def forward()

in deepseek_vl2/models/modeling_deepseek_vl_v2.py [0:0]

• 27 lines of code
• 5 McCabe index (conditional complexity)

    def forward(self, x):
        if self.cfg.token_pooling:
            batch_size, wxh, channels = x.shape
            w = h = int(wxh ** 0.5)
            x = x.view(batch_size, w, h, channels)
            x = x.permute(0, 3, 1, 2)
            # import ipdb; ipdb.set_trace()
            patches = x.unfold(2, 2, 2).unfold(3, 2, 2)
            batch_size, channels, h_patches, w_patches, _, _ = patches.size()
            # 在通道维度上拼接
            patches = patches.contiguous().view(batch_size, channels, h_patches * w_patches, -1)

            # 通过线性层
            patches = patches.permute(0, 2, 1, 3).contiguous()
            patches = patches.view(batch_size, h_patches * w_patches, channels * 4)

            x = self.token_pooling_layer(patches)

        elif self.cfg.projector_type == 'downsample_mlp_gelu':
            bs, hw, input_dim = x.shape
            h = w = int((hw) ** 0.5)

            """compute padding"""
            if h % self.cfg.downsample_ratio:
                pad = self.cfg.downsample_ratio - h % self.cfg.downsample_ratio
            else:
                pad = 0
            x = x.reshape(bs, h, w, input_dim)
            if pad > 0:
                x = F.pad(x, (0, 0, 0, pad, 0, pad), "constant", 0)

            """4 to 1 concat"""
            x = x.permute(0, 3, 1, 2)  # B, C, H, W
            x = F.unfold(x, kernel_size=self.cfg.downsample_ratio, stride=self.cfg.downsample_ratio,
                         padding=0)  # B, C*4, HW // 4
            x = x.permute(0, 2, 1)

        return self.layers(x)