Skip to content

[PyTorch] Share CUDA graph memory across dynamic CP variants #3353

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鈥檒l occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
xiaoyao0115 wants to merge 1 commit into
base: main
Choose a base branch
from
Draft

[PyTorch] Share CUDA graph memory across dynamic CP variants #3353

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
6 changes: 6 additions & 0 deletions docs/envvars.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -505,6 +505,12 @@ Debugging and Profiling
:Default: ``0``
:Description: **(JAX only)** Enable verbose printing of tensor numerics for debugging purposes.

.. envvar:: NVTE_CUDA_GRAPH_SLOT_DEBUG

:Type: ``int`` (0 or 1)
:Default: ``0``
:Description: **(PyTorch only)** When set to ``1`` before ``make_graphed_callables`` captures graphs with the private ``_graph_memory_slots`` plan, each replay checks on the host that no other live graph owns the saved-tensor arena or output family it is about to reuse, from forward until the matching backward.

Testing
^^^^^^^

Expand Down
Loading
Loading