nihui
/
ncnn
Not watched
1
0
Fork 0
Code
Releases 46 Wiki evaluate Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain HPC
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.
2769 Commits
5 Branches
354 MB
2 Download
 
 
 
 
 
 
Tree: 057b5bb515
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from '057b5bb515'
${ noResults }
ncnn/tools/pnnx/src/pass_ncnn
nihui aed05aa851
pnnx fuse more function to module (#4351) 		3 years ago
..
F_adaptive_avg_pool1d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_adaptive_avg_pool2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_adaptive_avg_pool3d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_adaptive_max_pool1d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_adaptive_max_pool2d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_adaptive_max_pool3d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_avg_pool1d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_avg_pool2d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_avg_pool3d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_batch_norm.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_conv1d.cpp pnnx fuse more function to module (#4351) 3 years ago
F_conv2d.cpp pnnx fuse more function to module (#4351) 3 years ago
F_conv3d.cpp pnnx fuse more function to module (#4351) 3 years ago
F_conv_transpose1d.cpp pnnx fuse more function to module (#4351) 3 years ago
F_conv_transpose2d.cpp pnnx fuse more function to module (#4351) 3 years ago
F_conv_transpose3d.cpp pnnx fuse more function to module (#4351) 3 years ago
F_elu.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_embedding.cpp pnnx cmdline argument inputshape with type (#3419) 4 years ago
F_fold.cpp implement ncnn fold and unfold (#4326) 3 years ago
F_gelu.cpp pnnx support torch 1.12 (#3981) 4 years ago
F_glu.cpp pnnx glu batchindex aware conversion (#4285) 3 years ago
F_grid_sample.cpp gridsample op support (#4288) 3 years ago
F_group_norm.cpp pnnx fuse more function to module (#4351) 3 years ago
F_hardsigmoid.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_hardswish.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_hardtanh.cpp fix pnnx hardtanh ncnn conversion 4 years ago
F_instance_norm.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_interpolate.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_layer_norm.cpp pnnx fuse more function to module (#4351) 3 years ago
F_leaky_relu.cpp fix pnnx hardtanh ncnn conversion 4 years ago
F_linear.cpp pnnx fuse more function to module (#4351) 3 years ago
F_local_response_norm.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_max_pool1d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_max_pool2d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_max_pool3d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
F_mish.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_normalize.cpp fix pnnx softmax/normalize/slice negative axis conversion to ncnn (#4284) 3 years ago
F_pad.cpp apply code-format changes 4 years ago
F_pixel_shuffle.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_pixel_unshuffle.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_prelu.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_relu.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_relu6.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_selu.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_sigmoid.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_silu.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_softmax.cpp fix pnnx softmax/normalize/slice negative axis conversion to ncnn (#4284) 3 years ago
F_tanh.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_unfold.cpp implement ncnn fold and unfold (#4326) 3 years ago
F_upsample.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_upsample_bilinear.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
F_upsample_nearest.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
Tensor_contiguous.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
Tensor_repeat.cpp layer tile (#3491) 4 years ago
Tensor_reshape.cpp handle reshape 5d with batch index 0 4 years ago
Tensor_slice.cpp fix slice end index, fix fp16 model weight alignment (#4317) 3 years ago
Tensor_view.cpp handle reshape 5d with batch index 0 4 years ago
chain_multi_output.cpp add pnnx ncnn pass for chain dict output (#3836) 4 years ago
chain_multi_output.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
convert_Tensor_select.cpp convert Tensor.select to ncnn, fuse multiheadattention unpack, add vit_b_32 end2end test (#3763) 4 years ago
convert_Tensor_select.h convert Tensor.select to ncnn, fuse multiheadattention unpack, add vit_b_32 end2end test (#3763) 4 years ago
convert_attribute.cpp layernorm for vector and mat along w, pnnx convnext end2end test (#3764) 4 years ago
convert_attribute.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
convert_custom_op.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
convert_custom_op.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
convert_half_to_float.cpp handle pnnx half weight when converting to ncnn 4 years ago
convert_half_to_float.h handle pnnx half weight when converting to ncnn 4 years ago
convert_input.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
convert_input.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
convert_torch_cat.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
convert_torch_cat.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
convert_torch_chunk.cpp convert pnnx clone normal 4 years ago
convert_torch_chunk.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
convert_torch_einsum.cpp implement einsum layer and pnnx conversion (#3768) 4 years ago
convert_torch_einsum.h implement einsum layer and pnnx conversion (#3768) 4 years ago
convert_torch_split.cpp convert some pnnx reduction family 4 years ago
convert_torch_split.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
convert_torch_tensor_split.cpp pnnx convert torch.tensor_split, fuse full dim size slice to tensor_split (#3988) 4 years ago
convert_torch_tensor_split.h pnnx convert torch.tensor_split, fuse full dim size slice to tensor_split (#3988) 4 years ago
convert_torch_unbind.cpp convert pnnx torch.unbind and torch.ones/ones_like family (#3583) 4 years ago
convert_torch_unbind.h convert pnnx torch.unbind and torch.ones/ones_like family (#3583) 4 years ago
eliminate_noop.cpp dynamic convolution weight (#3408) 4 years ago
eliminate_noop.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
eliminate_output.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
eliminate_output.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
eliminate_tail_reshape_permute.cpp eliminate pnnx tail reshape permute 4 years ago
eliminate_tail_reshape_permute.h apply code-format changes 4 years ago
expand_expression.cpp pnnx math operation (#4251) 3 years ago
expand_expression.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_binaryop_eltwise.cpp pnnx fuse binaryop eltwise as weighted sum (#4000) 4 years ago
fuse_binaryop_eltwise.h pnnx fuse binaryop eltwise as weighted sum (#4000) 4 years ago
fuse_convert_shufflechannel_slice.cpp pnnx fuse megvii style shufflechannel slice (#3999) 4 years ago
fuse_convert_shufflechannel_slice.h pnnx fuse megvii style shufflechannel slice (#3999) 4 years ago
fuse_convolution1d_activation.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_convolution1d_activation.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_convolution_activation.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_convolution_activation.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_convolutiondepthwise1d_activation.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_convolutiondepthwise1d_activation.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_convolutiondepthwise_activation.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_convolutiondepthwise_activation.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_deconvolution_activation.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_deconvolution_activation.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_deconvolutiondepthwise_activation.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_deconvolutiondepthwise_activation.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_innerproduct_activation.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_innerproduct_activation.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
fuse_transpose_matmul.cpp convert torch.matmul, eliminate noop pad and identity op, fuse transpose matmul, fuse select to unbind (#3554) 4 years ago
fuse_transpose_matmul.h convert torch.matmul, eliminate noop pad and identity op, fuse transpose matmul, fuse select to unbind (#3554) 4 years ago
insert_reshape_linear.cpp insert reshape for nn.Linear with 4d/5d inputs (#3959) 4 years ago
insert_reshape_linear.h insert reshape for nn.Linear with 4d/5d inputs (#3959) 4 years ago
insert_reshape_pooling.cpp insert reshape for nn.Linear with 4d/5d inputs (#3959) 4 years ago
insert_reshape_pooling.h add reshape before and after pooling 123d with no batch dimension (#3566) 4 years ago
insert_split.cpp implement einsum layer and pnnx conversion (#3768) 4 years ago
insert_split.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_AdaptiveAvgPool1d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_AdaptiveAvgPool2d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_AdaptiveAvgPool3d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_AdaptiveMaxPool1d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_AdaptiveMaxPool2d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_AdaptiveMaxPool3d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_AvgPool1d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_AvgPool2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_AvgPool3d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_BatchNorm1d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_BatchNorm2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_BatchNorm3d.cpp relu3d, batchnorm3d, reshape4d, flatten4d, permute4d (#3397) 4 years ago
nn_ChannelShuffle.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_ConstantPad1d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_ConstantPad2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_ConstantPad3d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_Conv1d.cpp pnnx write padding_mode zeros for conv 4 years ago
nn_Conv2d.cpp pnnx write padding_mode zeros for conv 4 years ago
nn_Conv3d.cpp pnnx write padding_mode zeros for conv 4 years ago
nn_ConvTranspose1d.cpp deconv1d deconv3d (#3584) 4 years ago
nn_ConvTranspose2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_ConvTranspose3d.cpp deconv1d deconv3d (#3584) 4 years ago
nn_ELU.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_Embedding.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_Fold.cpp implement ncnn fold and unfold (#4326) 3 years ago
nn_GELU.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_GLU.cpp pnnx glu batchindex aware conversion (#4285) 3 years ago
nn_GRU.cpp pnnx save ncnn bin with fp16 storage (#3715) 4 years ago
nn_GroupNorm.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_Hardsigmoid.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_Hardswish.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_Hardtanh.cpp convert pnnx clone normal 4 years ago
nn_InstanceNorm2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_LSTM.cpp implement lstm proj_size (#4263) 3 years ago
nn_LayerNorm.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_LeakyReLU.cpp fix pnnx hardtanh ncnn conversion 4 years ago
nn_Linear.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_LocalResponseNorm.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_MaxPool1d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_MaxPool2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_MaxPool3d.cpp initial data structure changes for 3dcnn, conv3d, pooling3d (#3378) 4 years ago
nn_Mish.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_MultiheadAttention.cpp implement MultiheadAttention kdim vdim (#4347) 3 years ago
nn_PReLU.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_PixelShuffle.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_PixelUnshuffle.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_RNN.cpp pnnx save ncnn bin with fp16 storage (#3715) 4 years ago
nn_ReLU.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_ReLU6.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_ReflectionPad1d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_ReflectionPad2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_ReplicationPad1d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_ReplicationPad2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_SELU.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_SiLU.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_Sigmoid.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_Softmax.cpp fix pnnx softmax/normalize/slice negative axis conversion to ncnn (#4284) 3 years ago
nn_Softmax2d.cpp pnnx convert nn.Softmax2d (#4324) 3 years ago
nn_Tanh.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_Unfold.cpp implement ncnn fold and unfold (#4326) 3 years ago
nn_Upsample.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_UpsamplingBilinear2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_UpsamplingNearest2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
nn_ZeroPad2d.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
solve_batch_index.cpp gridsample op support (#4288) 3 years ago
solve_batch_index.h PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
torch_addmm.cpp pnnx convert torch.addmm (#3634) 4 years ago
torch_amax.cpp treat old reduction axes param as failure 4 years ago
torch_amin.cpp treat old reduction axes param as failure 4 years ago
torch_bmm.cpp pnnx torch bmm max min (#3736) 4 years ago
torch_clamp.cpp convert pnnx clone normal 4 years ago
torch_clone.cpp convert pnnx clone normal 4 years ago
torch_flatten.cpp PNNX is an open standard for PyTorch model interoperability (#3262) 4 years ago
torch_logsumexp.cpp treat old reduction axes param as failure 4 years ago
torch_matmul.cpp convert torch.matmul, eliminate noop pad and identity op, fuse transpose matmul, fuse select to unbind (#3554) 4 years ago
torch_max.cpp pnnx torch bmm max min (#3736) 4 years ago
torch_mean.cpp pnnx torch bmm max min (#3736) 4 years ago
torch_min.cpp pnnx torch bmm max min (#3736) 4 years ago
torch_norm.cpp convert torch.norm to ncnn, fix F.normalize vector (#3828) 4 years ago
torch_permute.cpp dynamic convolution weight (#3408) 4 years ago
torch_prod.cpp pnnx torch bmm max min (#3736) 4 years ago
torch_squeeze.cpp squeeze and expanddims 4d (#4346) 3 years ago
torch_sum.cpp pnnx torch bmm max min (#3736) 4 years ago
torch_transpose.cpp convert torch.matmul, eliminate noop pad and identity op, fuse transpose matmul, fuse select to unbind (#3554) 4 years ago
torch_unsqueeze.cpp squeeze and expanddims 4d (#4346) 3 years ago
torchvision_DeformConv2d.cpp Add DeformableConv2D (#4070) 3 years ago
Delete task

Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.

cancel
confirm

No Description

C++ C Python Text Protocol Buffer other