103 open source projects by google-research

Fast Differentiable Sorting and Ranking

Check out the new game server:

ALBERT: A Lite BERT for Self-supervised Learning of Language Representations

Long Range Arena for Benchmarking Efficient Transformers

Offline Reinforcement Learning (aka Batch Reinforcement Learning) on Atari 2600 games

End-to-end training of sparse deep neural networks with little-to-no performance loss.

Real-World RL Benchmark Suite

arXiv LaTeX Cleaner: Easily clean the LaTeX code of your paper to submit to arXiv

SimCLRv2 - Big Self-Supervised Models are Strong Semi-Supervised Learners

Semi-supervised learning for object detection

No description, website, or topics provided.

Tensorflow/Keras code and trained models for Episodic Curiosity Through Reachability

Transformers for Longer Sequences

ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators

Find shape errors before you run your code!

Train robotic agents to learn pick and place with deep learning for vision-based manipulation in PyBullet. Transporter Nets, CoRL 2020.

Unsupervised Data Augmentation (UDA)

A two-level morphological analyzer for Turkish.

Neural reparameterization improves structural optimization

A collection of tools for reverse engineering neural networks.

Tensorflow implementation of S4L: Self-Supervised Semi-Supervised Learning

A library of GPU kernels for sparse matrix operations.

Official repository for the "Big Transfer (BiT): General Visual Representation Learning" paper.

An implementation of soft-DTW divergences.

Learning Latent Dynamics for Planning from Pixels

disentanglement_lib is an open-source library for research on learning disentangled representations.

Efficient Batched Reinforcement Learning in TensorFlow

Research language for array processing in the Haskell/ML family

Shared repository for open-sourced projects from the Google AI Language team.

Fast & Simple Resource-Constrained Learning of Deep Network Structure

TensorFlow code and pre-trained models for BERT

Differentiable SDE solvers with GPU support and efficient sensitivity analysis.

AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty

A simple method to perform semi-supervised learning with limited data.

A reimplementation of "The Lottery Ticket Hypothesis" (Frankle and Carbin) on MNIST.

End-to-end neural table-text understanding models.