* Initial implementation of OffloadedCache
* enable usage via cache_implementation
* Address feedback, add tests, remove legacy methods.
* Remove flash-attn, discover synchronization bugs, fix bugs
* Prevent usage in CPU only mode
* Add a section about offloaded KV cache to the docs
* Fix typos in docs
* Clarifications and better explanation of streams
* Fix conflicting key in init kwargs in PreTrainedTokenizerBase
* Update code to check for callable key in save_pretrained
* Apply PR suggestions
* Invoke CI
* Updates based on PR suggestion
* Remove user-defined tokens which can be obtained through merges
* Remove debug line
* formatting
* Refactor spm slow -> fast converter
* revert unnecessary refactor
* set comprehension
* remove test files
* Use `vocab_scores`
* Always replace spiece underline with space in decode
* we no longer need token filtering
* Add save fast load slow unit test
* Remove tokenizers version check
* Remove duplicate code
* Make `<start_of_turn>` and `<end_of_turn>` special tokens
* Bias merge priority with length if score is the same
* Add unit test for merge priority
* CI
* mvp
* added test (a few models need fixes)
* fix a few test cases
* test nits
* harder test 😈
* revert changes in stablelm
* test with improved condition
* add todo
* tmp commit
* merged with main
* nits
* add todo
* final corrections
* add docs for generation compilation
* docs nits
* add tip
* PR suggestions
* add more details to the compilation docs
* fix cache positions
* cache is now init in generate; update docs
* tag test as flaky
* docs
* post rebase make fixup and other nits
* remove unintended changes
* whisper (encoder-decoder) not supported
* move token default updates to ; add tests for token defaults
* push changes
* manual rebase
* chameleon doesn't support this
* fix test_static_cache_mha_mqa_gqa (broken in another PR)
* docs: dynamic is better with end-to-end compilation
* No more default chat templates
* Add the template to the GPT-SW3 tests since it's not available by default now
* Fix GPT2 test
* Fix Bloom test
* Fix Bloom test
* Remove default templates again
* Updated ruff version and fixed the required code accorindg to the latest version.
* Updated ruff version and fixed the required code accorindg to the latest version.
* Added noqa directive to ignore 1 error shown by ruff
* add DataCollatorBatchFlattening
* Update data_collator.py
* change name
* new FA2 flow if position_ids is provided
* add comments
* minor fix
* minor fix data collator
* add test cases for models
* add test case for data collator
* remove extra code
* formating for ruff check and check_repo.py
* ruff format
ruff format tests src utils
* custom_init_isort.py
* gguf conversion forces add_prefix_space=False for llama3, this is not required and forces from_slow, which fails. changing to None + test
* typo
* clean test
* Change resize_token_embeddings to make it return same Class that is passed to it
* Add explanatory comment as requested in review
* Add explanatory comments for add resizing function in lxmert
* Add comment for padding_idx and moving _resize_bias in lxmert to LxmertForPreTraining
---------
Co-authored-by: Prashanth Sateesh <prasatee@Prashanths-MBP.attlocal.net>
Co-authored-by: Prashanth Sateesh <prasatee@Prashanths-MacBook-Pro.local>
* Add YaRN and Dynamic-YaRN RoPE Scaling Methods
YaRN (Yet another RoPE extension method) combines the NTK-By-Parts
Interpolation and Attention Scaling methods, improving upon existing
RoPE interpolation methods for longer context window sizes.
Fine-tuned models maintain their original performance across benchmarks
while enabling efficient extrapolation and transfer learning for
quicker convergence, especially in compute-limited environments.
We implement YaRN and Dynamic-YaRN for the following list of models:
- LLaMA
- Falcon
- GPT-NeoX
- Olmo
- Persimmon
- Phi
- StableLM
- OpenLLaMA
New unit tests are added to assert YaRN's correct behavior on both
short and long sequence inputs.
For more details, please refer to https://arxiv.org/abs/2309.00071.
Co-authored-by: Miguel Almeida <miguel.pessanha.almeida@tecnico.ulisboa.pt>
* Refactor YaRN implementation for LLaMA
Iterate on YaRN implementation for LLaMA and remove diff from remaining
models for increased PR modularity.
This commit includes the following changes:
- Merge 'yarn_rope_scaling' and 'rope_scaling' dictionaries
- Remove unnecessary attributes ('extrapolation_factor' and 'finetuned')
from YaRN classes
- Inherit 'forward' method in YaRN classes from superclass
- Rename 'yarn' method to 'compute_yarn_scaling'
- Extend YaRN tests with further assertions
- Fix style inconsistencies
Co-authored-by: Miguel Monte e Freitas <miguelmontefreitas@tecnico.ulisboa.pt>
* Refactor Tensor Building Logic for YaRN
- Comply with the the tensor building logic introduced in #30743
- Add referencing to the optimized Attention Factor equation
- Remove Dynamic YaRN for a more agile deployment
Co-authored-by: mig-mfreitas <mig-mfreitas@users.noreply.github.com>
* remove unwanted file
---------
Co-authored-by: Miguel Almeida <miguel.pessanha.almeida@tecnico.ulisboa.pt>
Co-authored-by: mig-mfreitas <mig-mfreitas@users.noreply.github.com>
Co-authored-by: Joao Gante <joao@huggingface.co>
* fix mask creation of gpt2 and gpt_neox caused by me
* forgot the reshape of masks when shape > 2
* add tests for gpt neox and gpt2
* nit on a comment
