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

test-backend-ops : add performance eval mode + improve CUDA repeat and binary broadcast ops performance #636

Merged
merged 18 commits into from
Dec 7, 2023
Merged

test-backend-ops : add performance eval mode + improve CUDA repeat and binary broadcast ops performance #636

merged 18 commits into from
Dec 7, 2023

Commits on Dec 5, 2023

  1. ggml-cuda : implement repeat with bin_bcast

    @slaren
    slaren committed Dec 5, 2023
    Configuration menu
    Copy the full SHA
    4dd5370 View commit details
    Browse the repository at this point in the history

  2. ggml-cuda : change supports_op for mul_mat to match compute_forward

    @slaren
    slaren committed Dec 5, 2023
    Configuration menu
    Copy the full SHA
    0866d84 View commit details
    Browse the repository at this point in the history

  3. test-backend-ops : add performance eval mode

    @slaren
    slaren committed Dec 5, 2023
    Configuration menu
    Copy the full SHA
    2699dd7 View commit details
    Browse the repository at this point in the history

  4. improve formatting

    @slaren
    slaren committed Dec 5, 2023
    Configuration menu
    Copy the full SHA
    1f6c60d View commit details
    Browse the repository at this point in the history

Commits on Dec 6, 2023

  1. add sd test cases

    @slaren
    slaren committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    2ffe1a2 View commit details
    Browse the repository at this point in the history

  2. fix test case

    @slaren
    slaren committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    5c74195 View commit details
    Browse the repository at this point in the history

  3. ggml-cuda : bin_bcast: better block sizes, two elements per thread

    @slaren
    slaren committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    a5d7a1d View commit details
    Browse the repository at this point in the history

  4. metal : add dim3 broadcast support for mul mat

    @ggerganov
    ggerganov committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    69d844c View commit details
    Browse the repository at this point in the history

  5. cleanup

    @slaren
    slaren committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    8eb145a View commit details
    Browse the repository at this point in the history

  6. typo

    @slaren
    slaren committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    171a091 View commit details
    Browse the repository at this point in the history

  7. metal : enable mul mat-vec for dim2 > 1

    @ggerganov
    ggerganov committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    cca2854 View commit details
    Browse the repository at this point in the history

  8. metal : mul mat-vec support dim3 broadcasts 

    ggml-ci
    @ggerganov
    ggerganov committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    3ebcec1 View commit details
    Browse the repository at this point in the history

  9. ggml-cuda : fix bin_bcast for ne0=1 

    ggml-ci
    @slaren
    slaren committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    45fd1a2 View commit details
    Browse the repository at this point in the history

  10. ggml-cuda : limit block size z dim to 64

    @slaren
    slaren committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    d5f2fc9 View commit details
    Browse the repository at this point in the history

  11. test-backend-ops : add test cases

    @slaren
    slaren committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    da6ca50 View commit details
    Browse the repository at this point in the history

Commits on Dec 7, 2023

  1. test-backend-ops : add warmup run, print test type before trying to c… 

    …ompute
    @slaren
    slaren committed Dec 7, 2023
    Configuration menu
    Copy the full SHA
    ebdc505 View commit details
    Browse the repository at this point in the history

  2. ggml-cuda : bin_bcast: collapse dimensions when possible, add fallbac… 

    …k kernel for large tensors

ggml-ci
    @slaren
    slaren committed Dec 7, 2023
    Configuration menu
    Copy the full SHA
    fcdaaa2 View commit details
    Browse the repository at this point in the history

  3. test-backend-ops : avoid division by zero

    @slaren
    slaren committed Dec 7, 2023
    Configuration menu
    Copy the full SHA
    8ad1e85 View commit details
    Browse the repository at this point in the history