Skip to content

Pull requests: ROCm/AMDMIGraphX

New pull request New
51 Open 2,490 Closed
51 Open 2,490 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
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Normalize standard input shapes for rocBLAS (#3786)
#3827 opened Feb 19, 2025 by causten Loading…
1
Add permutation flag to python shape constructor
#3826 opened Feb 18, 2025 by pfultz2 Loading…
3
Small fixes simple small or simple changes
#3825 opened Feb 18, 2025 by CharlieL7 Loading…
@CharlieL7
3
fix compilation on HIP SDK 6.3 for Windows - undefined behavior Windows Related changes for Windows Environments
#3824 opened Feb 18, 2025 by apwojcik Loading…
4
HIP SDK 6.3 for Windows bug workaround Windows Related changes for Windows Environments
#3823 opened Feb 18, 2025 by apwojcik Loading…
4
Make hipblaslt the default instead of rocBlas
#3818 opened Feb 14, 2025 by causten Loading…
1
Add Attention Microsoft Contrib Operator Onnx Operators Adding or modifying an Onnx Operator in the MIGraphX codebase onnxruntime PR changes interaction between MIGraphX and Onnxruntime roadmap Tasks to finish for a release
#3816 opened Feb 14, 2025 by TedThemistokleous • Draft
@TedThemistokleous
Use fill_argument for literals that have the same value
#3815 opened Feb 14, 2025 by pfultz2 Loading…
@pfultz2
3
Add RotaryEmbedding contrib op
#3812 opened Feb 13, 2025 by turneram Loading…
3
WIP sparse_attention
#3805 opened Feb 10, 2025 by TedThemistokleous • Draft
@marko-fabo-htec
2
Set warmup and tuning iterations through env variable for hipBLASLt
#3792 opened Jan 31, 2025 by ahsan-ca • Draft
@ahsan-ca
6
ConvInteger: fix parsing for x_zero_point and w_zero_point (#3763)
#3781 opened Jan 24, 2025 by causten Loading…
1
Fix: Driver --batch option sets Window Dimensions.
#3770 opened Jan 20, 2025 by lakhinderwalia Loading…
@lakhinderwalia
61
Remove rocmlir unsupported reduce types
#3766 opened Jan 17, 2025 by dhernandez0 Loading…
@dhernandez0 @richagadgil
8
Experimental output fusion
#3759 opened Jan 12, 2025 by shivadbhavsar • Draft
@shivadbhavsar
1
Intial commit to target all reshaper opsops after mlir_op
#3754 opened Jan 10, 2025 by TedThemistokleous • Draft
1
Propagate layout in reshape operator and broadcasting in binary operators
#3753 opened Jan 9, 2025 by pfultz2 • Draft
@kahmed10
1
Fuse multiple outputs for pointwise and reductions
#3752 opened Jan 9, 2025 by pfultz2 • Draft
Exhaustive tune reduce operators
#3751 opened Jan 9, 2025 by pfultz2 Loading…
2
Tile channels for group norm and also fuse output reshapes in mlir
#3750 opened Jan 9, 2025 by pfultz2 • Draft
1
[In progress] Free nodes as they're parsed in program.from_value()
#3749 opened Jan 9, 2025 by kahmed10 • Draft
@kahmed10
1
Add a trim size flag
#3746 opened Jan 7, 2025 by pfultz2 Loading…
3
Layout convolution as NHWC or NCHW only roadmap Tasks to finish for a release
#3729 opened Dec 20, 2024 by pfultz2 Loading…
@pfultz2
1
Issue with int8 for MaxPool
#3725 opened Dec 19, 2024 by taylding-amd • Draft
1
1
Introduce export feature to TensorRT JSON format
#3721 opened Dec 18, 2024 by mirza-halilcevic • Draft
1
ProTip! Follow long discussions with comments:>50.