Skip to content

Issues: NVIDIA/cudnn-frontend

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

19 Open 46 Closed
19 Open 46 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

[Question] What is the graph API replacement for deprecated cudnnTransformTensor()?
#119 opened Oct 23, 2024 by caelunshun
[Question] Mismatch between Cudnn graph API DOC and FE code implementation
#117 opened Oct 22, 2024 by xiang1guo
[Question] "Error: No execution plans support the graph."
#116 opened Oct 20, 2024 by ahnseunghae
1
Cudnn frontend sdpa wrong assertion in attention module with d_bias
#115 opened Oct 8, 2024 by Skylion007
2
[Question] How to perform backpropagation for a conv + sigmoid layer?
#113 opened Sep 24, 2024 by zhewenhu
3
sdpa_fp8 having different seqlen_q and seqlen_k
#112 opened Sep 21, 2024 by MustafaFayez
cudnn-frontend crashes in case of MAX_OPGRAPH_OPS violation
#108 opened Sep 7, 2024 by gritukan
@Anerudhan
1
Wrong result of tensor addition with broadcasting
#107 opened Sep 7, 2024 by gritukan
2
Which Visual Studio 2022 BuildTools MSVC is the best version for Cuda 11.8 and Cuda 12.4 and so
#106 opened Sep 6, 2024 by FurkanGozukara
Extremely slow fp8 conv2d wgrad operation
#103 opened Aug 27, 2024 by jimgao1
4
Support batch size 0
#101 opened Aug 18, 2024 by Birch-san
[Question] Making dO contiguous affects output?
#100 opened Aug 14, 2024 by vedantroy
2
[Question] How to match Flash Attention 2 performance?
#98 opened Aug 14, 2024 by vedantroy
4
Is performance better when not using padding?
#96 opened Aug 13, 2024 by vedantroy
Matmul test failure
#78 opened May 16, 2024 by shiwenloong
2
cudnn._compiled_module.cudnnGraphNotSupportedError: [cudnn_frontend] Error: No execution plans built successfully.
#75 opened May 13, 2024 by ifromeast
2
Windows build error
#66 opened Mar 1, 2024 by tianleiwu
@Anerudhan
3
CUDNN_FRONTEND_BUILD_UNIT_TESTS option is broken
#62 opened Feb 19, 2024 by iskunk
@Anerudhan
2
What's the difference of flash attention implement between cudnn and Dao-AILab?
#52 opened Dec 20, 2023 by MoFHeka
19
ProTip! no:milestone will show everything without a milestone.