张壹 zhangyiss
  • Joined on 2024-09-10
zhangyiss synced commits to refs/pull/2491/head at zhangyiss/mlx from mirror 2025-08-18 00:16:41 +08:00
f433c9a421 Revert back to old rms norm kernel
3a2b90fc1a Implement forward rms_norm with cuDNN
290b45eba3 Add RAII managed CudaGraph class
c422050ca7 Update cuDNN Frontend to v1.14 (#2505)
1ba18ff7d9 [CUDA] Fix conv grads with groups (#2495)
Compare 9 commits »
zhangyiss synced commits to refs/pull/2491/merge at zhangyiss/mlx from mirror 2025-08-18 00:16:41 +08:00
f433c9a421 Revert back to old rms norm kernel
3a2b90fc1a Implement forward rms_norm with cuDNN
290b45eba3 Add RAII managed CudaGraph class
c422050ca7 Update cuDNN Frontend to v1.14 (#2505)
Compare 5 commits »
zhangyiss synced and deleted reference refs/tags/refs/pull/2505/merge at zhangyiss/mlx from mirror 2025-08-18 00:16:40 +08:00
zhangyiss synced and deleted reference refs/tags/refs/pull/2506/merge at zhangyiss/mlx from mirror 2025-08-18 00:16:40 +08:00
zhangyiss synced commits to cuda-gemm-conv at zhangyiss/mlx from mirror 2025-08-18 00:16:40 +08:00
68ac75ad3c Ensure wt is contiguous
4833e45ac4 Enable gemm_conv for 1D and 3D
73027c0b71 Unfolder needs contiguous input
634eb50de3 Use fallback when execution failed
435e49ece3 Fix reshaping wt
Compare 10 commits »
zhangyiss synced commits to custom_kernel_test at zhangyiss/mlx from mirror 2025-08-18 00:16:40 +08:00
zhangyiss synced new reference custom_kernel_test to zhangyiss/mlx from mirror 2025-08-18 00:16:40 +08:00
zhangyiss synced commits to main at zhangyiss/mlx from mirror 2025-08-18 00:16:40 +08:00
1df9887998 Ensure no oob read in gemv_masked (#2508)
73f22d6226 Ensure small sort doesn't use indices if not argsort (#2506)
c422050ca7 Update cuDNN Frontend to v1.14 (#2505)
Compare 3 commits »
zhangyiss synced and deleted reference refs/tags/fix-sort-null-read at zhangyiss/mlx from mirror 2025-08-18 00:16:39 +08:00
zhangyiss synced and deleted reference refs/tags/gemv-masked-oob-read at zhangyiss/mlx from mirror 2025-08-18 00:16:39 +08:00
zhangyiss synced commits to refs/pull/1094/head at zhangyiss/FTXUI from mirror 2025-08-17 22:57:20 +08:00
143b24c6a5 Add opt-in piped input support for POSIX systems
40e1fac3d4 Warn against Microsoft <windows.h> min and max macro (#1084)
8ef18ab647 Remove pthread dependency
Compare 3 commits »
zhangyiss synced commits to refs/pull/1094/merge at zhangyiss/FTXUI from mirror 2025-08-17 22:57:20 +08:00
143b24c6a5 Add opt-in piped input support for POSIX systems
f3448f49f1 Improve example style (#1097)
40e1fac3d4 Warn against Microsoft <windows.h> min and max macro (#1084)
8ef18ab647 Remove pthread dependency
Compare 5 commits »
zhangyiss synced commits to refs/pull/1070/merge at zhangyiss/FTXUI from mirror 2025-08-17 22:57:19 +08:00
40e1fac3d4 Warn against Microsoft <windows.h> min and max macro (#1084)
Compare 2 commits »
zhangyiss synced commits to refs/pull/1084/head at zhangyiss/FTXUI from mirror 2025-08-17 22:57:19 +08:00
ff984016d7 Format
c3a3b2f0e5 Add warning
5035ef95f9 Revert "fix: using max() min() collides in windows build with predefined macros"
54c1053ca3 fix: using max() min() collides in windows build with predefined macros
8ef18ab647 Remove pthread dependency
Compare 6 commits »
zhangyiss synced commits to gh-pages at zhangyiss/FTXUI from mirror 2025-08-17 22:57:08 +08:00
zhangyiss synced commits to main at zhangyiss/FTXUI from mirror 2025-08-17 22:57:08 +08:00
f3448f49f1 Improve example style (#1097)
40e1fac3d4 Warn against Microsoft <windows.h> min and max macro (#1084)
Compare 2 commits »
zhangyiss synced and deleted reference refs/tags/refs/pull/1084/merge at zhangyiss/FTXUI from mirror 2025-08-17 22:57:07 +08:00
zhangyiss synced new reference fix-sort-null-read to zhangyiss/mlx from mirror 2025-08-17 15:46:44 +08:00
zhangyiss synced commits to gemv-masked-oob-read at zhangyiss/mlx from mirror 2025-08-17 15:46:44 +08:00
zhangyiss synced new reference gemv-masked-oob-read to zhangyiss/mlx from mirror 2025-08-17 15:46:44 +08:00