Skip to content

Issues: NVIDIA/Fuser

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

265 Open 379 Closed
265 Open 379 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

RoPE-like slice-concat patterns may result in cyclic exact graphs rope
#3072 opened Oct 1, 2024 by naoyam
2
Unable to reshard non-contiguous inputs.
#3071 opened Sep 30, 2024 by wujingyue
@cowanmeg
1
Inlining RoPE-like patterns require an analysis akin to ComputeAtLogicalDomainMap rope
#3068 opened Sep 30, 2024 by naoyam
2
CollectiveBasedOverlapTest.ReduceScatterBasedPipeliningATenImplementation fails in nightly CI.
#3051 opened Sep 29, 2024 by wujingyue
3
FusionExecutorCache not synchronizing with multiple devices Multidevice
#3042 opened Sep 27, 2024 by cowanmeg
SdpaFwdOp::toString doesn't print all outputs.
#3037 opened Sep 26, 2024 by wujingyue
@Priya2698
1
Dynamic allocations in local memory error Thunder
#3027 opened Sep 26, 2024 by riccardofelluga
1
@jacobhinkle
6
Enable codegen for matmul/linear in nvfuser through Thunder matmul Python API Issues related to the Python API Top-Down Matmul Dev
#3022 opened Sep 26, 2024 by Priya2698
@Priya2698
Reduce test verbosity. good first issue Good for newcomers
#3020 opened Sep 25, 2024 by csarofeen
CI build check is slow
#3012 opened Sep 24, 2024 by naoyam
14
Reconsider TensorDomain::setLoopDomain, TensorDomain::broadcast, TendorDomain::additional_ids_
#3000 opened Sep 23, 2024 by naoyam
Fix ir_utils::compareDomains with extra loop IDs
#2985 opened Sep 21, 2024 by naoyam
@naoyam
Self-mapping error when compiling 1D bias linear fusion with smem-epilogue and no split-K bug Something isn't working matmul
#2979 opened Sep 20, 2024 by jacobhinkle
@jacobhinkle
1
Make sure setLoopDomain to use reduction domains for reduction logical domains
#2974 opened Sep 20, 2024 by naoyam
@naoyam
Update DistributedTransformerTest to use GPT-3 sizes.
#2962 opened Sep 18, 2024 by wujingyue
3
Avoid propagating invalid mappings by replicating persistent tensors
#2954 opened Sep 18, 2024 by naoyam
Enable the IdModel-based indexer by default idmodel
#2951 opened Sep 17, 2024 by naoyam
@naoyam
Clarify container methods for collections of statements cleanup good first issue Good for newcomers
#2949 opened Sep 17, 2024 by jacobhinkle
FusionReductionWithTrivialReduction_CUDA fails with compute-sanitizer
#2930 opened Sep 11, 2024 by xwang233
1
@naoyam
10
Investigate pytest hooks for running common functions before benchmarking enhancement New feature or request Python Benchmarks
#2927 opened Sep 10, 2024 by Priya2698
@Priya2698
Bad error message from illegal alias
#2910 opened Sep 5, 2024 by jjsjann123
1
@jjsjann123
Avoiding having to manually specify the comparison threshold between a sharded implementation and its reference.
#2906 opened Sep 4, 2024 by wujingyue
@Priya2698
1
testValidate uses half tolerances for fp32 accumulation.
#2904 opened Sep 4, 2024 by wujingyue
@Priya2698
1
Adding CPU scalar tensor to scalar value produces CUDA tensor. bug Something isn't working
#2853 opened Aug 27, 2024 by Priya2698
Improving Scheduling and Lowering of Slice and Concat (Part 1) rope
#2851 opened Aug 26, 2024 by naoyam
@naoyam
ProTip! Updated in the last three days: updated:>2024-09-29.