-
Dear Devs,
thank you for this library. I am currently writing my model using it, but I am a bit confused about the usage. I am working with sparse matrices but I find myself struggling with which typ…
-
### Issue summary
I’m measuring the performance of the inference of a single sparse convolutional layer using SkimCaffe. However, the results show that:
1. The effective flops achieved is only aro…
-
1. It has been mentioned in the paper that `or the first layer, instead of a 7×7 2D convolution, we use a 5×5×5×1 generalized sparse convolution. However, for the rest of the networks, we follow the o…
-
I have some confusion regarding the future frames mentioned in your paper. How are "ours-0" and "ours-7" defined? The convolution across the temporal dimension in the code uses padding of seven frames…
-
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…
-
I am trying to implement a tensorRT plugin for a single sparse convolution layer and need to write the convolution result back to output pointer defined in the ```enqueue()``` interface from TensorRT.…
hygxy updated
9 months ago
-
[SuiteSparseGraphBLAS.jl](https://github.com/JuliaSparse/SuiteSparseGraphBLAS.jl) provide the sparse representation of a graph and have better performance on algebraic operations, relative to `SparseM…
-
### 🚀 The feature, motivation and pitch
I'm working on tensors that most of the cells are 0's and some cells are 1's, say `100` 1's out of `50^3` 0's.
I tried to use [turn tensor into sparse tensor]…
-
From a code perspective, sparse convolution operations only appear in test_ forward. Does this mean that the model only trains traditional classification and regression heads? And the process shown in…
-
I have installed all the packages and fix all the runtime Error and Keyerror, but the training is terminated quickly. (I use one A100 GPU to drive the code.)
Experiment name: semseg-pt-v3m1-0-base
…