Skip to content

Fix the retrieval of overwrite_main_grad #2764

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
shjwudp wants to merge 2 commits into
base: main
Choose a base branch
from
Draft

Fix the retrieval of overwrite_main_grad #2764

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a0d4199
Fix the retrieval of overwrite_main_grad: fetch it from the original …
shjwudp Mar 15, 2026
d8285e1
Correct weight name
shjwudp Mar 16, 2026
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 transformer_engine/pytorch/module/grouped_linear.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -476,7 +476,7 @@ def backward(ctx, grad_output: torch.Tensor) -> Tuple[Union[torch.Tensor, None],
use_split_accumulator=wgrad_gemm_use_split_accumulator,
accumulate=(
accumulate_wgrad_into_param_main_grad
if not getattr(weights[0], "overwrite_main_grad", False)
if not getattr(origin_weights[0], "overwrite_main_grad", False)
else False
),
)
Expand Down
2 changes: 1 addition & 1 deletion transformer_engine/pytorch/module/layernorm_linear.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -864,7 +864,7 @@ def backward(
"quantization_params": ctx.grad_weight_quantizer,
"accumulate": (
accumulate_wgrad_into_param_main_grad
if not getattr(weight, "overwrite_main_grad", False)
if not getattr(origin_weight, "overwrite_main_grad", False)
else False
),
"layout": "NT",
Expand Down
4 changes: 2 additions & 2 deletions transformer_engine/pytorch/module/layernorm_mlp.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1224,7 +1224,7 @@ def backward(
"quantization_params": ctx.fc2_grad_weight_quantizer, # wgrad in high precision
"accumulate": (
accumulate_wgrad_into_param_main_grad
if not getattr(fc1_weight, "overwrite_main_grad", False)
if not getattr(origin_fc2_weight, "overwrite_main_grad", False)
else False
),
"layout": "NT",
Expand Down Expand Up @@ -1471,7 +1471,7 @@ def fc2_wgrad_gemm(
"quantization_params": ctx.fc1_grad_weight_quantizer,
"accumulate": (
accumulate_wgrad_into_param_main_grad
if not getattr(fc2_weight, "overwrite_main_grad", False)
if not getattr(origin_fc1_weight, "overwrite_main_grad", False)
else False
),
"layout": "NT",
Expand Down
Loading