GitLab

* Clean up deprecation warnings

Notes:
Changed some strings in tests to raw strings, which will change the literal content of the strings as they are fed into whatever machine handles them.
Test cases for past in the past/past_key_values switch changed/removed due to warning of impending removal

* Add PILImageResampling abstraction for PIL.Image.Resampling

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

This PR (https://github.com/huggingface/transformers/pull/19367) introduced a few breaking changes:

- Removed an argument `mask_threshold`.
- Broke the default behavior (instance vs panoptic in the function call)
  https://github.com/huggingface/transformers/pull/19367/files#diff-60f846b86fb6a21d4caf60f5b3d593a04accb8f248de3029cccae2ff898c5bc3R119-R120
- Broke the actual masks: https://github.com/huggingface/transformers/pull/1961

This PR is the start of a handful that will aim at bringing back the old
behavior(s).

- tests should not have to specify `task` by default, unless we want to
  modify the behavior and have a lower form of segmentation running)
- `test_small_model_pt` should be working.

This specific PR starts with adding more information to the masks hash
because missing the actual mask was actual easy to miss (the hashes do
change, but it was easy to miss that one code path wasn't properly
updated).

So we go from a simple `hash` to
```
{"hash": #smaller hash, "shape": (h, w), "white_pixels": n}
```

The `shape` should help make sure the interpolation of the mask works
correctly, the `white_pixels` hopefully helps detect big regressions in
their amount when the hash gets modified.

* add return_tensors parameter for feature_extraction  w/ test

add return_tensor parameter for feature extraction

Revert "Merge branch 'feature-extraction-return-tensor' of https://github.com/ajsanjoaquin/transformers

 into feature-extraction-return-tensor"

This reverts commit d559da743b87914e111a84a98ba6dbb70d08ad88, reversing
changes made to bbef89278650c04c090beb65637a8e9572dba222.

call parameter directly
Co-authored-by: Nicolas Patry <patry.nicolas@protonmail.com>

Fixup.

Update src/transformers/pipelines/feature_extraction.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Fix the imports.

* Fixing the test by not overflowing the model capacity.
Co-authored-by: AJ San Joaquin <ajsanjoaquin@gmail.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

🚨 🚨 🚨

 [Breaking change] Deformable DETR intermediate representations (#19678)

* [Breaking change] Deformable DETR intermediate representations

- Fixes naturally the `object-detection` pipeline.
- Moves from `[n_decoders, batch_size, ...]` to `[batch_size,
  n_decoders, ...]` instead.

* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

This reverts commit 35bd089a.

* add return_tensors parameter for feature_extraction  w/ test

add return_tensor parameter for feature extraction

Revert "Merge branch 'feature-extraction-return-tensor' of https://github.com/ajsanjoaquin/transformers

 into feature-extraction-return-tensor"

This reverts commit d559da743b87914e111a84a98ba6dbb70d08ad88, reversing
changes made to bbef89278650c04c090beb65637a8e9572dba222.

* call parameter directly
Co-authored-by: Nicolas Patry <patry.nicolas@protonmail.com>

* Fixup.

* Update src/transformers/pipelines/feature_extraction.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Nicolas Patry <patry.nicolas@protonmail.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Fixes

* update expected values

* style

* fix
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

* Partial TF port for ESM model

* Add ESM-TF tests

* Add the various imports for TF-ESM

* TF weight conversion almost ready

* Stop ignoring the decoder weights in PT

* Add tests and lots of fixes

* fix-copies

* Fix imports, add model docs

* Add get_vocab() to tokenizer

* Fix vocab links for pretrained files

* Allow multiple inputs with a sep

* Use EOS as SEP token because ESM vocab lacks SEP

* Correctly return special tokens mask from ESM tokenizer

* make fixup

* Stop testing unsupported embedding resizing

* Handle TF bias correctly

* Skip all models with slow tokenizers in the token classification test

* Fixing the batch/unbatcher of pipelines to accomodate the `None` being

passed around.

* Fixing pipeline bug caused by slow tokenizer  being different.

* Update src/transformers/models/esm/modeling_tf_esm.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* Update src/transformers/models/esm/modeling_tf_esm.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* Update src/transformers/models/esm/modeling_tf_esm.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* Update set_input_embeddings and the copyright notices
Co-authored-by: Your Name <you@example.com>
Co-authored-by: Nicolas Patry <patry.nicolas@protonmail.com>
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* Remove key word argument X from pipeline predict and transform methods

As __call__ of pipeline clasees require one positional argument, passing
the input as a keyword argument inside predict, transform methods, causing
__call__ to fail. Hence in this commit the keyword argument is modified
into positional argument.

* Implement basic tests for scikitcompat pipeline interface

* Seperate tests instead of running with parameterized based on framework as both frameworks will not be active at the same time

* Improve error messaging for ASR pipeline.

- Raise error early (in `_sanitize`) so users don't waste time trying to
  run queries with invalid params.

- Fix the error was after using `config.inputs_to_logits_ratio` so our
  check was masked by the failing property does not exist.

- Added some manual check on s2t for the error message.
  No non ctc model seems to be used by the default runner (they are all
  skipped).

* Removing pdb.

* Stop the early error it doesn't really work :(.

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

* Cast masks to np.unit8 before converting to PIL.Image.Image

* Update tests

* Fixup

* Add initial files for depth estimation pipelines

* Add test file for depth estimation pipeline

* Update model mapping names

* Add updates for depth estimation output

* Add generic test

* Hopefully fixing the tests.

* Check if test passes

* Add make fixup and make fix-copies changes after rebase with main

* Rebase with main

* Fixing up depth pipeline.

* This is not used anymore.

* Fixing the test. `Image` is a module `Image.Image` is the type.

* Update docs/source/en/main_classes/pipelines.mdx
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Nicolas Patry <patry.nicolas@protonmail.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Added tokenize keyword arguments to feature extraction pipeline

* Reverted truncation parameter

* Import numpy moved to top

* Implement multiple span support

* Address comments

* Add tests + fix bugs

* update feature extractor params

* update attention mask handling

* fix doc and pipeline test

* add warning when skipping test

* add whisper translation and transcription test

* fix build doc test

* Rework pipeline tests

* Try to fix Flax tests

* Try to put it before

* Use a new decorator instead

* Remove ignore marker since it doesn't work

* Filter pipeline tests

* Woopsie

* Use the fitlered list

* Clean up and fake modif

* Remove init

* Revert fake modif

- Fixes the image segmentation pipeline test failures caused by changes to the postprocessing methods of supported models
- Updates the ImageSegmentationPipeline tests
- Improves docs, adds 'task' argument to optionally perform semantic, instance or panoptic segmentation

* Add ZeroShotObjectDetectionPipeline (#18445)

* Add AutoModelForZeroShotObjectDetection task

This commit also adds the following

- Add explicit _processor method for ZeroShotObjectDetectionPipeline.
  This is necessary as pipelines don't auto infer processors yet and
  `OwlVitProcessor` wraps tokenizer and feature_extractor together, to
  process multiple images at once

- Add auto tests and other tests for ZeroShotObjectDetectionPipeline

* Add AutoModelForZeroShotObjectDetection task

This commit also adds the following

- Add explicit _processor method for ZeroShotObjectDetectionPipeline.
  This is necessary as pipelines don't auto infer processors yet and
  `OwlVitProcessor` wraps tokenizer and feature_extractor together, to
  process multiple images at once

- Add auto tests and other tests for ZeroShotObjectDetectionPipeline

* Add batching for ZeroShotObjectDetectionPipeline

* Fix doc-string ZeroShotObjectDetectionPipeline

* Fix output format: ZeroShotObjectDetectionPipeline

* Fix pipeline tests for Roberta-like tokenizers

* Fix fix

Co-authored-by: Ankur Goyal <ankur@impira.com>

* Updated test values

The image segmentation pipeline tests - tests/pipelines/test_pipelines_image_segmentation.py - were failing after the merging of #1849  (49e44b21). This was due to the difference in rescaling. Previously the images were rescaled by `image = image / 255`. In the new commit, a `rescale` method was added, and images rescaled using `image = image * scale`. This was known to cause small differences in the processed images (see
[PR comment](https://github.com/huggingface/transformers/pull/18499#discussion_r940347575)).

Testing locally, changing the `rescale` method to divide by a scale factor (255) resulted in the tests passing. It was therefore decided the test values could be updated, as there was no logic difference between the commits.

* Use double quotes, like previous example

* Fix up

* Fix DocumentQuestionAnsweringPipelineTests
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

* First draft

* More improvements

* Improve model, add custom CUDA code

* Import torch before

* Add script that imports custom layer

* Add everything in new ops directory

* Import custom layer in modeling file

* Fix ARCHIVE_MAP typo

* Creating the custom kernel on the fly.

* Import custom layer in modeling file

* More improvements

* Fix CUDA loading

* More improvements

* Improve conversion script

* Improve conversion script

* Make it work until encoder_outputs

* Make forward pass work

* More improvements

* Make logits match original implementation

* Make implementation also support single_scale model

* Add support for single_scale and dilation checkpoint

* Add support for with_box_refine model

* Support also two stage model

* Improve tests

* Fix more tests

* Make more tests pass

* Upload all models to the hub

* Clean up some code

* Improve decoder outputs

* Rename intermediate hidden states and reference points

* Improve model outputs

* Move tests to dedicated folder

* Improve model outputs

* Fix retain_grad test

* Improve docs

* Clean up and make test_initialization pass

* Improve variable names

* Add copied from statements

* Improve docs

* Fix style

* Improve docs

* Improve docs, move tests to model folder

* Fix rebase

* Remove DetrForSegmentation from auto mapping

* Apply suggestions from code review

* Improve variable names and docstrings

* Apply some more suggestions from code review

* Apply suggestion from code review

* better docs and variables names

* hint to num_queries and two_stage confusion

* remove asserts and code refactor

* add exception if two_stage is True and with_box_refine is False

* use f-strings

* Improve docs and variable names

* Fix code quality

* Fix rebase

* Add require_torch_gpu decorator

* Add pip install ninja to CI jobs

* Apply suggestion of @sgugger

* Remove DeformableDetrForObjectDetection from auto mapping

* Remove DeformableDetrModel from auto mapping

* Add model to toctree

* Add model back to mappings, skip model in pipeline tests

* Apply @sgugger's suggestion

* Fix imports in the init

* Fix copies

* Add CPU implementation

* Comment out GPU function

* Undo previous change

* Apply more suggestions

* Remove require_torch_gpu annotator

* Fix quality

* Add logger.info

* Fix logger

* Fix variable names

* Fix initializaztion

* Add missing initialization

* Update checkpoint name

* Add model to doc tests

* Add CPU/GPU equivalence test

* Add Deformable DETR to pipeline tests

* Skip model for object detection pipeline
Co-authored-by: Nicolas Patry <patry.nicolas@protonmail.com>
Co-authored-by: Nouamane Tazi <nouamane98@gmail.com>
Co-authored-by: Sylvain Gugger <Sylvain.gugger@gmail.com>

* [WIP] Skeleton of VisualQuestionAnweringPipeline extended to support LayoutLM-like models

* Fixup

* Use the full encoding

* Basic refactoring to DocumentQuestionAnsweringPipeline

* Cleanup

* Improve args, docs, and implement preprocessing

* Integrate OCR

* Refactor question_answering pipeline

* Use refactored QA code in the document qa pipeline

* Fix tests

* Some small cleanups

* Use a string type annotation for Image.Image

* Update encoding with image features

* Wire through the basic docs

* Handle invalid response

* Handle empty word_boxes properly

* Docstring fix

* Integrate Donut model

* Fixup

* Incorporate comments

* Address comments

* Initial incorporation of tests

* Address Comments

* Change assert to ValueError

* Comments

* Wrap `score` in float to make it JSON serializable

* Incorporate AutoModeLForDocumentQuestionAnswering changes

* Fixup

* Rename postprocess function

* Fix auto import

* Applying comments

* Improve docs

* Remove extra assets and add copyright

* Address comments
Co-authored-by: Ankur Goyal <ankur@impira.com>

* Further reduce the number of alls to head for cached models/tokenizers/pipelines

* Fix tests

* Address review comments

Co-authored-by: Olivier Dehaene <olivier@huggingface.co>

* Add Image2TextGenerationPipeline to supported pipelines

* Add Flax and Tensorflow support

* Add Flax and Tensorflow small tests

* Add default model for Tensorflow

* Add docstring

* Fix doc style

* Add tiny models for pytorch and flax

* Remove flax from pipeline.
Fix tests

* Use ydshieh/vit-gpt2-coco-en as a default for both PyTorch and Tensorflow

* Fix Tensorflow support
Co-authored-by: Olivier Dehaene <olivier@huggingface.co>

* Use commit hash to look in cache instead of calling head

* Add tests

* Add attr for local configs too

* Stupid typos

* Fix tests

* Update src/transformers/utils/hub.py
Co-authored-by: Julien Chaumond <julien@huggingface.co>

* Address Julien's comments
Co-authored-by: Julien Chaumond <julien@huggingface.co>

* Adding a new `align_to_words` param to qa pipeline.

* Update src/transformers/pipelines/question_answering.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Import protection.
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* [DX fix] Fixing QA pipeline streaming a dataset.

QuestionAnsweringArgumentHandler would iterate over the whole dataset
effectively killing all properties of the pipeline.
This restores nice properties when using `Dataset` or `Generator` since
those are meant to be consumed lazily.

* Handling TF better.

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

* Fix pipeline tests

* Make sure all pipelines tests run with init changes

* Adding a better error message when the model is improperly configured

within transformers.

* Update src/transformers/pipelines/__init__.py

* Black version.

* Overriding task aliases so that tokenizer+feature_extractor

values are correct.

* Fixing task aliases by overriding their names early

* X.

* Fixing feature-extraction.

* black again.

* Normalizing `translation` too.

* Fixing last few corner cases.

translation need to use its non normalized name (translation_XX_to_YY,
so that the task_specific_params are correctly overloaded).
This can be removed and cleaned up in a later PR.

`speech-encode-decoder` actually REQUIRES to pass a `tokenizer` manually
so the error needs to be discarded when the `tokenizer` is already
there.

* doc-builder fix.

* Fixing the real issue.

* Removing dead code.

* Do not import the actual config classes.