Point Transformer v2的参数量怎么计算

代码地址：https://github.com/Pointcept/Pointcept/blob/main/pointcept/models/point_transformer_v2/point_transformer_v2m2_base.py

参数设置：
enc_depths=(2, 2, 6, 2)
enc_channels=(96, 192, 384, 512)
enc_groups=(12, 24, 48, 64)
enc_neighbours=(16, 16, 16, 16)

Block的定义：

class Block(nn.Module):
    def __init__(
        self,
        embed_channels,
        groups,
        qkv_bias=True,
        pe_multiplier=False,
        pe_bias=True,
        attn_drop_rate=0.0,
        drop_path_rate=0.0,
        enable_checkpoint=False,
    ):

设计主函数输出每一层encoder的参数量，不包含transition layer而且只有1个block：

if __name__ == '__main__':

    enc_channels=(96, 192, 384, 512)
    enc_groups=(12, 24, 48, 64)

    for i in range(4):

        model = Block(enc_channels[i], enc_groups[i])

        pytorch_total_params = sum(p.numel() for p in model.parameters() if p.requires_grad)

        print(str(i) + ": " + str(pytorch_total_params/1024.0/1024.0) + "M")

结果：

0: 0.054828643798828125M
1: 0.21540069580078125M
2: 0.8537750244140625M
3: 1.51434326171875M