Templating tool with boiler plate code for building robust machine learning projects in python.
Project description
ML Template
ML template is an easy-to-use tool to automate the boilerplate code for most machine learning projects.
This tool creates a user-oriented project architecture for machine learning projects.
Modify the code under #TODO
comments in the template project repository to easily adapt the template to your use-case.
How to use it?
- Install the package as -
pip install mltemplate
- Then, simply run
mltempate
from your terminal and follow the prompts
And Voila!
This creates a project directory in your current folder similar to -
template
├── Dockerfile.cpu
├── Dockerfile.gpu
├── LICENSE.md
├── Makefile
├── README.md
├── jupyter.sh
├── requirements.txt
└── template
├── __init__.py
├── __main__.py
├── cli
│ ├── __init__.py
│ ├── predict.py
│ └── train.py
├── notebooks
└── src
├── __init__.py
├── models.py
├── datasets.py
└── transforms.py
All you have to do next is -
- Update python frameworks and versions in
template/requirements.txt
as need for your project - Head to
template/datasets.py
and modify create a new dataset that will work for your use case - Navigate to
template/models.py
and create a new model class with your sota (or not) architecture - In
template/transforms.py
add transforms such as Normalizer, Denormalize etc. - Follow the
TODO
steps intemplate/cli/train.py
andtemplate/cli/predict.py
to make the necessary changes
Checkout the README.md
in the template
directory for further instructions on how to train, predict and also monitor your loss plots using tensor board.
Future Work
Currently, this package only supports boilerplate creation for ML projects in pytorch
We plan to support tensorflow
in the future.
Development
To create a new version of the framework, update the version in pyproject.toml
file
Merge your changes to main and then publish git tag to trigger the release ci
git tag <x.x.x>
git push origin <x.x.x>
License
Copyright © 2020 Sowmya Yellapragada
Distributed under the MIT License (MIT).
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file mltemplate-1.0.2.tar.gz
.
File metadata
- Download URL: mltemplate-1.0.2.tar.gz
- Upload date:
- Size: 3.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.6.1 CPython/3.10.13 Linux/6.2.0-1012-azure
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 99abbcecbce1c68bacbcb8f59ced0bca2455c2a6bee4155976dc0f8414d9e13e |
|
MD5 | 1d5d289ecaa2bd75975782586a1f413a |
|
BLAKE2b-256 | c7ceea4fcf477f17e1348f7e61feedf7b6d2b11585cfd50d2b753842afed4b4c |
File details
Details for the file mltemplate-1.0.2-py3-none-any.whl
.
File metadata
- Download URL: mltemplate-1.0.2-py3-none-any.whl
- Upload date:
- Size: 4.8 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.6.1 CPython/3.10.13 Linux/6.2.0-1012-azure
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 2ad2ef75674b446250aa889e110de66a16795043aca9602d111b0fac3a8b8892 |
|
MD5 | 44de9d467e56d6ac53986caf6bbabaaa |
|
BLAKE2b-256 | 3ffdfd73c5df4f178800c1c6f8e29c452c73a1324d0d3d688a425acf8e44defd |