DeepSpeed
Fix ZeRO parameter initialization for tensors with `requires_grad=True`
#4138
Merged

Fix ZeRO parameter initialization for tensors with `requires_grad=True` #4138

XuehaiPan
XuehaiPan Fix ZeRO parameter initialization for tensors with `requires_grad=True`
05d97f9e
XuehaiPan XuehaiPan requested a review from jeffra jeffra 2 years ago
XuehaiPan XuehaiPan requested a review from tjruwase tjruwase 2 years ago
XuehaiPan XuehaiPan requested a review from samyam samyam 2 years ago
XuehaiPan XuehaiPan requested a review from mrwyattii mrwyattii 2 years ago
tjruwase tjruwase requested a review from tohtana tohtana 2 years ago
tjruwase tjruwase removed review request from jeffra jeffra 2 years ago
tjruwase tjruwase removed review request from samyam samyam 2 years ago
tjruwase tjruwase removed review request from mrwyattii mrwyattii 2 years ago
tohtana
XuehaiPan Simplify detach logic
85ce73bc
XuehaiPan Merge branch 'master' into detach-dtype-cast
f67410d5
XuehaiPan
tohtana
tohtana approved these changes on 2023-08-17
loadams Merge branch 'master' into detach-dtype-cast
430ae909
loadams loadams enabled auto-merge 2 years ago
XuehaiPan Merge branch 'master' into detach-dtype-cast
e27ca495
loadams loadams merged 426810a2 into master 2 years ago
XuehaiPan XuehaiPan deleted the detach-dtype-cast branch 2 years ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone