New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
global reasoning unit #181
global reasoning unit #181
Conversation
This pull request was exported from Phabricator. Differential Revision: D18344566 |
e285314
to
cc4f38b
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
cc4f38b
to
2bbe46f
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
2bbe46f
to
5abf6cf
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
5abf6cf
to
243c5dc
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
243c5dc
to
1fa2702
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
1fa2702
to
00706d7
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
00706d7
to
5ebfda4
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
5ebfda4
to
8401250
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
8401250
to
1120927
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
1120927
to
91c0214
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
91c0214
to
f2a6eee
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
Summary: Pull Request resolved: facebookresearch/ClassyVision#181 - Implement Global Reasoning Networks (`glore`) (https://arxiv.org/abs/1811.12814). The reference implementation is on Github (https://github.com/facebookresearch/GloRe) - For now, we don't plan to open source `glore`. Thus, we implement a `ProdResNeXt3D` model in `prod_resnext3d.py`. Refactor `ResNeXt3D` to reduce the duplicate code between `ResNeXt3D` and `ProdResNeXt3D`. - In the long-run, we are more flexible to add more implementation to `ProdResNeXt3D` to meet prod needs while keeping the open source `ResNeXt3D` implementation minimal. - Rename the json configs to tell whether glore is `on` or `off`. Therefore - `train_kinetics400_glore_off_i3d50` means `glore` is off, but we use the baseline I3D-50 model in `glore` paper. - `train_kinetics400_glore_on_i3d50` means `glore` is on and `glore` units are inserted into baseline I3D-50 model. - Similar changes are made to SlowFast `sf` json config. Since we only implement baseline I3D model used in SF, which is a slightly different I3D model from I3D model in `glore`, it is always `off`. Reviewed By: vreis Differential Revision: D18344566 fbshipit-source-id: fac0fe090f7814e5aa41d5be5ff29f198147a27d
f2a6eee
to
b6d453a
Compare
This pull request was exported from Phabricator. Differential Revision: D18344566 |
Summary: Pull Request resolved: #181 - Implement Global Reasoning Networks (`glore`) (https://arxiv.org/abs/1811.12814). The reference implementation is on Github (https://github.com/facebookresearch/GloRe) - For now, we don't plan to open source `glore`. Thus, we implement a `ProdResNeXt3D` model in `prod_resnext3d.py`. Refactor `ResNeXt3D` to reduce the duplicate code between `ResNeXt3D` and `ProdResNeXt3D`. - In the long-run, we are more flexible to add more implementation to `ProdResNeXt3D` to meet prod needs while keeping the open source `ResNeXt3D` implementation minimal. - Rename the json configs to tell whether glore is `on` or `off`. Therefore - `train_kinetics400_glore_off_i3d50` means `glore` is off, but we use the baseline I3D-50 model in `glore` paper. - `train_kinetics400_glore_on_i3d50` means `glore` is on and `glore` units are inserted into baseline I3D-50 model. - Similar changes are made to SlowFast `sf` json config. Since we only implement baseline I3D model used in SF, which is a slightly different I3D model from I3D model in `glore`, it is always `off`. Reviewed By: vreis Differential Revision: D18344566 fbshipit-source-id: ccf27f732a0839d3dfe93afb54c9ce05d1833cf6
Summary:
glore
) (https://arxiv.org/abs/1811.12814). The reference implementation is on Github (https://github.com/facebookresearch/GloRe)on
oroff
. Thereforetrain_kinetics400_glore_off_i3d50
meansglore
is off, but we use the baseline I3D-50 model inglore
paper.train_kinetics400_glore_on_i3d50
meansglore
is on andglore
units are inserted into baseline I3D-50 model.sf
json config. Since we only implement baseline I3D model used in SF, which is a slightly different I3D model from I3D model inglore
, it is alwaysoff
.Differential Revision: D18344566