mlx/mlx/backend/metal
Awni Hannun 881615b072
Faster metal compiled kernels + some fixes (#1486)
* bump mac tests to use py39

* work per thread for compiled kernels

* fixe for large arrays

* fix
2024-10-14 12:45:38 -07:00
..
jit More fixes for arrays with large sizes (#1405) 2024-09-17 12:46:31 -07:00
kernels Faster metal compiled kernels + some fixes (#1486) 2024-10-14 12:45:38 -07:00
allocator.cpp Allow querying the allocator for the buffer size (#1404) 2024-09-11 21:02:16 -07:00
allocator.h Allow querying the allocator for the buffer size (#1404) 2024-09-11 21:02:16 -07:00
binary.cpp Faster metal compiled kernels + some fixes (#1486) 2024-10-14 12:45:38 -07:00
binary.h Fixes for large arrays with a few ops (#1299) 2024-07-30 17:18:39 -07:00
CMakeLists.txt Chore: add pre-commit hook for cmake (#1362) 2024-09-16 12:53:01 -07:00
compiled.cpp Faster metal compiled kernels + some fixes (#1486) 2024-10-14 12:45:38 -07:00
conv.cpp Conv grad with groups + bugfix (#1449) 2024-10-06 07:08:53 -07:00
copy.cpp Faster metal compiled kernels + some fixes (#1486) 2024-10-14 12:45:38 -07:00
copy.h Fix copying scalars by adding fill_gpu (#1402) 2024-09-09 15:54:08 -07:00
custom_kernel.cpp Make the GPU device more thread safe (#1478) 2024-10-12 17:49:15 -07:00
device.cpp Make the GPU device more thread safe (#1478) 2024-10-12 17:49:15 -07:00
device.h Make the GPU device more thread safe (#1478) 2024-10-12 17:49:15 -07:00
distributed.cpp Adds send/recv ops in distributed (#1366) 2024-08-26 23:01:37 -07:00
event.cpp Fix array is_available race cases (#1468) 2024-10-07 19:13:50 -07:00
fft.cpp Fix normalization check_input (#1452) 2024-10-03 13:26:56 -07:00
hadamard.cpp Make the GPU device more thread safe (#1478) 2024-10-12 17:49:15 -07:00
indexing.cpp Make the GPU device more thread safe (#1478) 2024-10-12 17:49:15 -07:00
jit_kernels.cpp Faster metal compiled kernels + some fixes (#1486) 2024-10-14 12:45:38 -07:00
kernels.h fix jit reduce (#1395) 2024-09-04 14:03:10 -07:00
make_compiled_preamble.sh fix compiling with space in paths (#1332) 2024-08-15 16:39:24 -07:00
matmul.cpp Conv grad with groups + bugfix (#1449) 2024-10-06 07:08:53 -07:00
matmul.h Conv grad with groups + bugfix (#1449) 2024-10-06 07:08:53 -07:00
metal_impl.h Add synchronize function (#1006) 2024-04-22 08:25:46 -07:00
metal.cpp Fix array is_available race cases (#1468) 2024-10-07 19:13:50 -07:00
metal.h Reset peak memory (#1074) 2024-05-03 17:12:51 -07:00
nojit_kernels.cpp fix jit reduce (#1395) 2024-09-04 14:03:10 -07:00
normalization.cpp Fix normalization check_input (#1452) 2024-10-03 13:26:56 -07:00
primitives.cpp Avoid io timeout for large arrays (#1442) 2024-09-27 13:32:14 -07:00
quantized.cpp Fix normalization check_input (#1452) 2024-10-03 13:26:56 -07:00
reduce.cpp Fix normalization check_input (#1452) 2024-10-03 13:26:56 -07:00
reduce.h Further reduction tuning (#1349) 2024-08-23 10:35:25 -07:00
rope.cpp Xcode 160 (#1384) 2024-09-10 15:15:17 -07:00
scaled_dot_product_attention.cpp Metal shaders for memory efficient self attention on large sequences (#964) 2024-06-03 09:16:19 -07:00
scan.cpp Fix normalization check_input (#1452) 2024-10-03 13:26:56 -07:00
slicing.cpp Fix copying scalars by adding fill_gpu (#1402) 2024-09-09 15:54:08 -07:00
slicing.h Fix slice data size (#1394) 2024-09-04 19:10:43 -07:00
softmax.cpp Fix normalization check_input (#1452) 2024-10-03 13:26:56 -07:00
sort.cpp Fix normalization check_input (#1452) 2024-10-03 13:26:56 -07:00
ternary.cpp Faster metal compiled kernels + some fixes (#1486) 2024-10-14 12:45:38 -07:00
ternary.h Add some internal GPU apis (#1177) 2024-06-04 09:24:26 -07:00
unary.cpp Faster metal compiled kernels + some fixes (#1486) 2024-10-14 12:45:38 -07:00
unary.h Add some internal GPU apis (#1177) 2024-06-04 09:24:26 -07:00
utils.cpp Add gemv masked to JIT plus some fixes (#1310) 2024-08-07 13:38:07 -07:00
utils.h Add gemv masked to JIT plus some fixes (#1310) 2024-08-07 13:38:07 -07:00