Morph MORPH
SIGN IN SIGN UP
dhruvhead / Paddle UNCLAIMED
0 0 1 C++

COMMITS

develop
July 17, 2026
L
[Build] Compile FlashAttention on arm64 platforms (#79488)
liuhao2638 committed
31c7cfe
July 16, 2026
Z
[Memory] Guard CUDA VMM sources in CPU builds (#79486)
Zhang Ting committed
f92f172
N
[Build] Ensure load nvidia libraries from `site-packages` on arm64 build (#79483)
Nyakku Shigure committed
92c7ffd
Z
[Memory] Improve VMM V2 Remap Safety (#79479)
Zhang Ting committed
d6be615
苍
[Operator Mechanism] Reject `token_dim == 0` in `moe_permute` / `moe_unpermute` (#79470)
苍天荒 committed
4f91f53
F
[Operator Mechanism] fix BroadcastConfig do not support big_tensor (#79439)
feixi committed
9424709
F
[Operator Mechanism] Replace VCOPY/SCAL with direct copy and ForRange scale in CPU/GPU kernels (#79468)
feixi committed
1202d35
A
[Performance] support muon comm opt with ep >8 (#79448)
AlAuAu committed
a35b002
July 15, 2026
L
[Execute Infrastructure] ci: add bot approval required check gated on risemeup1111 approval (#79473)
liuhao2638 committed
a842085
July 14, 2026
Z
[CINN] Fix `inplace_map` bug in `register_custom_python_op` (#79454)
Zx committed
b6fb7ea
Z
Add VMM V2 remap and IPC support (#79427)
Zhang Ting committed
e126869
G
[Profiler] Export common host event gather (#79426)
gouzil committed
bd37fde
July 13, 2026
A
[API Compatibility] `paddle.optim` related apis (#79341)
ALGO1832 committed
c002606
N
[CI][H-Coverage] Increase Fleet test timeouts (#79465)
Nyako Shigure committed
206a21d
July 10, 2026
D
feat(memory): add large/small pool separate query APIs to avoid small pool traversal overhead in subbatch (#79438)
Difer committed
8817599
B
align rtruediv with torch 2.12.0 when use dynamic graph (#79429)
Bohao Fan committed
d53ee6a
W
Revert "Revert "[Operator Mechanism]Align cuBLAS workspace size for SM10 GPUs…" (#79443)
wanghuancoder committed
245e7ed
F
[Operator Mechanism]move fp8_fast_transpose_kernel to eliminate duplicate CUDA kernel error (#79433)
feixi committed
6af5ca8
July 9, 2026
Z
Improve VMM OOM diagnostics (#79435)
Zhang Ting committed
55178e2
July 8, 2026
Z
[API Compatibility] instance_norm/to_empty/linalg.det/gumbel_softmax/vstack/addcdiv/batch_norm/set_default_device/set_grad_enabled/new_tensor/_Loss/_pair/histc/hstack/true_divide_/linalg.eigh/Tensor.H/ELU/relu6/clamp_max/qr/logdet/linalg.cholesky/elu/tensordot/prelu/expand_copy/linalg.qr/clamp_ (#79383)
zhwesky2010-bot committed
034a838
July 7, 2026
X
[API Compatibility] distributions support sample_shape (#79414)
Xiaochun Yang committed
3abf104
G
[Windows] export operator kernel registry (#79419)
gouzil committed
9717bfc
W
[API Compatibility] Fix paddle.cuda.set_device(int) raising when the current place is CPU (#79392)
Wenfei (Charles) Qi committed
7b741d6
Z
[Memory] Remove VMM V2 parts from hot allocation path (#79401)
Zhang Ting committed
75c87b0
July 5, 2026
N
[Build][CI] Switch pocketfft submodule URL to GitHub mirror (#79420)
Nyako Shigure committed
2235ce4
July 4, 2026
N
[CodeStyle][Typos] Bump typos to v1.48.0 (#79416)
Nyako Shigure committed
18c11b8
July 3, 2026
G
[Build] Move test-only code out of production paths (#79403)
gouzil committed
37617cb
July 2, 2026
Z
Revert "[Operator Mechanism]Align cuBLAS workspace size for SM10 GPUs (#79373)" (#79405)
zhengshengning committed
bcd9ea6
X
[API Compatibility] add compat distributions in setup (#79398)
Xiaochun Yang committed
6c4c089
B
fix rmsnorm grad when input size is zero (#79397)
Bohao Fan committed
0f92bdc