Skip to content

Commit

Permalink
fix lion8b
Browse files Browse the repository at this point in the history
  • Loading branch information
mvpatel2000 committed Dec 29, 2023
1 parent 5c4f3e7 commit c04dc39
Showing 1 changed file with 1 addition and 1 deletion.
2 changes: 1 addition & 1 deletion llmfoundry/optim/lion8b.py
Original file line number Diff line number Diff line change
Expand Up @@ -277,7 +277,7 @@ def set_data(self, data: torch.Tensor) -> None:
f'on device {data.device} with shape {data.shape}.')
self.data = None
assert self._f_encode is not None # pyright
self.quantized, self.scales = self._f_encode(data)
self.quantized, self.scales, _ = self._f_encode(data)
else:
self.data = data.to(dtype=torch.float32)
self.quantized = None
Expand Down

0 comments on commit c04dc39

Please sign in to comment.