SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Enable PyTorch nightly build CI (#17335)

Browse Source 
* nightly build pytorch CI

* fix working dir

* change time and event name

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
This commit is contained in:
Yih-Dar
2022-06-17 16:42:27 +02:00
committed by GitHub
parent 3c7e56fbb1
commit ca169dbdf1
7 changed files with 321 additions and 227 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
docker/transformers-pytorch-deepspeed-latest-gpu/Dockerfile
View File

@@ -3,6 +3,9 @@ LABEL maintainer="Hugging Face"
ARG DEBIAN_FRONTEND=noninteractive
# Example: `cu102`, `cu113`, etc.
ARG CUDA='cu113'
RUN apt -y update
RUN apt install -y libaio-dev
RUN python3 -m pip install --no-cache-dir --upgrade pip
@@ -13,13 +16,16 @@ RUN git clone https://github.com/huggingface/transformers && cd transformers &&
# Install latest release PyTorch
# (PyTorch must be installed before pre-compiling any DeepSpeed c++/cuda ops.)
# (https://www.deepspeed.ai/tutorials/advanced-install/#pre-install-deepspeed-ops)
RUN python3 -m pip install --no-cache-dir -U torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu113
RUN python3 -m pip install --no-cache-dir -U torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/$CUDA
RUN python3 -m pip install --no-cache-dir ./transformers[deepspeed-testing]
# Pre-build DeepSpeed, so it would be ready for testing (otherwise, the 1st deepspeed test will timeout)
# Pre-build **latest** DeepSpeed, so it would be ready for testing (otherwise, the 1st deepspeed test will timeout)
RUN python3 -m pip uninstall -y deepspeed
RUN DS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 python3 -m pip install deepspeed --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check
# This has to be run (again) inside the GPU VMs running the tests.
# The installation works here, but some tests fail, if we don't pre-build deepspeed again in the VMs running the tests.
# TODO: Find out why test fail.
RUN DS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 python3 -m pip install deepspeed --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check 2>&1
# When installing in editable mode, `transformers` is not recognized as a package.
# this line must be added in order for python to be aware of transformers.