mirror of https://github.com/hpcaitech/ColossalAI
ca6e75bc28
* Add handson to ColossalAI. * Change names of handsons and edit sequence parallel example. * Edit wrong folder name * resolve conflict * delete readme |
||
---|---|---|
.. | ||
README.md | ||
config.py | ||
train.py |
README.md
Handson 4: Comparison of Large Batch Training Optimization
Prepare Dataset
We use CIFAR10 dataset in this example. The dataset will be downloaded to ../data
by default.
If you wish to use customized directory for the dataset. You can set the environment variable DATA
via the following command.
export DATA=/path/to/data
Run on 2*2 device mesh
colossalai run --nproc_per_node 4 train.py --config config.py