GitLab

Update optimization.py

* stash commit (will discard all of this)

* stash commit

* First commit - needs a lot of testing!

* Add a test

* Fix imports and make the tests actually test something

* Tests pass!

* Rearrange test

* Add comments (but it's still a bit confusing)

* Stop storing the tokenizer

* Comment fixup

* Fix for input_ids with a single sequence

* Update tests to test single sequences

* make fixup

* Fix incorrect use of isin()

* Expand tests to catch more cases

* Expand tests to catch more cases

* make fixup

* Fix length calculation and update tests

* Handle Ġ as a space replacement too

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

* Add optimizations from Joao's suggestion

* Remove TODO

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

* Update tests/generation/test_stopping_criteria.py
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* make fixup

* Rename some variables and remove some debugging clauses for clarity

* Add tests for the sub-methods

* Clarify one test slightly

* Add stop_strings to GenerationConfig

* generate() supports stop_string arg, asks for tokenizer if not provided

* make fixup

* Cleanup code and rename variables for clarity

* Update tokenizer error

* Update tokenizer passing, handle generation on GPU

* Slightly more explanation cleanup

* More comment cleanup

* Factor out the token cleanup so it's more obvious what we're doing, and we can change it later

* Careful with that cleanup!

* Cleanup + optimizations to _get_matching_positions

* More minor performance tweaks

* Implement caching and eliminate some expensive ops (startup time: 200ms -> 9ms)

* Remove the pin_memory call

* Parallelize across all stop strings!

* Quick fix for tensor devices

* Update embeddings test for the new format

* Fix test imports

* Manual patching for BERT-like tokenizers

* Return a bool vector instead of a single True/False

* Better comment

* Better comment

* Add tests from @zucchini-nlp

* Amy's list creation nit

* tok_list -> token_list

* Push a big expanded docstring (should we put it somewhere else?)

* Expand docstrings

* Docstring fixups

* Rebase

* make fixup

* Make a properly general method for figuring out token strings

* Fix naming throughout the functions

* Move cache, refactor, fix tests

* Add comment

* Remove finished TODO

* Remove finished TODO

* make fixup

* Update src/transformers/generation/stopping_criteria.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update and shorten docstring

* Update tests to be shorter/clearer and test specific cases

---------
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update docstrings for text generation pipeline

* Fix docstring arg

* Update docstring to explain chat mode

* Fix doctests

* Fix doctests

* nit to make sure cache positions are not sliced

* fix other models

* nit

* style

* Add FSDP config for CPU RAM efficient loading

* Style fix

* Update src/transformers/training_args.py
Co-authored-by: Zach Mueller <muellerzr@gmail.com>

* Update src/transformers/training_args.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Add sync_module_states and cpu_ram_efficient_loading validation logic

* Update src/transformers/training_args.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Style

---------
Co-authored-by: Zach Mueller <muellerzr@gmail.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* warn if pad token is negative

* Update src/transformers/generation/configuration_utils.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

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

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

---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* feat: support for vitmatte

* feat: support for vivit

* feat: support for beit

* feat: support for blip :D

* feat: support for data2vec

* Update llava_next.md

* Update seggpt.md

* Add resources

* Address comments

* Apply suggestions from code review
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

---------
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Duplicate swiftformer

* Convert SwiftFormerPatchEmbedding

* Convert SwiftFormerEmbeddings

* Convert TFSwiftFormerMlp

* Convert TFSwiftFormerConvEncoder

* Convert TFSwiftFormerLocalRepresentation

* convert TFSwiftFormerEncoderBlock

* Convert SwiftFormerStage

* Convert SwiftFormerEncoder

* Add TFSWiftFormerPreTrainedModel

* Convert SwiftFormerForImageClassification

* Add kwargs and start drop path

* Fix syntax

* Change Model class name

* Add TFSwiftFormer to __init__

* Duplicate test_modeling_swiftformer

* First test conversions

* Change require_torch to require_tf

* Add exports to swiftformer __init__

* Add TFSwiftFormerModel wrapper

* Fix __init__ and run black

* Remove docstring from MainLayer, fix padding

* Use keras.layers.Activation on keras.Sequential

* Fix swiftformer exports

* Fix activation layer from config

* Remove post_inits

* Use tf.keras.layers.ZeroPadding2D

* Convert torch normalize

* Change tf test input shape

* Fix softmax and reduce_sum

* Convert expand_dims and repeat

* Add missing reshape and tranpose

* Simplify TFSwiftFormerEncoderBlock.call

* Fix mismatch in patch embeddings

* Fix expected output shape to match channels last

* Fix swiftformer typo

* Disable test_onnx

* Fix TFSwiftFormerForImageClassification call

* Add unpack inputs

* Convert flatten(2).mean(-1)

* Change vision dummy inputs (to be reviewed)

* Change test_forward_signature to use .call

* Fix @unpack_inputs

* Set return_tensors="tf" and rename class

* Rename wrongly named patch_embeddings layer

* Add serving_output and change dummy_input shape

* Make dimensions BCHW and transpose inside embedding layer

* Change SwiftFormerEncoderBlock

* Fix ruff problems

* Add image size to swiftformer config

* Change tranpose to MainLayer and use -1 for reshape

* Remove serving_outputs and dummy_inputs

* Remove test_initialization test from tf model

* Make Sequential component a separate layer

* Fix layers' names

* Tranpose encoder outputs

* Fix tests and check if hidden states is not None

* Fix TFSwiftFormerForImageClassification

* Run make fixup

* Run make fix-copies

* Update modeling_tf_auto

* Update docs

* Fix modeling auto mapping

* Update modelint_tf_swiftformer docs

* Fill image_size doc and type

* Add reduction=None to loss computation

* Update docs

* make style

* Debug: Delete the tip to see if that changes anything

* Re-add tip

* Remove add_code_sample_docstrings

* Remove unused import

* Get the debug to actually tell us the problem it has with the docs

* Try a substitution to match the PyTorch file?

* Add swiftformer to ignore list

* Add build() methods

* Update copyright year
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Remove FIXME comment

* Remove from_pt

* Update copyright year
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Rename one-letter variables

* Remove FIXMEs related to momentum

* Remove old TODO comment

* Remove outstanding FIXME comments

* Get dropout rate from config

* Add specific dropout config for MLP

* Add convencoder dropout to config

* Pass config to SwiftFormerDropPath layer

* Fix drop_path variable name and add Adapted from comment

* Run ruff

* Removed copied from comment

* Run fix copies

* Change drop_path to identity to match pt

* Cleanup build() methods and move to new keras imports

* Update docs/source/en/model_doc/swiftformer.md
Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>

* Raise error if drop_path_rate > 0.0

* Apply suggestions from code review

Replace (self.dim), with self.dim,
Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>

* Remove drop_path function

* Add training to TFSwiftFormerEncoder

* Set self.built = True last
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Should have been added to previous commit
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Apply suggestions from code review
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Change default_feature_extractor to default_image_processor
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Import Keras from modeling_tf_utils

* Remove relative import

* Run ruff --fix

* Move import keras to tf_available

* Add copied from comment to test_forward_signature

* Reduce batch size and num_labels

* Extract loss logic to hf_compute_loss

* Run ruff format

---------
Co-authored-by: Matt <rocketknight1@gmail.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>

* Update modeling_utils.py

* Update test_modeling_utils.py

* Update test_modeling_utils.py

* Update test_modeling_utils.py

* remove seq length from generation tests

* style and quality

* [test_all] & PR suggestion
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

* Update tests/generation/test_utils.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* [test all] remove unused variables

---------
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Use unwrap with the one in accelerate

* oups

* update unwrap

* fix

* wording

* raise error instead

* comment

* doc

* Update src/transformers/modeling_utils.py
Co-authored-by: Zach Mueller <muellerzr@gmail.com>

* style

* put else

---------
Co-authored-by: Zach Mueller <muellerzr@gmail.com>

* fix Parameter dtype in audio models

* restore casting of masked_spec_embed

* restore casting of masked_spec_embed

* initial commit, remove warnings on default chat templates

* stash commit

* Raise a much sterner warning for default chat templates, and prepare for depreciation

* Update the docs

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

* fix tests

* style

* more fixes

* move model to device

* move logits to cpu

* update expected values

* use ungated dataset

* fix

* fix

* update

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

fix again

fixes

* feat: upgrade wandb callback with new features

* fix: ci issues with imports and run fixup

* feat: multidevice for resnet

* feat: yes! resnet

* fix: compare all elements in tuple

* feat: support for regnet

* feat: support for convnextv2

* feat: support for bit

* feat: support for cvt

* feat: add support for focalnet

* feat: support for yolos

* feat: support for glpn

* feat: support for imagegpt

* feat: support for levit

* feat: support for mgp_str

* feat: support for mobilnet_v1

* feat: support for mobilnet_v2

* feat: support for mobilevit

* feat: support for mobilevitv2

* feat: support for poolformer

* fix: copies

* fix: code quality check

* update: upstream changes from main

* fix: consistency check

* feat: support for sam

* feat: support for switchformer

* feat: support for swin

* feat: support for swinv2

* feat: support for timesformer

* feat: suport for trocr

* feat: support for upernet

* fix: check copies

* update: rerun CI

* update: rerun again, maybe

* update: one more rerun

---------
Co-authored-by: Jacky Lee <jackylee328@gmail.com>

* Add special tokens

* Add special tokens

* Use fmt

* Uncomment code

* Add test

* Remove scripts

* Address comments

* Improve tests

* Address comment

* Remove flag

* fix

* fix

* fix

* update comments

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

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

* Alias

* Note alias

* Tests and src

* Rest

* Clean

* Change typing?

* Fix tests

* Deprecation versions

* Fix test with exif_transpose image

* Replace datasets with PIL to load image in tests

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

FIX: Fixes unexpected behaviour for Llava / LLama & AWQ Fused modules + revert #30070 at the same time (#30317)

* Update awq.py

* style

* revert felix PR

* fix

* add felix comments

* wip

* fix __init__.py

* add docs

* Apply suggestions from code review
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* address comments 1

* work on make fixup

* pass configs down

* add sdpa attention

* remove DbrxBlock

* add to configuration_auto

* docstring now passes formatting test

* fix style

* update READMEs

* add dbrx to modeling_auto

* make fix-copies generated this

* add DBRX_PRETRAINED_CONFIG_ARCHIVE_MAP

* config docstring passes formatting test

* rename moe_loss_weight to router_aux_loss_coef

* add to flash-attn documentation

* fix model-path in tests

* Explicitly make `"suli"` the default `ffn_act_fn`
Co-authored-by: Wing Lian <wing.lian@gmail.com>

* default to using router_aux_loss_coef over ffn_config[moe_loss_weight]

* fix _flash_attn_uses_top_left_mask and is_causal

* fix tests path

* don't use token type IDs

* follow Llama and remove token_type_ids from test

* init ConfigTester differently so tests pass

* remove multiple choice test

* remove question + answer test

* remove sequence classification test

* remove token classification test

* copy Llama tests and remove token_type_ids from test inputs

* do not test pruning or headmasking; style code

* add _tied_weights_keys parameter to pass test

* add type hints

* fix type check

* update config tester

* remove masked_lm test

* remove encoder tests

* initialize DbrxModelTester with correct params

* style

* torch_dtype does not rely on torch

* run make fixup, fix-copies

* use https://huggingface.co/v2ray/dbrx-base-fixed/blob/main/modeling_dbrx.py



* add copyright info

* fix imports and DbrxRotaryEmbedding

* update DbrxModel docstring

* use copies

* change model path in docstring

* use config in DbrxFFN

* fix flashattention2, sdpaattention

* input config to DbrXAttention, DbrxNormAttentionNorm

* more fixes

* fix

* fix again!

* add informative comment

* fix ruff?

* remove print statement + style

* change doc-test

* fix doc-test

* fix docstring

* delete commented out text

* make defaults match dbrx-instruct

* replace `router_aux_loss_coef` with `moe_loss_weight`

* is_decoder=True

* remove is_decoder from configtester

* implement sdpa properly

* make is_decoder pass tests

* start on the GenerationTesterMixin tests

* add dbrx to sdpa documentation

* skip weight typing test

* style

* initialize smaller model
Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>

* Add DBRX to toctree

* skip test_new_cache_format

* make config defaults smaller again

* add pad_token_id

* remove pad_token_id from config

* Remove all references to DBRX_PRETRAINED_CONFIG_ARCHIVE_MAP

* Update src/transformers/models/dbrx/__init__.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update src/transformers/models/dbrx/modeling_dbrx.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update docs/source/en/model_doc/dbrx.md
Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>

* Update src/transformers/models/dbrx/configuration_dbrx.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Update docs/source/en/model_doc/dbrx.md
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* fix typo

* Apply suggestions from code review
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* update docs, fix configuration_auto.py

* address pr comments

* remove is_decoder flag

* slice

* fix requires grad

* remove grad

* disconnect differently

* remove grad

* enable grads

* patch

* detach expert

* nissan al ghaib

* Update modeling_dbrx.py

* Update src/transformers/models/dbrx/modeling_dbrx.py
Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>

* replace "Gemma" with "Dbrx"

* remove # type: ignore

* don't hardcode vocab_size

* remove ToDo

* Re-add removed idefics2 line

* Update test to use tiny-random!

* Remove TODO

* Remove one more case of loading the entire dbrx-instruct in the tests

* Update src/transformers/models/dbrx/modeling_dbrx.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* address some comments

* small model

* add dbrx to tokenization_auto

* More docstrings with add_start_docstrings

* Dbrx for now

* add PipelineTesterMixin

* Update src/transformers/models/dbrx/configuration_dbrx.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* remove flash-attn2 import error

* fix docstring
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* add useage example

* put on one line
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* fix ffn_act_fn
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* change "dbrx" to "DBRX" for display purposes.

* fix __init__.py?

* fix __init__.py

* fix README

* return the aux_loss

* remove extra spaces

* fix configuration_auto.py

* fix format in tokenization_auto

* remove new line

* add more useage examples

---------
Co-authored-by: Abhi Venigalla <abhi.venigalla@databricks.com>
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>
Co-authored-by: Eitan Turok <eitan.turok@databricks.com>
Co-authored-by: Eitan Turok <150733043+eitanturok@users.noreply.github.com>
Co-authored-by: Wing Lian <wing.lian@gmail.com>
Co-authored-by: Eitan Turok <eitanturok@gmail.com>
Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>
Co-authored-by: Matt <rocketknight1@gmail.com>
Co-authored-by: Your Name <you@example.com>
Co-authored-by: Mihir Patel <mihir.v.patel7@gmail.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* overlooked

* style

* cleaner

* Revert "Re-enable SDPA's FA2 path (#30070)"

This reverts commit 05bdef16.

* Revert "Fix quality Olmo + SDPA (#30302)"

This reverts commit ec92f983.

* Switch to non persistant buffer

* fix device mismatch issue due to cache

* style

atol for sliding window test

* Add jamba arch

* apply "make fix-copies" changes

* fix link to model in JambaConfig docstring

* Add n_ctx in modeling file because repo-consistency wants that

* Add jamba to flash attention and sdpa documentation

* mamba dt_proj quant fix now works for LoRA as well

* override test_left_padding_compatibility and use a more permissive tolerance. left padding numerical difference are accentuated by mamba layers

* add jamba to tokenization auto

* fix comments of shape (PR #24 in the model page: https://huggingface.co/ai21labs/Jamba-v0.1/discussions/24)

* simple PR fixes

* remove unnecessary kwargs from JambaAttentionDecoderLayer and JambaMambaDecoderLayer

* remove the LoRA hack for the mamba dt_proj bias. It was solved in huggingface/peft#1530 (https://github.com/huggingface/peft/pull/1530)

* Add copied comment on JambaMLP (it's the same as MixtralMLP)

* remove padding_mask warnings. It's not supported anymore

* fix docstring. Float instead of int

* A few more minor PR fixes

* (1) lowercase names for mamba layernorms (2) remove _apply_inner_layernorms and do it directly in the forward pass

* Return None attention weights from mamba layers. Append to all attentions only if not None.

* remove some leftover jamba archive lists

* Better separation between expert vs non-expert layers. non-expert layers return None as router_logits, and it is not concatenated to all_router_logits returned from JambaModel

* no need to take router_logits at config.expert_layer_offset anymore. result.router_logits now holds results only for expert layers

* Add Jamba paper on READMEs

* (1) rename n_ctx -> max_position_embeddings (2) don't use it in the modeling file since it's not needed (set it as an exception to check_config_attributes)

* Add copied from comment

* remove the code path for apply_inner_layernorms=False. Jamba always has the inner mamba layernorms

* clearer docstring for _convert_to_standard_cache

* style fixes

* Change calc_logits_for_entire_prompt (bool) to num_logits_to_keep (int). Adapt assisted decoding code tp use it. Also small change in low memory beam search decoding path to support this new int value in model_inputs

* rename test so it still overrides what its meant to override

* draft

* oups

* nit

* remove more complexe logic

* fix names used in config

* fix fix fix

* style

* fix some more failing tests

* generate did not init the cache 🙃



* more small nits

* typo

* config.mamba_expand * config.hidden_size for the intermediate size of the mamba shapes

* fix init of pkv with torch.tensor()

* empty tensor

* fix some init issues

* stupid changes required by generate because it does not even support it's own DynamicCache class

* more fixes

* fix general assisted gen cache_position bug

* tests passing

* Add offsets and periods as SPECIAL_CASES_TO_ALLOW in check_config_attributes.py

* fix reorder_cache to reorder mamba states and override some more functions in HybridMambaAttentionDynamicCache

* no need to override test_past_key_values_format() and _check_past_key_values_for_generate() in tests anymore

* fix docstrings and typehints for past_key_values

* style fixes

* fix docs

* change typehint due to copy from Mixtral

* forgot import

* import order

* Add configuration_jamba and modeling_jamba to not_doctested because the model is too big to download (in docstring of JambaForCausalLM.forward)

* Add integration test with tiny tandom Jamba model on hub

* fix flash attention cache shapes

* bring back forgotten hidden states

* rename HybridMambaAttentionDynamicCache.seqlen_offset to has_previous_state (and make bool) and bugfix - it should be set to True after a finished forward pass of the entire model

* align integration test after modeling fixes

* bugfix - mamba can use precomputed states only of forward pass is on a single token

* bugfix - mamba can use precomputed states only if they match the batch size

* typo

* remove making _prepare_4d_causal_attention_mask a leaf function

* stop using past_seq_len.get_seq_length(). Use cache positions instead. Adjust test (test_decoder_model_past_with_large_inputs) accordingly

---------
Co-authored-by: Arthur Zucker <arthur.zucker@gmail.com>
Co-authored-by: Joao Gante <joao@huggingface.co>

* fix

* fix

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