fix loss normalization for ddp training

haoneng.lhn

2023-12-07 4152cf4615e234b9892703a8e088cdd51937fc13

 funasr/layers/stft.py

@@ -4,8 +4,11 @@
from typing import Union

import torch
from torch_complex.tensor import ComplexTensor

try:
    from torch_complex.tensor import ComplexTensor
except:
    print("Please install torch_complex firstly")
from funasr.modules.nets_utils import make_pad_mask
from funasr.layers.complex_utils import is_complex
from funasr.layers.inversible_interface import InversibleInterface