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

support half-precision models #206

Merged
merged 13 commits into from
Mar 1, 2023
Merged

support half-precision models #206

merged 13 commits into from
Mar 1, 2023

Commits on Mar 1, 2023

  1. test half precision on all diffusion models

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    565873b View commit details
    Browse the repository at this point in the history

  2. fix(api): reuse ORT session params for each device

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    b532570 View commit details
    Browse the repository at this point in the history

  3. only convert inner nodes with ORT conversion helpers

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    2210ee8 View commit details
    Browse the repository at this point in the history

  4. reload model from proto file before converting

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    7e65e21 View commit details
    Browse the repository at this point in the history

  5. extract tensors after conversion

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    9ef89db View commit details
    Browse the repository at this point in the history

  6. better error logging during conversion

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    86984be View commit details
    Browse the repository at this point in the history

  7. fix(api): run shape inference before converting models to fp16 

    per discussion in microsoft/onnxruntime#14827
    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    dbf9eaf View commit details
    Browse the repository at this point in the history

  8. fix(api): write external weights into same directory as optimized model

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    74aae1b View commit details
    Browse the repository at this point in the history

  9. use filename for tensors

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    2f4ab20 View commit details
    Browse the repository at this point in the history

  10. apply lint, remove unused

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    3f9f94f View commit details
    Browse the repository at this point in the history

  11. restore launch script

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    bf6cf4c View commit details
    Browse the repository at this point in the history

  12. fix(scripts): pass protocol and port to all test scripts

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    b44e644 View commit details
    Browse the repository at this point in the history

  13. fix(api): mark all convert methods as no_grad

    @ssube
    ssube committed Mar 1, 2023
    Configuration menu
    Copy the full SHA
    21fc7c5 View commit details
    Browse the repository at this point in the history