-
Notifications
You must be signed in to change notification settings - Fork 6k
Pull requests: PaddlePaddle/Paddle
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[CP] No longer perform supplementary checks for the specified GPU architecture
#78734
opened Apr 20, 2026 by
swgu98
Member
Loading…
[release/3.4] fix: use ThrustAllocator in argsort 1D path to avoid implicit cudaStr…
contributor
External developers
#78733
opened Apr 20, 2026 by
risemeup1111
Loading…
No longer perform supplementary checks for the specified GPU architecture
#78732
opened Apr 20, 2026 by
swgu98
Member
Loading…
CI测试不review fix [phi_ops]
contributor
External developers
#78731
opened Apr 20, 2026 by
co63oc
Contributor
Loading…
[release/3.4][CINN] Make CINN compilable with arm CPU and add missing External developers
skip-ci: slice
nvidia-* dependencies for arm wheel
contributor
#78730
opened Apr 20, 2026 by
risemeup1111
Loading…
support to detect the recompute context
#78728
opened Apr 20, 2026 by
liufengwei0103
Contributor
Loading…
Include cutlass/version.h for version compatibility
#78724
opened Apr 20, 2026 by
xuanyuanminzheng
Contributor
Loading…
feat(HIP): register bfloat16 kernels for conv_transpose on ROCm
contributor
External developers
PaddlePaddle Hackathon
#78721
opened Apr 19, 2026 by
Dryoung95
Loading…
[WIP][CUDA13.2] Environment Adaptation support Paddle on CUDA 13.2
contributor
External developers
#78720
opened Apr 19, 2026 by
gouzil
Member
Loading…
CI测试不review update select_impl.cu.h [phi_ops]
contributor
External developers
#78717
opened Apr 19, 2026 by
co63oc
Contributor
Loading…
[ROCm] 适配 HIP BF16: 注册 BF16 layer_norm、绕开 MIOpen BF16 softmax、HIP 跳过 cuDNN-only conv2d 融合 pass
contributor
External developers
PaddlePaddle Hackathon
#78711
opened Apr 18, 2026 by
austin1997
Loading…
[FA] Update FA version for CP32 overlap hang fix
#78709
opened Apr 18, 2026 by
Enigmatisms
Contributor
Loading…
[Cpp API Compatibility] Align ScalarType and some other APIs
contributor
External developers
#78707
opened Apr 17, 2026 by
youge325
Contributor
Loading…
[XPU] Fix precision for paddle.nn.functional.max_unpool2d
#78701
opened Apr 17, 2026 by
YqGe585
Member
Loading…
do not merge test: pull PaddleFleet PR #798 into coverage CI
#78695
opened Apr 16, 2026 by
liuhao2638
Contributor
Loading…
[XPU] Fix precision for paddle.nn.functional.max_pool2d — correct index tie-breaking
#78693
opened Apr 16, 2026 by
YqGe585
Member
Loading…
[XPU] Fix bicubic interpolate precision: use half_pixel coordinate transform to match GPU
#78692
opened Apr 16, 2026 by
YqGe585
Member
Loading…
[Precision Depth Alignment] [Adamw] Add torch-compatible AdamW kernel
#78691
opened Apr 16, 2026 by
zhengshengning
Contributor
Loading…
[Cherry-Pick] fix SharedLayerDesc lose attr is_firstly_shared (#78680)
#78689
opened Apr 16, 2026 by
Wennie396
Contributor
Loading…
CINN: Add hardware parameter logging for CustomDevice tile config
#78683
opened Apr 15, 2026 by
YuhanXu
Contributor
Loading…
[ROCm][Fix] Adapt cupti.cmake and interpolate kernels for DTK 25.04
#78682
opened Apr 15, 2026 by
DongBaiYue
Contributor
Loading…
do not merge test: pull PaddleFleet PR #772 into coverage CI
#78677
opened Apr 14, 2026 by
liuhao2638
Contributor
Loading…
[XPU] Optimize the number of Sharding communication operations in the EB model.
XPU
#78674
opened Apr 14, 2026 by
ZibinGuo
Contributor
Loading…
Previous Next
ProTip!
Adding no:label will show everything without a label.