GitLab

* Script & Manual edition

* Update

* add sdpa

* wip

* cleaning

* add ref

* yet more cleaning

* and more :)

* wip llama

* working llama

* add output_attentions=True support

* bigcode sdpa support

* fixes

* gpt-bigcode support, require torch>=2.1.1

* add falcon support

* fix conflicts falcon

* style

* fix attention_mask definition

* remove output_attentions from attnmaskconverter

* support whisper without removing any Copied from statement

* fix mbart default to eager renaming

* fix typo in falcon

* fix is_causal in SDPA

* check is_flash_attn_2_available in the models init as well in case the model is not initialized through from_pretrained

* add warnings when falling back on the manual implementation

* precise doc

* wip replace _flash_attn_enabled by config.attn_implementation

* fix typo

* add tests

* style

* add a copy.deepcopy on the config in from_pretrained, as we do not want to modify it inplace

* obey to config.attn_implementation if a config is passed in from_pretrained

* fix is_torch_sdpa_available when torch is not installed

* remove dead code

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

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

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

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

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

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

* remove duplicate pretraining_tp code

* add dropout in llama

* precise comment on attn_mask

* add fmt: off for _unmask_unattended docstring

* precise num_masks comment

* nuke pretraining_tp in LlamaSDPAAttention following Arthur's suggestion

* cleanup modeling_utils

* backward compatibility

* fix style as requested

* style

* improve documentation

* test pass

* style

* add _unmask_unattended tests

* skip meaningless tests for idefics

* hard_check SDPA requirements when specifically requested

* standardize the use if XXX_ATTENTION_CLASSES

* fix SDPA bug with mem-efficient backend on CUDA when using fp32

* fix test

* rely on SDPA is_causal parameter to handle the causal mask in some cases

* fix FALCON_ATTENTION_CLASSES

* remove _flash_attn_2_enabled occurences

* fix test

* add OPT to the list of supported flash models

* improve test

* properly test on different SDPA backends, on different dtypes & properly handle separately the pad tokens in the test

* remove remaining _flash_attn_2_enabled occurence

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

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

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

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

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

* remove use_attn_implementation

* fix docstring & slight bug

* make attn_implementation internal (_attn_implementation)

* typos

* fix tests

* deprecate use_flash_attention_2=True

* fix test

* add back llama that was removed by mistake

* fix tests

* remove _flash_attn_2_enabled occurences bis

* add check & test that passed attn_implementation is valid

* fix falcon torchscript export

* fix device of mask in tests

* add tip about torch.jit.trace and move bt doc below sdpa

* fix parameterized.expand order

* move tests from test_modeling_attn_mask_utils to test_modeling_utils as a relevant test class is already there

* update sdpaattention class with the new cache

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

* Update src/transformers/models/bark/modeling_bark.py

* address review comments

* WIP torch.jit.trace fix. left: test both eager & sdpa

* add test for torch.jit.trace for both eager/sdpa

* fix falcon with torch==2.0 that needs to use sdpa

* fix doc

* hopefully last fix

* fix key_value_length that has no default now in mask converter

* is it flacky?

* fix speculative decoding bug

* tests do pass

* fix following #27907

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

* first batch of structure improvements for model_docs

* second batch of structure improvements for model_docs

* more structure improvements for model_docs

* more structure improvements for model_docs

* structure improvements for cv model_docs

* more structural refactoring

* addressed feedback about image processors

* added flash attention of gpt_bigcode

* changed docs

* Update src/transformers/models/gpt_bigcode/modeling_gpt_bigcode.py

* add FA-2 docs

* oops

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

* fix

* oops

* remove padding_mask

* change getattr->hasattr logic

* changed .md file

---------
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>
Co-authored-by: younesbelkada <younesbelkada@gmail.com>
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>

* Rename index.mdx to index.md

* With saved modifs

* Address review comment

* Treat all files

* .mdx -> .md

* Remove special char

* Update utils/tests_fetcher.py
Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>

---------
Co-authored-by: Lysandre Debut <lysandre.debut@reseau.eseo.fr>

* Add model with cli tool

* Remove unwanted stuff

* Add new code

* Remove inference runner

* Style

* Fix checks

* Test updates

* make fixup

* fix docs

* fix doc

* fix test

* hopefully fix pipeline tests

* refactor

* fix CIs

* add comment

* rename to `GPTBigCodeForCausalLM`

* correct readme

* make fixup + docs

* make fixup

* fixes

* fixes

* Remove pruning

* Remove import

* Doc updates

* More pruning removal

* Combine copies

* Single MQA implementation, remove kv cache pre-allocation and padding

* Update doc

* Revert refactor to match gpt2 style

* Merge back key and value caches, fix some type hints

* Update doc

* Fix position ids pith padding (PR 21080)

* Add conversion script temporarily

* Update conversion script

* Remove checkpoint conversion

* New model

* Fix MQA test

* Fix copies

* try fix tests

* FIX TEST!!

* remove  `DoubleHeadsModel`

* add MQA tests

* add slow tests

* clean up

* add CPU checker

* final fixes

* fixes

- fix GPU issue
- fixed slow tests
- skip disk offload

* fix final issue

* Simplify and comment baddbmm fix

* Remove unnecessary code

* Transpose tweaks

* Use beta=1 on cpu, improve tests

---------
Co-authored-by: younesbelkada <younesbelkada@gmail.com>