Skip to content
Open
Changes from all commits
Commits
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 marlin/__init__.py
Original file line number Diff line number Diff line change
Expand Up @@ -102,7 +102,7 @@ def forward(self, A):
def pack(self, linear, scales):
"""Pack a fake-quantized linear layer into this actual Marlin representation.
@linear: fake-quantized `torch.nn.Linear` layer to convert (must be of type `torch.half`)
@scales: corresponding quantization scales of shape `(infeatures, groups)`
@scales: corresponding quantization scales of shape `(outfeatures, groups)`
"""
if linear.weight.dtype != torch.half:
raise ValueError('Only `torch.half` weights are supported.')
Expand Down