at:: to torch::

README.md

@@ -94,14 +94,6 @@ locations = torch.LongTensor(locations)
 features = torch.FloatTensor(features)
 input.set_locations(locations, features, 0)
 
-# Optional: allow metadata preprocessing to be done in batch preparation threads
-# to improve GPU utilization.
-#
-# Parameter:
-#    3 if using MP3/2 pooling or C3/2 convolutions for downsizing,
-#    2 if using MP2 pooling for downsizing.
-input.precomputeMetadata(3)
-
 model.train()
 if use_gpu:
     input.cuda()

examples/hello-world.py

@@ -58,14 +58,6 @@ locations = torch.LongTensor(locations)
 features = torch.FloatTensor(features)
 input.set_locations(locations, features, 0)
 
-# Optional: allow metadata preprocessing to be done in batch preparation threads
-# to improve GPU utilization.
-#
-# Parameter:
-#    3 if using MP3/2 pooling or C3/2 convolutions for downsizing,
-#    2 if using MP2 pooling for downsizing.
-input.precomputeMetadata(3)
-
 model.train()
 if use_cuda:
     input.cuda()

sparseconvnet/SCN/Metadata/Metadata.cpp

@@ -145,15 +145,13 @@ void Metadata<dimension>::setInputSpatialLocations(
 }
 
 template <Int dimension>
-void Metadata<dimension>::getSpatialLocations(/*long*/ at::Tensor spatialSize,
-                                              /*long*/ at::Tensor locations) {
+at::Tensor
+Metadata<dimension>::getSpatialLocations(/*long*/ at::Tensor spatialSize) {
   Int nActive = getNActive(spatialSize);
   auto &SGs = getSparseGrid(spatialSize);
   Int batchSize = SGs.size();
 
-  locations.resize_({(int)nActive, dimension + 1});
-  locations.zero_();
-
+  auto locations = torch::zeros({(int)nActive, dimension + 1}, at::kLong);
   auto lD = locations.data<long>();
 
   for (Int i = 0; i < batchSize; i++) {
@@ -166,6 +164,7 @@ void Metadata<dimension>::getSpatialLocations(/*long*/ at::Tensor spatialSize,
       lD[(it->second + offset) * (dimension + 1) + dimension] = i;
     }
   }
+  return locations;
 }
 template <Int dimension>
 void Metadata<dimension>::createMetadataForDenseToSparse(
@@ -260,7 +259,7 @@ Metadata<dimension>::sparsifyCompare(Metadata<dimension> &mReference,
                                      Metadata<dimension> &mSparsified,
                                      /*long*/ at::Tensor spatialSize) {
   auto p = LongTensorToPoint<dimension>(spatialSize);
-  at::Tensor delta = at::zeros({nActive[p]}, at::kFloat);
+  at::Tensor delta = torch::zeros({nActive[p]}, at::kFloat);
   float *deltaPtr = delta.data<float>();
   auto &sgsReference = mReference.grids[p];
   auto &sgsFull = grids[p];
@@ -588,13 +587,13 @@ Metadata<dimension>::compareSparseHelper(Metadata<dimension> &mR,
       }
     }
   }
-  at::Tensor cL_ = at::empty({(long)cL.size()}, at::CPU(at::kLong));
+  at::Tensor cL_ = torch::empty({(long)cL.size()}, at::CPU(at::kLong));
   std::memcpy(cL_.data<long>(), &cL[0], cL.size() * sizeof(long));
-  at::Tensor cR_ = at::empty({(long)cR.size()}, at::CPU(at::kLong));
+  at::Tensor cR_ = torch::empty({(long)cR.size()}, at::CPU(at::kLong));
   std::memcpy(cR_.data<long>(), &cR[0], cR.size() * sizeof(long));
-  at::Tensor L_ = at::empty({(long)L.size()}, at::CPU(at::kLong));
+  at::Tensor L_ = torch::empty({(long)L.size()}, at::CPU(at::kLong));
   std::memcpy(L_.data<long>(), &L[0], L.size() * sizeof(long));
-  at::Tensor R_ = at::empty({(long)R.size()}, at::CPU(at::kLong));
+  at::Tensor R_ = torch::empty({(long)R.size()}, at::CPU(at::kLong));
   std::memcpy(R_.data<long>(), &R[0], R.size() * sizeof(long));
   return {cL_, cR_, L_, R_};
 }

sparseconvnet/SCN/Metadata/Metadata.h

@@ -92,8 +92,7 @@ public:
                                 /*long*/ at::Tensor locations,
                                 /*float*/ at::Tensor vecs, bool overwrite);
 
-  void getSpatialLocations(/*long*/ at::Tensor spatialSize,
-                           /*long*/ at::Tensor locations);
+  at::Tensor getSpatialLocations(/*long*/ at::Tensor spatialSize);
   void createMetadataForDenseToSparse(/*long*/ at::Tensor spatialSize,
                                       /*long*/ at::Tensor nz_, long batchSize);
 

sparseconvnet/sparseConvNetTensor.py

@@ -19,8 +19,7 @@ class SparseConvNetTensor(object):
         "Coordinates and batch index for the active spatial locations"
         if spatial_size is None:
             spatial_size = self.spatial_size
-        t = torch.LongTensor()
-        self.metadata.getSpatialLocations(spatial_size, t)
+        t = self.metadata.getSpatialLocations(spatial_size)
         return t
 
     def type(self, t=None):