ShardedOptimStateDictConfig

class tensorplay.distributed.fsdp.ShardedOptimStateDictConfig(offload_to_cpu: bool = True, _use_dtensor: bool = False)[source]