Spaces:

BFZD
/

Diving-into-the-Fusion-of-Monocular-Priors-for-Generalized-Stereo-Matching-Demo

Running

BFZD233 commited on Jul 23

Commit

87c7cb4

1 Parent(s): f2e0b8a

update

Files changed (1) hide show

Depth-Anything-V2/depth_anything_v2/dinov2_layers/attention.py CHANGED Viewed

@@ -74,7 +74,7 @@ class MemEffAttention(Attention):
         q, k, v = unbind(qkv, 2)
         # x = memory_efficient_attention(q, k, v, attn_bias=attn_bias)
-        x = scaled_dot_product_attention(q, k, v, attn_bias=attn_bias)
         x = x.reshape([B, N, C])
         x = self.proj(x)

         q, k, v = unbind(qkv, 2)
         # x = memory_efficient_attention(q, k, v, attn_bias=attn_bias)
+        x = scaled_dot_product_attention(q, k, v)
         x = x.reshape([B, N, C])
         x = self.proj(x)