• L
    [bf16] pten matmul cuda kernel support bf16 (#39485) · d5a0d31a
    Leo Chen 提交于
    * pten matmul cuda kernel support bf16
    
    * fix pten kernel name
    
    * add matmul_grad bf16 kernel
    
    * add emptylike bf16 kernel
    
    * fix compile
    
    * suppport rocm
    
    * fix error
    
    * fix rocm
    
    * add bf16 header file
    
    * fix compile
    d5a0d31a
op_test.py 85.4 KB