torch.xpu.memory.max\_memory\_allocated ======================================= .. currentmodule:: torch.xpu.memory .. autofunction:: max_memory_allocated