HuPengsheet
HuPengsheet
``` const int packn = csrr_vlenb() / 2; //8 const size_t vl = vsetvl_e16m1(packn); //8 // Mat bottom_im2col(size, maxk, inch, 2u * packn, packn, opt.workspace_allocator); const int size = bottom_im2col.w;...
## error log | 日志或报错信息 | ログ ``` [root@milkv-duo]/home/github/benchmark# ./benchncnn syscall error -1 loop_count = 4 num_threads = 1 powersave = 0 gpu_device = -1 cooling_down = 1 Illegal instructio...
nihui,您好! 我将ncnn移植在milkv-duo上,在运行benchncnn时,发现在proxylessnasnet出现了segment fault错误。 我把运行的算子打印出来了,确定是F0_expand这个layer报错 ``` [root@milkv-duo]/home/github/benchmark# ../bin/benchncnn syscall error -1 loop_count = 4 num_threads = 1 powersave = 0 gpu_device = -1 cooling_down = 1 A0_linear B0_expand B0_linear B1_linear C0_expand...
程序段错误
UserWarning: The .grad attribute of a Tensor that is not a leaf Tensor is being accessed. Its .grad attribute won't be populated during autograd.backward(). If you indeed want the .grad...
您好你可以提供一下以及训练好的模型,供下载吗,DeiT-T
ncnn中的Mat类,是在channel维度上对齐的,有一个c_step参数。 我在b站看过作者的讲解,说这个对齐主要考虑到ARMV7和openmp下的cache miss导致的性能原因。 想请问一下,这个cache miss是怎么理解,我想学习一下这个内存对齐对于性能的影响。 感谢!