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

Numa #1556

Merged
merged 20 commits into from
Jun 26, 2023
Merged

Numa #1556

merged 20 commits into from
Jun 26, 2023

Commits on May 21, 2023

  1. detect NUMA systems and pin work threads to nodes (linux)

    @zrm
    zrm committed May 21, 2023
    Configuration menu
    Copy the full SHA
    6fc5f17 View commit details
    Browse the repository at this point in the history

  2. disable mmap prefetch/readahead for NUMA systems

    @zrm
    zrm committed May 21, 2023
    Configuration menu
    Copy the full SHA
    0d23f8c View commit details
    Browse the repository at this point in the history

  3. avoid sending finalize op to thread pool if it does nothing

    @zrm
    zrm committed May 21, 2023
    Configuration menu
    Copy the full SHA
    9d058c2 View commit details
    Browse the repository at this point in the history

Commits on May 23, 2023

  1. silence robot

    @zrm
    zrm committed May 23, 2023
    Configuration menu
    Copy the full SHA
    2c1b5ae View commit details
    Browse the repository at this point in the history

  2. fix args

    @zrm
    zrm committed May 23, 2023
    Configuration menu
    Copy the full SHA
    8502d51 View commit details
    Browse the repository at this point in the history

Commits on Jun 17, 2023

  1. make --numa a param

    @zrm
    zrm committed Jun 17, 2023
    Configuration menu
    Copy the full SHA
    bf83dcb View commit details
    Browse the repository at this point in the history

  2. recommendation that n_nodes evenly divide n_threads did not warrant s… 

    …uch aggressive enforcement
    @zrm
    zrm committed Jun 17, 2023
    Configuration menu
    Copy the full SHA
    b71dfe6 View commit details
    Browse the repository at this point in the history

Commits on Jun 18, 2023

  1. lower synchronization overhead

    @zrm
    zrm committed Jun 18, 2023
    Configuration menu
    Copy the full SHA
    adaad10 View commit details
    Browse the repository at this point in the history

  2. statically allocate

    @zrm
    zrm committed Jun 18, 2023
    Configuration menu
    Copy the full SHA
    c31d51d View commit details
    Browse the repository at this point in the history

  3. move numa state to g_state

    @zrm
    zrm committed Jun 18, 2023
    Configuration menu
    Copy the full SHA
    2f5bb46 View commit details
    Browse the repository at this point in the history

  4. add description for --numa

    @zrm
    zrm committed Jun 18, 2023
    Configuration menu
    Copy the full SHA
    4b94582 View commit details
    Browse the repository at this point in the history

Commits on Jun 19, 2023

  1. Merge branch 'master' into HEAD

    @ggerganov
    ggerganov committed Jun 19, 2023
    Configuration menu
    Copy the full SHA
    90a0e65 View commit details
    Browse the repository at this point in the history

  2. ggml : minor style changes

    @ggerganov
    ggerganov committed Jun 19, 2023
    Configuration menu
    Copy the full SHA
    d0e3596 View commit details
    Browse the repository at this point in the history

  3. ggml : minor style + try fix sanitizer build

    @ggerganov
    ggerganov committed Jun 19, 2023
    Configuration menu
    Copy the full SHA
    67ba34e View commit details
    Browse the repository at this point in the history

Commits on Jun 26, 2023

  1. Merge branch 'master' into HEAD

    @ggerganov
    ggerganov committed Jun 26, 2023
    Configuration menu
    Copy the full SHA
    8f98035 View commit details
    Browse the repository at this point in the history

  2. llama : allow to initialize backend with NUMA support

    @ggerganov
    ggerganov committed Jun 26, 2023
    Configuration menu
    Copy the full SHA
    0fe4b00 View commit details
    Browse the repository at this point in the history

  3. llama : avoid ggml include in llama-util.h

    @ggerganov
    ggerganov committed Jun 26, 2023
    Configuration menu
    Copy the full SHA
    875a1e1 View commit details
    Browse the repository at this point in the history

  4. ggml : style / formatting

    @ggerganov
    ggerganov committed Jun 26, 2023
    Configuration menu
    Copy the full SHA
    4a555b4 View commit details
    Browse the repository at this point in the history

  5. ggml : fix handling of ops with n_threads > n_tasks > 1

    @ggerganov
    ggerganov committed Jun 26, 2023
    Configuration menu
    Copy the full SHA
    81a40e9 View commit details
    Browse the repository at this point in the history

  6. server : utilize numa parameter

    @ggerganov
    ggerganov committed Jun 26, 2023
    Configuration menu
    Copy the full SHA
    9aec2b7 View commit details
    Browse the repository at this point in the history