A Last weapon to save Data scientist
Project description
mlVajra
A framework or best practices to develop end to end machine learning pipeline (also has some tips for ML-management people ) Aim : To built robust depoyment pipeline strategies using Open source stack planning to add as many strategies in this repo pertaining to ML-deployment
Installation :
pip install mlvajra
- only installs mlvajra binaries
To install complete dependencies: (for time being)
git clone https://github.com/rajagurunath/mlvajra.git
create virtualenv
virtualenv -p python3 vajra_env
source vajra_env\bin\activate
cd mlvajra
Install all required dependencies from repo
pip install -r requirements.txt
TODO list
Deploy
- Mlflow
- Tensorflow serving
model-Training /distribuited
- mlflow -generic classification metrics (done)
- nnictl-automl -tensorflow /pytorch
Feature Engineering
- pandas
- pyspark-Flint
preprocessing
- cyclic features (done)
- lag features
- window features
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Filename, size | File type | Python version | Upload date | Hashes |
---|---|---|---|---|
Filename, size mlvajra-0.1.4.3-py2.py3-none-any.whl (2.4 MB) | File type Wheel | Python version py2.py3 | Upload date | Hashes View |
Close
Hashes for mlvajra-0.1.4.3-py2.py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 1e40defcb2e1d7b816de2cc657cf27936ae5ee4890e9013c686ba2afb2e1b0a7 |
|
MD5 | 803b50b04f6fb79d67c0d562a54af980 |
|
BLAKE2-256 | 24424dbc538b65874acf9629d02be0e0be30c2bd5880c0e6dcd8c768de4fcdc3 |