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

Update to use torch.nn.attention.sdpa_kernel #131

Open
wants to merge 14 commits into
base: main
Choose a base branch
from
Open

Update to use torch.nn.attention.sdpa_kernel #131

wants to merge 14 commits into from

Commits on Feb 29, 2024

  1. Merge pull request pytorch-labs#118 from yanboliang/cleanup 

    Clean up mixtral-moe
    @yanboliang
    yanboliang committed Feb 29, 2024
    Configuration menu
    Copy the full SHA
    f08f0dd View commit details
    Browse the repository at this point in the history

  2. Adding Mistral-7B support (pytorch-labs#116)

    @Artyom17
    Artyom17 committed Feb 29, 2024
    Configuration menu
    Copy the full SHA
    f121b47 View commit details
    Browse the repository at this point in the history

  3. Minor fix for generate.py (pytorch-labs#117)

    @Artyom17
    Artyom17 committed Feb 29, 2024
    Configuration menu
    Copy the full SHA
    1c23b94 View commit details
    Browse the repository at this point in the history

Commits on Mar 4, 2024

  1. add weight only quantization support for cpu device

    @mingfeima
    mingfeima committed Mar 4, 2024
    Configuration menu
    Copy the full SHA
    3ad26cc View commit details
    Browse the repository at this point in the history

  2. update error log

    @mingfeima
    mingfeima committed Mar 4, 2024
    Configuration menu
    Copy the full SHA
    fba5d25 View commit details
    Browse the repository at this point in the history

Commits on Mar 7, 2024

  1. Merge pull request pytorch-labs#123 from mingfeima/pr_weight_only_qua… 

    …ntization_cpu

Add weight only quantization support for cpu device
    @mikekgfb
    mikekgfb committed Mar 7, 2024
    Configuration menu
    Copy the full SHA
    f68e81e View commit details
    Browse the repository at this point in the history

Commits on Mar 9, 2024

  1. transposed w2 to have reduction dim be innermost dim

    @Chillee
    Chillee committed Mar 9, 2024
    Configuration menu
    Copy the full SHA
    635db73 View commit details
    Browse the repository at this point in the history

  2. fix converting checkpoint and tp

    @yanboliang
    yanboliang committed Mar 9, 2024
    Configuration menu
    Copy the full SHA
    776b733 View commit details
    Browse the repository at this point in the history

  3. Update perf number

    @yanboliang
    yanboliang committed Mar 9, 2024
    Configuration menu
    Copy the full SHA
    ca10839 View commit details
    Browse the repository at this point in the history

Commits on Mar 10, 2024

  1. Update perf number

    @yanboliang
    yanboliang committed Mar 10, 2024
    Configuration menu
    Copy the full SHA
    4f98fe0 View commit details
    Browse the repository at this point in the history

  2. Update

    @yanboliang
    yanboliang committed Mar 10, 2024
    Configuration menu
    Copy the full SHA
    7e50fcc View commit details
    Browse the repository at this point in the history

  3. Merge pull request pytorch-labs#128 from yanboliang/mixtral_improvements 

    Mixtral MoE improvements: transposed w2 to have reduction dim be innermost dim
    @yanboliang
    yanboliang committed Mar 10, 2024
    Configuration menu
    Copy the full SHA
    873723b View commit details
    Browse the repository at this point in the history

Commits on Mar 11, 2024

  1. Merge branch 'main' of https://github.com/pytorch-labs/gpt-fast

    @yanboliang
    yanboliang committed Mar 11, 2024
    Configuration menu
    Copy the full SHA
    52625f8 View commit details
    Browse the repository at this point in the history

  2. Update to torch.nn.attention.sdpa_kernel

    @yanboliang
    yanboliang committed Mar 11, 2024
    Configuration menu
    Copy the full SHA
    eac291b View commit details
    Browse the repository at this point in the history