-
Notifications
You must be signed in to change notification settings - Fork 663
Issues: iree-org/iree
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.
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
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
Make StripCompilationInfoPass to support the attention op
enhancement ➕
New feature or request
#20053
opened Feb 21, 2025 by
bangtianliu
Duplicated code for existing layout checks in CPU/GPU attribute interface implementation
codegen
Shared code generation infrastructure and dialects
good first issue 🌱
Good for newcomers
quality of life 😊
Nice things are nice; let's have some
#20050
opened Feb 20, 2025 by
hanhanW
Adding default tuning specs for SDXL MI300 + MI308
enhancement ➕
New feature or request
#20044
opened Feb 20, 2025 by
geomin12
[hip] Improvements for the hip async allocator.
enhancement ➕
New feature or request
#20043
opened Feb 20, 2025 by
AWoloszyn
Llama 8b FP8 fails to compile after first attention output to linear layer
bug 🐞
Something isn't working
#20038
opened Feb 20, 2025 by
aviator19941
[Codegen][AMDGPU] Vectorization of stores in matmuls causes regressions
codegen/rocm
ROCm code generation compiler backend (HIP/HSA)
#20032
opened Feb 19, 2025 by
qedawkins
Retire LinalgExt::PackOp and LinalgExt::UnPackOp
codegen
Shared code generation infrastructure and dialects
good first issue 🌱
Good for newcomers
quality of life 😊
Nice things are nice; let's have some
#20030
opened Feb 19, 2025 by
hanhanW
Pointless GPU barrier removal
codegen
Shared code generation infrastructure and dialects
enhancement ➕
New feature or request
good first issue 🌱
Good for newcomers
help wanted
Extra attention is needed
#20029
opened Feb 19, 2025 by
krzysz00
Performance regression for slow_memcpy
bug 🐞
Something isn't working
hal/hip
Runtime HIP HAL backend
#20028
opened Feb 19, 2025 by
Alex-Vasile
Add Relating to the IREE compiler dialects (flow, hal, vm)
AffinityAttr::areExecutionCompatible
interface.
compiler/dialects
#20026
opened Feb 19, 2025 by
benvanik
[gpu][rocm] Numeric mismatch with/out Something isn't working
pad-to-intrinsics
for convolution with trailing fusion
bug 🐞
#20024
opened Feb 18, 2025 by
jerryyin
[DT] Implement host/serialized encoding interface for GPUEncodingLayoutAttr
#20021
opened Feb 18, 2025 by
hanhanW
[DispatchGeneration] Can we fuse two linalg.generic into one dispatch?
#20019
opened Feb 18, 2025 by
yzhang93
[CPU] Incorrect result with CPU for onnx ReverseSequence
bug 🐞
Something isn't working
#20011
opened Feb 17, 2025 by
IanWood1
Implement
gpu.subgroup_reduce
with DPP intrinsics on AMD GPUs
#20007
opened Feb 17, 2025 by
krzysz00
2 tasks
[Codegen] llama 8b fp8 with attention vector distribute fail
bug 🐞
Something isn't working
#19991
opened Feb 14, 2025 by
AmosLewis
[GPU][Codegen] SDXL value mismatch without spec file on MI300
bug 🐞
Something isn't working
#19984
opened Feb 13, 2025 by
jerryyin
Fix flaky newly migrated MI300 workflow
infrastructure
Relating to build systems, CI, or testing
#19955
opened Feb 11, 2025 by
yamiyysu
Accuracy Issue for Sharded Llama
bug 🐞
Something isn't working
#19948
opened Feb 10, 2025 by
stbaione
Previous Next
ProTip!
Adding no:label will show everything without a label.