MCPcopy Create free account
hub / github.com/SLDGroup/EMCAD / Block

Class Block

lib/pvtv2.py:112–149  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

110
111
112class Block(nn.Module):
113
114 def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0.,
115 drop_path=0., act_layer=nn.GELU, norm_layer=nn.LayerNorm, sr_ratio=1):
116 super().__init__()
117 self.norm1 = norm_layer(dim)
118 self.attn = Attention(
119 dim,
120 num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale,
121 attn_drop=attn_drop, proj_drop=drop, sr_ratio=sr_ratio)
122 # NOTE: drop path for stochastic depth, we shall see if this is better than dropout here
123 self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
124 self.norm2 = norm_layer(dim)
125 mlp_hidden_dim = int(dim * mlp_ratio)
126 self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)
127
128 self.apply(self._init_weights)
129
130 def _init_weights(self, m):
131 if isinstance(m, nn.Linear):
132 trunc_normal_(m.weight, std=.02)
133 if isinstance(m, nn.Linear) and m.bias is not None:
134 nn.init.constant_(m.bias, 0)
135 elif isinstance(m, nn.LayerNorm):
136 nn.init.constant_(m.bias, 0)
137 nn.init.constant_(m.weight, 1.0)
138 elif isinstance(m, nn.Conv2d):
139 fan_out = m.kernel_size[0] * m.kernel_size[1] * m.out_channels
140 fan_out //= m.groups
141 m.weight.data.normal_(0, math.sqrt(2.0 / fan_out))
142 if m.bias is not None:
143 m.bias.data.zero_()
144
145 def forward(self, x, H, W):
146 x = x + self.drop_path(self.attn(self.norm1(x), H, W))
147 x = x + self.drop_path(self.mlp(self.norm2(x), H, W))
148
149 return x
150
151
152class OverlapPatchEmbed(nn.Module):

Callers 1

__init__Method · 0.85

Calls

no outgoing calls

Tested by

no test coverage detected