Fix meta tensor device mismatch by also patching torch.arange

Patch both residual_fsq and finite_scalar_quantization tensor refs

Fix meta tensor: patch residual_fsq module's tensor reference, not torch.tensor

Fix meta tensor: monkey-patch ResidualFSQ to force CPU tensor creation

Fix meta tensor crash: force CPU device context during model init

Try _fast_init=False to prevent meta device initialization

Fix dtype parameter: use dtype instead of torch_dtype, remove device_map

Fix meta tensor: add device_map=None and use torch_dtype parameter

Fix meta tensor error: disable low_cpu_mem_usage for model loading

Deploy ACE-Step Custom Edition with bug fixes