awesome-list

The research awesome list can be found here

List of curated machine learning frameworks and tools, inspired by awesome-machine-learning.

Contribute

Contributions welcome! Read the contribution guidelines first.

Table of Contents

• Datasets

  • Speach

• Environments

  • Graphical
  • Hybrid

• Libraries

  • Deep Learning
  • Dimensionality Reduction
  • Multipurpose
  • Natural Language Processing
  • Optimization
  • Probabilistic Programming
  • Recommender Systems
  • Speech Processing
  • Model deployment

• Tools

  • Compilers
  • Data Adapters
  • Data Gathering
  • Data Management
  • Feature engineering
  • Hardware Management
  • Job Management
  • Parallelization
  • Data Visualization
  • Reporting

• License

Datasets

Speach

• Common-Voice - Multi language, open source database with voice samples that anyone can use to train speech-enabled applications.

Environments

Graphical

• AI-Blocks - a powerful and intuitive WYSIWYG interface that allows anyone to create Machine Learning models.

Hybrid

• Luna Studio - Hybrid textual and visual functional programming

Libraries

Deep Learning

• fast.ai - The fastai library simplifies training fast and accurate neural nets using modern best practices
• PyTorch - Tensors and Dynamic neural networks in Python with strong GPU acceleration
• tensorflow - An Open Source Machine Learning Framework for Everyone by Google
• neon - Intel® Nervana™ reference deep learning framework committed to best performance on all hardware
• cleverhans - An adversarial example library for constructing attacks, building defenses, and benchmarking both
• Netron - a viewer for neural network, deep learning and machine learning models.
• Online viewer -
• List of conversion tools for DNN models - list of many libraries (github projects) that provides options for converting DNN models between different frameworks

Dimensionality Reduction

• umap - dimension reduction technique that can be used for visualisation similarly to t-SNE, but also for general non-linear dimension reduction

Multipurpose

• DALI - A library containing both highly optimized building blocks and an execution engine for data pre-processing in deep learning applications docs
• gin-config - Gin provides a lightweight configuration framework for Python, by Google.
• imbalanced-learn - A python package offering a number of re-sampling techniques. Compatible with scikit-learn, is part of scikit-learn-contrib projects.
• mlxtend - A library of extension and helper modules for Python's data analysis and machine learning libraries.
• numpy - The fundamental package for scientific computing with Python.
• PyOD - Outlier detection library
• RAPIDS - Open GPU Data Science. More here or in cheatsheet
• scikit-learn - machine learning in Python
• scikit-learn-laboratory (SKLL) - CLI for sklearn, working with configuration files
• scipy - open-source software for mathematics, science, and engineering. It includes modules for statistics, optimization, integration, linear algebra, Fourier transforms, signal and image processing, ODE solvers, and more.
• statsmodels - statistical modeling and econometrics in Python: time-series analysis, survival analysis, discrete models, Generalized Linear Models
• SymPy - A computer algebra system written in pure Python, library for symbolic mathematics
• Vaex - Out-of-Core DataFrames for Python, visualize and explore big tabular data at a billion rows per second. Project page

Natural Language Processing

• Allen NLP - An open-source NLP research library, built on PyTorch.
• PyText - A natural language modeling framework based on PyTorch by Facebook Research
• pytorch-transformers - A library of state-of-the-art pretrained models for (NLP) including BERT, GPT, GPT-2, Transformer-XL, XLNet and XLM with multiple pre-trained model weights
• flair - A very simple framework for state-of-the-art Natural Language Processing (NLP) by Zalando Research
• gensim - Topic modeling for humans. Enables analysis of plain-text documents for semantic structure. Compatible with Word2Vec, FastText and other NLP models.
• spaCy - spaCy is a library for advanced Natural Language Processing in Python and Cython. spaCy comes with pretrained statistical models and word vectors, and supports tokenization for 50+ languages.
• TextBlob - Simple, Pythonic, text processing--Sentiment analysis, part-of-speech tagging, noun phrase extraction, translation, and more.

Optimization

• hyperopt - Distributed Asynchronous Hyperparameter Optimization in Python
• nevergrad - A Python toolbox for performing gradient-free optimization by Facebook Research

Probabilistic Programming

• pyro - Deep universal probabilistic programming with Python and PyTorch by Uber
• pgmpy - Python Library for Probabilistic Graphical Models

Recommender systems

• surprise - A Python scikit for building and analyzing recommender systems

Speech Processing

• warp-ctc - loss function to train on misaligned data and labels by Baidu Research
• DeepSpeech - A TensorFlow implementation of Baidu's DeepSpeech architecture
• speech-to-text-wavenet - End-to-end sentence level English speech recognition based on DeepMind's WaveNet and tensorflow
• pykaldi - A Python wrapper for Kaldi - a toolkit for speech recognition
• pytorch-kaldi - pytorch-kaldi is a project for developing state-of-the-art DNN/RNN hybrid speech recognition systems. The DNN part is managed by pytorch, while feature extraction, label computation, and decoding are performed with the kaldi toolkit.

Model Deployment

• gradio - Library to easily integrate models into existing python (web) apps.

Tools

Compilers

• glow - Compiler for Neural Network hardware accelerators by PyTorch
• jax - Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more by Google
• numba - NumPy aware dynamic Python compiler using LLVM

Data Adapters

• csvkit - A suite of utilities for converting to and working with CSV, the king of tabular file formats.
• redash - Connect to any data source, easily visualize, dashboard and share your data.
• odo - Odo migrates between many formats. These include in-memory structures like list, pd.DataFrame and np.ndarray and also data outside of Python like CSV/JSON/HDF5 files, SQL databases, data on remote machines, and the Hadoop File System.

Data Annotation

• doccano - Open source text annotation tool for machine learning practitioners.
• snorkel - A system for quickly generating training data with weak supervision

Data Gathering

• scrapy - high-level library to write crawlers and spiders.

Data Management

• Quilt - Quilt versions and deploys data

Data Visualization

• matplotlib - plotting with Python
• bokeh - Interactive Web Plotting for Python
• plotly - An open-source, interactive graphing library for Python
• dash - Analytical Web Apps for Python. No JavaScript Required.
• Jupyter Dashboards - Jupyter layout extension
• vega - visualization grammar, a declarative format for creating, saving, and sharing interactive visualization designs. With Vega you can describe data visualizations in a JSON format, and generate interactive views using either HTML5 Canvas or SVG.
• schema crawler - a tool to visualize database schema
• scikit-plot - sklearn wrapper to automate frequently used machine learning visualizations.

Feature engineering

• featuretools - an open source python framework for automated feature engineering.

Hardware Management

• nvtop - a (h)top like task monitor for NVIDIA GPUs. It can handle multiple GPUs and print information about them in a htop familiar way.

• s2i - Source-to-Image (S2I) is a toolkit and workflow for building reproducible container images from source code. S2I produces ready-to-run images by injecting source code into a container image and letting the container prepare that source code for execution. By creating self-assembling builder images, you can version and control your build environments exactly like you use container images to version your runtime environments.

Job Management

• luigi - Luigi is a Python module that helps you build complex pipelines of batch jobs. It handles dependency resolution, workflow management, visualization etc. It also comes with Hadoop support built in. By Spotify.

Parallelization

• pywren - parfor on AWS Lambda
• horovod - Distributed training framework for TensorFlow, Keras, PyTorch, and MXNet by Uber.
• dask - library for parallel computing in Python with dynamic task scheduling: numpy computation graphs.

Reporting

• shap - A unified approach to explain the output of any machine learning model
• tensorboardX - tensorboard for pytorch (and chainer, mxnet, numpy, ...)
• Weights and Biases - Experiment Tracking for Deep Learning
• pandas-profiling - tool for generating exploratory data analysis for the provided DataFrame - presenting results in the form of HTML report

License

To the extent possible under law, Netguru has waived all copyright and related or neighboring rights to this work. See LICENSE.