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

Added support of NANOO fp8. #3993

Merged
merged 4 commits into from
Jul 2, 2024
Merged

Added support of NANOO fp8. #3993

merged 4 commits into from
Jul 2, 2024

Commits on Jun 28, 2024

  1. Added support of NANOO fp8. 

    There are several different genres of fp8 formats used by different
HW vendors. Two popular genres include
- OCP fp8, which is used natively on NVIDIA H100
- NANOO fp8, which is used natively on AMD MI300 and Graphcore HW.

These two genres of fp8 formats work very similarly. This PR is to
enable support of NANOO fp8 as it is also now supported in JAX and XLA.

References:
- OCP fp8 paper: https://arxiv.org/abs/2209.05433
- NANOO fp8 paper: https://arxiv.org/abs/2206.02915
- JAX PR: jax-ml/jax#21376
- XLA PR: openxla/xla#9531
    @wenchenvincent
    wenchenvincent committed Jun 28, 2024
    Configuration menu
    Copy the full SHA
    28bdb79 View commit details
    Browse the repository at this point in the history

  2. Addressed the reviewer's comment. Avoided changing the existing 

    Fp8DotGeneralOp API.
    @wenchenvincent
    wenchenvincent committed Jun 28, 2024
    Configuration menu
    Copy the full SHA
    541bddf View commit details
    Browse the repository at this point in the history

  3. Fixed test_fp8_meta_dtype.

    @wenchenvincent
    wenchenvincent committed Jun 28, 2024
    Configuration menu
    Copy the full SHA
    bc083b8 View commit details
    Browse the repository at this point in the history

Commits on Jun 29, 2024

  1. Minor refactor.

    @wenchenvincent
    wenchenvincent committed Jun 29, 2024
    Configuration menu
    Copy the full SHA
    a6f52ae View commit details
    Browse the repository at this point in the history