3 Commits

Author	SHA1	Message	Date
Laurent Mazare	d4b6f6eef6	Add a minimal test for the metal bf16 matmul. (#2381)	2024-08-01 11:22:46 +02:00
Laurent Mazare	e6a5b82ba6	Fix the matmul layout for accelerate & mkl. (#2011) ... * Fix the matmul layout for accelerate & mkl. * Reduce the required precision for pow (because of accelerate). * And a fix the gelu f16 test.	2024-04-04 19:18:03 +02:00
Laurent Mazare	08c049def3	Improve the handling of matmul with squeezed layouts. (#1998) ... * Improve the handling of matmul with squeezed layouts. * Fix for the cuda backend. * Revert the temporary fix.	2024-04-02 23:17:05 +02:00