|
---
|
|
license: apache-2.0
|
|
---
|
|
|
|
Set the Cookiecutter-MLOps in Hugging Face
|
|
==============================================
|
|
|
|
1 - Create Model repository in Hugging Face (e.g. myHFrepo)
|
|
2 - Clone your Hugging face repo to your local directory:
|
|
|
|
cd /path/to/parent directory of project folder
|
|
git clone git@hf.co:USERNAME/myHFrepo
|
|
|
|
For ssh connection check [here](https://huggingface.co/docs/hub/security-git-ssh#add-a-ssh-key-to-your-account)
|
|
|
|
3 - Create your virtual environment (e.g. jointvenv)
|
|
|
|
cd myHFrepo
|
|
python -m venv jointvenv
|
|
source jointvenv/bin/activate
|
|
|
|
4 - Transfer and set DagsHub's cookiecutter template employing MLOps best practices to your Huggingface repo
|
|
|
|
git clone https://dagshub.com/DagsHub/Cookiecutter-MLOps.git
|
|
|
|
4.1 - Delete git files cloned from Cookiecutter-MLOps repo
|
|
|
|
rm -r /path/to/myHFrepo/Cookiecutter-MLOps/.git
|
|
|
|
4.2 - dResolve conflicts with .gitattributes and README.md
|
|
|
|
cat /path/to/myHFrepo/Cookiecutter-MLOps/.gitattributes >> /path/to/myHFrepo/.gitattributes
|
|
rm /path/to/myHFrepo/Cookiecutter-MLOps/.gitattributes
|
|
git add .gitattributes
|
|
git commit -m "Paste .gitattributes info from DagsHub/Cookiecutter-MLOps"
|
|
|
|
cat /path/to/myHFrepo/Cookiecutter-MLOps/README.md >> /path/to/myHFrepo/README.md
|
|
rm /path/to/myHFrepo/Cookiecutter-MLOps/README.md
|
|
git add README.md
|
|
git commit -m "Paste README info from DagsHub/Cookiecutter-MLOps"
|
|
|
|
4.3 - Move remaining files from DagsHub/Cookiecutter-MLOps yo your Hugging Face repo .gitattributes and README.md
|
|
|
|
cd /path/to/myHFrepo/Cookiecutter-MLOps
|
|
mv * .[^.]* ..
|
|
cd /path/to/myHFrepo
|
|
rm -r /path/to/myHFrepo/Cookiecutter-MLOps
|
|
|
|
5 - Add venv folder to.gitignore
|
|
|
|
echo '' >> .gitignore
|
|
echo '#'Virtual Environment >> .gitignore
|
|
echo jointvenv/ >> .gitignore
|
|
git add .
|
|
git commit -m "add remaining DagsHub/Cookiecutter-MLOps repo content"
|
|
|
|
6 - Run step 2 from DagsHub/Cookiecutter-MLOps
|
|
|
|
make dirs
|
|
|
|
7 - Run step 4 from DagsHub/Cookiecutter-MLOps
|
|
|
|
make requirements
|
|
|
|
8 - Keep record of your own requirements
|
|
|
|
mv requirements.txt requirementsCookiecutter-MLOps.txt
|
|
git add requirementsCookiecutter-MLOps.txt
|
|
git commit -m "external requirements from Cookiecutter-MLOps"
|
|
|
|
pip freeze > requirements.txt
|
|
git add requirements.txt
|
|
git commit -m "First report venv requirements"
|
|
|
|
9 - Push your changes to the remote Hugging face repository
|
|
|
|
git push origin main
|
|
|
|
10 - Optional Create Model repository in your Hugging Face organization (e.g. myHFrepo)
|
|
|
|
git remote add dcc git@hf.co:MYORG/mywslHFrepo
|
|
git pull dcc main --allow-unrelated-histories
|
|
|
|
Resolve conflicts in .gitattributes and README.md
|
|
|
|
git add .
|
|
git commit -m "Merge HuggingFace individual and organization repos"
|
|
git push dcc main
|
|
|
|
==============================
|
|
Cookiecutter-MLOps
|
|
==============================
|
|
|
|
A cookiecutter template employing MLOps best practices, so you can focus on building machine learning products while
|
|
having MLOps best practices applied.
|
|
|
|
Instructions
|
|
------------
|
|
1. Clone the repo.
|
|
2. Run `make dirs` to create the missing parts of the directory structure described below.
|
|
3. *Optional:* Run `make virtualenv` to create a python virtual environment. Skip if using conda or some other env manager.
|
|
1. Run `source env/bin/activate` to activate the virtualenv.
|
|
4. Run `make requirements` to install required python packages.
|
|
5. Put the raw data in `data/raw`.
|
|
6. To save the raw data to the DVC cache, run `dvc add data/raw`
|
|
7. Edit the code files to your heart's desire.
|
|
8. Process your data, train and evaluate your model using `dvc repro` or `make reproduce`
|
|
9. To run the pre-commit hooks, run `make pre-commit-install`
|
|
10. For setting up data validation tests, run `make setup-setup-data-validation`
|
|
11. For **running** the data validation tests, run `make run-data-validation`
|
|
12. When you're happy with the result, commit files (including .dvc files) to git.
|
|
|
|
Project Organization
|
|
------------
|
|
|
|
├── LICENSE
|
|
├── Makefile <- Makefile with commands like `make dirs` or `make clean`
|
|
├── README.md <- The top-level README for developers using this project.
|
|
├── data
|
|
│ ├── processed <- The final, canonical data sets for modeling.
|
|
│ └── raw <- The original, immutable data dump
|
|
│
|
|
├── models <- Trained and serialized models, model predictions, or model summaries
|
|
│
|
|
├── notebooks <- Jupyter notebooks. Naming convention is a number (for ordering),
|
|
│ the creator's initials, and a short `-` delimited description, e.g.
|
|
│ `1.0-jqp-initial-data-exploration`.
|
|
├── references <- Data dictionaries, manuals, and all other explanatory materials.
|
|
├── reports <- Generated analysis as HTML, PDF, LaTeX, etc.
|
|
│ └── figures <- Generated graphics and figures to be used in reporting
|
|
│ └── metrics.txt <- Relevant metrics after evaluating the model.
|
|
│ └── training_metrics.txt <- Relevant metrics from training the model.
|
|
│
|
|
├── requirements.txt <- The requirements file for reproducing the analysis environment, e.g.
|
|
│ generated with `pip freeze > requirements.txt`
|
|
│
|
|
├── setup.py <- makes project pip installable (pip install -e .) so src can be imported
|
|
├── src <- Source code for use in this project.
|
|
│ ├── __init__.py <- Makes src a Python module
|
|
│ │
|
|
│ ├── data <- Scripts to download or generate data
|
|
│ │ ├── great_expectations <- Folder containing data integrity check files
|
|
│ │ ├── make_dataset.py
|
|
│ │ └── data_validation.py <- Script to run data integrity checks
|
|
│ │
|
|
│ ├── models <- Scripts to train models and then use trained models to make
|
|
│ │ │ predictions
|
|
│ │ ├── predict_model.py
|
|
│ │ └── train_model.py
|
|
│ │
|
|
│ └── visualization <- Scripts to create exploratory and results oriented visualizations
|
|
│ └── visualize.py
|
|
│
|
|
├── .pre-commit-config.yaml <- pre-commit hooks file with selected hooks for the projects.
|
|
├── dvc.lock <- constructs the ML pipeline with defined stages.
|
|
└── dvc.yaml <- Traing a model on the processed data.
|
|
|
|
|
|
--------
|
|
|
|
<p><small>Project based on the <a target="_blank" href="https://drivendata.github.io/cookiecutter-data-science/">cookiecutter data science project template</a>. #cookiecutterdatascience</small></p>
|
|
|
|
|
|
---
|
|
|
|
To create a project like this, just go to https://dagshub.com/repo/create and select the **Cookiecutter DVC** project template.
|
|
|
|
Made with 🐶 by [DAGsHub](https://dagshub.com/).
|
|
|
