Library of deep learning models and datasets designed to make deep learning more accessible and accelerate ML research.
Minor fix to 1.13.3, please see release notes there.
TODO(afrozm): Document more.
PRs: fixed get_standardized_layers spelling, thanks @cbockman in #1529 serving utils fixes - Thanks @Drunkar ! in #1495 Fixing a checkpoint name bug in #1487, thanks @lzhang10
Enhancements:
Bugs: Correct flat CIFAR modality to not consider 0 as padding
** Modalities refactor: Thanks to Dustin, all modalities are now an enum and just functions, making it easier to understand what's happening in the model. Thanks Dustin!
Model-Based Reinforcement Learning for Atari using T2T, please find a nice writeup in at https://github.com/tensorflow/tensor2tensor/blob/master/tensor2tensor/rl/README.md -- thanks a lot to all the authors! @lukaszkaiser @mbz @piotrmilos @blazejosinski Roy Campbell @konradczechowski @doomie Chelsea Finn @koz4k Sergey Levine @rsepassi George Tucker and @henrykmichalewski !
TRAX = T2T + [JAX](https://github.com/google/jax) - please try out and give us feedback at #1478
New Models:
Documentation and Logging:
Thanks again @cwbeitel and @lgeiger -- good docs and logging goes a long way for understandability.
Bugs fixed:
Many many thanks @wanqizhu @lgeiger @hbrylkowski @artitw @googlehjx and @qixiuai for finding and fixing these and sorry for missing anyone else -- this is really really helpful.
Code Cleanups:
Many thanks for the cleanups @jackd and @lgeiger -- and sorry if I missed anyone else.
Bug Fixes:
Summary of changes:
PRs:
New Model and Problems:
New Metrics:
Enhancements:
RL: Lots of commits to Model Based Reinforcement Learning code by @konradczechowski @koz4k @blazejosinski @piotrmilos - thanks all !
PRs:
New Problems:
RL:
Video Models:
NOTE:
New Problems:
New layers, models, etc:
Usability:
PRs accepted: Cleaning up the code for gru/lstm as transition function for universal transformer. Thanks @MostafaDehghani ! Clipwrapper by @piotrmilos ! Corrected transformer spelling mistake - Thanks @jurasofish! Fix to universal transformer update weights - Thanks @cbockman and @cyvius96 ! Common Voice problem fixes and refactoring - Thanks @tlatkowski ! Infer observation datatype and shape from the environment - Thanks @koz4k !
New Problems / Models:
StanfordNLI
in stanford_nli.py. Thanks @urvashik !Text2TextRemotedir
added for problems with a persistent remote directory. Thanks @rsepassi !SummarizeWikiPretrainSeqToSeq32k
and Text2textElmo
added.AutoencoderResidualVAE
added, thanks @lukaszkaiser !Code Cleanups:
Bug Fixes:
Enhancements to MTF, Video Models and much more!
Introducing MeshTensorFlow - this enables training really big models O(Billions) of parameters.
Models/Layers:
models/research/vqa_*
Weight Normalization
layer from https://arxiv.org/abs/1602.07868.Datasets/Problems:
VideoBairRobotPushingWithActions
by @mbz !Usability:
Bug Fixes:
StackWrapper
eliminates problem with repeating actions. Thanks @blazejosinski !Testing:
Many more fixes, tests and work on RL, Glow, SAVP, Video and other models and problems.