bluemist 0.1.3

Bluemist AI is a low code machine learning library written in Python to develop, evaluate and deploy automated ML pipleines.

Bluemist AI

Bluemist AI is an advanced, open-source, low-code machine learning library developed in Python. Its primary purpose is to facilitate the development, evaluation, and deployment of automated machine learning models and pipelines.

The library acts as a convenient wrapper service, integrating seamlessly with popular Python libraries such as scikit-learn, NumPy, pandas, mlflow, and FastAPI. For visualization purposes, Bluemist AI leverages the capabilities of pandas-profiling, sweetviz, dtale, and autoviz.

Key Features

Bluemist AI offers a range of powerful features, including:

• Native Data Integration: Seamlessly extract data from various sources, including MySQL, PostgreSQL, MS SQL, Oracle, MariaDB, Amazon Aurora, and Amazon S3, providing a unified interface for working with diverse datasets.
• Exploratory Data Analysis (EDA): Conduct thorough EDA on your data, enabling comprehensive insights into the characteristics, distributions, and relationships within the dataset.
• Data Preprocessing: Streamline the preprocessing pipeline by leveraging Bluemist AI's built-in preprocessing capabilities, allowing you to handle missing data, perform feature scaling, encoding, and other essential preprocessing tasks efficiently.
• Algorithm Selection and Comparison: Bluemist AI facilitates training data across multiple algorithms, enabling you to compare and evaluate various models using comprehensive metrics and insights.
• Hyperparameter Tuning: Optimize model performance with built-in hyperparameter tuning functionalities, automatically searching for the optimal combination of hyperparameters to achieve the best results.
• Experiment Tracking: Keep track of your experiments effortlessly, storing essential metadata, parameters, and results, allowing for easy reproducibility and experimentation management.
• API Deployment: Bluemist AI simplifies the process of deploying machine learning models as APIs, making it easier to integrate your models into production systems and applications.

These features make Bluemist AI a powerful tool for developing, evaluating, and deploying machine learning models efficiently and effectively.

Getting Started

For the detailed list of supported and upcoming features, visit https://www.bluemist-ai.one

Full documentation is available @ https://bluemist-ai.readthedocs.io

User installation

Method 1

To install minimal version of the package with hard dependencies listed in requirements.txt

pip install -U bluemist

To install the complete package including optional dependencies listed in requirements-optional.txt. Refer Minimal package vs Full package for more details.

pip install -U bluemist[complete]

Method 2 (recommended)

It is advised to set up a separate Python environment to avoid conflicts with package dependencies. Follow the steps below:

• Install the package virtualenv

pip install virtualenv

• Create a separate directory where bluemist environment will be created

mkdir /path/to/bluemist-ai
cd /path/to/bluemist-ai

• Create the bluemist environment

virtualenv bluemist-env

• Activate the environment and install bluemist

source bluemist-env/bin/activate
pip install -U bluemist

Method 3

bluemist package can be installed using pipx utility. It automatically creates an isolated environment to run the bluemist package

pip install pipx
pipx install bluemist
pipx upgrade bluemist

Support for GPU/CPU Acceleration

GPU Acceleration

• If your system is equipped with an NVIDIA GPU and you want to utilize GPU acceleration for model training, you can install the RAPIDS cuML package by executing the following command:

pip install cudf-cu11 cuml-cu11 --extra-index-url=https://pypi.nvidia.com

• The cuml package provides GPU-accelerated implementations of various machine learning algorithms. By installing this package, you can leverage the power of your GPU to significantly improve the speed and performance of your machine learning workflows.

• For the list of supported algorithms, please refer https://docs.rapids.ai/api/cuml/stable/api/#regression-and-classification

• Acceleration extensions can ben enabled by passing the parameter enable_acceleration_extensions=True during the initialize phase.

CPU Acceleration

• If your system is equipped with an Intel CPU and you want to utilize CPU acceleration for model training, you can install the Intel® Extension for Scikit-learn package by executing the following command:

pip install -U scikit-learn-intelex

• The scikit-learn-intelex package provides CPU-accelerated implementations of Scikit-learn algorithms. By installing this package, you can leverage the power of your Intel CPU to significantly improve the speed and performance of your machine learning workflows.

• For the list of supported algorithms, please refer to the Intel® Extension for Scikit-learn documentation.

• Acceleration extensions can ben enabled by passing the parameter enable_acceleration_extensions=True during the initialize phase.

Minimal package vs Full package

Below functionalities are available only with complete package installation

• Data extraction from RDBMS or cloud
• EDA Visualizations using dtale and sweetviz
• CPU/GPU Acceleration

Alternatively a single optional package can be installed using the pip command. For example, if you would like to extract data from Amazon S3 but do not wish to install other optional packages :

pip install boto3

License

This project is licensed under the MIT License.

See Third Party Libraries for license details of third party libraries included in the distribution.

MIT License

Copyright (c) 2022-2023 Bluemist AI

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.