: [ cuda11.8, cuda12.1 ]
|
||||
|
||||
container:
|
||||
image: ${{ matrix.container-image }}
|
||||
image: "ghcr.io/microsoft/mscclpp/mscclpp:base-${{ matrix.cuda }}"
|
||||
options: --privileged --ipc=host --gpus=all --ulimit memlock=-1:-1
|
||||
|
||||
steps:
|
||||
- name: Checkout
|
||||
uses: actions/checkout@v2
|
||||
uses: actions/checkout@v4
|
||||
|
||||
- name: Install CMake
|
||||
run: |
|
||||
@@ -39,35 +42,35 @@ jobs:
|
||||
run: |
|
||||
set -e
|
||||
export PATH=/usr/local/mpi/bin:$PATH
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allgather_test_perf -b 1K -e 1G -f 2 -o output.jsonl
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allgather_test_perf -b 1K -e 1G -f 2 -k 1 -o output.jsonl
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allgather_test_perf -b 1K -e 1G -f 2 -k 2 -o output.jsonl
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allgather_test_perf -b 1K -e 1G -f 2 -k 3 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allgather_test_perf -b 1K -e 1G -f 2 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allgather_test_perf -b 1K -e 1G -f 2 -k 1 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allgather_test_perf -b 1K -e 1G -f 2 -k 2 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allgather_test_perf -b 1K -e 1G -f 2 -k 3 -o output.jsonl
|
||||
|
||||
- name: Run mscclpp SendRecv test
|
||||
run: |
|
||||
set -e
|
||||
export PATH=/usr/local/mpi/bin:$PATH
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/sendrecv_test_perf -b 1K -e 1G -f 2 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/sendrecv_test_perf -b 1K -e 1G -f 2 -o output.jsonl
|
||||
|
||||
- name: Run mscclpp AllReduce test
|
||||
run: |
|
||||
set -e
|
||||
export PATH=/usr/local/mpi/bin:$PATH
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 1K -e 1G -f 2 -o output.jsonl
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 1K -e 1G -f 2 -k 1 -o output.jsonl
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 1K -e 1G -f 2 -k 2 -o output.jsonl
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 1K -e 1G -f 2 -k 3 -o output.jsonl
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 1K -e 1G -f 2 -k 4 -o output.jsonl
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 12M -e 48M -i 3145728 2 -k 5 -o output.jsonl
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 24K -e 768K -i 24576 -k 6 -w 100 -n 100 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 1K -e 1G -f 2 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 1K -e 1G -f 2 -k 1 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 1K -e 1G -f 2 -k 2 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 1K -e 1G -f 2 -k 3 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 1K -e 1G -f 2 -k 4 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 12M -e 48M -i 3145728 2 -k 5 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/allreduce_test_perf -b 24K -e 768K -i 24576 -k 6 -w 100 -n 100 -o output.jsonl
|
||||
|
||||
- name: Run mscclpp AllToAll test
|
||||
run: |
|
||||
set -e
|
||||
export PATH=/usr/local/mpi/bin:$PATH
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/alltoall_test_perf -b 1K -e 1G -f 2 -o output.jsonl
|
||||
mpirun -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/alltoall_test_perf -b 1K -e 1G -f 2 -k 1 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/alltoall_test_perf -b 1K -e 1G -f 2 -o output.jsonl
|
||||
mpirun --allow-run-as-root -np 8 --bind-to numa -x MSCCLPP_DEBUG=WARN ./build/test/mscclpp-test/alltoall_test_perf -b 1K -e 1G -f 2 -k 1 -o output.jsonl
|
||||
|
||||
- name: Check collective primitives performance
|
||||
run: |
|
||||
|
||||
23
.github/workflows/ut-backup.yml
vendored
23
.github/workflows/ut-backup.yml
vendored
@@ -5,18 +5,21 @@ on: workflow_dispatch
|
||||
jobs:
|
||||
UnitTest:
|
||||
runs-on: self-hosted
|
||||
defaults:
|
||||
run:
|
||||
shell: bash
|
||||
timeout-minutes: 30
|
||||
strategy:
|
||||
matrix:
|
||||
container-image: [ghcr.io/microsoft/mscclpp/mscclpp:base-cuda11.8, ghcr.io/microsoft/mscclpp/mscclpp:base-cuda12.1]
|
||||
cuda: [ cuda11.8, cuda12.1 ]
|
||||
|
||||
container:
|
||||
image: ${{ matrix.container-image }}
|
||||
image: "ghcr.io/microsoft/mscclpp/mscclpp:base-${{ matrix.cuda }}"
|
||||
options: --privileged --ipc=host --gpus=all --ulimit memlock=-1:-1
|
||||
|
||||
steps:
|
||||
- name: Checkout
|
||||
uses: actions/checkout@v2
|
||||
uses: actions/checkout@v4
|
||||
|
||||
- name: Build
|
||||
run: |
|
||||
@@ -44,9 +47,9 @@ jobs:
|
||||
run: |
|
||||
set -e
|
||||
export PATH=/usr/local/mpi/bin:$PATH
|
||||
mpirun -tag-output -np 2 ./build/test/mp_unit_tests
|
||||
mpirun -tag-output -np 4 ./build/test/mp_unit_tests
|
||||
mpirun -tag-output -np 8 ./build/test/mp_unit_tests
|
||||
mpirun --allow-run-as-root -tag-output -np 2 ./build/test/mp_unit_tests
|
||||
mpirun --allow-run-as-root -tag-output -np 4 ./build/test/mp_unit_tests
|
||||
mpirun --allow-run-as-root -tag-output -np 8 ./build/test/mp_unit_tests
|
||||
working-directory: ${{ github.workspace }}
|
||||
|
||||
- name: PyTests
|
||||
@@ -54,10 +57,10 @@ jobs:
|
||||
set -e
|
||||
export PATH=/usr/local/mpi/bin:$PATH
|
||||
cd build && make pylib-copy
|
||||
if [[ '${{ matrix.container-image }}' == *'cuda11'* ]]; then
|
||||
pip3 install -r ../python/test/requirements_cu11.txt
|
||||
if [[ '${{ matrix.cuda }}' == 'cuda11'* ]]; then
|
||||
python3 -m pip install -r ../python/test/requirements_cu11.txt
|
||||
else
|
||||
pip3 install -r ../python/test/requirements_cu12.txt
|
||||
python3 -m pip install -r ../python/test/requirements_cu12.txt
|
||||
fi
|
||||
mpirun -tag-output -np 8 ~/.local/bin/pytest ../python/test/test_mscclpp.py -x
|
||||
mpirun --allow-run-as-root -tag-output -np 8 $(which pytest) ../python/test/test_mscclpp.py -x
|
||||
working-directory: ${{ github.workspace }}
|
||||
|
||||
Reference in New Issue
Block a user