MegEngine

History

Megvii Engine Team b3e54eade1 feat(dnn/bn): use new cudnn BN kernel to support NHWC GitOrigin-RevId: `9d80f2009d`		4 years ago
..
argmxx	chore(mge): update copyright years	5 years ago

cond_take	chore(mge): update copyright years	5 years ago

cv	feat(bazel/windows/xp/sp2/inference): implement inference on windows xp	4 years ago

elemwise	feat(bazel/windows/xp/sp2/inference): implement inference on windows xp	4 years ago

elemwise_multi_type	feat(bazel/windows/xp/sp2/inference): implement inference on windows xp	4 years ago

local	feat(log): opt log	4 years ago

local_share	feat(log): opt log	4 years ago

megcore	feat(imperative): add support for quantized conv transpose2d	4 years ago

pooling	chore(mge): update copyright years	5 years ago

simd_macro	chore(mge): update copyright years	5 years ago

winograd	feat(log): opt log	4 years ago

adaptive_pooling.cpp	chore(mge): update copyright years	5 years ago

add_update.cpp	chore(mge): update copyright years	5 years ago

add_update_helper.h	chore(mge): update copyright years	5 years ago

algo_base.cpp	feat(mgb/dnn): add accuracy_depend_on_batch attribute	4 years ago

algo_base.h	feat(dnn/opencl): add heuristic rule for batched matmul	4 years ago

algo_chooser.h	feat(dnn): add an get_all_algorithms_safe interface	4 years ago

api_cache.h	fix(api_cache): fix serialization for conv_desc	4 years ago

argmxx_helper.h	chore(mge): update copyright years	5 years ago

argsort.cpp	chore(mge): update copyright years	5 years ago

asm_common_defs.h	chore(mge): update copyright years	5 years ago

basic_types.cpp	fix(dnn/cuda): fix elemwise and relayout int4 bug when last shape is 1	4 years ago

batch_conv_bias.cpp	chore(mge): update copyright years	5 years ago

batch_normalization.cpp	feat(dnn/bn): use new cudnn BN kernel to support NHWC	4 years ago

batched_matrix_mul.cpp	feat(log): opt log	4 years ago

check_has_inf.cpp	feat(mgb/opr): add check_has_inf kernel and opr	4 years ago

checksum.cpp	chore(mge): update copyright years	5 years ago

concat_split.cpp	feat(log): opt log	4 years ago

conv_bias.cpp	fix(mgb/gopt): fix testcase for enable nchw64 pass	4 years ago

conv_bias.h	refactor(mgb): make conv handle noncontiguous tensors	4 years ago

conv_pooling.cpp	chore(mge): update copyright years	5 years ago

convolution.cpp	feat(dnn/cuda): add nhwc int8 imma conv and conv fuse typecvt	4 years ago

convolution3d.cpp	refactor(mgb): make convolution3D handle noncontiguous tensors	4 years ago

correlation.cpp	feat(dnn/cuda): add correlation kernel	4 years ago

cpuinfo_arch_vendor.cpp	chore(mge): update copyright years	5 years ago

cpuinfo_arch_vendor.h	chore(mge): update copyright years	5 years ago

cumsum.cpp	chore(mge): update copyright years	5 years ago

cvt_color.cpp	chore(mge): update copyright years	5 years ago

dct.cpp	chore(mge): update copyright years	5 years ago

deformable_conv.cpp	feat(dnn): add heuristic cache	4 years ago

deformable_ps_roi_pooling.cpp	chore(mge): update copyright years	5 years ago

dot.cpp	chore(mge): update copyright years	5 years ago

dtype.cpp	feat(log): opt log	4 years ago

elemwise_helper.cpp	feat(mgb/opr): add lsq opr	4 years ago

elemwise_helper.cuh	chore(mge): update copyright years	5 years ago

eye.cpp	chore(mge): update copyright years	5 years ago

fake_quant.cpp	chore(mge): update copyright years	5 years ago

fill.cpp	feat(dnn): add fill kernel	4 years ago

flag_warn.cpp	feat(imperative): add support for quantized conv transpose2d	4 years ago

flip.cpp	chore(mge): update copyright years	5 years ago

gaussian_blur.cpp	chore(mge): update copyright years	5 years ago

gaussian_blur_helper.h	chore(mge): update copyright years	5 years ago

group_local.cpp	feat(log): opt log	4 years ago

handle.cpp	feat(imperative): add support for quantized conv transpose2d	4 years ago

handle_impl.h	feat(mgb/shuffle): add shuffle opr	4 years ago

hash_ct.h	chore(mge): update copyright years	5 years ago

heuristic_cache.cpp	feat(dnn): add heuristic cache	4 years ago

images2neibs.cpp	feat(mge/module): add python wrapper for unfold	4 years ago

indexing_multi_axis_vec.cpp	chore(mge): update copyright years	5 years ago

indexing_multi_axis_vec_kdef.h	chore(mge): update copyright years	5 years ago

indexing_one_hot.cpp	feat(log): opt log	4 years ago

linspace.cpp	chore(mge): update copyright years	5 years ago

lrn.cpp	chore(mge): update copyright years	5 years ago

lsq.cpp	feat(mgb/opr): add lsq opr	4 years ago

macro_helper.h	chore(mge): update copyright years	5 years ago

macro_helper_epilogue.h	chore(mge): update copyright years	5 years ago

mask_conv.cpp	chore(mge): update copyright years	5 years ago

matrix_inverse.cpp	feat(log): opt log	4 years ago

matrix_mul.cpp	feat(log): opt log	4 years ago

max_tensor_diff.cpp	chore(mge): update copyright years	5 years ago

mesh_indexing.cpp	chore(mge): update copyright years	5 years ago

metahelper.h	chore(mge): update copyright years	5 years ago

named_tensor.cpp	refactor(mgb/gopt): refactor tensor reformat opt pass	4 years ago

nchw_nchwxx_valid.cpp	chore(mge): update copyright years	5 years ago

nchw_nchwxx_valid.h	feat(dnn/arm): nchw_nchw44 conv support 1x1s1	4 years ago

opr_delegate.cpp	chore(mge): update copyright years	5 years ago

opr_delegate.h	feat(bazel/windows/xp/sp2/inference): implement inference on windows xp	4 years ago

opr_trait.h	feat(dnn/bn): use new cudnn BN kernel to support NHWC	4 years ago

padding.cpp	feat(mgb): add padding opr for megbrain	4 years ago

param_pack.cpp	chore(mge): update copyright years	5 years ago

pooling.cpp	feat(dnn/cuda): add cuda int4 pooling	4 years ago

postprocess.h	chore(mge): update copyright years	5 years ago

postprocess_helper.h	chore(mge): update copyright years	5 years ago

powc.cpp	chore(mge): update copyright years	5 years ago

reduce.cpp	chore(mge): update copyright years	5 years ago

reduce_helper.cpp	chore(mge): update copyright years	5 years ago

reduce_helper.h	fix(dnn/reduce): fix reduce_mean o16c32 is incorrect for large tensor	4 years ago

relayout.cpp	chore(mge): update copyright years	5 years ago

relayout_format.cpp	feat(dnn/cuda): add relayout format kernel for nchw <-> nhwc	4 years ago

relayout_helper.h	ci(gitlab-ci): add inline lineno checking in copybara linter	4 years ago

remap.cpp	feat(log): opt log	4 years ago

resize.cpp	feat(mgb): add arm resize nchwxx and naive nearest interp	4 years ago

resize.cuh	refactor(dnn): move arch-dependant code to arch.h	4 years ago

rng.cpp	feat(mgb/shuffle): add shuffle opr	4 years ago

roi_align.cpp	chore(mge): update copyright years	5 years ago

roi_align_helper.h	chore(mge): update copyright years	5 years ago

roi_copy.cpp	chore(mge): update copyright years	5 years ago

roi_pooling.cpp	chore(mge): update copyright years	5 years ago

roi_pooling_helper.h	chore(mge): update copyright years	5 years ago

rotate.cpp	chore(mge): update copyright years	5 years ago

rounding_converter.cuh	refactor(dnn): move arch-dependant code to arch.h	4 years ago

separableConv.cpp	feat(log): opt log	4 years ago

separableFilter.cpp	feat(log): opt log	4 years ago

sliding_window_transpose.cpp	feat(mge): add SlidingWindowTranspose opr	4 years ago

small_vector.cpp	chore(mge): update copyright years	5 years ago

svd.cpp	chore(mge): update copyright years	5 years ago

tensor_format.cpp	feat(bazel/windows/xp/sp2/inference): implement inference on windows xp	4 years ago

tensor_iter.cpp	chore(mge): update copyright years	5 years ago

tensor_remap.cpp	chore(mge): update copyright years	5 years ago

tile_repeat.cpp	chore(mge): update copyright years	5 years ago

tile_repeat_helper.cpp	chore(mge): update copyright years	5 years ago

tile_repeat_helper.h	chore(mge): update copyright years	5 years ago

topk.cpp	chore(mge): update copyright years	5 years ago

tqt.cpp	chore(mge): update copyright years	5 years ago

transpose.cpp	chore(mge): update copyright years	5 years ago

type_cvt.cpp	chore(mge): update copyright years	5 years ago

unroll_macro.h	chore(mge): update copyright years	5 years ago

utils.cpp	fix(dnn/cuda): fix testcase for fallback nchw qs8 conv	4 years ago

utils.cuh	feat(log): opt log	4 years ago

utils.h	feat(bazel/windows/xp/sp2/inference): implement inference on windows xp	4 years ago

version.cpp	chore(mge): update copyright years	5 years ago

version_symbol.h	chore(mge): update copyright years	5 years ago

warp_affine.cpp	feat(log): opt log	4 years ago

warp_common.cpp	chore(mge): update copyright years	5 years ago

warp_common.h	feat(imperative): add support for quantized conv transpose2d	4 years ago

warp_perspective.cpp	feat(dnn/cuda): add nhwc 4bit warp perspective	4 years ago

warp_perspective_helper.cpp	chore(mge): update copyright years	5 years ago

warp_perspective_helper.h	chore(mge): update copyright years	5 years ago

MegEngine 安装包中集成了使用 GPU 运行代码所需的 CUDA 环境，不用区分 CPU 和 GPU 版。如果想要运行 GPU 程序，请确保机器本身配有 GPU 硬件设备并安装好驱动。如果你想体验在云端 GPU 算力平台进行深度学习开发的感觉，欢迎访问 MegStudio 平台

模型部署模型训练

C++ Cuda Python C SVG other

megengine@megvii.com luzzyzhang@gmail.com huangxinda@megvii.com 1041563782@qq.com acdoge.cao@gmail.com xxr@megvii.com 61608041+megvii-mge@users.noreply.github.com wenjuan@megvii.com lijiansong@ict.ac.cn mediosrity@gmail.com huangxinda9434@126.com chenjiahui@megvii.com 44659878+flashrunrun@users.noreply.github.com jia.kai66@gmail.com khj.application@aliyun.com liuke@megvii.com qazazsw@gmail.com 59949790+stoneMo@users.noreply.github.com wkcn@live.cn 942002795@qq.com

Contributors (20) All

Contributors (20)
All