Skip to content

[QwixTileSize] Use more descriptive variable name in dot_general_qt back propagation. #229

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

Open
copybara-service wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[QwixTileSize] Use more descriptive variable name in dot_general_qt back propagation. #229

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
2 changes: 1 addition & 1 deletion qwix/_src/core/dot_general_qt.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -215,7 +215,7 @@ def _compute_gradient_for_operand(g: jax.Array, *, for_dlhs: bool):

if g_qtype and numerics.should_quantize(g.dtype):
if isinstance(y, qarray.QArray) and not any(
v > 1 for v in qarray.get_tiled_axes(y).values()
tile_size > 1 for tile_size in qarray.get_tiled_axes(y).values()
):
# Apply the scale of y to g, this trick avoids requantizing y because
# the y from fwd pass has different channelwise_axes.
Expand Down