Error when using DDP

[smartbarbarian]

I used ZeroRedundancyOptimizer to wrap Ranger21 and running it in a 4-GPU machine. But its performace is much worse compared with the simple AdamW and it showed the error.
loerror in warmdown pct calc. new pct = 67.11272727272727���������������������������������������������������������������� | 19156/19505 [8:57:59<04:55, 1.18it/auto handled but please report issue
error in warmdown - lr below min lr. current lr = 2.999999999999997e-05
auto handling but please report issue!
error in warmdown pct calc. new pct = 67.11272727272727
auto handled but please report issue
error in warmdown - lr below min lr. current lr = 2.999999999999997e-05
auto handling but please report issue!
error in warmdown pct calc. new pct = 67.11272727272727
auto handled but please report issue
error in warmdown - lr below min lr. current lr = 2.999999999999997e-05
auto handling but please report issue!
error in warmdown pct calc. new pct = 67.11272727272727
auto handled but please report issue
error in warmdown - lr below min lr. current lr = 2.999999999999997e-05
auto handling but please report issue!

[TolerantChief]

Did you find a solution to this? I have the same problem.