Skip to content

Improvement for torch.compile support on Params4bit #1673

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
matthewdouglas merged 1 commit into from
Jun 8, 2025
Merged

Improvement for torch.compile support on Params4bit #1673

Show file tree
Hide file tree
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
7 changes: 0 additions & 7 deletions bitsandbytes/nn/modules.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -291,13 +291,6 @@ def from_prequantized(

return self

@classmethod
def __torch_function__(cls, func, types, args=(), kwargs=None):
if kwargs is None:
kwargs = {}
with torch._C.DisableTorchFunctionSubclass():
return func(*args, **kwargs)

def _quantize(self, device):
w = self.data.contiguous().to(device)
w_4bit, quant_state = bnb.functional.quantize_4bit(
Expand Down
5 changes: 1 addition & 4 deletions tests/test_linear4bit.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -270,10 +270,7 @@ def test_params4bit_real_serialization(device, quant_type, blocksize, compress_s
@pytest.mark.parametrize("mode", ["default", "reduce-overhead"], ids=id_formatter("mode"))
@pytest.mark.skipif(torch.__version__ < (2, 4), reason="Not supported in torch < 2.4")
def test_linear4bit_torch_compile(device, quant_type, compute_dtype, compress_statistics, bias, fullgraph, mode):
if device == "cpu" and quant_type == "fp4":
pytest.skip("FP4 is not supported for CPU")

if fullgraph and torch.__version__ < (2, 8):
if fullgraph and torch.__version__ < (2, 8, 0, "dev"):
pytest.skip("fullgraph mode requires torch 2.8 or higher")

if device == "cuda" and platform.system() == "Windows":
Expand Down