GitLab

* Initial addition of t5forsequenceclassification

* Adding imports and adding tests

* Formatting

* Running make fix-copies

* Adding mt5forseq

* Formatting

* run make fix-copies

* Adding to docs

* Add model_parallel

* Fix bug

* Fix

* Remove TODO

* Fixing tests for T5ForSequenceClassification

* Undo changes to dependency_versions_table.py

* Change classification head to work with T5Config directly

* Change seq length to let tests pass

* PR comments for formatting

* Formatting

* Initial addition of UMT5ForSequenceClassification

* Adding to inits and formatting

* run make fix-copies

* Add doc for UMT5ForSeqClass

* Update UMT5 config

* Fix docs

* Skip torch fx test for SequenceClassification

* Formatting

* Add skip to UMT5 tests as well

* Fix umt5 tests

* Running make fix-copies

* PR comments

* Fix for change to sentence_representation

* Rename seq_len to hidden_size since that's what it is

* Use base_model to follow format of the rest of the library

* Update docs

* Extract the decoder_input_ids changes and make one liner

* Make one-liner

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

* draft solution

* use `setdefault`

* nits

* add tests and fix truncation issue

* fix test

* test passes locally

* quality

* updates

* update tsets

* tf versions

* apply changes to other models

* 3 models slipped through the cracks

* support left padding

* nit

* Update src/transformers/models/gpt_neox/modeling_gpt_neox.py

* Update src/transformers/models/gpt_neox/modeling_gpt_neox.py

* fix

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

* update

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

* add example NoBadWordsLogitsProcessor

* fix L764 & L767

* make style

* draft add new model like

* some cleaning of the config

* nits

* add nested configs

* nits

* update

* update

* added layer norms + triton kernels

* consider only LPLayerNorm for now.

* update

* all keys match.

* Update

* fixing nits here and there

* working forward pass.

* removed einops dependency

* nits

* format

* add alibi

* byebye head mask

* refactor attention

* nits.

* format

* fix nits.

* nuke ande updates

* nuke tokenizer test

* don't reshape query with kv heads

* added a bit of documentation.

* remove unneeded things

* nuke more stuff

* nit

* logits match - same generations

* rm unneeded methods

* 1 remaining failing CI test

* nit

* fix nits

* fix docs

* fix docs

* rm tokenizer

* fixup

* fixup

* fixup and fix tests

* fixed configuration object.

* use correct activation

* few minor fixes

* clarify docs a bit

* logits match à 1e-12

* skip and unskip a test

* added some slow tests.

* fix readme

* add more details

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

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

* fix configuration issues

* more fixes in config

* added more models

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

* remove unneeded position ids

* fix some  comments

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

* revert suggestion

* mpt alibi + added batched generation

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

* remove init config

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

* fix nit

* add another slow test

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

* fits in one line

* some refactor because make fixup doesn't pass

* add ft notebook

* update md

* correct doc path

---------
Co-authored-by: younesbelkada <younesbelkada@gmail.com>
Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

Repeat per sample for SAM image embeddings

[`generate`]  Only warn users if the `generation_config`'s `max_length` is set to the default value (#25030)

* check max length is default

* nit

* update warning: no-longer deprecate

* comment in the configuration_utils in case max length's default gets changed in the futur

* Fix last models for common tests that are too big.

* Remove print statement

fix rope_scaling doc string

* Better error message when signal is not supported on OS

* Address review comments

fix 8bit corner case with Blip2 8bit

compute_loss in trainer failing to label shift for PEFT model when label smoothing enabled. (#25044)

* added PeftModelForCausalLM to MODEL_FOR_CAUSAL_LM_MAPPING_NAMES dict

* check for PEFT model in compute_loss section

---------
Co-authored-by: Nathan Brake <nbrake3@mmm.com>

* pull and push updates

* add docs

* fix modeling

* Add and run test

* make copies

* add task

* fix tests and fix small issues

* Checks on a Pull Request

* fix docs

* add desc pvt.md

* Dispatch batches

* Copy items

* Better handling missing SYS in llama conversation tokenizer

The existing code failed to add SYS if the conversation has history
without SYS, but did modify the passed conversation as it did.

Rearrange the code so modification to the conversation object are taken
into account for token id generation.

* Fix formatting with black

* Avoid one-liners

* Also fix fast tokenizer

* Drop List decl

* Support GatedRepoError + use raise from

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

* Use token instead of use_auth_token in error messages

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

fix: store training args to wandb config without sanitization.

Allows resuming runs by reusing the wandb config.
Co-authored-by: Bharat Ramanathan <ramanathan.parameshwaran@gohuddl.com>

set default logger

* improve from_pretrained for zero3 multi gpus mode

* Add check if torch.distributed.is_initialized

* Revert torch.distributed

---------
Co-authored-by: Stas Bekman <stas@stason.org>

remove persistent tensor

add simple check for bnb

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

* Avoid importing all models when instantiating a pipeline

* Remove sums that don't work

* pad token should be None by default

* fix tests

* nits

* Update tokenization_llama.py

* Update tokenization_llama_fast.py

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

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

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

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

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

* fix fsdp prepare to remove the warnings and fix excess memory usage

* Update training_args.py

* parity for FSDP+XLA

* Update trainer.py

* fix: cast input pixels to appropriate dtype for image_to_text tasks

* fix: add casting to pixel inputs of additional models after running copy checks

* fix fsdp load

* Update trainer.py

* remove saving duplicate state_dict

* [trainer] fallback for deepspeed param count

* [trainer] more readable numel count

Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>

add GC support for RWKV

* testing

* example script

* fix typehinting

* some tests

* make test

* optional update

* Union of arguments

* does this fix the issue

* remove reports

* set default to False

* documentation change

* None support

* does not need None

* Fix typing annotations for FSDP and DeepSpeed in TrainingArguments (#24549)

* Fix typing annotations for FSDP and DeepSpeed in TrainingArguments

* Change dict to Dict

* Revert "Fix typing annotations for FSDP and DeepSpeed in TrainingArguments" (#24574)

Revert "Fix typing annotations for FSDP and DeepSpeed in TrainingArguments (#24549)"

This reverts commit c5e29d43

.

* Fix typing annotations for FSDP and DeepSpeed in TrainingArguments (#24549)

* Fix typing annotations for FSDP and DeepSpeed in TrainingArguments

* Change dict to Dict

* merge

* hacky fix

* fixup

---------
Co-authored-by: Max Ryabinin <mryabinin0@gmail.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>