| /third_party/mindspore/mindspore/ccsrc/backend/kernel_compiler/gpu/cuda_impl/ |
| D | layer_norm_grad_grad_impl.cu | 193 … const T &epsilon, T *sum5, T *sum6, T *sum7, T *share_mem, in InputThreadReduceOuterMean() argument
223 sum7[0] -= part3; in InputThreadReduceOuterMean()
231 … const half &epsilon, half *sum5, half *sum6, half *sum7, in InputThreadReduceOuterMean() argument
261 sum7[0] -= part3; in InputThreadReduceOuterMean()
268 inline __device__ void InputWarpReduceOuterMean(T *sum5, T *sum6, T *sum7) { in InputWarpReduceOuterMean() argument
272 sum7[0] += __shfl_down_sync(0xffffffff, sum7[0], delta); in InputWarpReduceOuterMean()
277 inline __device__ void InputBlockReduceOuterMean(const int &col_dim, T *sum5, T *sum6, T *sum7, T *… in InputBlockReduceOuterMean() argument
285 share_mem[offset + 6] = sum7[0]; in InputBlockReduceOuterMean()
371 T sum7 = 0; in InputPropKernel() local
379 …InputThreadReduceOuterMean(row, col_dim, param_dim, epsilon, &sum5, &sum6, &sum7, share_mem.addr()… in InputPropKernel()
[all …]
|
| /third_party/cmsis/CMSIS/DSP/Source/MatrixFunctions/ |
| D | arm_mat_mult_f32.c | 549 float32_t sum0,sum1, sum2,sum3, sum4, sum5 , sum6, sum7; in arm_mat_mult_f32() local
599 sum7 = 0.0f; in arm_mat_mult_f32()
689 sum7 += vget_lane_f32(accum, 0) + vget_lane_f32(accum, 1); in arm_mat_mult_f32()
705 sum7 += *pIn1H++ * (*pIn2); in arm_mat_mult_f32()
720 *pxH++ = sum7; in arm_mat_mult_f32()
|
| /third_party/ffmpeg/libavfilter/ |
| D | vf_convolution.c | 195 int sum7 = c0[x] * 5 + c1[x] * 5 + c2[x] * -3 + in filter16_kirsch() local
202 sum6 = FFMAX(sum6, sum7); in filter16_kirsch()
298 int sum7 = c0[x] * 5 + c1[x] * 5 + c2[x] * -3 + in filter_kirsch() local
305 sum6 = FFMAX(sum6, sum7); in filter_kirsch()
|
| /third_party/ffmpeg/libavcodec/mips/ |
| D | hpeldsp_msa.c | 670 v8u16 sum0, sum1, sum2, sum3, sum4, sum5, sum6, sum7; in common_hv_bil_no_rnd_8x8_msa() local
699 sum7 = add7 + add8 + 1; in common_hv_bil_no_rnd_8x8_msa()
702 SRA_4V(sum4, sum5, sum6, sum7, 2); in common_hv_bil_no_rnd_8x8_msa()
705 PCKEV_B2_SB(sum5, sum4, sum7, sum6, out0, out1); in common_hv_bil_no_rnd_8x8_msa()
|
| /third_party/mindspore/tests/st/ops/gpu/ |
| D | test_layer_norm_grad_grad_op.py | 92 sum7 = np.mean(-part3, tuple(norm_axis), keepdims=True)
94 d_x = part3 + part4 + sum7
|
| /third_party/ffmpeg/libpostproc/ |
| D | postprocess_altivec_template.c | 658 vector signed int sum5, sum6, sum7, sum8, sum9; in dering_altivec() local
695 const vector signed int sump57 = vec_mergel(sum5, sum7); in dering_altivec()
|