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.

Sign up for GitHub

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 docker params for the rocm case to the multi-node test #126

Draft
wants to merge 2 commits into
base: master
Choose a base branch
from
Draft

Add docker params for the rocm case to the multi-node test #126

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
7135f65
Add docker params for the rocm case
abergeron May 16, 2023
a9f3301
Sync rocm docker with the cuda one to add libibverbs there too
abergeron May 18, 2023
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
11 changes: 9 additions & 2 deletions benchmarks/accelerate_opt/benchfile.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -70,9 +70,16 @@ async def run(self):
# XXX: this doesn't participate in the process timeout
for i, worker in enumerate(self.config.get('worker_addrs', [])):
command = ["docker", "run", "-i", "--rm",
"--ipc", "host",
"--network", "host",
"--privileged",
"--gpus", "all"]
"--privileged"]
if self.config['arch'] == 'cuda':
command.extend(["--gpus", "all"])
elif self.config['arch'] == 'rocm':
command.extend(["--security-opt", "seccomp=unconfined",
"--group-add", "video",
"-v", "/opt/amdgpu/share/libdrm/amdgpu.ids:/opt/amdgpu/share/libdrm/amdgpu.ids",
"-v", "/opt/rocm:/opt/rocm"])
env = self.make_env()
for var in ('MILABENCH_CONFIG', 'XDG_CACHE_HOME', 'OMP_NUM_THREADS'):
command.append("--env")
Expand Down
23 changes: 12 additions & 11 deletions docker/Dockerfile-rocm
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,7 @@ ENV MILABENCH_GPU_ARCH=$ARCH

ARG CONFIG=standard.yaml
ENV MILABENCH_CONFIG_NAME=$CONFIG

ENV MILABENCH_DOCKER=1

# Paths
# -----
Expand All @@ -35,22 +35,23 @@ COPY . /milabench/milabench/
# rustc: used by BERT models inside https://pypi.org/project/tokenizers/
# build-essential: for rust

RUN apt-get update &&\
apt-get install -y git build-essential curl &&\
# Use ofed_info -s to get your local version
ARG MOFED_VERSION=5.4-3.4.0.0

RUN apt-get update -y &&\
apt-get install -y git build-essential curl python3 python-is-python3 python3-pip &&\
curl -o /etc/apt/trusted.gpg.d/mellanox.asc https://content.mellanox.com/ofed/RPM-GPG-KEY-Mellanox &&\
curl -o /etc/apt/sources.list.d/mellanox.list https://linux.mellanox.com/public/repo/mlnx_ofed/${MOFED_VERSION}/ubuntu22.04/mellanox_mlnx_ofed.list &&\
apt-get update -y &&\
apt-get install -y libibverbs1 &&\
apt-get clean &&\
rm -rf /var/lib/apt/lists/*

# Install Rust

RUN curl https://sh.rustup.rs -sSf | sh -s -- -y
ENV PATH="/root/.cargo/bin:${PATH}"

# Install Python
# --------------

# Install anaconda because milabench will need it later anyway
RUN curl https://repo.anaconda.com/miniconda/Miniconda3-py39_23.1.0-1-Linux-x86_64.sh -o ~/miniconda.sh && \
/bin/bash ~/miniconda.sh -b -p $CONDA_PATH && rm ~/miniconda.sh
ENV PATH=$CONDA_PATH/bin:$PATH


# Install Milabench
# -----------------
Expand Down