'''Copyright The Microsoft DeepSpeed Team'''

import torch.nn as nn


class DeepSpeedTransformerBase(nn.module):
    def __init__(self):
        pass

    # this would be the new clean base class that will replace DeepSpeedTransformerInference.
    # we currently don't know how this will look like but keeping it here as a placeholder.