Awni Hannun 42afe27e12 std and expm1 (#973) ... * std and expm1 * actually add expm1 * fix linux * fix vjp * relax tol for linux test * Add it to the compilable primitives --------- Co-authored-by: Angelos Katharopoulos <a_katharopoulos@apple.com>		2024-04-08 14:26:01 -07:00
..
kernels	std and expm1 (#973)	2024-04-08 14:26:01 -07:00
mps	copyright + ack	2023-11-30 11:12:53 -08:00
allocator.cpp	Improve profiling with gpu tracing (#969)	2024-04-07 21:47:43 -07:00
allocator.h	Some fixes in cache / thread safety (#777)	2024-03-05 13:30:50 -08:00
CMakeLists.txt	Adds mx.fast.layer_norm (#870)	2024-03-21 13:55:51 -07:00
compiled_preamble.h	Kernel generation (#614)	2024-02-07 13:15:59 -08:00
compiled.cpp	Fix cpu compile (#934)	2024-04-01 17:37:12 -07:00
conv.cpp	No copy gems (#801)	2024-03-12 13:13:41 -07:00
copy.cpp	Fix copy donation and add partial rope (#881)	2024-03-22 17:28:26 -07:00
copy.h	Add a SliceUpdate op and primitive (#850)	2024-03-20 10:39:25 -07:00
device.cpp	Improve profiling with gpu tracing (#969)	2024-04-07 21:47:43 -07:00
device.h	Kernel generation (#614)	2024-02-07 13:15:59 -08:00
fft.cpp	copyright + ack	2023-11-30 11:12:53 -08:00
indexing.cpp	Some C++ code are not needed (#841)	2024-03-18 17:04:10 -07:00
make_compiled_preamble.sh	quote file name (#670)	2024-02-11 10:33:30 -08:00
matmul.cpp	add numeric type hierarchy and issubdtype as well as a set_dtype meth… (#427)	2024-03-25 12:32:59 -07:00
matmul.h	No copy gems (#801)	2024-03-12 13:13:41 -07:00
metal_impl.h	Improve profiling with gpu tracing (#969)	2024-04-07 21:47:43 -07:00
metal.cpp	Improve profiling with gpu tracing (#969)	2024-04-07 21:47:43 -07:00
metal.h	Improve profiling with gpu tracing (#969)	2024-04-07 21:47:43 -07:00
normalization.cpp	segfaut layer norm grad (#955)	2024-04-04 10:59:15 -07:00
primitives.cpp	std and expm1 (#973)	2024-04-08 14:26:01 -07:00
quantized.cpp	Fix nan and improve speed for qvm (#903)	2024-03-26 10:41:45 -07:00
reduce.cpp	Implement vjps for some primitives in the fast namespace (#883)	2024-03-26 16:35:34 -07:00
reduce.h	Implement vjps for some primitives in the fast namespace (#883)	2024-03-26 16:35:34 -07:00
rope.cpp	Implement vjps for some primitives in the fast namespace (#883)	2024-03-26 16:35:34 -07:00
scaled_dot_product_attention.cpp	add numeric type hierarchy and issubdtype as well as a set_dtype meth… (#427)	2024-03-25 12:32:59 -07:00
scan.cpp	copyright + ack	2023-11-30 11:12:53 -08:00
softmax.cpp	Option for precise softmax (#953)	2024-04-04 08:32:35 -07:00
sort.cpp	Fix multiblock sort limits (#906)	2024-03-26 14:00:00 -07:00
utils.h	Improve profiling with gpu tracing (#969)	2024-04-07 21:47:43 -07:00