Skip to content

Add conversion options #3051

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

Merged
merged 1 commit into from
Sep 24, 2025
Merged

Add conversion options #3051

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
13 changes: 10 additions & 3 deletions torchao/prototype/tensor_conversion/api.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -124,9 +124,16 @@ def _find_tied_params(model):


def _convert_model_for_aarch64(
model, *, tensor_type="auto", intx_packing_format="opaque_torchao_auto"
model,
*,
tensor_type="auto",
intx_packing_format="opaque_torchao_auto",
convert_tied_embedding=True,
convert_linear=True,
):
module_name_to_tied_param = _find_tied_params(model)
module_name_to_tied_param = (
_find_tied_params(model) if convert_tied_embedding else {}
)

# Iterate through modules in model and convert IntxUnpackedToInt8Tensor tensors to Int8LutTensor
for name, module in model.named_modules():
Expand All @@ -138,7 +145,7 @@ def _convert_model_for_aarch64(
print("Skipping converting nn.Embedding {name} because it is not tied")
continue

if not isinstance(module, nn.Linear):
if not (convert_linear and isinstance(module, nn.Linear)):
continue

weight = module.weight
Expand Down
Loading