metadata
license: apache-2.0
language:
- en
Preparing the dataset
NOTICE:
All code is owned by Hugging Face and uses the Apache 2.0 Licence. While I clean and strip the dataset for processing, do note that this dataset is under the same scruteny as the original Apache 2.0 License.
Clone Repo
Data souce used is the accelerate repository. I'm using the latest version, v0.25.0
git clone https://github.com/huggingface/accelerate
cd accelerate
git checkout v0.25.0
cd ..
mkdir docs src
mv accelerate/src/accelerate/* src
mv accelerate/docs/* docs
cd src
rm __init__.py commands/__init__.py test_utils/__init__.py utils/__init__.py
Cleaning the dataset
Using regex in VSCODE, use the following replacement:
# Copyright(.*\n)+# limitations under the license.
<!--Copyright(.*\n)+-->
In the source:
"""
To:
"""
Then remove all import statements (as we only care about the content).
Strip all blank spaces/whitespace:
^(?:[\t ]*(?:\r?\n|\r))+
WARNING: It is known that this will seperate out the _inner() in the source code and use it as a seperate function losing the context. Trying out with this issue for now.