address cosine sim distributed #51

Author: lucidrains

setup.py

@@ -3,7 +3,7 @@
 setup(
   name = 'vector_quantize_pytorch',
   packages = find_packages(),
-  version = '1.6.5',
+  version = '1.6.6',
   license='MIT',
   description = 'Vector Quantization - Pytorch',
   long_description_content_type = 'text/markdown',

vector_quantize_pytorch/vector_quantize_pytorch.py

@@ -592,7 +592,7 @@ def forward(
             self.cluster_size.data.lerp_(bins, 1 - self.decay)
 
             embed_sum = einsum('h n d, h n c -> h c d', flatten, embed_onehot)
-            self.all_reduce_fn(embed_sum)
+            self.all_reduce_fn(embed_sum.contiguous())
             self.embed_avg.data.lerp_(embed_sum, 1 - self.decay)
 
             cluster_size = laplace_smoothing(self.cluster_size, self.codebook_size, self.eps) * self.cluster_size.sum(dim = -1, keepdim = True)