7B bloom cuda oom run step3 with 80 T4(16G) when hybridEngine is on, but ok when hybridengine is off

[wang990099]

10 machines, 80 T4(16G)s, actor(7b bloom) with Zero3, critic(1.3b bloom) with Zero3, open both gradient_checkpointing

why HybridEngine using more gpu memory？

thank you in advance