Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Refactor flash attention implementation in transformers #31446

Merged
merged 62 commits into from
Jul 11, 2024
Merged

Refactor flash attention implementation in transformers #31446

merged 62 commits into from
Jul 11, 2024

Commits on May 31, 2024

  1. dumb commit

    @ArthurZucker
    ArthurZucker committed May 31, 2024
    Configuration menu
    Copy the full SHA
    b66fdb0 View commit details
    Browse the repository at this point in the history

Commits on Jun 17, 2024

  1. nit

    @ArthurZucker
    ArthurZucker committed Jun 17, 2024
    Configuration menu
    Copy the full SHA
    029ee11 View commit details
    Browse the repository at this point in the history

  2. Merge branch 'main' into backend-compatible

    @ArthurZucker
    ArthurZucker committed Jun 17, 2024
    Configuration menu
    Copy the full SHA
    9a7885d View commit details
    Browse the repository at this point in the history

  3. update

    @ArthurZucker
    ArthurZucker committed Jun 17, 2024
    Configuration menu
    Copy the full SHA
    ac3e5b5 View commit details
    Browse the repository at this point in the history

  4. something like this

    @ArthurZucker
    ArthurZucker committed Jun 17, 2024
    Configuration menu
    Copy the full SHA
    a7c48bd View commit details
    Browse the repository at this point in the history

  5. unpack in modeling utils

    @ArthurZucker
    ArthurZucker committed Jun 17, 2024
    Configuration menu
    Copy the full SHA
    682f221 View commit details
    Browse the repository at this point in the history

  6. safe import

    @ArthurZucker
    ArthurZucker committed Jun 17, 2024
    Configuration menu
    Copy the full SHA
    2201178 View commit details
    Browse the repository at this point in the history

Commits on Jun 21, 2024

  1. oups

    @ArthurZucker
    ArthurZucker committed Jun 21, 2024
    Configuration menu
    Copy the full SHA
    55a3503 View commit details
    Browse the repository at this point in the history

Commits on Jun 26, 2024

  1. update

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    b5cbaef View commit details
    Browse the repository at this point in the history

  2. nits

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    7c6fdd7 View commit details
    Browse the repository at this point in the history

  3. diff convert gemma

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    08d7e1e View commit details
    Browse the repository at this point in the history

  4. update

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    27044da View commit details
    Browse the repository at this point in the history

  5. start propagating

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    ca316a0 View commit details
    Browse the repository at this point in the history

  6. udpate other modeling code as well

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    ea93267 View commit details
    Browse the repository at this point in the history

  7. update for sliding window models

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    4b67223 View commit details
    Browse the repository at this point in the history

  8. nits

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    d59ac0c View commit details
    Browse the repository at this point in the history

  9. more init cleanups

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    a1d3866 View commit details
    Browse the repository at this point in the history

  10. styling

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    aea7f03 View commit details
    Browse the repository at this point in the history

  11. fixup

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    f1bedd0 View commit details
    Browse the repository at this point in the history

  12. noice

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    86e2edc View commit details
    Browse the repository at this point in the history

  13. pass fixup

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    e90a944 View commit details
    Browse the repository at this point in the history

  14. typo typing_extension -> typing_extensions

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    093fbf5 View commit details
    Browse the repository at this point in the history

  15. torch.nn.functionnal -> torch.nn.functional

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    a1c56d2 View commit details
    Browse the repository at this point in the history

  16. add to import structure

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    1aad4a2 View commit details
    Browse the repository at this point in the history

  17. unpack

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    10bc1fa View commit details
    Browse the repository at this point in the history

  18. simplify a bit more for this first version

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    9f08ddb View commit details
    Browse the repository at this point in the history

  19. nut

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    2e65e57 View commit details
    Browse the repository at this point in the history

  20. update

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    f8622e6 View commit details
    Browse the repository at this point in the history

  21. update

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    2bb4347 View commit details
    Browse the repository at this point in the history

  22. nit

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    9be7579 View commit details
    Browse the repository at this point in the history

  23. ease the import of Unpack

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    889cbf8 View commit details
    Browse the repository at this point in the history

  24. remove useless use_sliding_window

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    070af2d View commit details
    Browse the repository at this point in the history

  25. no qua please

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    80057a0 View commit details
    Browse the repository at this point in the history

  26. protect import?

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    c0b024d View commit details
    Browse the repository at this point in the history

  27. style

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    8f7d1c1 View commit details
    Browse the repository at this point in the history

  28. [run-slow]

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    46b77f9 View commit details
    Browse the repository at this point in the history

  29. [run slow] llama,gemma,mistral,mixtral

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    4a98ee7 View commit details
    Browse the repository at this point in the history

  30. remove extra kwargs

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    25b2c10 View commit details
    Browse the repository at this point in the history

  31. Merge branch 'main' of github.com:huggingface/transformers into backe… 

    …nd-compatible
    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    8c3780d View commit details
    Browse the repository at this point in the history

  32. fix llama

    @ArthurZucker
    ArthurZucker committed Jun 26, 2024
    Configuration menu
    Copy the full SHA
    1d38dab View commit details
    Browse the repository at this point in the history

Commits on Jul 1, 2024

  1. address review comments

    @fxmarty
    fxmarty committed Jul 1, 2024
    Configuration menu
    Copy the full SHA
    f64864a View commit details
    Browse the repository at this point in the history

  2. apply diff_model_converter to modeling_gemma.py

    @fxmarty
    fxmarty committed Jul 1, 2024
    Configuration menu
    Copy the full SHA
    565c5dc View commit details
    Browse the repository at this point in the history

  3. Merge branch 'main' into backend-compatible

    @fxmarty
    fxmarty committed Jul 1, 2024
    Configuration menu
    Copy the full SHA
    2403ce5 View commit details
    Browse the repository at this point in the history

Commits on Jul 2, 2024

  1. remove cache_position 1

    @fxmarty
    fxmarty committed Jul 2, 2024
    Configuration menu
    Copy the full SHA
    c89571d View commit details
    Browse the repository at this point in the history

  2. remove cache_position 2

    @fxmarty
    fxmarty committed Jul 2, 2024
    Configuration menu
    Copy the full SHA
    32c2df8 View commit details
    Browse the repository at this point in the history

  3. some cleaning

    @fxmarty
    fxmarty committed Jul 2, 2024
    Configuration menu
    Copy the full SHA
    54a9fb0 View commit details
    Browse the repository at this point in the history

  4. refactor gemma2 as well

    @fxmarty
    fxmarty committed Jul 2, 2024
    Configuration menu
    Copy the full SHA
    206731e View commit details
    Browse the repository at this point in the history

  5. Merge branch 'main' into backend-compatible

    @fxmarty
    fxmarty committed Jul 2, 2024
    Configuration menu
    Copy the full SHA
    7c65fc7 View commit details
    Browse the repository at this point in the history

Commits on Jul 3, 2024

  1. apply review comments

    @fxmarty
    fxmarty committed Jul 3, 2024
    Configuration menu
    Copy the full SHA
    1be8c31 View commit details
    Browse the repository at this point in the history

  2. rename file to modeling_flash_attention_utils.py

    @fxmarty
    fxmarty committed Jul 3, 2024
    Configuration menu
    Copy the full SHA
    8d181ea View commit details
    Browse the repository at this point in the history

Commits on Jul 8, 2024

  1. Merge branch 'main' into backend-compatible

    @fxmarty
    fxmarty committed Jul 8, 2024
    Configuration menu
    Copy the full SHA
    3a30cb6 View commit details
    Browse the repository at this point in the history

  2. siglip refactor

    @fxmarty
    fxmarty committed Jul 8, 2024
    Configuration menu
    Copy the full SHA
    c92028a View commit details
    Browse the repository at this point in the history

  3. remove dead code

    @fxmarty
    fxmarty committed Jul 8, 2024
    Configuration menu
    Copy the full SHA
    7243993 View commit details
    Browse the repository at this point in the history

  4. is the hub down?

    @fxmarty
    fxmarty committed Jul 8, 2024
    Configuration menu
    Copy the full SHA
    8b077d8 View commit details
    Browse the repository at this point in the history

Commits on Jul 9, 2024

  1. still down?

    @fxmarty
    fxmarty committed Jul 9, 2024
    Configuration menu
    Copy the full SHA
    a9796bc View commit details
    Browse the repository at this point in the history

Commits on Jul 10, 2024

  1. fix siglip

    @fxmarty
    fxmarty committed Jul 10, 2024
    Configuration menu
    Copy the full SHA
    6752a9c View commit details
    Browse the repository at this point in the history

Commits on Jul 11, 2024

  1. Merge branch 'main' into backend-compatible

    @fxmarty
    fxmarty committed Jul 11, 2024
    Configuration menu
    Copy the full SHA
    3a9cf1b View commit details
    Browse the repository at this point in the history

  2. fix gemma2

    @fxmarty
    fxmarty committed Jul 11, 2024
    Configuration menu
    Copy the full SHA
    b4d1df5 View commit details
    Browse the repository at this point in the history

  3. fatal: Could not read from remote repository.

    @fxmarty
    fxmarty committed Jul 11, 2024
    Configuration menu
    Copy the full SHA
    1e1bc2f View commit details
    Browse the repository at this point in the history

  4. fix typo in softcap implem

    @fxmarty
    fxmarty committed Jul 11, 2024
    Configuration menu
    Copy the full SHA
    c79ca83 View commit details
    Browse the repository at this point in the history

  5. flacky

    @fxmarty
    fxmarty committed Jul 11, 2024
    Configuration menu
    Copy the full SHA
    30dc123 View commit details
    Browse the repository at this point in the history

  6. Failed: Timeout >120.0s

    @fxmarty
    fxmarty committed Jul 11, 2024
    Configuration menu
    Copy the full SHA
    fae6843 View commit details
    Browse the repository at this point in the history