Skip to content

Fix: Handle FusedLAMB Fails with LowLevelZeroPlugin When Using CPU Offload #6418

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

Open
Truong5724 wants to merge 3 commits into
base: main
Choose a base branch
from
+7 −0
Open

Fix: Handle FusedLAMB Fails with LowLevelZeroPlugin When Using CPU Offload #6418

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
5d687c2
Fix: Move Tensor from CPU to GPU when using FusedLAMB Optimizer
Truong5724 Apr 15, 2026
a8139d5
Merge branch 'hpcaitech:main' into Truong-Fix-FusedLAMB
Truong5724 Apr 15, 2026
eca76f6
[pre-commit.ci] auto fixes from pre-commit.com hooks
pre-commit-ci[bot] Apr 15, 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
7 changes: 7 additions & 0 deletions colossalai/utils/multi_tensor_apply/multi_tensor_apply.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,7 @@
class MultiTensorApply(object):
"""
Apply an operation to a list of tensors efficiently.
Move tensors to CUDA if they are on CPU.

Args:
chunk_size (int): Size of a chunk.
Expand Down Expand Up @@ -32,4 +33,10 @@ def check_avail(self):
def __call__(self, op, noop_flag_buffer, tensor_lists, *args):
self.check_avail()

# Move tensors to GPU if not already on GPU
for i, tensor_list in enumerate(tensor_lists):
for j, tensor in enumerate(tensor_list):
if tensor.device.type == "cpu":
tensor_lists[i][j] = tensor.to("cuda")

return op(self.chunk_size, noop_flag_buffer, tensor_lists, *args)