Change tagline, cleanup README

README.rst

@@ -1,70 +1,14 @@
 ================================================================================
-               Toys: Friendly machine learning for rapid research
+                         Toys: The data science toolbox
 ================================================================================
 
-Toys is a framework for machine learning experiments inspired by Scikit-learn and built with PyTorch.
+Toys is a toolbox for data science, built with PyTorch, and designed for rapid research.
 
-
-Development Roadmap
-===================
-The implementation of Toys is just getting started. My initial effort is being put towards implementing training algorithms which wrap PyTorch modules, implementing model selection algorithms like ``GridSearchCV``, and implementing the most important scoring metrics. In the long run, I'd like to support Dask arrays for compatibility with that ecosystem of out-of-core datasets and distributed computation, but currently there are upstream compatibility issues between Dask and PyTorch. The hope is to one day reach feature parity with Scikit-learn, but this will certainly require a healthy community of contributors. To reach that point, I first need to develop the infrastructure and design guidelines to foster distributed development.
-
-Listed below are tentative project goals roughly organized by priority.
-
-Highest Priority
-----------------
-- Core API design
-- Infrastructure (setup.py, Sphinx, documentation, hosting)
-- Generic PyTorch wrappers
-- Parameter tuning, grid search
-- Common metrics, model validation
-- Serialization
-
-High Priority
--------------
-- Higher level neural net layers (see `tf.layers`_ and `tf.variable_scope`_)
-- Common neural net architectures (ResNet, U-Net)
-- Linear classifiers and regressors (linear, softmax, SVM)
-- Decision trees and ensembles (CART, random forest, boosted trees)
-- Naive Bayes
-- Preprocessing (standard scaling, TF-IDF)
-- Common datasets (MNIST, CIFAR)
-
-Medium Priority
----------------
-- `Dask`_ support
-- Bayesian inference, probabilistic graphical models (see `Pyro`_)
-- Reinforcement learning (see `Gym`_)
-- Unsupervised learning (K-NN, K-Means, spectral clustering)
-- Matrix decomposition / feature extraction (PCA, NMF, etc)
-- Feature selection
-
-Low Priority
-------------
-- A tool to extract MyPy type annotation stubs from docstrings
-- Scikit-learn feature parity
-
-
-.. _Dask: https://dask.pydata.org/en/latest/
-.. _Gym: https://gym.openai.com/
-.. _Pyro: http://pyro.ai/
-.. _PyTorch: http://pytorch.org/
-.. _tf.layers: https://www.tensorflow.org/api_guides/python/contrib.layers
-.. _tf.variable_scope: https://www.tensorflow.org/api_docs/python/tf/variable_scope
-
-
-Contributing
-============
-
-This is project is a large endeavor and all are welcome to contribute. But because the project is so young, coordination is key. Please reach out on the issue tracker, or in person if you are around UGA, if you are interested in contributing.
-
-The `contributing file`_ contains style guides and other useful guidelines for contributing to the project.
-
-.. _contributing file: https://github.com/cbarrick/toys/tree/master/CONTRIBUTING.rst
+Coming soon to an interpreter near you.
 
 
 License
-=======
+==================================================
 
 MIT License
 

docs/index.rst

@@ -1,7 +1,8 @@
-Toys: The data science toolbox
-==================================================
+================================================================================
+                         Toys: The data science toolbox
+================================================================================
 
-Toys is a toolbox for machine learning and data science, built with PyTorch, and designed for rapid research.
+Toys is a toolbox for data science, built with PyTorch, and designed for rapid research.
 
 
 Documentation