Spaces:
Runtime error
Runtime error
static __global__ void argmax_f32( | |
const float * x, int32_t * dst, const int64_t ncols, const int64_t nrows) { | |
int argmax_thread = 0; | |
const int64_t row0 = (int64_t)blockIdx.x*WARP_SIZE; | |
for (int64_t row1 = 0; row1 < WARP_SIZE; ++row1) { | |
const int64_t row = row0 + row1; | |
if (row >= nrows) { | |
break; | |
} | |
float maxval = -FLT_MAX; | |
int argmax = -1; | |
for (int32_t col = threadIdx.x; col < ncols; col += WARP_SIZE) { | |
const float val = x[row*ncols + col]; | |
const int bigger = val > maxval; | |
const int not_bigger = bigger ^ 0x00000001; | |
maxval = maxval*not_bigger + val*bigger; | |
argmax = argmax*not_bigger + col*bigger; | |
} | |
for (int mask = 16; mask > 0; mask >>= 1) { | |
const float val = __shfl_xor_sync(0xFFFFFFFF, maxval, mask, WARP_SIZE); | |
const int col = __shfl_xor_sync(0xFFFFFFFF, argmax, mask, WARP_SIZE); | |
const int bigger = val > maxval; | |
const int not_bigger = bigger ^ 0x00000001; | |
maxval = maxval*not_bigger + val*bigger; | |
argmax = argmax*not_bigger + col*bigger; | |
} | |
const int store = row1 == threadIdx.x; | |
argmax_thread += store*argmax; | |
} | |
const int row = row0 + threadIdx.x; | |
if (row >= nrows) { | |
return; | |
} | |
dst[row] = argmax_thread; | |
} | |
void ggml_cuda_argmax(ggml_backend_cuda_context & ctx, ggml_tensor * dst) { | |
const ggml_tensor * src0 = dst->src[0]; | |
GGML_ASSERT(src0->type == GGML_TYPE_F32); | |
GGML_ASSERT( dst->type == GGML_TYPE_I32); | |
GGML_ASSERT(ggml_is_contiguous(src0)); | |
const int64_t ne00 = src0->ne[0]; | |
const int64_t nrows = ggml_nrows(src0); | |
const float * src0_d = (const float *) src0->data; | |
int32_t * dst_d = (int32_t *) dst->data; | |
cudaStream_t stream = ctx.stream(); | |
const int64_t num_blocks = (nrows + WARP_SIZE - 1) / WARP_SIZE; | |
const dim3 blocks_dim(WARP_SIZE, 1, 1); | |
const dim3 blocks_num(num_blocks, 1, 1); | |
argmax_f32<<<blocks_num, blocks_dim, 0, stream>>>(src0_d, dst_d, ne00, nrows); | |
} | |