-
To speed up SD model more, will more conv kernel be supported?
-
## 🚀 Feature
The ability to force layer's weights to be sparse with some input mask so that for example a 1d convolution that has a weight matrix with mXn weights would have a sparse matrix of size m…
-
run python test_timeit.py:
compare time with nn.Conv3d()
nn.Conv3D() forward time is :0.000066
SparseConv3d() forward time is :0.001239
nn.Conv3D() backward time is :0.000870
SparseConv3d() backw…
-
Could you please release the code for calculating the FLOPS of sparse convolution?
-
I want to use nn.DataParallel for multiple-gpu training of sparse convolution in structed point cloud. However, the spc class contain spc.octrees, which is of data type uint8. So It will say "Unconve…
-
Grouped convolution is supported well in `pytorch`'s convolution layers / ops. If possible, it would a great to add that ability to `torchsparse`.
-
We want to test PineTree to constrain the halo biasing problem across the Quijote latin hypercube. It has currently only been tested on a single cosmology.
PineTree is a Neural Physical Engine, i.e. …
maho3 updated
2 months ago
-
I recently had the opportunity to read your paper "SpENCNN: Orchestrating Encoding and Sparsity for Fast Homomorphically Encrypted Neural Network Inference". In the paper you mentioned that for convol…
-
The implementation of `unfoldNd` relies on one-hot convolution. This means the convolution kernels are highly sparse. Hence, the code could run faster when using sparse tensors.
Open questions:
…
-
## 🚀 Feature
I suggest torch teams to add 'Sparse Convolution' as default nn module.
## Motivation
I want to use Sparse Convolution without external libraries like 'SPConv', 'Mincowski Engine',…
demul updated
2 years ago