You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
 
nihuini 74276314bb unroll size 4 for conv1x1s1 pack4 7 years ago
..
layer unroll size 4 for conv1x1s1 pack4 7 years ago
CMakeLists.txt shader fp16 packed 7 years ago
allocator.cpp bottom blob memory never alias, reuse blob memory more elegantly relying on refcount 7 years ago
allocator.h move mutex class family to platform.h 7 years ago
benchmark.cpp new int8 implement,better accuracy (#749) 7 years ago
benchmark.h decouple modelbin of different input types, simplify timestamp function 8 years ago
blob.cpp layer holds bottom blob scale, depthwise convolution read group scales 8 years ago
blob.h layer holds bottom blob scale, depthwise convolution read group scales 8 years ago
command.cpp more fix for fp16p, still disabled by default 7 years ago
command.h do not retrieve timestamp availabitliy bits 7 years ago
cpu.cpp Support for Raspberry Pi 3 compilation (#640) 7 years ago
cpu.h add ncnn source qwq 9 years ago
gpu.cpp more fix for fp16p, still disabled by default 7 years ago
gpu.h shader fp16 packed 7 years ago
layer.cpp move vulkan code to subdir, new layer interface create_pipeline and destroy_pipeline for post-loading works 7 years ago
layer.h move vulkan code to subdir, new layer interface create_pipeline and destroy_pipeline for post-loading works 7 years ago
layer_declaration.h.in fix: avoid rebuilding after cmake 7 years ago
layer_registry.h.in fix: avoid rebuilding after cmake 7 years ago
layer_shader_registry.h.in [WIP] vulkan compute (#618) 7 years ago
layer_shader_spv_data.h.in [WIP] vulkan compute (#618) 7 years ago
layer_type.h move layer type enum to separate header 8 years ago
layer_type_enum.h.in fix: avoid rebuilding after cmake 7 years ago
mat.cpp resize bicubic 7 years ago
mat.h Mat row should be elemsize-aware 7 years ago
mat_pixel.cpp the widely used yuv420sp2rgb, the fast approximate version 7 years ago
mat_pixel_resize.cpp resize bicubic 7 years ago
modelbin.cpp core int8 inference, quantize and dequantize, net using flag, caffe2ncnn reads int8 scale table 8 years ago
modelbin.h core int8 inference, quantize and dequantize, net using flag, caffe2ncnn reads int8 scale table 8 years ago
net.cpp more fix for fp16p, still disabled by default 7 years ago
net.h bind wait barrier count member to memory, fix #932 7 years ago
opencv.cpp optional image pixel api, fix #434 8 years ago
opencv.h optional image pixel api, fix #434 8 years ago
paramdict.cpp move vulkan code to subdir, new layer interface create_pipeline and destroy_pipeline for post-loading works 7 years ago
paramdict.h move vulkan code to subdir, new layer interface create_pipeline and destroy_pipeline for post-loading works 7 years ago
pipeline.cpp shader fp16 packed 7 years ago
pipeline.h set fixed local size 7 years ago
platform.h.in move mutex class family to platform.h 7 years ago