Skip to content
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

Fix RTN supported layer checking condition #1705

Merged
merged 3 commits into from
Apr 2, 2024
Merged

Fix RTN supported layer checking condition #1705

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
0be34ab
fix comparison conditions
Kaihui-intel Apr 1, 2024
35f211f
set supported layers to tuple
Kaihui-intel Apr 1, 2024
d85d001
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Apr 1, 2024
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
4 changes: 2 additions & 2 deletions neural_compressor/adaptor/torch_utils/weight_only.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -398,7 +398,7 @@ def rtn_quantize(
model: fake quantized torch module
"""
assert isinstance(model, torch.nn.Module), "only support torch module"
supported_layers = ["Linear"]
supported_layers = (torch.nn.Linear,)
if return_int:
compression_dtype = kwargs.get("compression_dtype", torch.int32)
compression_dim = kwargs.get("compression_dim", 1)
Expand All @@ -407,7 +407,7 @@ def rtn_quantize(
use_optimum_format = kwargs.get("use_optimum_format", True)
with torch.no_grad():
for name, m in model.named_modules():
if m.__class__.__name__ not in supported_layers:
if not isinstance(m, supported_layers):
continue
orig_dtype = next(m.parameters()).dtype
if orig_dtype != torch.float:
Expand Down
Loading