caffe模型报错

I Set caffe proto to caffe
I Load caffe model ./Li_C_deploy_new.prototxt
I Parsing net parameters ...
D import clients finished
I Parsing layer parameters ...
D Convert layer input
D Convert layer conv1
D Convert layer slice1
D Convert layer etlwise1
D Convert layer pool1
D Convert layer conv2a
D Convert layer slice2a
D Convert layer etlwise2a
D Convert layer conv2
D Convert layer slice2
D Convert layer etlwise2
D Convert layer pool2
D Convert layer conv3a
D Convert layer slice3a
D Convert layer etlwise3a
D Convert layer conv3
D Convert layer slice3
D Convert layer etlwise3
D Convert layer pool3
D Convert layer conv4a
D Convert layer slice4a
D Convert layer etlwise4a
D Convert layer conv4
D Convert layer slice4
D Convert layer etlwise4
D Convert layer conv5a
D Convert layer slice5a
D Convert layer etlwise5a
D Convert layer conv5
D Convert layer slice5
D Convert layer etlwise5
D Convert layer pool4
D Convert layer fc1
D Convert layer slice_fc1
D Convert layer etlwise_fc1
D Convert layer drop1
D Convert layer fc2
D Convert layer softmax
I Parsing connections ...
D Connect: input_0,0 to conv1_1,0
D Connect: conv1_1,0 to slice1_2,0
D Connect: slice1_2,0 to etlwise1_3,0
D Connect: slice1_2,1 to etlwise1_3,1
D Connect: etlwise1_3,0 to pool1_4,0
D Connect: pool1_4,0 to conv2a_5,0
D Connect: conv2a_5,0 to slice2a_6,0
D Connect: slice2a_6,0 to etlwise2a_7,0
D Connect: slice2a_6,1 to etlwise2a_7,1
D Connect: etlwise2a_7,0 to conv2_8,0
D Connect: conv2_8,0 to slice2_9,0
D Connect: slice2_9,0 to etlwise2_10,0
D Connect: slice2_9,1 to etlwise2_10,1
D Connect: etlwise2_10,0 to pool2_11,0
D Connect: pool2_11,0 to conv3a_12,0
D Connect: conv3a_12,0 to slice3a_13,0
D Connect: slice3a_13,0 to etlwise3a_14,0
D Connect: slice3a_13,1 to etlwise3a_14,1
D Connect: etlwise3a_14,0 to conv3_15,0
D Connect: conv3_15,0 to slice3_16,0
D Connect: slice3_16,0 to etlwise3_17,0
D Connect: slice3_16,1 to etlwise3_17,1
D Connect: etlwise3_17,0 to pool3_18,0
D Connect: pool3_18,0 to conv4a_19,0
D Connect: conv4a_19,0 to slice4a_20,0
D Connect: slice4a_20,0 to etlwise4a_21,0
D Connect: slice4a_20,1 to etlwise4a_21,1
D Connect: etlwise4a_21,0 to conv4_22,0
D Connect: conv4_22,0 to slice4_23,0
D Connect: slice4_23,0 to etlwise4_24,0
D Connect: slice4_23,1 to etlwise4_24,1
D Connect: etlwise4_24,0 to conv5a_25,0
D Connect: conv5a_25,0 to slice5a_26,0
D Connect: slice5a_26,0 to etlwise5a_27,0
D Connect: slice5a_26,1 to etlwise5a_27,1
D Connect: etlwise5a_27,0 to conv5_28,0
D Connect: conv5_28,0 to slice5_29,0
D Connect: slice5_29,0 to etlwise5_30,0
D Connect: slice5_29,1 to etlwise5_30,1
D Connect: etlwise5_30,0 to pool4_31,0
D Connect: pool4_31,0 to fc1_32,0
D Connect: fc1_32,0 to slice_fc1_33,0
D Connect: slice_fc1_33,0 to etlwise_fc1_34,0
D Connect: slice_fc1_33,1 to etlwise_fc1_34,1
D Connect: etlwise_fc1_34,0 to drop1_35,0
D Connect: drop1_35,0 to fc2_36,0
D Connect: fc2_36,0 to softmax_37,0
D Connect: softmax_37,0 to output_38,0,
I Load net complete.
D Process input_0 ...
D RKNN output shape(input): (0 128 128 1)
D Process conv1_1 ...
D RKNN output shape(convolution): (0 128 128 96)
D Process slice1_2 ...
D RKNN output shape(split): (0 128 128 48) (0 128 128 48)
D Process etlwise1_3 ...
D RKNN output shape(eltwise): (0 128 128 48)
D Process pool1_4 ...
D RKNN output shape(Pooling): (0 64 64 48)
D Process conv2a_5 ...
D RKNN output shape(convolution): (0 64 64 96)
D Process slice2a_6 ...
D RKNN output shape(split): (0 64 64 48) (0 64 64 48)
D Process etlwise2a_7 ...
D RKNN output shape(eltwise): (0 64 64 48)
D Process conv2_8 ...
D RKNN output shape(convolution): (0 64 64 192)
D Process slice2_9 ...
D RKNN output shape(split): (0 64 64 96) (0 64 64 96)
D Process etlwise2_10 ...
D RKNN output shape(eltwise): (0 64 64 96)
D Process pool2_11 ...
D RKNN output shape(Pooling): (0 32 32 96)
D Process conv3a_12 ...
D RKNN output shape(convolution): (0 32 32 192)
D Process slice3a_13 ...
D RKNN output shape(split): (0 32 32 96) (0 32 32 96)
D Process etlwise3a_14 ...
D RKNN output shape(eltwise): (0 32 32 96)
D Process conv3_15 ...
D RKNN output shape(convolution): (0 32 32 384)
D Process slice3_16 ...
D RKNN output shape(split): (0 32 32 192) (0 32 32 192)
D Process etlwise3_17 ...
D RKNN output shape(eltwise): (0 32 32 192)
D Process pool3_18 ...
D RKNN output shape(Pooling): (0 16 16 192)
D Process conv4a_19 ...
D RKNN output shape(convolution): (0 16 16 384)
D Process slice4a_20 ...
D RKNN output shape(split): (0 16 16 192) (0 16 16 192)
D Process etlwise4a_21 ...
D RKNN output shape(eltwise): (0 16 16 192)
D Process conv4_22 ...
D RKNN output shape(convolution): (0 16 16 256)
D Process slice4_23 ...
D RKNN output shape(split): (0 16 16 128) (0 16 16 128)
D Process etlwise4_24 ...
D RKNN output shape(eltwise): (0 16 16 128)
D Process conv5a_25 ...
D RKNN output shape(convolution): (0 16 16 256)
D Process slice5a_26 ...
D RKNN output shape(split): (0 16 16 128) (0 16 16 128)
D Process etlwise5a_27 ...
D RKNN output shape(eltwise): (0 16 16 128)
D Process conv5_28 ...
D RKNN output shape(convolution): (0 16 16 256)
D Process slice5_29 ...
D RKNN output shape(split): (0 16 16 128) (0 16 16 128)
D Process etlwise5_30 ...
D RKNN output shape(eltwise): (0 16 16 128)
D Process pool4_31 ...
D RKNN output shape(Pooling): (0 8 8 128)
D Process fc1_32 ...
D RKNN output shape(fullconnect): (0 512)
D Process slice_fc1_33 ...
D RKNN output shape(split): (0 256) (0 256)
D Process etlwise_fc1_34 ...
D RKNN output shape(eltwise): (0 256)
D Process drop1_35 ...
D RKNN output shape(dropout): (0 256)
D Process fc2_36 ...
D RKNN output shape(fullconnect): (0 99891)
D Process softmax_37 ...
D RKNN output shape(softmax): (0 99891)
D Process output_38 ...
D RKNN output shape(output): (0 99891)
I Build DeepFace_set003_net complete.
I Load blobs from caffe model ./LightenedCNN_C_new.caffemodel
I Parsing net blobs ...
D Load blobs of conv1
D Load blobs of conv2a
D Load blobs of conv2
D Load blobs of conv3a
D Load blobs of conv3
D Load blobs of conv4a
D Load blobs of conv4
D Load blobs of conv5a
D Load blobs of conv5
D Load blobs of fc1
W Skip blobs of fc2_ms
I Load blobs complete.
D Optimizing network with force_1d_tensor, swapper, merge_layer, auto_fill_bn, avg_pool_transform, auto_fill_zero_bias, proposal_opt_import, import_strip_op
done
--> Building model
D import clients finished
I Loading network...
I Load net...
D Load layer input_0 ...
D Load layer conv1_1 ...
D Load layer slice1_2 ...
D Load layer etlwise1_3 ...
D Load layer pool1_4 ...
D Load layer conv2a_5 ...
D Load layer slice2a_6 ...
D Load layer etlwise2a_7 ...
D Load layer conv2_8 ...
D Load layer slice2_9 ...
D Load layer etlwise2_10 ...
D Load layer pool2_11 ...
D Load layer conv3a_12 ...
D Load layer slice3a_13 ...
D Load layer etlwise3a_14 ...
D Load layer conv3_15 ...
D Load layer slice3_16 ...
D Load layer etlwise3_17 ...
D Load layer pool3_18 ...
D Load layer conv4a_19 ...
D Load layer slice4a_20 ...
D Load layer etlwise4a_21 ...
D Load layer conv4_22 ...
D Load layer slice4_23 ...
D Load layer etlwise4_24 ...
D Load layer conv5a_25 ...
D Load layer slice5a_26 ...
D Load layer etlwise5a_27 ...
D Load layer conv5_28 ...
D Load layer slice5_29 ...
D Load layer etlwise5_30 ...
D Load layer pool4_31 ...
D Load layer fc1_32 ...
D Load layer slice_fc1_33 ...
D Load layer etlwise_fc1_34 ...
D Load layer drop1_35 ...
D Load layer fc2_36 ...
D Load layer softmax_37 ...
D Load layer output_38 ...
I Load net complete...
I Load data...
D Process input_0 ...
D RKNN output shape(input): (0 128 128 1)
D Process conv1_1 ...
D RKNN output shape(convolution): (0 128 128 96)
D Process slice1_2 ...
D RKNN output shape(split): (0 128 128 48) (0 128 128 48)
D Process etlwise1_3 ...
D RKNN output shape(eltwise): (0 128 128 48)
D Process pool1_4 ...
D RKNN output shape(Pooling): (0 64 64 48)
D Process conv2a_5 ...
D RKNN output shape(convolution): (0 64 64 96)
D Process slice2a_6 ...
D RKNN output shape(split): (0 64 64 48) (0 64 64 48)
D Process etlwise2a_7 ...
D RKNN output shape(eltwise): (0 64 64 48)
D Process conv2_8 ...
D RKNN output shape(convolution): (0 64 64 192)
D Process slice2_9 ...
D RKNN output shape(split): (0 64 64 96) (0 64 64 96)
D Process etlwise2_10 ...
D RKNN output shape(eltwise): (0 64 64 96)
D Process pool2_11 ...
D RKNN output shape(Pooling): (0 32 32 96)
D Process conv3a_12 ...
D RKNN output shape(convolution): (0 32 32 192)
D Process slice3a_13 ...
D RKNN output shape(split): (0 32 32 96) (0 32 32 96)
D Process etlwise3a_14 ...
D RKNN output shape(eltwise): (0 32 32 96)
D Process conv3_15 ...
D RKNN output shape(convolution): (0 32 32 384)
D Process slice3_16 ...
D RKNN output shape(split): (0 32 32 192) (0 32 32 192)
D Process etlwise3_17 ...
D RKNN output shape(eltwise): (0 32 32 192)
D Process pool3_18 ...
D RKNN output shape(Pooling): (0 16 16 192)
D Process conv4a_19 ...
D RKNN output shape(convolution): (0 16 16 384)
D Process slice4a_20 ...
D RKNN output shape(split): (0 16 16 192) (0 16 16 192)
D Process etlwise4a_21 ...
D RKNN output shape(eltwise): (0 16 16 192)
D Process conv4_22 ...
D RKNN output shape(convolution): (0 16 16 256)
D Process slice4_23 ...
D RKNN output shape(split): (0 16 16 128) (0 16 16 128)
D Process etlwise4_24 ...
D RKNN output shape(eltwise): (0 16 16 128)
D Process conv5a_25 ...
D RKNN output shape(convolution): (0 16 16 256)
D Process slice5a_26 ...
D RKNN output shape(split): (0 16 16 128) (0 16 16 128)
D Process etlwise5a_27 ...
D RKNN output shape(eltwise): (0 16 16 128)
D Process conv5_28 ...
D RKNN output shape(convolution): (0 16 16 256)
D Process slice5_29 ...
D RKNN output shape(split): (0 16 16 128) (0 16 16 128)
D Process etlwise5_30 ...
D RKNN output shape(eltwise): (0 16 16 128)
D Process pool4_31 ...
D RKNN output shape(Pooling): (0 8 8 128)
D Process fc1_32 ...
D RKNN output shape(fullconnect): (0 512)
D Process slice_fc1_33 ...
D RKNN output shape(split): (0 256) (0 256)
D Process etlwise_fc1_34 ...
D RKNN output shape(eltwise): (0 256)
D Process drop1_35 ...
D RKNN output shape(dropout): (0 256)
D Process fc2_36 ...
D RKNN output shape(fullconnect): (0 99891)
D Process softmax_37 ...
D RKNN output shape(softmax): (0 99891)
D Process output_38 ...
D RKNN output shape(output): (0 99891)
I Build DeepFace_set003_net complete.
I Initialzing network optimizer by Default ...
D Optimizing network with add_lstmunit_io, auto_fill_zero_bias, conv_kernel_transform, twod_op_transform, conv_1xn_transform, strip_op, extend_add_to_conv2d, extend_fc_to_conv2d, extend_unstack_split, swapper, merge_layer, transform_layer, proposal_opt, strip_op, auto_fill_reshape_zero, adjust_output_attrs
D Strip layer drop1_35(dropout)
D Transform conv1_1 to convolutionrelu.
D Transform etlwise1_3 to max.
D Transform conv2a_5 to convolutionrelu.
D Transform etlwise2a_7 to max.
D Transform conv2_8 to convolutionrelu.
D Transform etlwise2_10 to max.
D Transform conv3a_12 to convolutionrelu.
D Transform etlwise3a_14 to max.
D Transform conv3_15 to convolutionrelu.
D Transform etlwise3_17 to max.
D Transform conv4a_19 to convolutionrelu.
D Transform etlwise4a_21 to max.
D Transform conv4_22 to convolutionrelu.
D Transform etlwise4_24 to max.
D Transform conv5a_25 to convolutionrelu.
D Transform etlwise5a_27 to max.
D Transform conv5_28 to convolutionrelu.
D Transform etlwise5_30 to max.
D Transform fc1_32 to fullconnectrelu.
D Transform etlwise_fc1_34 to max.
D Transform fc2_36 to fullconnectrelu.
D Optimizing network with c2drv_convert_axis, c2drv_convert_shape, c2drv_convert_array, c2drv_cast_dtype
I Building data ...
D Packing trans_conv1_1 ...
D Packing trans_conv2_8 ...
D Packing trans_conv2a_5 ...
D Packing trans_conv3_15 ...
D Packing trans_conv3a_12 ...
D Packing trans_conv4_22 ...
D Packing trans_conv4a_19 ...
D Packing trans_conv5_28 ...
D Packing trans_conv5a_25 ...
D Packing trans_fc1_32 ...
D nn_param.conv2d.ksize[0]=5
D nn_param.conv2d.ksize[1]=5
D nn_param.conv2d.weights=96
D nn_param.conv2d.stride[0]=1
D nn_param.conv2d.stride[1]=1
D nn_param.conv2d.pad[0]=2
D nn_param.conv2d.pad[1]=2
D nn_param.conv2d.pad[2]=2
D nn_param.conv2d.pad[3]=2
D nn_param.conv2d.group=1
D nn_param.conv2d.dilation[0]=1
D nn_param.conv2d.dilation[1]=1
D nn_param.conv2d.multiplier=0
D vx_param.has_relu=FALSE
D vx_param.overflow_policy=VX_CONVERT_POLICY_WRAP
D vx_param.rounding_policy=VX_ROUND_POLICY_TO_ZERO
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_FLOOR
D nn_param.split.axis=2
D nn_param.split.slices=slices_1
D nn_param.split.slices_num=2
D nn_param.pool.ksize[0]=2
D nn_param.pool.ksize[1]=2
D nn_param.pool.stride[0]=2
D nn_param.pool.stride[1]=2
D nn_param.pool.pad[0]=0
D nn_param.pool.pad[1]=0
D nn_param.pool.pad[2]=0
D nn_param.pool.pad[3]=0
D nn_param.pool.type=VX_CONVOLUTIONAL_NETWORK_POOLING_MAX
D nn_param.pool.round_type=RK_NN_ROUND_CEIL
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_CEILING
D nn_param.conv2d.ksize[0]=1
D nn_param.conv2d.ksize[1]=1
D nn_param.conv2d.weights=96
D nn_param.conv2d.stride[0]=1
D nn_param.conv2d.stride[1]=1
D nn_param.conv2d.pad[0]=0
D nn_param.conv2d.pad[1]=0
D nn_param.conv2d.pad[2]=0
D nn_param.conv2d.pad[3]=0
D nn_param.conv2d.group=1
D nn_param.conv2d.dilation[0]=1
D nn_param.conv2d.dilation[1]=1
D nn_param.conv2d.multiplier=0
D vx_param.has_relu=FALSE
D vx_param.overflow_policy=VX_CONVERT_POLICY_WRAP
D vx_param.rounding_policy=VX_ROUND_POLICY_TO_ZERO
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_FLOOR
D nn_param.split.axis=2
D nn_param.split.slices=slices_2
D nn_param.split.slices_num=2
D nn_param.conv2d.ksize[0]=3
D nn_param.conv2d.ksize[1]=3
D nn_param.conv2d.weights=192
D nn_param.conv2d.stride[0]=1
D nn_param.conv2d.stride[1]=1
D nn_param.conv2d.pad[0]=1
D nn_param.conv2d.pad[1]=1
D nn_param.conv2d.pad[2]=1
D nn_param.conv2d.pad[3]=1
D nn_param.conv2d.group=1
D nn_param.conv2d.dilation[0]=1
D nn_param.conv2d.dilation[1]=1
D nn_param.conv2d.multiplier=0
D vx_param.has_relu=FALSE
D vx_param.overflow_policy=VX_CONVERT_POLICY_WRAP
D vx_param.rounding_policy=VX_ROUND_POLICY_TO_ZERO
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_FLOOR
D nn_param.split.axis=2
D nn_param.split.slices=slices_3
D nn_param.split.slices_num=2
D nn_param.pool.ksize[0]=2
D nn_param.pool.ksize[1]=2
D nn_param.pool.stride[0]=2
D nn_param.pool.stride[1]=2
D nn_param.pool.pad[0]=0
D nn_param.pool.pad[1]=0
D nn_param.pool.pad[2]=0
D nn_param.pool.pad[3]=0
D nn_param.pool.type=VX_CONVOLUTIONAL_NETWORK_POOLING_MAX
D nn_param.pool.round_type=RK_NN_ROUND_CEIL
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_CEILING
D nn_param.conv2d.ksize[0]=1
D nn_param.conv2d.ksize[1]=1
D nn_param.conv2d.weights=192
D nn_param.conv2d.stride[0]=1
D nn_param.conv2d.stride[1]=1
D nn_param.conv2d.pad[0]=0
D nn_param.conv2d.pad[1]=0
D nn_param.conv2d.pad[2]=0
D nn_param.conv2d.pad[3]=0
D nn_param.conv2d.group=1
D nn_param.conv2d.dilation[0]=1
D nn_param.conv2d.dilation[1]=1
D nn_param.conv2d.multiplier=0
D vx_param.has_relu=FALSE
D vx_param.overflow_policy=VX_CONVERT_POLICY_WRAP
D vx_param.rounding_policy=VX_ROUND_POLICY_TO_ZERO
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_FLOOR
D nn_param.split.axis=2
D nn_param.split.slices=slices_4
D nn_param.split.slices_num=2
D nn_param.conv2d.ksize[0]=3
D nn_param.conv2d.ksize[1]=3
D nn_param.conv2d.weights=384
D nn_param.conv2d.stride[0]=1
D nn_param.conv2d.stride[1]=1
D nn_param.conv2d.pad[0]=1
D nn_param.conv2d.pad[1]=1
D nn_param.conv2d.pad[2]=1
D nn_param.conv2d.pad[3]=1
D nn_param.conv2d.group=1
D nn_param.conv2d.dilation[0]=1
D nn_param.conv2d.dilation[1]=1
D nn_param.conv2d.multiplier=0
D vx_param.has_relu=FALSE
D vx_param.overflow_policy=VX_CONVERT_POLICY_WRAP
D vx_param.rounding_policy=VX_ROUND_POLICY_TO_ZERO
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_FLOOR
D nn_param.split.axis=2
D nn_param.split.slices=slices_5
D nn_param.split.slices_num=2
D nn_param.pool.ksize[0]=2
D nn_param.pool.ksize[1]=2
D nn_param.pool.stride[0]=2
D nn_param.pool.stride[1]=2
D nn_param.pool.pad[0]=0
D nn_param.pool.pad[1]=0
D nn_param.pool.pad[2]=0
D nn_param.pool.pad[3]=0
D nn_param.pool.type=VX_CONVOLUTIONAL_NETWORK_POOLING_MAX
D nn_param.pool.round_type=RK_NN_ROUND_CEIL
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_CEILING
D nn_param.conv2d.ksize[0]=1
D nn_param.conv2d.ksize[1]=1
D nn_param.conv2d.weights=384
D nn_param.conv2d.stride[0]=1
D nn_param.conv2d.stride[1]=1
D nn_param.conv2d.pad[0]=0
D nn_param.conv2d.pad[1]=0
D nn_param.conv2d.pad[2]=0
D nn_param.conv2d.pad[3]=0
D nn_param.conv2d.group=1
D nn_param.conv2d.dilation[0]=1
D nn_param.conv2d.dilation[1]=1
D nn_param.conv2d.multiplier=0
D vx_param.has_relu=FALSE
D vx_param.overflow_policy=VX_CONVERT_POLICY_WRAP
D vx_param.rounding_policy=VX_ROUND_POLICY_TO_ZERO
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_FLOOR
D nn_param.split.axis=2
D nn_param.split.slices=slices_6
D nn_param.split.slices_num=2
D nn_param.conv2d.ksize[0]=3
D nn_param.conv2d.ksize[1]=3
D nn_param.conv2d.weights=256
D nn_param.conv2d.stride[0]=1
D nn_param.conv2d.stride[1]=1
D nn_param.conv2d.pad[0]=1
D nn_param.conv2d.pad[1]=1
D nn_param.conv2d.pad[2]=1
D nn_param.conv2d.pad[3]=1
D nn_param.conv2d.group=1
D nn_param.conv2d.dilation[0]=1
D nn_param.conv2d.dilation[1]=1
D nn_param.conv2d.multiplier=0
D vx_param.has_relu=FALSE
D vx_param.overflow_policy=VX_CONVERT_POLICY_WRAP
D vx_param.rounding_policy=VX_ROUND_POLICY_TO_ZERO
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_FLOOR
D nn_param.split.axis=2
D nn_param.split.slices=slices_7
D nn_param.split.slices_num=2
D nn_param.conv2d.ksize[0]=1
D nn_param.conv2d.ksize[1]=1
D nn_param.conv2d.weights=256
D nn_param.conv2d.stride[0]=1
D nn_param.conv2d.stride[1]=1
D nn_param.conv2d.pad[0]=0
D nn_param.conv2d.pad[1]=0
D nn_param.conv2d.pad[2]=0
D nn_param.conv2d.pad[3]=0
D nn_param.conv2d.group=1
D nn_param.conv2d.dilation[0]=1
D nn_param.conv2d.dilation[1]=1
D nn_param.conv2d.multiplier=0
D vx_param.has_relu=FALSE
D vx_param.overflow_policy=VX_CONVERT_POLICY_WRAP
D vx_param.rounding_policy=VX_ROUND_POLICY_TO_ZERO
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_FLOOR
D nn_param.split.axis=2
D nn_param.split.slices=slices_8
D nn_param.split.slices_num=2
D nn_param.conv2d.ksize[0]=3
D nn_param.conv2d.ksize[1]=3
D nn_param.conv2d.weights=256
D nn_param.conv2d.stride[0]=1
D nn_param.conv2d.stride[1]=1
D nn_param.conv2d.pad[0]=1
D nn_param.conv2d.pad[1]=1
D nn_param.conv2d.pad[2]=1
D nn_param.conv2d.pad[3]=1
D nn_param.conv2d.group=1
D nn_param.conv2d.dilation[0]=1
D nn_param.conv2d.dilation[1]=1
D nn_param.conv2d.multiplier=0
D vx_param.has_relu=FALSE
D vx_param.overflow_policy=VX_CONVERT_POLICY_WRAP
D vx_param.rounding_policy=VX_ROUND_POLICY_TO_ZERO
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_FLOOR
D nn_param.split.axis=2
D nn_param.split.slices=slices_9
D nn_param.split.slices_num=2
D nn_param.pool.ksize[0]=2
D nn_param.pool.ksize[1]=2
D nn_param.pool.stride[0]=2
D nn_param.pool.stride[1]=2
D nn_param.pool.pad[0]=0
D nn_param.pool.pad[1]=0
D nn_param.pool.pad[2]=0
D nn_param.pool.pad[3]=0
D nn_param.pool.type=VX_CONVOLUTIONAL_NETWORK_POOLING_MAX
D nn_param.pool.round_type=RK_NN_ROUND_CEIL
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_CEILING
D nn_param.fcl.weights=512
D vx_param.has_relu=FALSE
D vx_param.overflow_policy=VX_CONVERT_POLICY_WRAP
D vx_param.rounding_policy=VX_ROUND_POLICY_TO_ZERO
D vx_param.down_scale_size_rounding=VX_CONVOLUTIONAL_NETWORK_DS_SIZE_ROUNDING_FLOOR
D nn_param.split.axis=0
D nn_param.split.slices=slices_10
D nn_param.split.slices_num=2
E Catch exception when building RKNN model!
T Traceback (most recent call last):
T   File "rknn/api/rknn_base.py", line 487, in rknn.api.rknn_base.RKNNBase.build
T   File "rknn/api/rknn_base.py", line 416, in rknn.api.rknn_base.RKNNBase._build
T   File "rknn/base/ovxconfiggenerator.py", line 156, in rknn.base.ovxconfiggenerator.generate_vx_config_from_files
T   File "rknn/base/rknnlib/app/code_generator/vxconfiggenerator.py", line 181, in rknn.base.rknnlib.app.code_generator.vxconfiggenerator.VXConfigGenerator.generate
T   File "rknn/base/rknnlib/app/code_generator/vxconfiggenerator.py", line 240, in rknn.base.rknnlib.app.code_generator.vxconfiggenerator.VXConfigGenerator._generate_node_definitions
T   File "rknn/base/rknnlib/app/code_generator/vxconfiggenerator.py", line 534, in rknn.base.rknnlib.app.code_generator.vxconfiggenerator.VXConfigGenerator._proc_vxapi_map
T   File "rknn/base/rknnlib/app/code_generator/vxnode_mapper.py", line 119, in rknn.base.rknnlib.app.code_generator.vxnode_mapper.VXNodeMapper.proc_node
T   File "rknn/base/rknnlib/app/code_generator/vxnode_mapper.py", line 635, in rknn.base.rknnlib.app.code_generator.vxnode_mapper.VXNodeMapper.vxFullconnectRelu

日志好像不完整，最后面是不是少了几行？麻烦也贴出来以下

这个我自己解决了