PArallel Distributed Deep LEarning: Machine Learning Framework from Industrial Practice （『飞桨』核心框架，深度学习&机器学习高性能单机、分布式训练和跨平台部署）
COMMITS
April 3, 2026
Z
fix layer_norm grad (#78567)
zhengshengning committed
S
[RMSNorm] remove invvar from rms_norm Python API return value (#78438)
SUN Dong committed
W
[API Compatibility] align paddle.nn.Layer.to/paddle.Tensor.to -part (#78301)
Wenfei (Charles) Qi committed
F
fix KGroupedGemmXPUFunction bug (#78500)
Fang Ru committed
S
N
[CI] Add approval gate for `.agents/skills` changes (#78569)
Nyako Shigure committed
N
[0 SIZE] Fix canonical strides for zero-size DenseTensor (#78565)
Nyako Shigure committed
R
[Cpp API Compatibility] Fix flashmla compile (#78550)
Runming Xie committed
R
[Cpp API Compatibility] Align misc apis (#78555)
Runming Xie committed
April 2, 2026
R
[Cpp API Compatibility] Fix arange default dtype (#78552)
Runming Xie committed
G
N
[Dy2St] Make `RandomRotation` static comparison deterministic (#78562)
Nyako Shigure committed
R
[Cpp API Compatibility] Align resize api (#78554)
Runming Xie committed
G
R
[Cpp API Compatibility] Add `torch/torch.h` (#77854)
Runming Xie committed
R
[Cpp API Compatibility] Align event api (#78553)
Runming Xie committed
A
W
disable tests for Conv2dTranspose (#78561)
Wenfei (Charles) Qi committed
R
[Cpp API Compatibility] Align device related APIs (#78551)
Runming Xie committed
A
add arg force and warning messege for paddle.Tensor.numpy() (#78491)
ALGO1832 committed
April 1, 2026
A
fix bug in recompute.py (#78531)
adam-xiaoyao committed
H
Fix out-of-bounds memory access in SetKernel for 0-size tensor (#78486)
HU Shenwei committed
A
[Precision Depth Alignment]support p_norm_grad (#78338)
AlAuAu committed
S
C
optimize Resize phi::make_ddim [phi_ops] (#78420)
co63oc committed
March 31, 2026
W
[API Compatibility] Add conv_transpose decorator and alias, test conv2d_transpose only -part (#78475)
Wenfei (Charles) Qi committed
Z
[PAA] Align tanhshrink and gelu precision with PyTorch (#78241)
Zhan Rongrui committed
Z
[fix] phi::Full in masked_scatter_grad kernel (#78530)
zhengshengning committed