History

Frank Lee 80eba05b0a [test] refactor tests with spawn (#3452 ) * [test] added spawn decorator * polish code * polish code * polish code * polish code * polish code * polish code		2023-04-06 14:51:35 +08:00
..
saved_solution	[autoparallel] integrate device mesh initialization into autoparallelize (#2393 )	2023-01-11 14:03:49 +08:00
README.md	[example] upload auto parallel gpt2 demo (#2354 )	2023-01-06 11:38:38 +08:00
auto_parallel_with_gpt.py	[test] refactor tests with spawn (#3452 )	2023-04-06 14:51:35 +08:00
gpt_modules.py	[example] upload auto parallel gpt2 demo (#2354 )	2023-01-06 11:38:38 +08:00
requirements.txt	[auto] fix requirements typo for issue #3125 (#3209 )	2023-03-23 10:22:08 +08:00

README.md

Auto-Parallelism with GPT2

Requirements

Before you can launch training, you need to install the following requirements.

Install PyTorch

#conda
conda install pytorch==1.12.0 torchvision==0.13.0 torchaudio==0.12.0 cudatoolkit=11.3 -c pytorch
#pip
pip install torch==1.12.0+cu113 torchvision==0.13.0+cu113 torchaudio==0.12.0 --extra-index-url https://download.pytorch.org/whl/cu113

Install Colossal-AI v0.2.0 From Official Website

pip install colossalai==0.2.0+torch1.12cu11.3 -f https://release.colossalai.org

Install transformers

pip install transformers

Install pulp and coin-or-cbc

pip install pulp
conda install -c conda-forge coin-or-cbc

Dataset

For simplicity, the input data is randonly generated here.

Training

#Run the auto parallel resnet example with 4 GPUs with a dummy dataset.
colossalai run --nproc_per_node 4 auto_parallel_with_gpt.py