github/ColossalAI
1
0
Fork 0
mirror of https://github.com/hpcaitech/ColossalAI.git synced 2025-09-12 12:47:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

[hotfix] adapt ProcessGroup and Optimizer to ColoTensor (#1388)

Browse Source
This commit is contained in:
HELSON
2022-07-29 19:33:24 +08:00
committed by GitHub
parent ad678921db
commit c7221cb2d4
7 changed files with 20 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
colossalai/nn/optimizer/fused_sgd.py
View File

@@ -104,7 +104,7 @@ class FusedSGD(Optimizer):
# momentum application can be skipped in the main kernel.
if 'momentum_buffer' not in param_state:
first_run = True
buf = param_state['momentum_buffer'] = torch.zeros_like(p.data)
buf = param_state['momentum_buffer'] = torch.zeros_like(p)
momentums.append(buf)
else:
first_run = False