Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / stanford-futuredata/stk issues and pull requests
#16 - Only block size 128 and fp16?
Issue -
State: closed - Opened by oleksost 5 months ago
- 1 comment
#15 - bump torch, stk versions
Pull Request -
State: closed - Opened by eitanturok 6 months ago
#14 - Add assert that K % BLOCK_K == 0
Pull Request -
State: closed - Opened by sashaDoubov 6 months ago
- 1 comment
#13 - Cleanup `setup.py`
Pull Request -
State: closed - Opened by eitanturok 7 months ago
- 5 comments
#12 - Add masked loads in SDD to support K that is not a multiple of BLOCK_K
Pull Request -
State: closed - Opened by sashaDoubov 7 months ago
- 11 comments
#11 - [6-line repro] Cloning input `x` to `stk.ops.sdd()` leads to drastically different outputs
Issue -
State: closed - Opened by cmsflash 8 months ago
- 2 comments
#10 - How to build stk?
Issue -
State: closed - Opened by hesingh 9 months ago
- 1 comment
#9 - Update matrix.py to include non blocking
Pull Request -
State: closed - Opened by mvpatel2000 10 months ago
#8 - torch.checkpoint backward issue
Issue -
State: open - Opened by grad0s 10 months ago
- 1 comment
#7 - ctx.saved_tensors should be unpacked only once
Pull Request -
State: closed - Opened by grad0s 10 months ago
- 1 comment
#6 - Conversion from CSR to BCSR
Issue -
State: open - Opened by jselvam11 10 months ago
- 3 comments
#5 - Fixing version numbering to match releases.
Pull Request -
State: closed - Opened by tgale96 about 1 year ago
#4 - Updt triton pin
Pull Request -
State: closed - Opened by vchiley about 1 year ago
#3 - Is CUDA implementation available
Issue -
State: closed - Opened by FengcunLi about 1 year ago
- 3 comments
#2 - Add elementwise multiplication
Pull Request -
State: closed - Opened by sashaDoubov about 1 year ago
- 3 comments
#1 - Triton implementation of block-sparse matmul kernels.
Pull Request -
State: closed - Opened by Priya2698 over 1 year ago