search

tensorflow / community

Stores documents used by the TensorFlow developer community
Apache License 2.0
1.26k stars 580 forks source link

RFC: Creating SIG Models #314

Closed jaeyounkim closed 4 years ago

jaeyounkim commented 4 years ago

This RFC will be open for comment until Wednesday, November 11st, 2020.

Creating SIG Models

We propose to create a TensorFlow SIG Models.

Status Accepted
RFC # 314
Author(s) Jaeyoun Kim (jaeyounkim@google.com), Jing Li (jingli@google.com), Mike Liang (mliang@google.com), Shuangfeng Li (shuangfeng@google.com)
Sponsor Thea Lamkin (thealamkin@google.com)
Updated 2020-10-23

What is this group for?

This group is for discussions and collaborations on enabling community contributions to TensorFlow Model Garden and Tensorflow Hub.

SIG Models will focus on empowering the community to contribute state-of-the-art model implementation in TensorFlow 2. It will benefit the whole community by providing recommended implementations and models with reproducible results.

SIG Models will have several subgroups (e.g., SIG Models Vision and SIG Models NLP) covering different machine learning areas. There are several SIG leads for each group to coordinate the contributions, run community events like contests, and maintain the code quality through the review process.

sub-mod commented 4 years ago

I would like to be part of the SIG.Thanks

bhack commented 4 years ago

In general I think that is great and strategic to have this SIG cause Models and Hub are really one of the unique places in the ecosystem where all of the pieces converges together to compose our end2end goal. Also we often suffer more then the Pytorch ecosystem about availability of third party papers official (and unofficial) reference implementations in the github "universe".

But we really need to take this SIG as an occasion to improve our ecosystem about component re-usability and ownership disambiguation. It could be nice to have a short grace period for the new components (layer, loss, optimizer, preprocessing, postprocessing, etc.) that would be introduced in TF model. As we know external members don't know the public roadmap of the different internal teams (and this could happen also between different internal teams) you can find a recap thread at https://github.com/tensorflow/community/issues/29. This could create duplicates, unaware supersedes, implementation divergences and so on.. In Addons we have tried to Draft an early ecosystem-review process to notify the ecosystem that probably could be standardized and improved for general use to improve the coherence between internal teams and external members. By this point of view I see this SIG one of the most critical points to achieve this goal in our ecosystem.