COMMITS
August 10, 2026
N
[CI] Upgrade `apache-tvm-ffi` to `v0.1.13.post3` (#79642)
Nyako Shigure committed
D
Fix kv server for put vaule waiting (#79633)
Difer committed
August 7, 2026
Z
[Memory] Enable VMM V2 allocator by default (#79605)
Zhang Ting committed
August 6, 2026
N
[CI] Upgrade `apache-tvm-ffi` to `v0.1.13.post2` (#79622)
Nyako Shigure committed
N
[Cpp API Compatibility] Support `at::scalar_tensor` (#79623)
Nyakku Shigure committed
G
[CI] Remove unused develop wheel fetching (#79625)
Gu Shiwei committed
G
update ops file name (#79615)
Gu Shiwei committed
August 4, 2026
August 3, 2026
N
[CI] Remove unused approval checks (#79607)
Nyako Shigure committed
G
[CI] Enable NPU C++20 toolchain baseline (#79567)
gouzil committed
July 31, 2026
Z
[Memory] Make VMM V2 backing release transactional and teardown safe (#79592)
Zhang Ting committed
July 30, 2026
O
O
fix int32 index overflow in concat/split GPU functor y-loop (#79588)
omoYang committed
A
F
[Security] Recheck CIFAR archive MD5 before pickle load (#79576)
feixi committed
F
č
đ Fix add cuda 700 (#79575)
č夊č committed
N
[Cpp API Compatibility] Support dtype caster and CUDA opt-in shared memory (#79570)
Nyakku Shigure committed
č
đ Fix int32 check (#79584)
č夊č committed
July 29, 2026
N
[Cpp API Compatibility] Unwrap `StorageHolderView` in `record_stream` (#79566)
Nyako Shigure committed
F
[Operator Mechanism] Support mm out_dtype for FP16 CUDA (#79549)
feixi committed
L
[CI] Rerun approval checks by head sha (#79572)
liuhao2638 committed
G
[CI] Update build image baseline to GCC 11 and CUDA 12.8 (#79340)
gouzil committed
Z
[Operator Mechanism] DenseTensor deserialization heap overflow write (#79561)
zhengshengning committed
O
align the precision of log2 api with torch+2.12.0 (#79541)
omoYang committed
L
[CI] Rerun approval checks after approve (#79569)
liuhao2638 committed
č
[User Experience] support copy/pickle for Place objects (#79558)
č夊č committed