huggingface pytorch_block_sparse issues

huggingface / pytorch_block_sparse

Fast Block Sparse Matrices for Pytorch

Other

545 stars 35 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Only blockshape=(32, 32) is supported

#20 Routhleck opened 1 year ago
0
`bfloat16` support

#19 lostmsu opened 2 years ago
0
Using a smaller block size

#18 pclucas14 opened 2 years ago
0
What is "custom" "custom-back" in dispatch_policies.h?

#17 ziyuhuang123 opened 2 years ago
0
Sparse pattern is not guaranteed to be full rank

#16 jrapin opened 3 years ago
0
No memory reduction observed in a simple sparse-dense multiplication

#15 x-zho14 opened 3 years ago
0
I get error on this simple code

#14 ebrahimHappy opened 3 years ago
4
can we used pre-defined sparse matrix?

#13 hsmills opened 4 years ago
0
could you provide some code for loading a saved sparsified model?

#12 DanqingZ opened 4 years ago
0
does this package allow multi-gpu training and distributed training?

#11 DanqingZ opened 4 years ago
1
Specify the sparsity for the BlockSparseLinear

#10 zheng-ningxin closed 3 years ago
1
Any way to bypass constraint of input and output features to be a multiple of block size

#9 shreyanshs opened 4 years ago
0
Incorrect error message (says in_features, should be out_features)

#8 ejohb opened 4 years ago
3
OSError: CUDA_HOME environment variable is not set

#7 osainz59 opened 4 years ago
7
change order readme

#6 clmnt opened 4 years ago
0
Block Sparse Ax=b Solver?

#5 twhughes opened 4 years ago
0
Adds a Warning when the Patcher failed to find matching Layers.

#4 Narsil closed 4 years ago
0
Suggestion to improve Basic Usage readability.

#3 Narsil closed 4 years ago
0
quick wording pass on the readme

#2 clmnt closed 4 years ago
0
Update README.md

#1 thomwolf closed 4 years ago
0