Skip to content

[PyTorch] Add MXFP8 dgrad and gradient 4over6 NVFP4 options #3573

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

Draft
vincentzed wants to merge 1 commit into
base: main
Choose a base branch
from
Draft

[PyTorch] Add MXFP8 dgrad and gradient 4over6 NVFP4 options #3573

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 13 additions & 1 deletion docs/envvars.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -347,7 +347,7 @@ Kernel Configuration

:Type: ``str`` (``none``, ``weights``, ``activations``, or ``all``)
:Default: ``all``
:Description: Select NVFP4 4over6 quantizers that use 256 instead of 448 as the global E4M3 scale bound. By default, all 4over6 quantizers use 256. Set the env var to ``none`` (or set ``NVFP4BlockScaling(nvfp4_4over6_e4m3_use_256="none")``) to use the standard NVFP4 448 bound for all 4over6 quantizers. This option is only meaningful for tensor roles that also enable :envvar:`NVTE_NVFP4_4OVER6`.
:Description: Select NVFP4 4over6 quantizers that use 256 instead of 448 as the global E4M3 scale bound. By default, all 4over6 quantizers use 256. Set the env var to ``none`` (or set ``NVFP4BlockScaling(nvfp4_4over6_e4m3_use_256="none")``) to use the standard NVFP4 448 bound for all 4over6 quantizers. This option is only meaningful for tensor roles that also enable :envvar:`NVTE_NVFP4_4OVER6` or :envvar:`NVTE_NVFP4_4OVER6_GRAD`. Gradient 4over6 quantizers use 256 only when this option is ``all``.

.. envvar:: NVTE_NVFP4_4OVER6_ERR_MODE

Expand All @@ -361,6 +361,18 @@ Kernel Configuration
:Default: ``0``
:Description: Use the faster NVFP4 4over6 candidate error path that compares candidates in the E4M3-scaled domain after the E2M1 x E4M3 product is rounded to FP16. Error differences and accumulation remain FP32. By default, 4over6 error comparison uses the original input-domain path; ``NVTE_USE_FAST_MATH`` does not control this error-comparison path.

.. envvar:: NVTE_NVFP4_4OVER6_GRAD

:Type: ``int`` (0 or 1)
:Default: ``0``
:Description: Apply 4over6 adaptive NVFP4 block scaling to gradient quantizers in the ``NVFP4BlockScaling`` recipe. When set to ``1`` (or when ``NVFP4BlockScaling(nvfp4_4over6_grad=True)`` is used), gradient blocks use the same map-to-4 versus map-to-6 candidate selection as :envvar:`NVTE_NVFP4_4OVER6`. Requires RHT and stochastic rounding to be disabled for gradient quantizers (``NVTE_NVFP4_DISABLE_RHT=1`` and ``NVTE_NVFP4_DISABLE_STOCHASTIC_ROUNDING=1``, or ``disable_rht=True`` and ``disable_stochastic_rounding=True``).

.. envvar:: NVTE_NVFP4_DGRAD_MXFP8

:Type: ``int`` (0 or 1)
:Default: ``0``
:Description: Run the dgrad GEMM of the ``NVFP4BlockScaling`` recipe with MXFP8 operands while the fprop and wgrad GEMMs keep NVFP4 operands. When set to ``1`` (or when ``NVFP4BlockScaling(dgrad_mxfp8=True)`` is used), the weight quantizer emits NVFP4 rowwise data for fprop and MXFP8 columnwise data for dgrad, and the ``grad_output`` quantizer emits MXFP8 rowwise data for dgrad and NVFP4 columnwise data for wgrad. Cannot be combined with ``NVTE_BACKWARD_OVERRIDE`` or with quantized primary weights.

Torch Compilation and Fusion
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Expand Down
Loading
Loading