Add dlclibrary for model download from zoo#2088
Merged
Merged
Conversation
AlexEMG
requested changes
Dec 14, 2022
Member
AlexEMG
left a comment
There was a problem hiding this comment.
the /tools files should stay. Let's also add a unit-test for downloading model zoo weights to DLC (so we catch problems with DLClib here)?
Member
|
also let's add dlclib to setup.py? |
Contributor
Author
why? it's part of dlclib too |
Member
|
Right, good to use CLI at DLClib, but I think testing it here too, will make sure that we catch it (as I assume we and others look more on DLC's CLI). |
MMathisLab
approved these changes
Dec 20, 2022
Member
|
ready to merge @jeylau ? @yeshaokai ? |
Collaborator
|
I haven't tested all cases. It works for my use cases. |
AlexEMG
approved these changes
Dec 22, 2022
jeylau
added a commit
that referenced
this pull request
Dec 23, 2022
* Remove tools folder * Add dlclibrary for model download from zoo * also remove old weight download function * Add dlclibrary to setup.py * Update branch name to "main" in GitHub's CI workflow (#2089) (#2092) * Update branch name * Remove GUI-related deps from requirements.txt * Update actions' versions Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * put tools back * updated headers. * update requirements for CLI * fix trailing """ * model zoo unit-test * DLC 2.3 Co-authored-by: Alexander Mathis <alexander@deeplabcut.org> Co-authored-by: Mackenzie Mathis <mathis@rowland.harvard.edu>
AlexEMG
added a commit
that referenced
this pull request
Mar 7, 2023
* Added inference code for incoming super models * handled empty prediction in multiple scale prediction. Formatting * Update deeplabcut/pose_estimation_tensorflow/predict_supermodel.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Update deeplabcut/pose_estimation_tensorflow/predict_supermodel.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Update deeplabcut/pose_estimation_tensorflow/predict_supermodel.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Update deeplabcut/pose_estimation_tensorflow/predict_supermodel.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * fixed DLCscorer in make_label_video for suffix * Added example document. Added top down inference * Fixed previous commit * Made the scorer convetion include DLC_ * handled the case where one of topdown prediction gives empty prediction * first docs started * Update _toc.yml * change scale_list nonoptional * Added functional test script and api docs * Fixed a formatting * safer way to parse num_kpts * Made the superanimal inference simpler * Simplify the API and added more docs * changed the relative path in the rst file * Fixed circular import * Fixed known issue * Further getting rid of the dependency on config for inference * make_labeled_video takes superanimal_name now * Update docs/ModelZoo.md Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Cleanup and removal of bbox file * Minor fixes * Run black and isort * Remove unused variables * Get rid of the additional VideoWriter * Fix test * Hide _video_inference from high-level API * Add test_get_multi_scale_frames * Clean configs * Accelerate pandas's code * Test projection back to original image size * Faster DataFrame creation * Remove tools folder * Add dlclibrary for model download from zoo * also remove old weight download function * Add dlclibrary to setup.py * Added modelzoo folder * removed domain matcher for now * fix last commit * Users define threshold and pcutoff for adaptation api * Added documentation. Only keep the without project way of using the api * Now video adaptation code uses video reader for training * reformatting * Update deeplabcut/modelzoo/apis/spatiotemporal_adapt.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Update deeplabcut/modelzoo/apis/spatiotemporal_adapt.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Update deeplabcut/modelzoo/apis/spatiotemporal_adapt.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Remove unused imports * Remove duplicate code * Fixed unused threshold * Avoid hardcoding supermodel names * Add missing `vid` attribute * Delay import * Add missing attribute * Add video adaptation to GUI * Update branch name to "main" in GitHub's CI workflow (#2089) * Update branch name * Remove GUI-related deps from requirements.txt * Update actions' versions * inference api and adaptation api work with hugging face weights * minor update * trying to avoid circular import * WIP fixing previous commit * fixed logic for getting image shape for video adaptation * smaller batch sizes in train configs * fix the path after enforcing abs path * Added colab example * adaptation api takes kwargs * updated docstring for inference code * Update ModelZoo.md * Add dlclibrary for model download from zoo (#2088) * Remove tools folder * Add dlclibrary for model download from zoo * also remove old weight download function * Add dlclibrary to setup.py * Update branch name to "main" in GitHub's CI workflow (#2089) (#2092) * Update branch name * Remove GUI-related deps from requirements.txt * Update actions' versions Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * put tools back * updated headers. * update requirements for CLI * fix trailing """ * model zoo unit-test * DLC 2.3 Co-authored-by: Alexander Mathis <alexander@deeplabcut.org> Co-authored-by: Mackenzie Mathis <mathis@rowland.harvard.edu> * Backwards comp * Add some multi-animal defaults * Add a missing parameter * Add empty paf graph * Change dataset_type to single-animal * Update ModelZoo.md * Minor fixes * Video adaptation api (up to date) (#2116) * Fixed previous commit * first docs started * Added functional test script and api docs * Simplify the API and added more docs * Add dlclibrary for model download from zoo * Now video adaptation code uses video reader for training * reformatting * Update ModelZoo.md * Add some multi-animal defaults * allow more custom changes * change model names in models.json to fit the official convention * updated test scripts * Added test script for adaptation api. Handled empty predictions * Get rid of get_nuances function * rename modelzoo/apis to modelzoo/api. Make it possible to choose video adaptation in the inference api * updated the doc string * addressed jessys comments * ran black * Update ModelZoo.md * Add back cute emojis * no .imports now * added back imgaug line * Update deeplabcut/modelzoo/api/spatiotemporal_adapt.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Update deeplabcut/modelzoo/api/spatiotemporal_adapt.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Update deeplabcut/modelzoo/api/spatiotemporal_adapt.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Update deeplabcut/modelzoo/api/spatiotemporal_adapt.py Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * adddressed jessys comments Co-authored-by: Mackenzie Mathis <mathis@rowland.harvard.edu> Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> * Fix imports * Fix multiprocessing RuntimeError * Fix model zoo GUI tab * Fix typo * Fix custom config path * Test video adaptation on a shorter video * Avoid empty folder creation if not adapting * Remove unused apply_filter flag * Avoid multiprocessing when started with spawn * Fix hardcoded paths * Option to pass in display and saveiters * Append video analysis success * Silence numpy RuntimeWarnings * Add Colab example and form fields * 30x speedup of averaging over multiple scales * Add Open in Colab badge * Minor change * Use newer API in modelzoo's GUI tab * Add trajectory plotting to superanimal inference * Update spatiotemporal_adapt.py * Rename variable * Add tools directory back in --------- Co-authored-by: shaokaiye <shaokaiyeah@gmail.com> Co-authored-by: Jessy Lauer <30733203+jeylau@users.noreply.github.com> Co-authored-by: Alexander Mathis <alexander@deeplabcut.org>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes #2086