Skip to content

Fix torchao tests #18953

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
metascroy merged 2 commits into from
Apr 16, 2026
Merged

Fix torchao tests #18953

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
caa85a8
up
metascroy Apr 16, 2026
8acb164
Merge branch 'main' into metascroy/fix-torchao-test
metascroy Apr 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
10 changes: 6 additions & 4 deletions examples/models/llama/export_llama_lib.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -794,10 +794,6 @@ def _prepare_for_llama_export(llm_config: LlmConfig) -> LLMEdgeManager:
)
)

# Now cast to the dtype override after quantization, so non-quantized
# components use the desired computation dtype.
edge_manager.model = edge_manager.model.to(dtype=dtype_override.to_torch_dtype())

return edge_manager


Expand Down Expand Up @@ -1857,6 +1853,12 @@ def _get_source_transforms( # noqa
)
)

# Cast to dtype_override after quantization transforms, so non-quantized
# components use the desired computation dtype. This must happen before
# _convert_model_for_aarch64 which converts IntxUnpackedToInt8Tensor to
# IntxOpaqueTensor (which doesn't support .to()).
transforms.append(lambda m: m.to(dtype=dtype_override.to_torch_dtype()))

if any([use_torchao_kernels_linear, use_torchao_kernels_tied_embedding]):
from torchao.prototype.tensor_conversion.api import _convert_model_for_aarch64

Expand Down
Loading