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

Add GroupQueryAttention with KV-Cache #3425

Open
wants to merge 44 commits into
base: develop
Choose a base branch
from
Open

Add GroupQueryAttention with KV-Cache #3425

wants to merge 44 commits into from
+3,384 −3

Commits on May 24, 2024

  1. Add rmsnorms, gqa

    @turneram
    turneram committed May 24, 2024
    Configuration menu
    Copy the full SHA
    194d5e9 View commit details
    Browse the repository at this point in the history

Commits on Jul 9, 2024

  1. Checkpoint

    @turneram
    turneram committed Jul 9, 2024
    Configuration menu
    Copy the full SHA
    157c576 View commit details
    Browse the repository at this point in the history

Commits on Aug 1, 2024

  1. JIT impl running

    @turneram
    turneram committed Aug 1, 2024
    Configuration menu
    Copy the full SHA
    06321e5 View commit details
    Browse the repository at this point in the history

  2. Merge

    @turneram
    turneram committed Aug 1, 2024
    Configuration menu
    Copy the full SHA
    d3ac4c8 View commit details
    Browse the repository at this point in the history

Commits on Aug 5, 2024

  1. Split gpu op

    @turneram
    turneram committed Aug 5, 2024
    Configuration menu
    Copy the full SHA
    dfcc73f View commit details
    Browse the repository at this point in the history

Commits on Aug 16, 2024

  1. Checkpoint

    @turneram
    turneram committed Aug 16, 2024
    Configuration menu
    Copy the full SHA
    692a404 View commit details
    Browse the repository at this point in the history

Commits on Aug 21, 2024

  1. Improve rotary embedding perf; add scale factor to sln

    @turneram
    turneram committed Aug 21, 2024
    Configuration menu
    Copy the full SHA
    dcaba12 View commit details
    Browse the repository at this point in the history

Commits on Aug 28, 2024

  1. Convert to float for reduce mean

    @turneram
    turneram committed Aug 28, 2024
    Configuration menu
    Copy the full SHA
    0d20b7f View commit details
    Browse the repository at this point in the history

Commits on Aug 29, 2024

  1. Merge remote-tracking branch 'origin/develop' into gqa-jit

    @turneram
    turneram committed Aug 29, 2024
    Configuration menu
    Copy the full SHA
    75788e1 View commit details
    Browse the repository at this point in the history

Commits on Sep 5, 2024

  1. Merge

    @turneram
    turneram committed Sep 5, 2024
    Configuration menu
    Copy the full SHA
    9040735 View commit details
    Browse the repository at this point in the history

  2. Merge remote-tracking branch 'origin/develop' into gqa-jit

    @turneram
    turneram committed Sep 5, 2024
    Configuration menu
    Copy the full SHA
    875dd2e View commit details
    Browse the repository at this point in the history

  3. Clean up

    @turneram
    turneram committed Sep 5, 2024
    Configuration menu
    Copy the full SHA
    9e72f3b View commit details
    Browse the repository at this point in the history

Commits on Sep 6, 2024

  1. Clean up

    @turneram
    turneram committed Sep 6, 2024
    Configuration menu
    Copy the full SHA
    5ece981 View commit details
    Browse the repository at this point in the history

  2. Formatting

    @turneram
    turneram committed Sep 6, 2024
    Configuration menu
    Copy the full SHA
    332fde6 View commit details
    Browse the repository at this point in the history

  3. Undo changes to format.py

    @turneram
    turneram committed Sep 6, 2024
    Configuration menu
    Copy the full SHA
    c79b495 View commit details
    Browse the repository at this point in the history

  4. Sync unchanged files

    @turneram
    turneram committed Sep 6, 2024
    Configuration menu
    Copy the full SHA
    695c4b9 View commit details
    Browse the repository at this point in the history

  5. Sync unchange files

    @turneram
    turneram committed Sep 6, 2024
    Configuration menu
    Copy the full SHA
    79efe42 View commit details
    Browse the repository at this point in the history

  6. Merge remote-tracking branch 'origin/develop' into gqa-jit

    @turneram
    turneram committed Sep 6, 2024
    Configuration menu
    Copy the full SHA
    442d055 View commit details
    Browse the repository at this point in the history

Commits on Sep 11, 2024

  1. Refactoring

    @turneram
    turneram committed Sep 11, 2024
    Configuration menu
    Copy the full SHA
    4fd0902 View commit details
    Browse the repository at this point in the history

  2. Formatting

    @turneram
    turneram committed Sep 11, 2024
    Configuration menu
    Copy the full SHA
    1e31f03 View commit details
    Browse the repository at this point in the history

  3. Merge branch 'develop' into gqa-jit

    @turneram
    turneram authored Sep 11, 2024
    Configuration menu
    Copy the full SHA
    cd194b2 View commit details
    Browse the repository at this point in the history

Commits on Sep 12, 2024

  1. Fix CI issues

    @turneram
    turneram committed Sep 12, 2024
    Configuration menu
    Copy the full SHA
    829324e View commit details
    Browse the repository at this point in the history

  2. Formatting

    @turneram
    turneram committed Sep 12, 2024
    Configuration menu
    Copy the full SHA
    979f4e2 View commit details
    Browse the repository at this point in the history

  3. Merge branch 'gqa-jit' of https://github.com/ROCm/AMDMIGraphX into gq… 

    …a-jit
    @turneram
    turneram committed Sep 12, 2024
    Configuration menu
    Copy the full SHA
    d6b60a2 View commit details
    Browse the repository at this point in the history

  4. Merge remote-tracking branch 'origin/develop' into gqa-jit

    @turneram
    turneram committed Sep 12, 2024
    Configuration menu
    Copy the full SHA
    90a2375 View commit details
    Browse the repository at this point in the history

  5. Remove unused onnx files

    @turneram
    turneram committed Sep 12, 2024
    Configuration menu
    Copy the full SHA
    736c0b2 View commit details
    Browse the repository at this point in the history

Commits on Sep 13, 2024

  1. Fix clang tidy

    @turneram
    turneram committed Sep 13, 2024
    Configuration menu
    Copy the full SHA
    a9d9f9d View commit details
    Browse the repository at this point in the history

  2. Tidy fixes

    @turneram
    turneram committed Sep 13, 2024
    Configuration menu
    Copy the full SHA
    1a21ee1 View commit details
    Browse the repository at this point in the history

  3. Formatting

    @turneram
    turneram committed Sep 13, 2024
    Configuration menu
    Copy the full SHA
    e8933cd View commit details
    Browse the repository at this point in the history

  4. Update simplified_layernorm tests

    @turneram
    turneram committed Sep 13, 2024
    Configuration menu
    Copy the full SHA
    c420f26 View commit details
    Browse the repository at this point in the history

  5. Formatting

    @turneram
    turneram committed Sep 13, 2024
    Configuration menu
    Copy the full SHA
    fb6b6de View commit details
    Browse the repository at this point in the history

  6. Merge remote-tracking branch 'origin/develop' into gqa-jit

    @turneram
    turneram committed Sep 13, 2024
    Configuration menu
    Copy the full SHA
    4b54ac5 View commit details
    Browse the repository at this point in the history

Commits on Sep 18, 2024

  1. Rename to instructions_tuple and add test

    @turneram
    turneram committed Sep 18, 2024
    Configuration menu
    Copy the full SHA
    541a406 View commit details
    Browse the repository at this point in the history

  2. Formatting

    @turneram
    turneram committed Sep 18, 2024
    Configuration menu
    Copy the full SHA
    496e213 View commit details
    Browse the repository at this point in the history

  3. Add parser tests

    @turneram
    turneram committed Sep 18, 2024
    Configuration menu
    Copy the full SHA
    bc4a240 View commit details
    Browse the repository at this point in the history

  4. Formatting

    @turneram
    turneram committed Sep 18, 2024
    Configuration menu
    Copy the full SHA
    eaa0a87 View commit details
    Browse the repository at this point in the history

  5. Merge remote-tracking branch 'origin/develop' into gqa-jit

    @turneram
    turneram committed Sep 18, 2024
    Configuration menu
    Copy the full SHA
    291ba66 View commit details
    Browse the repository at this point in the history

  6. Only use packed qkv format

    @turneram
    turneram committed Sep 18, 2024
    Configuration menu
    Copy the full SHA
    41622b0 View commit details
    Browse the repository at this point in the history

  7. Formatting

    @turneram
    turneram committed Sep 18, 2024
    Configuration menu
    Copy the full SHA
    141a1bf View commit details
    Browse the repository at this point in the history

  8. Formatting

    @turneram
    turneram committed Sep 18, 2024
    Configuration menu
    Copy the full SHA
    41d2af7 View commit details
    Browse the repository at this point in the history

  9. Formatting

    @turneram
    turneram committed Sep 18, 2024
    Configuration menu
    Copy the full SHA
    7617ae5 View commit details
    Browse the repository at this point in the history

Commits on Sep 24, 2024

  1. Clang tidy and codecov

    @turneram
    turneram committed Sep 24, 2024
    Configuration menu
    Copy the full SHA
    3a99379 View commit details
    Browse the repository at this point in the history

  2. Formatting

    @turneram
    turneram committed Sep 24, 2024
    Configuration menu
    Copy the full SHA
    20c0f15 View commit details
    Browse the repository at this point in the history

  3. Merge remote-tracking branch 'origin/develop' into gqa-jit

    @turneram
    turneram committed Sep 24, 2024
    Configuration menu
    Copy the full SHA
    54ff0e9 View commit details
    Browse the repository at this point in the history