张壹 zhangyiss
  • Joined on 2024-09-10
zhangyiss synced commits to sdpav-base at zhangyiss/mlx from mirror 2025-08-07 12:30:33 +08:00
7fa520e955 Remove batch sdpa
a22d0bf273 Add stricter condition to matrix sdpa
99d8de8445 Fix cudnn routing
c66b76a8c8 Update routing
f81edd184f Complete 2 pass sdpav
Compare 5 commits »
zhangyiss synced commits to main at zhangyiss/mlx from mirror 2025-08-07 12:30:32 +08:00
f2adb5638d Fix typo in metal command encoder (#2471)
728d4db582 Support destination arg in tree flatten/unflatten (#2450)
Compare 2 commits »
zhangyiss synced commits to sdpav-backup at zhangyiss/mlx from mirror 2025-08-07 12:30:32 +08:00
zhangyiss synced new reference sdpav-backup to zhangyiss/mlx from mirror 2025-08-07 12:30:32 +08:00
zhangyiss synced and deleted reference refs/tags/refs/pull/2450/merge at zhangyiss/mlx from mirror 2025-08-07 12:30:31 +08:00
zhangyiss synced commits to refs/pull/2300/merge at zhangyiss/mlx from mirror 2025-08-07 04:06:46 +08:00
db5c7efcf6 revert default cuda install (#2465)
7bb96e4249 fix cublas on h100 (#2466)
Compare 3 commits »
zhangyiss synced commits to refs/pull/2450/merge at zhangyiss/mlx from mirror 2025-08-07 04:06:46 +08:00
db5c7efcf6 revert default cuda install (#2465)
7bb96e4249 fix cublas on h100 (#2466)
Compare 3 commits »
zhangyiss synced commits to sdpav-base at zhangyiss/mlx from mirror 2025-08-07 04:06:45 +08:00
7f8ba2a003 [WIP] 2 pass sdpav
c28249b81a Add more nvtx range for debug
e74bcdc5e3 Add sdpa file
d8ed6c1aa3 Add base cudnn attention support
db5c7efcf6 revert default cuda install (#2465)
Compare 40 commits »
zhangyiss synced commits to refs/pull/2074/merge at zhangyiss/mlx from mirror 2025-08-07 04:06:45 +08:00
db5c7efcf6 revert default cuda install (#2465)
7bb96e4249 fix cublas on h100 (#2466)
Compare 3 commits »
zhangyiss synced commits to main at zhangyiss/mlx from mirror 2025-08-07 04:06:44 +08:00
db5c7efcf6 revert default cuda install (#2465)
7bb96e4249 fix cublas on h100 (#2466)
Compare 2 commits »
zhangyiss synced and deleted reference refs/tags/refs/pull/2215/merge at zhangyiss/mlx from mirror 2025-08-07 04:06:43 +08:00
zhangyiss synced and deleted reference refs/tags/refs/pull/2465/merge at zhangyiss/mlx from mirror 2025-08-07 04:06:43 +08:00
zhangyiss synced and deleted reference refs/tags/refs/pull/2466/merge at zhangyiss/mlx from mirror 2025-08-07 04:06:43 +08:00
zhangyiss synced and deleted reference refs/tags/revert_cuda_install at zhangyiss/mlx from mirror 2025-08-07 04:06:42 +08:00
zhangyiss synced commits to refs/pull/2104/merge at zhangyiss/mlx from mirror 2025-08-06 19:50:22 +08:00
fa89f0b150 faster gather qmm sorted test (#2463)
ca973d1e83 fix install tags (#2464)
828c5f1137 Use SmallVector for shapes and strides (#2454)
7d86a5c108 Feat: add USE_SYSTEM_FMT CMake option (#2219)
Compare 10 commits »
zhangyiss synced commits to refs/pull/1091/merge at zhangyiss/FTXUI from mirror 2025-08-06 14:16:39 +08:00
Compare 2 commits »
zhangyiss synced commits to remove-threads-2 at zhangyiss/FTXUI from mirror 2025-08-06 14:16:38 +08:00
dc3a6044ee Update
zhangyiss synced commits to refs/pull/1091/head at zhangyiss/FTXUI from mirror 2025-08-06 14:16:38 +08:00
dc3a6044ee Update
zhangyiss pushed to dev_yi at zhangyiss/gctl_tutorials 2025-08-06 13:03:07 +08:00
zhangyiss synced commits to refs/pull/2300/merge at zhangyiss/mlx from mirror 2025-08-06 11:36:48 +08:00
fa89f0b150 faster gather qmm sorted test (#2463)
Compare 2 commits »