Khadas vim3 Convert ONNX model

VIM3
1

Hi, I want to convert a onnx model in khadas Vim3. but I have following results:
./0_import_model.sh
I Current ONNX Model use ir_version 4 opset_version 9
I build output layer attach_Softmax_109:out0
I build output layer attach_Concat_108:out0
I Try match Concat_108:out0
I Match concat_4 [[‘Concat_108’]] [[‘Concat’]] to [[‘concat’]]
I Try match Softmax_109:out0
I Match r_softmax [[‘Softmax_109’]] [[‘Softmax’]] to [[‘softmax’]]
I Try match Reshape_54:out0
I Match r_rsp_v5 [[‘Reshape_54’, ‘Initializer_105’]] [[‘Reshape’, ‘Constant_0’]] to [[‘reshape’]]
I Try match Reshape_76:out0
I Match r_rsp_v5 [[‘Reshape_76’, ‘Initializer_107’]] [[‘Reshape’, ‘Constant_0’]] to [[‘reshape’]]
I Try match Reshape_94:out0
I Match r_rsp_v5 [[‘Reshape_94’, ‘Initializer_109’]] [[‘Reshape’, ‘Constant_0’]] to [[‘reshape’]]
I Try match Reshape_106:out0
I Match r_rsp_v5 [[‘Reshape_106’, ‘Initializer_111’]] [[‘Reshape’, ‘Constant_0’]] to [[‘reshape’]]
I Try match Concat_107:out0
I Match concat_4 [[‘Concat_107’]] [[‘Concat’]] to [[‘concat’]]
I Try match Transpose_53:out0
I Match r_transpose [[‘Transpose_53’]] [[‘Transpose’]] to [[‘permute’]]
I Try match Transpose_75:out0
I Match r_transpose [[‘Transpose_75’]] [[‘Transpose’]] to [[‘permute’]]
I Try match Transpose_93:out0
I Match r_transpose [[‘Transpose_93’]] [[‘Transpose’]] to [[‘permute’]]
I Try match Transpose_105:out0
I Match r_transpose [[‘Transpose_105’]] [[‘Transpose’]] to [[‘permute’]]
I Try match Reshape_49:out0
I Match r_rsp_v5 [[‘Reshape_49’, ‘Initializer_104’]] [[‘Reshape’, ‘Constant_0’]] to [[‘reshape’]]
I Try match Reshape_71:out0
I Match r_rsp_v5 [[‘Reshape_71’, ‘Initializer_106’]] [[‘Reshape’, ‘Constant_0’]] to [[‘reshape’]]
I Try match Reshape_89:out0
I Match r_rsp_v5 [[‘Reshape_89’, ‘Initializer_108’]] [[‘Reshape’, ‘Constant_0’]] to [[‘reshape’]]
I Try match Reshape_103:out0
I Match r_rsp_v5 [[‘Reshape_103’, ‘Initializer_110’]] [[‘Reshape’, ‘Constant_0’]] to [[‘reshape’]]
I Try match Conv_52:out0
I Match r_conv [[‘Conv_52’, ‘Initializer_23’, ‘Initializer_22’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_74:out0
I Match r_conv [[‘Conv_74’, ‘Initializer_27’, ‘Initializer_26’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_92:out0
I Match r_conv [[‘Conv_92’, ‘Initializer_31’, ‘Initializer_30’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_104:out0
I Match r_conv [[‘Conv_104’, ‘Initializer_33’, ‘Initializer_32’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Transpose_48:out0
I Match r_transpose [[‘Transpose_48’]] [[‘Transpose’]] to [[‘permute’]]
I Try match Transpose_70:out0
I Match r_transpose [[‘Transpose_70’]] [[‘Transpose’]] to [[‘permute’]]
I Try match Transpose_88:out0
I Match r_transpose [[‘Transpose_88’]] [[‘Transpose’]] to [[‘permute’]]
I Try match Transpose_102:out0
I Match r_transpose [[‘Transpose_102’]] [[‘Transpose’]] to [[‘permute’]]
I Try match Relu_51:out0
I Match r_relu [[‘Relu_51’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_73:out0
I Match r_relu [[‘Relu_73’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_91:out0
I Match r_relu [[‘Relu_91’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_100:out0
I Match r_relu [[‘Relu_100’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_47:out0
I Match r_conv [[‘Conv_47’, ‘Initializer_3’, ‘Initializer_2’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_69:out0
I Match r_conv [[‘Conv_69’, ‘Initializer_7’, ‘Initializer_6’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_87:out0
I Match r_conv [[‘Conv_87’, ‘Initializer_11’, ‘Initializer_10’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_101:out0
I Match r_conv [[‘Conv_101’, ‘Initializer_13’, ‘Initializer_12’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_50:out0
I Match r_conv [[‘Conv_50’, ‘Initializer_21’, ‘Initializer_20’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_72:out0
I Match r_conv [[‘Conv_72’, ‘Initializer_25’, ‘Initializer_24’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_90:out0
I Match r_conv [[‘Conv_90’, ‘Initializer_29’, ‘Initializer_28’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_99:out0
I Match r_conv [[‘Conv_99’, ‘Initializer_19’, ‘Initializer_18’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_46:out0
I Match r_relu [[‘Relu_46’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_68:out0
I Match r_relu [[‘Relu_68’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_86:out0
I Match r_relu [[‘Relu_86’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_85:out0
I Match r_conv [[‘Conv_85’, ‘Initializer_9’, ‘Initializer_8’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_44:out0
I Match r_relu [[‘Relu_44’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_66:out0
I Match r_relu [[‘Relu_66’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_84:out0
I Match r_relu [[‘Relu_84’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_98:out0
I Match r_relu [[‘Relu_98’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_45:out0
I Match r_conv [[‘Conv_45’, ‘Initializer_1’, ‘Initializer_0’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_67:out0
I Match r_conv [[‘Conv_67’, ‘Initializer_5’, ‘Initializer_4’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_97:out0
I Match r_conv [[‘Conv_97’, ‘Initializer_17’, ‘Initializer_16’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Add_43:out0
I Match r_add [[‘Add_43’]] [[‘Add’]] to [[‘add’]]
I Try match Conv_65:out0
I Match r_conv [[‘Conv_65’, ‘Initializer_94’, ‘Initializer_95’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_83:out0
I Match r_conv [[‘Conv_83’, ‘Initializer_102’, ‘Initializer_103’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_96:out0
I Match r_relu [[‘Relu_96’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_41:out0
I Match r_conv [[‘Conv_41’, ‘Initializer_80’, ‘Initializer_81’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_42:out0
I Match r_conv [[‘Conv_42’, ‘Initializer_82’, ‘Initializer_83’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_64:out0
I Match r_relu [[‘Relu_64’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_82:out0
I Match r_relu [[‘Relu_82’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_95:out0
I Match r_conv [[‘Conv_95’, ‘Initializer_15’, ‘Initializer_14’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Concat_40:out0
I Match concat_3 [[‘Concat_40’]] [[‘Concat’]] to [[‘concat’]]
I Try match Relu_25:out0
I Match r_relu [[‘Relu_25’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_63:out0
I Match r_conv [[‘Conv_63’, ‘Initializer_92’, ‘Initializer_93’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_81:out0
I Match r_conv [[‘Conv_81’, ‘Initializer_100’, ‘Initializer_101’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_80:out0
I Match r_relu [[‘Relu_80’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_29:out0
I Match r_conv [[‘Conv_29’, ‘Initializer_64’, ‘Initializer_65’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_33:out0
I Match r_conv [[‘Conv_33’, ‘Initializer_70’, ‘Initializer_71’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_39:out0
I Match r_conv [[‘Conv_39’, ‘Initializer_78’, ‘Initializer_79’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_24:out0
I Match r_conv [[‘Conv_24’, ‘Initializer_58’, ‘Initializer_59’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_62:out0
I Match r_relu [[‘Relu_62’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_79:out0
I Match r_conv [[‘Conv_79’, ‘Initializer_98’, ‘Initializer_99’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_28:out0
I Match r_relu [[‘Relu_28’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_32:out0
I Match r_relu [[‘Relu_32’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_38:out0
I Match r_relu [[‘Relu_38’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_23:out0
I Match r_relu [[‘Relu_23’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_61:out0
I Match r_conv [[‘Conv_61’, ‘Initializer_90’, ‘Initializer_91’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_78:out0
I Match r_relu [[‘Relu_78’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_27:out0
I Match r_conv [[‘Conv_27’, ‘Initializer_62’, ‘Initializer_63’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_31:out0
I Match r_conv [[‘Conv_31’, ‘Initializer_68’, ‘Initializer_69’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_37:out0
I Match r_conv [[‘Conv_37’, ‘Initializer_76’, ‘Initializer_77’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_22:out0
I Match r_conv [[‘Conv_22’, ‘Initializer_56’, ‘Initializer_57’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_60:out0
I Match r_relu [[‘Relu_60’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_77:out0
I Match r_conv [[‘Conv_77’, ‘Initializer_96’, ‘Initializer_97’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_26:out0
I Match r_conv [[‘Conv_26’, ‘Initializer_60’, ‘Initializer_61’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_30:out0
I Match r_conv [[‘Conv_30’, ‘Initializer_66’, ‘Initializer_67’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_36:out0
I Match r_relu [[‘Relu_36’]] [[‘Relu’]] to [[‘relu’]]
I Try match Relu_21:out0
I Match r_relu [[‘Relu_21’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_59:out0
I Match r_conv [[‘Conv_59’, ‘Initializer_88’, ‘Initializer_89’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_58:out0
I Match r_relu [[‘Relu_58’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_57:out0
I Match r_conv [[‘Conv_57’, ‘Initializer_86’, ‘Initializer_87’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_56:out0
I Match r_relu [[‘Relu_56’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_35:out0
I Match r_conv [[‘Conv_35’, ‘Initializer_74’, ‘Initializer_75’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_20:out0
I Match r_conv [[‘Conv_20’, ‘Initializer_54’, ‘Initializer_55’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_55:out0
I Match r_conv [[‘Conv_55’, ‘Initializer_84’, ‘Initializer_85’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Conv_34:out0
I Match r_conv [[‘Conv_34’, ‘Initializer_72’, ‘Initializer_73’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_19:out0
I Match r_relu [[‘Relu_19’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_18:out0
I Match r_conv [[‘Conv_18’, ‘Initializer_52’, ‘Initializer_53’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_17:out0
I Match r_relu [[‘Relu_17’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_16:out0
I Match r_conv [[‘Conv_16’, ‘Initializer_50’, ‘Initializer_51’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_15:out0
I Match r_relu [[‘Relu_15’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_14:out0
I Match r_conv [[‘Conv_14’, ‘Initializer_48’, ‘Initializer_49’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_13:out0
I Match r_relu [[‘Relu_13’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_12:out0
I Match r_conv [[‘Conv_12’, ‘Initializer_46’, ‘Initializer_47’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_11:out0
I Match r_relu [[‘Relu_11’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_10:out0
I Match r_conv [[‘Conv_10’, ‘Initializer_44’, ‘Initializer_45’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_9:out0
I Match r_relu [[‘Relu_9’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_8:out0
I Match r_conv [[‘Conv_8’, ‘Initializer_42’, ‘Initializer_43’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_7:out0
I Match r_relu [[‘Relu_7’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_6:out0
I Match r_conv [[‘Conv_6’, ‘Initializer_40’, ‘Initializer_41’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_5:out0
I Match r_relu [[‘Relu_5’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_4:out0
I Match r_conv [[‘Conv_4’, ‘Initializer_38’, ‘Initializer_39’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_3:out0
I Match r_relu [[‘Relu_3’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_2:out0
I Match r_conv [[‘Conv_2’, ‘Initializer_36’, ‘Initializer_37’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I Try match Relu_1:out0
I Match r_relu [[‘Relu_1’]] [[‘Relu’]] to [[‘relu’]]
I Try match Conv_0:out0
I Match r_conv [[‘Conv_0’, ‘Initializer_34’, ‘Initializer_35’]] [[‘Conv’, ‘Constant_0’, ‘Constant_1’]] to [[‘convolution’]]
I build input layer Input_0:out0
D connect Reshape_54_4 0 ~ Concat_108_2 0
D connect Reshape_76_5 0 ~ Concat_108_2 1
D connect Reshape_94_6 0 ~ Concat_108_2 2
D connect Reshape_106_7 0 ~ Concat_108_2 3
D connect Concat_107_8 0 ~ Softmax_109_3 0
D connect Transpose_53_9 0 ~ Reshape_54_4 0
D connect Transpose_75_10 0 ~ Reshape_76_5 0
D connect Transpose_93_11 0 ~ Reshape_94_6 0
D connect Transpose_105_12 0 ~ Reshape_106_7 0
D connect Reshape_49_13 0 ~ Concat_107_8 0
D connect Reshape_71_14 0 ~ Concat_107_8 1
D connect Reshape_89_15 0 ~ Concat_107_8 2
D connect Reshape_103_16 0 ~ Concat_107_8 3
D connect Conv_52_17 0 ~ Transpose_53_9 0
D connect Conv_74_18 0 ~ Transpose_75_10 0
D connect Conv_92_19 0 ~ Transpose_93_11 0
D connect Conv_104_20 0 ~ Transpose_105_12 0
D connect Transpose_48_21 0 ~ Reshape_49_13 0
D connect Transpose_70_22 0 ~ Reshape_71_14 0
D connect Transpose_88_23 0 ~ Reshape_89_15 0
D connect Transpose_102_24 0 ~ Reshape_103_16 0
D connect Relu_51_25 0 ~ Conv_52_17 0
D connect Relu_73_26 0 ~ Conv_74_18 0
D connect Relu_91_27 0 ~ Conv_92_19 0
D connect Relu_100_28 0 ~ Conv_104_20 0
D connect Conv_47_29 0 ~ Transpose_48_21 0
D connect Conv_69_30 0 ~ Transpose_70_22 0
D connect Conv_87_31 0 ~ Transpose_88_23 0
D connect Conv_101_32 0 ~ Transpose_102_24 0
D connect Conv_50_33 0 ~ Relu_51_25 0
D connect Conv_72_34 0 ~ Relu_73_26 0
D connect Conv_90_35 0 ~ Relu_91_27 0
D connect Conv_99_36 0 ~ Relu_100_28 0
D connect Relu_46_37 0 ~ Conv_47_29 0
D connect Relu_68_38 0 ~ Conv_69_30 0
D connect Relu_86_39 0 ~ Conv_87_31 0
D connect Relu_100_28 0 ~ Conv_101_32 0
D connect Relu_44_41 0 ~ Conv_50_33 0
D connect Relu_66_42 0 ~ Conv_72_34 0
D connect Relu_84_43 0 ~ Conv_90_35 0
D connect Relu_98_44 0 ~ Conv_99_36 0
D connect Conv_45_45 0 ~ Relu_46_37 0
D connect Conv_67_46 0 ~ Relu_68_38 0
D connect Conv_85_40 0 ~ Relu_86_39 0
D connect Relu_84_43 0 ~ Conv_85_40 0
D connect Add_43_48 0 ~ Relu_44_41 0
D connect Conv_65_49 0 ~ Relu_66_42 0
D connect Conv_83_50 0 ~ Relu_84_43 0
D connect Conv_97_47 0 ~ Relu_98_44 0
D connect Relu_44_41 0 ~ Conv_45_45 0
D connect Relu_66_42 0 ~ Conv_67_46 0
D connect Relu_96_51 0 ~ Conv_97_47 0
D connect Conv_41_52 0 ~ Add_43_48 0
D connect Conv_42_53 0 ~ Add_43_48 1
D connect Relu_64_54 0 ~ Conv_65_49 0
D connect Relu_82_55 0 ~ Conv_83_50 0
D connect Conv_95_56 0 ~ Relu_96_51 0
D connect Concat_40_57 0 ~ Conv_41_52 0
D connect Relu_25_58 0 ~ Conv_42_53 0
D connect Conv_63_59 0 ~ Relu_64_54 0
D connect Conv_81_60 0 ~ Relu_82_55 0
D connect Relu_84_43 0 ~ Conv_95_56 0
D connect Conv_29_62 0 ~ Concat_40_57 0
D connect Conv_33_63 0 ~ Concat_40_57 1
D connect Conv_39_64 0 ~ Concat_40_57 2
D connect Conv_24_65 0 ~ Relu_25_58 0
D connect Relu_62_66 0 ~ Conv_63_59 0
D connect Relu_80_61 0 ~ Conv_81_60 0
D connect Conv_79_67 0 ~ Relu_80_61 0
D connect Relu_28_68 0 ~ Conv_29_62 0
D connect Relu_32_69 0 ~ Conv_33_63 0
D connect Relu_38_70 0 ~ Conv_39_64 0
D connect Relu_23_71 0 ~ Conv_24_65 0
D connect Conv_61_72 0 ~ Relu_62_66 0
D connect Relu_78_73 0 ~ Conv_79_67 0
D connect Conv_27_74 0 ~ Relu_28_68 0
D connect Conv_31_75 0 ~ Relu_32_69 0
D connect Conv_37_76 0 ~ Relu_38_70 0
D connect Conv_22_77 0 ~ Relu_23_71 0
D connect Relu_60_78 0 ~ Conv_61_72 0
D connect Conv_77_79 0 ~ Relu_78_73 0
D connect Conv_26_80 0 ~ Conv_27_74 0
D connect Conv_30_81 0 ~ Conv_31_75 0
D connect Relu_36_82 0 ~ Conv_37_76 0
D connect Relu_21_83 0 ~ Conv_22_77 0
D connect Conv_59_84 0 ~ Relu_60_78 0
D connect Relu_66_42 0 ~ Conv_77_79 0
D connect Relu_25_58 0 ~ Conv_26_80 0
D connect Relu_25_58 0 ~ Conv_30_81 0
D connect Conv_35_88 0 ~ Relu_36_82 0
D connect Conv_20_89 0 ~ Relu_21_83 0
D connect Relu_58_85 0 ~ Conv_59_84 0
D connect Conv_57_86 0 ~ Relu_58_85 0
D connect Relu_56_87 0 ~ Conv_57_86 0
D connect Conv_55_90 0 ~ Relu_56_87 0
D connect Conv_34_91 0 ~ Conv_35_88 0
D connect Relu_19_92 0 ~ Conv_20_89 0
D connect Relu_44_41 0 ~ Conv_55_90 0
D connect Relu_25_58 0 ~ Conv_34_91 0
D connect Conv_18_93 0 ~ Relu_19_92 0
D connect Relu_17_94 0 ~ Conv_18_93 0
D connect Conv_16_95 0 ~ Relu_17_94 0
D connect Relu_15_96 0 ~ Conv_16_95 0
D connect Conv_14_97 0 ~ Relu_15_96 0
D connect Relu_13_98 0 ~ Conv_14_97 0
D connect Conv_12_99 0 ~ Relu_13_98 0
D connect Relu_11_100 0 ~ Conv_12_99 0
D connect Conv_10_101 0 ~ Relu_11_100 0
D connect Relu_9_102 0 ~ Conv_10_101 0
D connect Conv_8_103 0 ~ Relu_9_102 0
D connect Relu_7_104 0 ~ Conv_8_103 0
D connect Conv_6_105 0 ~ Relu_7_104 0
D connect Relu_5_106 0 ~ Conv_6_105 0
D connect Conv_4_107 0 ~ Relu_5_106 0
D connect Relu_3_108 0 ~ Conv_4_107 0
D connect Conv_2_109 0 ~ Relu_3_108 0
D connect Relu_1_110 0 ~ Conv_2_109 0
D connect Conv_0_111 0 ~ Relu_1_110 0
D connect input_112 0 ~ Conv_0_111 0
D connect Softmax_109_3 0 ~ attach_Softmax_109/out0_0 0
D connect Concat_108_2 0 ~ attach_Concat_108/out0_1 0
2020-04-05 21:30:04.405364: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
D Process input_112 …
D Acuity output shape(input): (1 240 320 3)
D Process Conv_0_111 …
D Acuity output shape(convolution): (1 120 160 16)
D Process Relu_1_110 …
D Acuity output shape(relu): (1 120 160 16)
D Process Conv_2_109 …
D Acuity output shape(convolution): (1 120 160 16)
D Process Relu_3_108 …
D Acuity output shape(relu): (1 120 160 16)
D Process Conv_4_107 …
D Acuity output shape(convolution): (1 120 160 32)
D Process Relu_5_106 …
D Acuity output shape(relu): (1 120 160 32)
D Process Conv_6_105 …
D Acuity output shape(convolution): (1 60 80 32)
D Process Relu_7_104 …
D Acuity output shape(relu): (1 60 80 32)
D Process Conv_8_103 …
D Acuity output shape(convolution): (1 60 80 32)
D Process Relu_9_102 …
D Acuity output shape(relu): (1 60 80 32)
D Process Conv_10_101 …
D Acuity output shape(convolution): (1 60 80 32)
D Process Relu_11_100 …
D Acuity output shape(relu): (1 60 80 32)
D Process Conv_12_99 …
D Acuity output shape(convolution): (1 60 80 32)
D Process Relu_13_98 …
D Acuity output shape(relu): (1 60 80 32)
D Process Conv_14_97 …
D Acuity output shape(convolution): (1 30 40 32)
D Process Relu_15_96 …
D Acuity output shape(relu): (1 30 40 32)
D Process Conv_16_95 …
D Acuity output shape(convolution): (1 30 40 64)
D Process Relu_17_94 …
D Acuity output shape(relu): (1 30 40 64)
D Process Conv_18_93 …
D Acuity output shape(convolution): (1 30 40 64)
D Process Relu_19_92 …
D Acuity output shape(relu): (1 30 40 64)
D Process Conv_20_89 …
D Acuity output shape(convolution): (1 30 40 64)
D Process Relu_21_83 …
D Acuity output shape(relu): (1 30 40 64)
D Process Conv_22_77 …
D Acuity output shape(convolution): (1 30 40 64)
D Process Relu_23_71 …
D Acuity output shape(relu): (1 30 40 64)
D Process Conv_24_65 …
D Acuity output shape(convolution): (1 30 40 64)
D Process Relu_25_58 …
D Acuity output shape(relu): (1 30 40 64)
D Process Conv_26_80 …
D Acuity output shape(convolution): (1 30 40 8)
D Process Conv_27_74 …
D Acuity output shape(convolution): (1 30 40 16)
D Process Relu_28_68 …
D Acuity output shape(relu): (1 30 40 16)
D Process Conv_29_62 …
D Acuity output shape(convolution): (1 32 42 16)
D Process Conv_30_81 …
D Acuity output shape(convolution): (1 30 40 8)
D Process Conv_31_75 …
D Acuity output shape(convolution): (1 30 40 16)
D Process Relu_32_69 …
D Acuity output shape(relu): (1 30 40 16)
D Process Conv_33_63 …
D Acuity output shape(convolution): (1 34 44 16)
D Process Conv_34_91 …
D Acuity output shape(convolution): (1 30 40 8)
D Process Conv_35_88 …
D Acuity output shape(convolution): (1 30 40 12)
D Process Relu_36_82 …
D Acuity output shape(relu): (1 30 40 12)
D Process Conv_37_76 …
D Acuity output shape(convolution): (1 30 40 16)
D Process Relu_38_70 …
D Acuity output shape(relu): (1 30 40 16)
D Process Conv_39_64 …
D Acuity output shape(convolution): (1 38 48 16)
D Process Concat_40_57 …
D Acuity output shape(concat): (1 32 42 48)
D Process Conv_41_52 …
D Acuity output shape(convolution): (1 32 42 64)
D Process Conv_42_53 …
D Acuity output shape(convolution): (1 30 40 64)
D Process Add_43_48 …
D Acuity output shape(add): (1 32 42 64)
D Process Relu_44_41 …
D Acuity output shape(relu): (1 32 42 64)
D Process Conv_45_45 …
D Acuity output shape(convolution): (1 32 42 64)
D Process Relu_46_37 …
D Acuity output shape(relu): (1 32 42 64)
D Process Conv_47_29 …
D Acuity output shape(convolution): (1 32 42 6)
D Process Transpose_48_21 …
Traceback (most recent call last):
File “convertonnx.py”, line 25, in
File “convertonnx.py”, line 20, in main
File “acuitylib/app/importer/import_onnx.py”, line 49, in run
File “acuitylib/acuitynetbuilder.py”, line 279, in build
File “acuitylib/acuitynetbuilder.py”, line 300, in build_layer
File “acuitylib/acuitynetbuilder.py”, line 300, in build_layer
File “acuitylib/acuitynetbuilder.py”, line 300, in build_layer
File “acuitylib/acuitynetbuilder.py”, line 300, in build_layer
File “acuitylib/acuitynetbuilder.py”, line 304, in build_layer
File “acuitylib/layer/acuitylayer.py”, line 262, in compute_shape
File “acuitylib/layer/permute.py”, line 21, in compute_out_shape
ValueError: invalid literal for int() with base 10: ‘[0,’
[9984] Failed to execute script convertonnx

please help me for debug this problem, thanks.

2

@Frank Please help on this

3

@NG1368 Give me a link to downloads this model , I will try , It just show me a invalid literal error in you log . I am not sure what is the problem . What platform files do you convert to onnx files ？

4

thanks for reply, I want to convert a face recognition onnx model (ultra light algorithm) for use in khadas NPU.

I used from ready onnx model from this link:

It should be noted, I have already used these models to implement in the cpu and worked great.

in the first step, when I run following command:

./0_import_model.sh

I encounter those errors.

5

@NG1368 Can you try other paltforms ? onnx model support worse than the others . Colleague tools are not open source. Sometimes it’s hard to judge what caused the errors in log.

6

dear Frank, I have to implement onnx model for evaluation and comparison of ultra light algorithm (onnx model) in CPU and NPU. if you can, help me about this.
thanks…

7

there is a caffe model for ultra-light face detector (download link is:

what do you think about conversion of this model?

8

@NG1368 I don’t know . But I have try the other caffe model which come from github . It work fine .

9

in caffe transmorfation, 0_import_model.sh and 1_quantize_model.sh truly work, but in 2_export_case_code.sh, I have been facing with this error:
E [vnn_CreateNeuralNetwork:174]CHECK PTR 174
E [main:210]CHECK PTR 210
mv: cannot stat ‘/home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/.nb’: No such file or directory
mv: cannot stat '/home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/.dat’: No such file or directory
Traceback (most recent call last):
File “ovxgenerator.py”, line 190, in
File “ovxgenerator.py”, line 181, in main
File “acuitylib/app/exporter/ovxlib_case/casegenerator.py”, line 362, in generate
File “acuitylib/app/exporter/ovxlib_case/casegenerator.py”, line 337, in _gen_special_case
File “acuitylib/app/exporter/ovxlib_case/casegenerator.py”, line 315, in _gen_nb_file
AttributeError: ‘CaseGenerator’ object has no attribute ‘nbg_graph_file_path’
[3798] Failed to execute script ovxgenerator
rm: cannot remove ‘*.lib’: No such file or directory
mv: cannot stat ‘nbg_unify’: No such file or directory
./2_export_case_code.sh: line 23: cd: nbg_unify_caffeface: No such file or directory
mv: cannot stat ‘network_binary.nb’: No such file or directory

please give me a hint about this issue.

10

It looks like can’t find the nb file .

11

After run this code, that .nb file must be generated. After generation, in code with mv command ، .nb file moved to another folder.

12

I have not made any change in 2_export_case_code.sh file. Why does it make such errors?

13

@NG1368 You should give me the full log info . I can’t analyze the problem without log

14

response of ./2_export_case_code.sh command is:
I Load net in caffeface.json
D Load layer input_0 …
D Load layer 185_1 …
D Load layer 187_2 …
D Load layer 188_3 …
D Load layer 190_4 …
D Load layer 191_5 …
D Load layer 193_6 …
D Load layer 194_7 …
D Load layer 196_8 …
D Load layer 197_9 …
D Load layer 374_93 …
D Load layer boxes_94 …
D Load layer scores_95 …
D Load layer output_96 …
D Load layer output_97 …
I Load caffeface.json complete.
I Load data in caffeface.data
I Load quantization tensor table caffeface.quantize
2020-04-17 21:41:45.560262: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
D Process input_0 …
D Acuity output shape(input): (1 240 320 3)
D Process 185_1 …
D Acuity output shape(convolution): (1 120 160 16)
D Process 187_2 …
D Acuity output shape(relu): (1 120 160 16)
D Process 188_3 …
D Acuity output shape(convolution): (1 120 160 16)
D Process 190_4 …
D Acuity output shape(relu): (1 120 160 16)
D Process 191_5 …
D Acuity output shape(convolution): (1 120 160 32)
D Process 193_6 …
D Process 212_19 …
D Acuity output shape(convolution): (1 30 40 64)
D Process 214_20 …
D Acuity output shape(relu): (1 30 40 64)
D Process 215_21 …
D Acuity output shape(convolution): (1 30 40 64)
D Process 217_22 …
D Acuity output shape(relu): (1 30 40 64)
D Process 218_23 …
D Acuity output shape(convolution): (1 30 40 64)
D Process 220_24 …
D Acuity output shape(relu): (1 30 40 64)
D Process 221_25 …
D Acuity output shape(convolution): (1 30 40 64)
D Process 223_26 …
D Acuity output shape(relu): (1 30 40 64)
D Process 224_27 …
D Acuity output shape(convolution): (1 30 40 64)
D Process 226_28 …
D Acuity output shape(relu): (1 30 40 64)
D Process 227_29 …
D Acuity output shape(convolution): (1 30 40 64)
D Process 229_30 …
D Acuity output shape(relu): (1 30 40 64)
D Process 244_36 …
D Acuity output shape(convolution): (1 30 40 64)
D Process 245_37 …
D Acuity output shape(relu): (1 30 40 64)
D Process 246_38 …
D Acuity output shape(convolution): (1 30 40 12)
D Process 247_39 …
D Acuity output shape(permute): (1 30 40 12)
D Process 257_40 …
D Acuity output shape(reshape): (1 4 3600)
D Process 258_41 …
D Acuity output shape(convolution): (1 15 20 64)
D Process 260_42 …
D Acuity output shape(relu): (1 15 20 64)
D Process 261_43 …
D Acuity output shape(convolution): (1 15 20 128)
D Process 263_44 …
D Acuity output shape(relu): (1 15 20 128)
D Process 264_45 …
D Acuity output shape(convolution): (1 15 20 128)
D Process 266_46 …
D Acuity output shape(relu): (1 15 20 128)
D Process 267_47 …
D Acuity output shape(convolution): (1 15 20 128)
D Process 269_48 …
D Acuity output shape(relu): (1 15 20 128)
D Process 270_49 …
D Acuity output shape(convolution): (1 15 20 128)
D Process 272_50 …
D Acuity output shape(relu): (1 15 20 128)
D Process 273_51 …
D Acuity output shape(convolution): (1 15 20 128)
D Process 275_52 …
D Acuity output shape(relu): (1 15 20 128)
D Process 290_58 …
D Acuity output shape(convolution): (1 15 20 128)
D Process 291_59 …
D Acuity output shape(relu): (1 15 20 128)
D Process 292_60 …
D Acuity output shape(convolution): (1 15 20 8)
D Process 293_61 …
D Acuity output shape(permute): (1 15 20 8)
D Process 303_62 …
D Acuity output shape(reshape): (1 4 600)
D Process 304_63 …
D Acuity output shape(convolution): (1 8 10 128)
D Process 306_64 …
D Acuity output shape(relu): (1 8 10 128)
D Process 307_65 …
D Acuity output shape(convolution): (1 8 10 256)
D Process 309_66 …
D Acuity output shape(relu): (1 8 10 256)
D Process 310_67 …
D Acuity output shape(convolution): (1 8 10 256)
D Process 312_68 …
D Acuity output shape(relu): (1 8 10 256)
D Process 313_69 …
D Acuity output shape(convolution): (1 8 10 256)
D Process 315_70 …
D Acuity output shape(relu): (1 8 10 256)
D Process 330_76 …
D Acuity output shape(convolution): (1 8 10 256)
D Process 331_77 …
D Acuity output shape(relu): (1 8 10 256)
D Process 332_78 …
D Acuity output shape(convolution): (1 8 10 8)
D Process 333_79 …
D Acuity output shape(permute): (1 8 10 8)
D Process 343_80 …
D Acuity output shape(reshape): (1 4 160)
D Process 344_81 …
D Acuity output shape(convolution): (1 8 10 64)
D Process 345_82 …
D Acuity output shape(relu): (1 8 10 64)
3600)
D Process output_96 …
D Acuity output shape(output): (1 16 3600)
D Process 230_31 …
D Acuity output shape(convolution): (1 30 40 64)
D Process 231_32 …
D Acuity output shape(relu): (1 30 40 64)
D Process 232_33 …
D Acuity output shape(convolution): (1 30 40 6)
D Process 233_34 …
D Acuity output shape(permute): (1 30 40 6)
D Process 243_35 …
D Acuity output shape(reshape): (1 2 3600)
D Process 276_53 …
D Acuity output shape(convolution): (1 15 20 128)
D Process 277_54 …
D Acuity output shape(relu): (1 15 20 128)
D Process 278_55 …
D Acuity output shape(convolution): (1 15 20 4)
D Process 279_56 …
D Acuity output shape(permute): (1 15 20 4)
D Process 289_57 …
D Acuity output shape(reshape): (1 2 600)
D Process 316_71 …
D Acuity output shape(convolution): (1 8 10 256)
D Process 317_72 …
D Acuity output shape(relu): (1 8 10 256)
D Process 318_73 …
D Acuity output shape(convolution): (1 8 10 4)
D Process 319_74 …
D Acuity output shape(permute): (1 8 10 4)
D Process 329_75 …
D Acuity output shape(reshape): (1 2 160)
D Process 350_87 …
D Acuity output shape(convolution): (1 4 5 6)
D Process 351_88 …
D Acuity output shape(permute): (1 4 5 6)
D Process 361_89 …
D Acuity output shape(reshape): (1 2 60)
D Process 374_93 …
D Acuity output shape(concat): (1 8 3600)
D Process scores_95 …
D Acuity output shape(softmax): (1 8 3600)
D Process output_97 …
D Acuity output shape(output): (1 8 3600)
I Build complete.
I Config File “VIPNANOQI_PID0X88” is missing. Try to find another path…
I Config File “/home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/…/bin/VIPNANOQI_PID0X88” found successfully
I Initialzing network optimizer by /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/…/bin/VIPNANOQI_PID0X88 …
D Optimizing network with avgpool_to_conv, multiply_transform, add_extra_io, format_input_ops, 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, extend_batchnormalize, swapper, merge_layer, transform_layer, proposal_opt, broadcast_op, strip_op, auto_fill_reshape_zero, adjust_output_attrs
I insert reshape layer before softmax scores_95
I insert reshape layer after softmax scores_95
D Optimizing network with c2drv_convert_axis, c2drv_convert_shape, c2drv_convert_array, c2drv_cast_dtype
I Building data …
…/bin/vcmdtools
…/bin/vcmdtools/lib
I Packing data …
D Packing 185_1 …
D Quantize @185_1:bias to dynamic_fixed_point.
D Quantize @185_1:weight to dynamic_fixed_point.
D Packing 188_3 …
D Quantize @188_3:bias to dynamic_fixed_point.
D Quantize @188_3:weight to dynamic_fixed_point.
D Packing 191_5 …
D Quantize @191_5:bias to dynamic_fixed_point.
D Quantize @191_5:weight to dynamic_fixed_point.
D Packing 194_7 …
D Quantize @194_7:bias to dynamic_fixed_point.
D Quantize @194_7:weight to dynamic_fixed_point.
D Packing 197_9 …
I Saving data to caffeface.export.data
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/vnn_caffeface.c
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/vnn_caffeface.h
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/vnn_post_process.c
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/vnn_post_process.h
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/vnn_pre_process.c
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/vnn_pre_process.h
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/vnn_global.h
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/main.c
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/BUILD
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/caffeface.vcxproj
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/makefile.linux
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/.cproject
I Save vx network source file to /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/.project
D Generate fake input /home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/input_0.tensor
gcc -Wall -std=c++0x -I. -I…/bin/vcmdtools/include/ -I…/bin/vcmdtools/include/CL -I…/bin/vcmdtools/include/VX -I…/bin/vcmdtools/include/ovxlib -D__linux__ -DLINUX -O3 -c vnn_pre_process.c
cc1: warning: command line option ‘-std=c++11’ is valid for C++/ObjC++ but not for C
gcc -Wall -std=c++0x -I. -I…/bin/vcmdtools/include/ -I…/bin/vcmdtools/include/CL -I…/bin/vcmdtools/include/VX -I…/bin/vcmdtools/include/ovxlib -D__linux__ -DLINUX -O3 -c vnn_post_process.c
cc1: warning: command line option ‘-std=c++11’ is valid for C++/ObjC++ but not for C
gcc -Wall -std=c++0x -I. -I…/bin/vcmdtools/include/ -I…/bin/vcmdtools/include/CL -I…/bin/vcmdtools/include/VX -I…/bin/vcmdtools/include/ovxlib -D__linux__ -DLINUX -O3 -c main.c
cc1: warning: command line option ‘-std=c++11’ is valid for C++/ObjC++ but not for C
gcc -Wall -std=c++0x -I. -I…/bin/vcmdtools/include/ -I…/bin/vcmdtools/include/CL -I…/bin/vcmdtools/include/VX -I…/bin/vcmdtools/include/ovxlib -D__linux__ -DLINUX -O3 -c vnn_caffeface.c
cc1: warning: command line option ‘-std=c++11’ is valid for C++/ObjC++ but not for C
gcc -Wall -std=c++0x -I. -I…/bin/vcmdtools/include/ -I…/bin/vcmdtools/include/CL -I…/bin/vcmdtools/include/VX -I…/bin/vcmdtools/include/ovxlib -D__linux__ -DLINUX -O3 -O3 vnn_pre_process.o vnn_post_process.o main.o vnn_caffeface.o -L…/bin/vcmdtools/lib -lOpenVX -lOpenVXU -lCLC -lVSC -lGAL -lLLVM_viv -lovxlib -lEmulator -lvdtproxy -L…/bin/vcmdtools/lib/x64_linux -lOpenVX -lOpenVXU -lCLC -lVSC -lGAL -lLLVM_viv -lovxlib -lEmulator -lvdtproxy …/bin/vcmdtools/lib/libjpeg.a -o gen_nbg
I [load_data:120]Read 432 data.
I [load_data:120]Read 64 data.
I [load_data:120]Read 144 data.
I [load_data:120]Read 64 data.
I [load_data:120]Read 512 data.
I [load_data:120]Read 128 data.
I [load_data:120]Read 288 data.
I [load_data:120]Read 128 data.
I [load_data:120]Read 1024 data.
I [load_data:120]Read 128 data.
I [load_data:120]Read 288 data.
I [load_data:120]Read 128 data.
I [load_data:120]Read 1024 data.
I [load_data:120]Read 128 data.
I [load_data:120]Read 288 data.
I [load_data:120]Read 128 data.
I [load_data:120]Read 2048 data.
I [load_data:120]Read 256 data.
I [load_data:120]Read 256 data.
I [load_data:120]Read 384 data.
I [load_data:120]Read 24 data.
I [load_data:120]Read 768 data.
I [load_data:120]Read 48 data.
I [load_data:120]Read 8192 data.
I [load_data:120]Read 512 data.
I [load_data:120]Read 1152 data.
I [load_data:120]Read 512 data.
I [load_data:120]Read 16384 data.
I [load_data:120]Read 512 data.
I [load_data:120]Read 1152 data.
I [load_data:120]Read 512 data.
I [load_data:120]Read 16384 data.
I [load_data:120]Read 512 data.
I [load_data:120]Read 1152 data.
I [load_data:120]Read 512 data.
I [load_data:120]Read 1152 data.
I [load_data:120]Read 512 data.
I [load_data:120]Read 1152 data.
I [load_data:120]Read 512 data.
I [load_data:120]Read 512 data.
I [load_data:120]Read 16 data.
I [load_data:120]Read 1024 data.
I [load_data:120]Read 32 data.
I [load_data:120]Read 32768 data.
I [load_data:120]Read 1024 data.
I [load_data:120]Read 2304 data.
I [load_data:120]Read 1024 data.
I [load_data:120]Read 65536 data.
I [load_data:120]Read 1024 data.
I [load_data:120]Read 32 data.
I [load_data:120]Read 576 data.
I [load_data:120]Read 256 data.
I [load_data:120]Read 16384 data.
I [load_data:120]Read 1024 data.
I [load_data:120]Read 13824 data.
I [load_data:120]Read 24 data.
I [load_data:120]Read 27648 data.
I [load_data:120]Read 48 data.
D [setup_node:367]Setup node id[0] uid[1] op[CONV2D]
D [print_tensor:129]in : id[ 0] shape[ 320, 240, 3, 1 ] fmt[i8 ] qnt[DFP fl= 7]
D [print_tensor:129]in : id[ 3] shape[ 3, 3, 3, 16 ] fmt[i8 ] qnt[DFP fl= 7]
D [print_tensor:129]in : id[ 4] shape[ 16 ] fmt[i32] qnt[DFP fl= 14]
D [print_tensor:129]out: id[ 87] shape[ 160, 120, 16, 1 ] fmt[i8 ] qnt[DFP fl= 6]
D [setup_node:367]Setup node id[1] uid[2] op[RELU]
D [print_tensor:129]in : id[ 87] shape[ 160, 120, 16, 1 ] fmt[i8 ] qnt[DFP fl= 6]
D [print_tensor:129]out: id[ 88] shape[ 160, 120, 16, 1 ] fmt[i8 ] qnt[DFP fl= 6]
D [setup_node:367]Setup node id[2] uid[3] op[CONV2D]
D [print_tensor:129]in : id[ 88] shape[ 160, 120, 16, 1 ] fmt[i8 ] qnt[DFP fl= 6]
D [print_tensor:129]in : id[ 5] shape[ 3, 3, 16, 1 ] fmt[i8 ] qnt[DFP fl= 4]
D [print_tensor:129]in : id[ 6] shape[ 16 ] fmt[i32] qnt[DFP fl= 10]
D [print_tensor:129]out: id[ 89] shape[ 160, 120, 16, 1 ] fmt[i8 ] qnt[DFP fl= 5]
D [setup_node:367]Setup node id[3] uid[4] op[RELU]
D [print_tensor:129]out: id[ 178] shape[ 4, 60, 1 ] fmt[i8 ] qnt[DFP fl= 4]
D [setup_node:367]Setup node id[92] uid[93] op[CONCAT]
D [print_tensor:129]in : id[ 129] shape[ 2, 3600, 1 ] fmt[i8 ] qnt[DFP fl= 5]
D [print_tensor:129]in : id[ 151] shape[ 2, 600, 1 ] fmt[i8 ] qnt[DFP fl= 5]
D [print_tensor:129]in : id[ 169] shape[ 2, 160, 1 ] fmt[i8 ] qnt[DFP fl= 5]
D [print_tensor:129]in : id[ 177] shape[ 2, 60, 1 ] fmt[i8 ] qnt[DFP fl= 5]
D [print_tensor:129]out: id[ 179] shape[ 2, 4420, 1 ] fmt[i8 ] qnt[DFP fl= 5]
D [setup_node:367]Setup node id[93] uid[94] op[CONCAT]
E [op_check:322]Concat output dims size(3600 vs 4)
E [setup_node:381]Check node[93] CONCAT fail
D [print_tensor:129]in : id[ 130] shape[ 4, 3600, 1 ] fmt[i8 ] qnt[DFP fl= 4]
D [print_tensor:129]in : id[ 152] shape[ 4, 600, 1 ] fmt[i8 ] qnt[DFP fl= 4]
D [print_tensor:129]in : id[ 170] shape[ 4, 160, 1 ] fmt[i8 ] qnt[DFP fl= 4]
D [print_tensor:129]in : id[ 178] shape[ 4, 60, 1 ] fmt[i8 ] qnt[DFP fl= 4]
D [print_tensor:129]out: id[ 1] shape[ 3600, 16, 1 ] fmt[i8 ] qnt[DFP fl= 4]
E [vnn_CreateNeuralNetwork:174]CHECK PTR 174
E [main:210]CHECK PTR 210
mv: cannot stat ‘/home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/.nb’: No such file or directory
mv: cannot stat '/home/sgi/Desktop/SDK/aml_npu_sdk/acuity-toolkit/conversion_scripts/.dat’: No such file or directory
Traceback (most recent call last):
File “ovxgenerator.py”, line 190, in
File “ovxgenerator.py”, line 181, in main
File “acuitylib/app/exporter/ovxlib_case/casegenerator.py”, line 362, in generate
File “acuitylib/app/exporter/ovxlib_case/casegenerator.py”, line 337, in _gen_special_case
File “acuitylib/app/exporter/ovxlib_case/casegenerator.py”, line 315, in _gen_nb_file
AttributeError: ‘CaseGenerator’ object has no attribute ‘nbg_graph_file_path’
[3170] Failed to execute script ovxgenerator
rm: cannot remove ‘*.lib’: No such file or directory
mv: cannot stat ‘nbg_unify’: No such file or directory
./2_export_case_code.sh: line 23: cd: nbg_unify_caffeface: No such file or directory
mv: cannot stat ‘network_binary.nb’: No such file or directory

15

@NG1368 I just found this error , .nb file not found . It should be create when conversion . But it doesn’t, and don’t show any info about why it doesn’t create nb file …

16

I dont know why .nb file not created. I do in accordance with khadas transform link:


my 2_export_case_code.sh file is:
#!/bin/bash

NAME=caffeface
ACUITY_PATH=…/bin/

export_ovxlib=${ACUITY_PATH}ovxgenerator

$export_ovxlib
–model-input ${NAME}.json
–data-input ${NAME}.data
–reorder-channel ‘2 1 0’
–channel-mean-value ‘0 0 0 256’
–export-dtype quantized
–model-quantize ${NAME}.quantize
–optimize VIPNANOQI_PID0X88
–viv-sdk …/bin/vcmdtools
–pack-nbg-unify

rm *.h *.c .project .cproject *.vcxproj *.lib BUILD *.linux

mv nbg_unify nbg_unify_${NAME}

cd nbg_unify_${NAME}

mv network_binary.nb ${NAME}.nb

I have not made any other changes to the file.

17

@NG1368 Maybe you can try to transform you model to other platforms . Then convert again.

18

Can I send you my caffe model for test it?

19

I used from models of this link:

please help me. thanks

20

@NG1368 I will try it when I finish the work at hand. Give me some time . I will test it .