[New Model] Add TimeSformer model#18908
Conversation
…into timesformer
There was a problem hiding this comment.
Thanks for your work, looks great to me. Pinging @amyeroberts for the feature extractor part; I guess we can't replace it with VideoProcessor for the moment?
Thanks for all the constructive feedback! |
sgugger
left a comment
There was a problem hiding this comment.
Thanks a lot for adding this new model! I have a couple of nits and the checkpoints should be put in the right place, but good to merge otherwise!
src/transformers/models/timesformer/configuration_timesformer.py
Outdated
Show resolved
Hide resolved
|
Hello @NielsRogge, thanks a lot for all the help you have provided. I have opened multiple PRs for the config and model files of all timesformer variants: Is there anything I should do about this PR? |
|
Looks great on my side and ready to merge! Will let @NielsRogge double-check on last time and merge if he's happy :-) |
|
Thanks for all your work! Feel free to share on social media, we'll amplify ;) |
|
Thanks a lot @NielsRogge, will share it after preparing a space :) |
* init timesformer * apply fix-copies * reformat style * revert back some incoorect style updates * init timesformer * apply fix-copies * reformat style * revert back some incoorect style updates * update timseformer doc * add some functions and classes * add new config params * implement multiple classes * update TimeSformerLayer * update TimeSformerModel, TimeSformerPreTrainedModel, TimeSformerEncoder * several fixes * reformat * temporary update * fix some typos * fix weight converter * more fixes * fix a typo * fix typo * remove redundant params * fix for latest hf-hub * merge fix * fix some checks * video classification works with einops * add paper info to docs * merge fix * remove redundant line * remove redundant docstring * update config * fix some typos * fix converter * update some test constants * refactor einops functions * reformat * fix a comment * remove redundat imports * reformat * fix a typo * remove comment * remove unused imports * remove redundant doc line * reformat * add missing line * fix docs * fix timesformer auto feat ext * add unittests * reformat * fix docs * some fixes and updates * fix readme * fix modeling * fix readme * update index * revert _toctree.yml changes * update timseformer.mdx * update drop_path_prob to drop_path_rate * add dosctring for drop_path_rate * update TimeSformerPatchEmbed naming * remove to_2tuple * explicit use of nn.functional * reformat * many updates from review comments * fix a typo * reformat * remove assert, better variable name * make variable names more explicit * add some adapted from * more explicit variable names * remove redundant docstring * fix initilaization * move permute inside embedding * update class names * remove unused imports * add test for video classification * update PretrainedModel with PreTrainedModel * remove double permute * update based on sylvain's review * aply auto fix * update image_processing_auto for timesformer * update hub urls * reformat * remove duplicate import * update doc link
What does this PR do?
Fixes #18724
Before submitting
Pull Request section?
to it if that's the case.
documentation guidelines, and
here are tips on formatting docstrings.
@NielsRogge