Ticket Name: TDA2PXEVM: Is Conv with no relue supported In TDA2? Query Text: Part Number: TDA2PXEVM Other Parts Discussed in Thread: TDA2 Hi I get some accuracy problems with caffemodel importing. Here is my steps: a) Use tidl_model_import.out.exe convert regnet.prototxt/caffemodel to tidl_param/bin b) Compare each layer's output : onnx vs trace_dump_idx_wxh.y c) I find the first conv layer(with bn & relu) matched well(error < 1%) d) But conv_layer with no relu(which are inputs of eltwise_layer) cannot match original model layer's outputs I put a snapshot below : conv layers in green rect match, but conv layer without relu in red circle not match. regnet_import.zip So, is relu strictedly demanded to place after conv layer? Or this is just bugs in import tool? I upload my model and tools for analysis. Responses: Hi, Can you share the import output log to check the issue ? Thanks, Praveen Hi Praveen Thanks for your replay! I uploaded LOG.txt. Geroge. 6675.LOG.txt .\tidl_model_import.out.exe .\ONNX_Reg200M_CIFAR\tidl_import.txt Caffe Network File : ONNX_Reg200M_CIFAR\trained\regnetx200mf_cifar_Relu96.prototxt Caffe Model File : ONNX_Reg200M_CIFAR\trained\regnetx200mf_cifar_Relu96.caffemodel TIDL Network File : ONNX_Reg200M_CIFAR\model\tidl_net_reg200cifar_relu96.bin TIDL Model File : ONNX_Reg200M_CIFAR\model\tidl_param_reg200cifar_relu96.bin Name of the Network : REG200MCIFAR-ONNX Num Inputs : 1 Num of Layer Detected : 71 0, TIDL_DataLayer , data 0, -1 , 1 , x , x , x , x , x , x , x , x , 0 , 0 , 0 , 0 , 0 , 1 , 3 , 32 , 32 , 0 , 1, TIDL_ConvolutionLayer , Conv_0 1, 1 , 1 , 0 , x , x , x , x , x , x , x , 1 , 1 , 3 , 32 , 32 , 1 , 32 , 32 , 32 , 884736 , 2, TIDL_ConvolutionLayer , Conv_2 1, 1 , 1 , 1 , x , x , x , x , x , x , x , 2 , 1 , 32 , 32 , 32 , 1 , 24 , 32 , 32 , 786432 , 3, TIDL_ConvolutionLayer , Conv_4 1, 1 , 1 , 2 , x , x , x , x , x , x , x , 3 , 1 , 24 , 32 , 32 , 1 , 24 , 16 , 16 , 442368 , 4, TIDL_ConvolutionLayer , Conv_6 1, 1 , 1 , 3 , x , x , x , x , x , x , x , 4 , 1 , 24 , 16 , 16 , 1 , 24 , 16 , 16 , 147456 , 5, TIDL_ConvolutionLayer , Conv_7 1, 1 , 1 , 1 , x , x , x , x , x , x , x , 5 , 1 , 32 , 32 , 32 , 1 , 24 , 16 , 16 , 196608 , 6, TIDL_EltWiseLayer , Add_8 1, 2 , 1 , 4 , 5 , x , x , x , x , x , x , 6 , 1 , 24 , 16 , 16 , 1 , 24 , 16 , 16 , 6144 , 7, TIDL_BatchNormLayer , Relu_9 1, 1 , 1 , 6 , x , x , x , x , x , x , x , 7 , 1 , 24 , 16 , 16 , 1 , 24 , 16 , 16 , 6144 , 8, TIDL_ConvolutionLayer , Conv_10 1, 1 , 1 , 7 , x , x , x , x , x , x , x , 8 , 1 , 24 , 16 , 16 , 1 , 56 , 16 , 16 , 344064 , 9, TIDL_ConvolutionLayer , Conv_12 1, 1 , 1 , 8 , x , x , x , x , x , x , x , 9 , 1 , 56 , 16 , 16 , 1 , 56 , 16 , 16 , 1032192 , 10, TIDL_ConvolutionLayer , Conv_14 1, 1 , 1 , 9 , x , x , x , x , x , x , x , 10 , 1 , 56 , 16 , 16 , 1 , 56 , 16 , 16 , 802816 , 11, TIDL_ConvolutionLayer , Conv_15 1, 1 , 1 , 7 , x , x , x , x , x , x , x , 11 , 1 , 24 , 16 , 16 , 1 , 56 , 16 , 16 , 344064 , 12, TIDL_EltWiseLayer , Add_16 1, 2 , 1 , 10 , 11 , x , x , x , x , x , x , 12 , 1 , 56 , 16 , 16 , 1 , 56 , 16 , 16 , 14336 , 13, TIDL_BatchNormLayer , Relu_17 1, 1 , 1 , 12 , x , x , x , x , x , x , x , 13 , 1 , 56 , 16 , 16 , 1 , 56 , 16 , 16 , 14336 , 14, TIDL_ConvolutionLayer , Conv_18 1, 1 , 1 , 13 , x , x , x , x , x , x , x , 14 , 1 , 56 , 16 , 16 , 1 , 152 , 16 , 16 , 2179072 , 15, TIDL_ConvolutionLayer , Conv_20 1, 1 , 1 , 14 , x , x , x , x , x , x , x , 15 , 1 , 152 , 16 , 16 , 1 , 152 , 8 , 8 , 700416 , 16, TIDL_ConvolutionLayer , Conv_22 1, 1 , 1 , 15 , x , x , x , x , x , x , x , 16 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 1478656 , 17, TIDL_ConvolutionLayer , Conv_23 1, 1 , 1 , 13 , x , x , x , x , x , x , x , 17 , 1 , 56 , 16 , 16 , 1 , 152 , 8 , 8 , 544768 , 18, TIDL_EltWiseLayer , Add_24 1, 2 , 1 , 16 , 17 , x , x , x , x , x , x , 18 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 9728 , 19, TIDL_BatchNormLayer , Relu_25 1, 1 , 1 , 18 , x , x , x , x , x , x , x , 19 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 9728 , 20, TIDL_ConvolutionLayer , Conv_26 1, 1 , 1 , 19 , x , x , x , x , x , x , x , 20 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 1478656 , 21, TIDL_ConvolutionLayer , Conv_28 1, 1 , 1 , 20 , x , x , x , x , x , x , x , 21 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 700416 , 22, TIDL_ConvolutionLayer , Conv_30 1, 1 , 1 , 21 , x , x , x , x , x , x , x , 22 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 1478656 , 23, TIDL_EltWiseLayer , Add_31 1, 2 , 1 , 22 , 19 , x , x , x , x , x , x , 23 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 9728 , 24, TIDL_BatchNormLayer , Relu_32 1, 1 , 1 , 23 , x , x , x , x , x , x , x , 24 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 9728 , 25, TIDL_ConvolutionLayer , Conv_33 1, 1 , 1 , 24 , x , x , x , x , x , x , x , 25 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 1478656 , 26, TIDL_ConvolutionLayer , Conv_35 1, 1 , 1 , 25 , x , x , x , x , x , x , x , 26 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 700416 , 27, TIDL_ConvolutionLayer , Conv_37 1, 1 , 1 , 26 , x , x , x , x , x , x , x , 27 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 1478656 , 28, TIDL_EltWiseLayer , Add_38 1, 2 , 1 , 27 , 24 , x , x , x , x , x , x , 28 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 9728 , 29, TIDL_BatchNormLayer , Relu_39 1, 1 , 1 , 28 , x , x , x , x , x , x , x , 29 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 9728 , 30, TIDL_ConvolutionLayer , Conv_40 1, 1 , 1 , 29 , x , x , x , x , x , x , x , 30 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 1478656 , 31, TIDL_ConvolutionLayer , Conv_42 1, 1 , 1 , 30 , x , x , x , x , x , x , x , 31 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 700416 , 32, TIDL_ConvolutionLayer , Conv_44 1, 1 , 1 , 31 , x , x , x , x , x , x , x , 32 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 1478656 , 33, TIDL_EltWiseLayer , Add_45 1, 2 , 1 , 32 , 29 , x , x , x , x , x , x , 33 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 9728 , 34, TIDL_BatchNormLayer , Relu_46 1, 1 , 1 , 33 , x , x , x , x , x , x , x , 34 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 9728 , 35, TIDL_ConvolutionLayer , Conv_47 1, 1 , 1 , 34 , x , x , x , x , x , x , x , 35 , 1 , 152 , 8 , 8 , 1 , 368 , 8 , 8 , 3579904 , 36, TIDL_ConvolutionLayer , Conv_49 1, 1 , 1 , 35 , x , x , x , x , x , x , x , 36 , 1 , 368 , 8 , 8 , 1 , 368 , 4 , 4 , 423936 , 37, TIDL_ConvolutionLayer , Conv_51 1, 1 , 1 , 36 , x , x , x , x , x , x , x , 37 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 38, TIDL_ConvolutionLayer , Conv_52 1, 1 , 1 , 34 , x , x , x , x , x , x , x , 38 , 1 , 152 , 8 , 8 , 1 , 368 , 4 , 4 , 894976 , 39, TIDL_EltWiseLayer , Add_53 1, 2 , 1 , 37 , 38 , x , x , x , x , x , x , 39 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 40, TIDL_BatchNormLayer , Relu_54 1, 1 , 1 , 39 , x , x , x , x , x , x , x , 40 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 41, TIDL_ConvolutionLayer , Conv_55 1, 1 , 1 , 40 , x , x , x , x , x , x , x , 41 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 42, TIDL_ConvolutionLayer , Conv_57 1, 1 , 1 , 41 , x , x , x , x , x , x , x , 42 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 423936 , 43, TIDL_ConvolutionLayer , Conv_59 1, 1 , 1 , 42 , x , x , x , x , x , x , x , 43 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 44, TIDL_EltWiseLayer , Add_60 1, 2 , 1 , 43 , 40 , x , x , x , x , x , x , 44 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 45, TIDL_BatchNormLayer , Relu_61 1, 1 , 1 , 44 , x , x , x , x , x , x , x , 45 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 46, TIDL_ConvolutionLayer , Conv_62 1, 1 , 1 , 45 , x , x , x , x , x , x , x , 46 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 47, TIDL_ConvolutionLayer , Conv_64 1, 1 , 1 , 46 , x , x , x , x , x , x , x , 47 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 423936 , 48, TIDL_ConvolutionLayer , Conv_66 1, 1 , 1 , 47 , x , x , x , x , x , x , x , 48 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 49, TIDL_EltWiseLayer , Add_67 1, 2 , 1 , 48 , 45 , x , x , x , x , x , x , 49 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 50, TIDL_BatchNormLayer , Relu_68 1, 1 , 1 , 49 , x , x , x , x , x , x , x , 50 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 51, TIDL_ConvolutionLayer , Conv_69 1, 1 , 1 , 50 , x , x , x , x , x , x , x , 51 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 52, TIDL_ConvolutionLayer , Conv_71 1, 1 , 1 , 51 , x , x , x , x , x , x , x , 52 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 423936 , 53, TIDL_ConvolutionLayer , Conv_73 1, 1 , 1 , 52 , x , x , x , x , x , x , x , 53 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 54, TIDL_EltWiseLayer , Add_74 1, 2 , 1 , 53 , 50 , x , x , x , x , x , x , 54 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 55, TIDL_BatchNormLayer , Relu_75 1, 1 , 1 , 54 , x , x , x , x , x , x , x , 55 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 56, TIDL_ConvolutionLayer , Conv_76 1, 1 , 1 , 55 , x , x , x , x , x , x , x , 56 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 57, TIDL_ConvolutionLayer , Conv_78 1, 1 , 1 , 56 , x , x , x , x , x , x , x , 57 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 423936 , 58, TIDL_ConvolutionLayer , Conv_80 1, 1 , 1 , 57 , x , x , x , x , x , x , x , 58 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 59, TIDL_EltWiseLayer , Add_81 1, 2 , 1 , 58 , 55 , x , x , x , x , x , x , 59 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 60, TIDL_BatchNormLayer , Relu_82 1, 1 , 1 , 59 , x , x , x , x , x , x , x , 60 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 61, TIDL_ConvolutionLayer , Conv_83 1, 1 , 1 , 60 , x , x , x , x , x , x , x , 61 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 62, TIDL_ConvolutionLayer , Conv_85 1, 1 , 1 , 61 , x , x , x , x , x , x , x , 62 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 423936 , 63, TIDL_ConvolutionLayer , Conv_87 1, 1 , 1 , 62 , x , x , x , x , x , x , x , 63 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 64, TIDL_EltWiseLayer , Add_88 1, 2 , 1 , 63 , 60 , x , x , x , x , x , x , 64 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 65, TIDL_BatchNormLayer , Relu_89 1, 1 , 1 , 64 , x , x , x , x , x , x , x , 65 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 66, TIDL_ConvolutionLayer , Conv_90 1, 1 , 1 , 65 , x , x , x , x , x , x , x , 66 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 67, TIDL_ConvolutionLayer , Conv_92 1, 1 , 1 , 66 , x , x , x , x , x , x , x , 67 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 423936 , 68, TIDL_ConvolutionLayer , Conv_94 1, 1 , 1 , 67 , x , x , x , x , x , x , x , 68 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 2166784 , 69, TIDL_EltWiseLayer , Add_95 1, 2 , 1 , 68 , 65 , x , x , x , x , x , x , 69 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , 70, TIDL_BatchNormLayer , Relu_96 1, 1 , 1 , 69 , x , x , x , x , x , x , x , 70 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 5888 , Total Giga Macs : 0.0567 已复制 1 个文件。 Processing config file .\tempDir\qunat_stats_config.txt ! 0, TIDL_DataLayer , 0, -1 , 1 , x , x , x , x , x , x , x , x , 0 , 0 , 0 , 0 , 0 , 1 , 3 , 32 , 32 , 1, TIDL_ConvolutionLayer , 1, 1 , 1 , 0 , x , x , x , x , x , x , x , 1 , 1 , 3 , 32 , 32 , 1 , 32 , 32 , 32 , 2, TIDL_ConvolutionLayer , 1, 1 , 1 , 1 , x , x , x , x , x , x , x , 2 , 1 , 32 , 32 , 32 , 1 , 24 , 32 , 32 , 3, TIDL_ConvolutionLayer , 1, 1 , 1 , 2 , x , x , x , x , x , x , x , 3 , 1 , 24 , 32 , 32 , 1 , 24 , 16 , 16 , 4, TIDL_ConvolutionLayer , 1, 1 , 1 , 3 , x , x , x , x , x , x , x , 4 , 1 , 24 , 16 , 16 , 1 , 24 , 16 , 16 , 5, TIDL_ConvolutionLayer , 1, 1 , 1 , 1 , x , x , x , x , x , x , x , 5 , 1 , 32 , 32 , 32 , 1 , 24 , 16 , 16 , 6, TIDL_EltWiseLayer , 1, 2 , 1 , 4 , 5 , x , x , x , x , x , x , 6 , 1 , 24 , 16 , 16 , 1 , 24 , 16 , 16 , 7, TIDL_BatchNormLayer , 1, 1 , 1 , 6 , x , x , x , x , x , x , x , 7 , 1 , 24 , 16 , 16 , 1 , 24 , 16 , 16 , 8, TIDL_ConvolutionLayer , 1, 1 , 1 , 7 , x , x , x , x , x , x , x , 8 , 1 , 24 , 16 , 16 , 1 , 56 , 16 , 16 , 9, TIDL_ConvolutionLayer , 1, 1 , 1 , 8 , x , x , x , x , x , x , x , 9 , 1 , 56 , 16 , 16 , 1 , 56 , 16 , 16 , 10, TIDL_ConvolutionLayer , 1, 1 , 1 , 9 , x , x , x , x , x , x , x , 10 , 1 , 56 , 16 , 16 , 1 , 56 , 16 , 16 , 11, TIDL_ConvolutionLayer , 1, 1 , 1 , 7 , x , x , x , x , x , x , x , 11 , 1 , 24 , 16 , 16 , 1 , 56 , 16 , 16 , 12, TIDL_EltWiseLayer , 1, 2 , 1 , 10 , 11 , x , x , x , x , x , x , 12 , 1 , 56 , 16 , 16 , 1 , 56 , 16 , 16 , 13, TIDL_BatchNormLayer , 1, 1 , 1 , 12 , x , x , x , x , x , x , x , 13 , 1 , 56 , 16 , 16 , 1 , 56 , 16 , 16 , 14, TIDL_ConvolutionLayer , 1, 1 , 1 , 13 , x , x , x , x , x , x , x , 14 , 1 , 56 , 16 , 16 , 1 , 152 , 16 , 16 , 15, TIDL_ConvolutionLayer , 1, 1 , 1 , 14 , x , x , x , x , x , x , x , 15 , 1 , 152 , 16 , 16 , 1 , 152 , 8 , 8 , 16, TIDL_ConvolutionLayer , 1, 1 , 1 , 15 , x , x , x , x , x , x , x , 16 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 17, TIDL_ConvolutionLayer , 1, 1 , 1 , 13 , x , x , x , x , x , x , x , 17 , 1 , 56 , 16 , 16 , 1 , 152 , 8 , 8 , 18, TIDL_EltWiseLayer , 1, 2 , 1 , 16 , 17 , x , x , x , x , x , x , 18 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 19, TIDL_BatchNormLayer , 1, 1 , 1 , 18 , x , x , x , x , x , x , x , 19 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 20, TIDL_ConvolutionLayer , 1, 1 , 1 , 19 , x , x , x , x , x , x , x , 20 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 21, TIDL_ConvolutionLayer , 1, 1 , 1 , 20 , x , x , x , x , x , x , x , 21 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 22, TIDL_ConvolutionLayer , 1, 1 , 1 , 21 , x , x , x , x , x , x , x , 22 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 23, TIDL_EltWiseLayer , 1, 2 , 1 , 22 , 19 , x , x , x , x , x , x , 23 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 24, TIDL_BatchNormLayer , 1, 1 , 1 , 23 , x , x , x , x , x , x , x , 24 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 25, TIDL_ConvolutionLayer , 1, 1 , 1 , 24 , x , x , x , x , x , x , x , 25 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 26, TIDL_ConvolutionLayer , 1, 1 , 1 , 25 , x , x , x , x , x , x , x , 26 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 27, TIDL_ConvolutionLayer , 1, 1 , 1 , 26 , x , x , x , x , x , x , x , 27 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 28, TIDL_EltWiseLayer , 1, 2 , 1 , 27 , 24 , x , x , x , x , x , x , 28 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 29, TIDL_BatchNormLayer , 1, 1 , 1 , 28 , x , x , x , x , x , x , x , 29 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 30, TIDL_ConvolutionLayer , 1, 1 , 1 , 29 , x , x , x , x , x , x , x , 30 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 31, TIDL_ConvolutionLayer , 1, 1 , 1 , 30 , x , x , x , x , x , x , x , 31 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 32, TIDL_ConvolutionLayer , 1, 1 , 1 , 31 , x , x , x , x , x , x , x , 32 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 33, TIDL_EltWiseLayer , 1, 2 , 1 , 32 , 29 , x , x , x , x , x , x , 33 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 34, TIDL_BatchNormLayer , 1, 1 , 1 , 33 , x , x , x , x , x , x , x , 34 , 1 , 152 , 8 , 8 , 1 , 152 , 8 , 8 , 35, TIDL_ConvolutionLayer , 1, 1 , 1 , 34 , x , x , x , x , x , x , x , 35 , 1 , 152 , 8 , 8 , 1 , 368 , 8 , 8 , 36, TIDL_ConvolutionLayer , 1, 1 , 1 , 35 , x , x , x , x , x , x , x , 36 , 1 , 368 , 8 , 8 , 1 , 368 , 4 , 4 , 37, TIDL_ConvolutionLayer , 1, 1 , 1 , 36 , x , x , x , x , x , x , x , 37 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 38, TIDL_ConvolutionLayer , 1, 1 , 1 , 34 , x , x , x , x , x , x , x , 38 , 1 , 152 , 8 , 8 , 1 , 368 , 4 , 4 , 39, TIDL_EltWiseLayer , 1, 2 , 1 , 37 , 38 , x , x , x , x , x , x , 39 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 40, TIDL_BatchNormLayer , 1, 1 , 1 , 39 , x , x , x , x , x , x , x , 40 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 41, TIDL_ConvolutionLayer , 1, 1 , 1 , 40 , x , x , x , x , x , x , x , 41 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 42, TIDL_ConvolutionLayer , 1, 1 , 1 , 41 , x , x , x , x , x , x , x , 42 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 43, TIDL_ConvolutionLayer , 1, 1 , 1 , 42 , x , x , x , x , x , x , x , 43 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 44, TIDL_EltWiseLayer , 1, 2 , 1 , 43 , 40 , x , x , x , x , x , x , 44 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 45, TIDL_BatchNormLayer , 1, 1 , 1 , 44 , x , x , x , x , x , x , x , 45 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 46, TIDL_ConvolutionLayer , 1, 1 , 1 , 45 , x , x , x , x , x , x , x , 46 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 47, TIDL_ConvolutionLayer , 1, 1 , 1 , 46 , x , x , x , x , x , x , x , 47 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 48, TIDL_ConvolutionLayer , 1, 1 , 1 , 47 , x , x , x , x , x , x , x , 48 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 49, TIDL_EltWiseLayer , 1, 2 , 1 , 48 , 45 , x , x , x , x , x , x , 49 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 50, TIDL_BatchNormLayer , 1, 1 , 1 , 49 , x , x , x , x , x , x , x , 50 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 51, TIDL_ConvolutionLayer , 1, 1 , 1 , 50 , x , x , x , x , x , x , x , 51 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 52, TIDL_ConvolutionLayer , 1, 1 , 1 , 51 , x , x , x , x , x , x , x , 52 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 53, TIDL_ConvolutionLayer , 1, 1 , 1 , 52 , x , x , x , x , x , x , x , 53 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 54, TIDL_EltWiseLayer , 1, 2 , 1 , 53 , 50 , x , x , x , x , x , x , 54 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 55, TIDL_BatchNormLayer , 1, 1 , 1 , 54 , x , x , x , x , x , x , x , 55 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 56, TIDL_ConvolutionLayer , 1, 1 , 1 , 55 , x , x , x , x , x , x , x , 56 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 57, TIDL_ConvolutionLayer , 1, 1 , 1 , 56 , x , x , x , x , x , x , x , 57 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 58, TIDL_ConvolutionLayer , 1, 1 , 1 , 57 , x , x , x , x , x , x , x , 58 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 59, TIDL_EltWiseLayer , 1, 2 , 1 , 58 , 55 , x , x , x , x , x , x , 59 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 60, TIDL_BatchNormLayer , 1, 1 , 1 , 59 , x , x , x , x , x , x , x , 60 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 61, TIDL_ConvolutionLayer , 1, 1 , 1 , 60 , x , x , x , x , x , x , x , 61 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 62, TIDL_ConvolutionLayer , 1, 1 , 1 , 61 , x , x , x , x , x , x , x , 62 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 63, TIDL_ConvolutionLayer , 1, 1 , 1 , 62 , x , x , x , x , x , x , x , 63 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 64, TIDL_EltWiseLayer , 1, 2 , 1 , 63 , 60 , x , x , x , x , x , x , 64 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 65, TIDL_BatchNormLayer , 1, 1 , 1 , 64 , x , x , x , x , x , x , x , 65 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 66, TIDL_ConvolutionLayer , 1, 1 , 1 , 65 , x , x , x , x , x , x , x , 66 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 67, TIDL_ConvolutionLayer , 1, 1 , 1 , 66 , x , x , x , x , x , x , x , 67 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 68, TIDL_ConvolutionLayer , 1, 1 , 1 , 67 , x , x , x , x , x , x , x , 68 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 69, TIDL_EltWiseLayer , 1, 2 , 1 , 68 , 65 , x , x , x , x , x , x , 69 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 70, TIDL_BatchNormLayer , 1, 1 , 1 , 69 , x , x , x , x , x , x , x , 70 , 1 , 368 , 4 , 4 , 1 , 368 , 4 , 4 , 71, TIDL_DataLayer , 0, 1 , -1 , 70 , x , x , x , x , x , x , x , 0 , 1 , 368 , 4 , 4 , 0 , 0 , 0 , 0 , Layer ID ,inBlkWidth ,inBlkHeight ,inBlkPitch ,outBlkWidth ,outBlkHeight,outBlkPitch ,numInChs ,numOutChs ,numProcInChs,numLclInChs ,numLclOutChs,numProcItrs ,numAccItrs ,numHorBlock ,numVerBlock ,inBlkChPitch,outBlkChPitc,alignOrNot 1 40 34 40 32 32 32 3 32 3 1 8 1 3 1 1 1360 1024 1 2 32 32 32 32 32 32 32 24 32 7 8 1 5 1 1 1024 1024 1 3 40 36 40 16 16 16 8 8 8 4 8 1 2 1 1 1440 256 1 4 16 16 16 16 16 16 24 24 24 8 8 1 3 1 1 256 256 1 5 32 32 32 16 16 16 32 24 32 7 8 1 5 1 1 1024 256 1 8 16 16 16 16 16 16 24 56 24 8 8 1 3 1 1 256 256 1 9 24 18 24 16 16 16 8 8 8 4 8 1 2 1 1 432 256 1 10 16 16 16 16 16 16 56 56 56 8 8 1 7 1 1 256 256 1 11 16 16 16 16 16 16 24 56 24 8 8 1 3 1 1 256 256 1 14 16 16 16 16 16 16 56 152 56 8 8 1 7 1 1 256 256 1 15 40 20 40 16 8 16 8 8 8 4 8 1 2 1 1 800 128 1 16 16 8 16 16 8 16 152 152 152 8 8 1 19 1 1 128 128 1 17 32 16 32 16 8 16 56 152 56 8 8 1 7 1 1 512 128 1 20 16 8 16 16 8 16 152 152 152 8 8 1 19 1 1 128 128 1 21 24 10 24 16 8 16 8 8 8 4 8 1 2 1 1 240 128 1 22 16 8 16 16 8 16 152 152 152 8 8 1 19 1 1 128 128 1 25 16 8 16 16 8 16 152 152 152 8 8 1 19 1 1 128 128 1 26 24 10 24 16 8 16 8 8 8 4 8 1 2 1 1 240 128 1 27 16 8 16 16 8 16 152 152 152 8 8 1 19 1 1 128 128 1 30 16 8 16 16 8 16 152 152 152 8 8 1 19 1 1 128 128 1 31 24 10 24 16 8 16 8 8 8 4 8 1 2 1 1 240 128 1 32 16 8 16 16 8 16 152 152 152 8 8 1 19 1 1 128 128 1 35 16 8 16 16 8 16 152 368 152 8 8 1 19 1 1 128 128 1 36 40 12 40 16 4 16 8 8 8 4 8 1 2 1 1 480 64 1 37 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 38 32 8 32 16 4 16 152 368 152 8 8 1 19 1 1 256 64 1 41 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 42 24 6 24 16 4 16 8 8 8 4 8 1 2 1 1 144 64 1 43 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 46 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 47 24 6 24 16 4 16 8 8 8 4 8 1 2 1 1 144 64 1 48 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 51 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 52 24 6 24 16 4 16 8 8 8 4 8 1 2 1 1 144 64 1 53 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 56 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 57 24 6 24 16 4 16 8 8 8 4 8 1 2 1 1 144 64 1 58 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 61 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 62 24 6 24 16 4 16 8 8 8 4 8 1 2 1 1 144 64 1 63 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 66 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 67 24 6 24 16 4 16 8 8 8 4 8 1 2 1 1 144 64 1 68 16 4 16 16 4 16 368 368 368 8 8 1 46 1 1 64 64 1 Processing Frame Number : 0 Layer 1 : Out Q : 239 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.88, 1.17, Sparsity : -32.41 Layer 2 : Out Q : 324 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.79, 0.88, Sparsity : -12.50 Layer 3 : Out Q : 212 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.44, 0.44, Sparsity : 0.00 Layer 4 : Out Q : 95 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.15, 0.15, Sparsity : 0.00 Layer 5 : Out Q : 142 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.20, 0.22, Sparsity : -12.50 Layer 6 : Out Q : 87 , TIDL_EltWiseLayer, PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 7 : Out Q : 175 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 8 : Out Q : 452 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.34, 0.34, Sparsity : 0.00 Layer 9 : Out Q : 544 , TIDL_ConvolutionLayer, PASSED #MMACs = 1.03, 1.03, Sparsity : 0.00 Layer 10 : Out Q : 239 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.80, 0.80, Sparsity : 0.00 Layer 11 : Out Q : 148 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.34, 0.34, Sparsity : 0.00 Layer 12 : Out Q : 139 , TIDL_EltWiseLayer, PASSED #MMACs = 0.03, 0.03, Sparsity : 0.00 Layer 13 : Out Q : 279 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 14 : Out Q : 512 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.18, 2.18, Sparsity : 0.00 Layer 15 : Out Q : 570 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.70, 0.70, Sparsity : 0.00 Layer 16 : Out Q : 236 , TIDL_ConvolutionLayer, PASSED #MMACs = 1.48, 1.48, Sparsity : 0.00 Layer 17 : Out Q : 235 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.54, 0.54, Sparsity : 0.00 Layer 18 : Out Q : 124 , TIDL_EltWiseLayer, PASSED #MMACs = 0.02, 0.02, Sparsity : 0.00 Layer 19 : Out Q : 249 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 20 : Out Q : 621 , TIDL_ConvolutionLayer, PASSED #MMACs = 1.48, 1.48, Sparsity : 0.00 Layer 21 : Out Q : 670 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.70, 0.70, Sparsity : 0.00 Layer 22 : Out Q : 267 , TIDL_ConvolutionLayer, PASSED #MMACs = 1.48, 1.48, Sparsity : 0.00 Layer 23 : Out Q : 131 , TIDL_EltWiseLayer, PASSED #MMACs = 0.02, 0.02, Sparsity : 0.00 Layer 24 : Out Q : 263 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 25 : Out Q : 680 , TIDL_ConvolutionLayer, PASSED #MMACs = 1.48, 1.48, Sparsity : 0.00 Layer 26 : Out Q : 271 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.70, 0.70, Sparsity : 0.00 Layer 27 : Out Q : 224 , TIDL_ConvolutionLayer, PASSED #MMACs = 1.48, 1.48, Sparsity : 0.00 Layer 28 : Out Q : 158 , TIDL_EltWiseLayer, PASSED #MMACs = 0.02, 0.02, Sparsity : 0.00 Layer 29 : Out Q : 317 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 30 : Out Q : 884 , TIDL_ConvolutionLayer, PASSED #MMACs = 1.48, 1.48, Sparsity : 0.00 Layer 31 : Out Q : 665 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.70, 0.70, Sparsity : 0.00 Layer 32 : Out Q : 556 , TIDL_ConvolutionLayer, PASSED #MMACs = 1.48, 1.48, Sparsity : 0.00 Layer 33 : Out Q : 150 , TIDL_EltWiseLayer, PASSED #MMACs = 0.02, 0.02, Sparsity : 0.00 Layer 34 : Out Q : 301 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 35 : Out Q : 870 , TIDL_ConvolutionLayer, PASSED #MMACs = 3.58, 3.58, Sparsity : 0.00 Layer 36 : Out Q : 409 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.42, 0.42, Sparsity : 0.00 Layer 37 : Out Q : 230 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 38 : Out Q : 199 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.89, 0.89, Sparsity : 0.00 Layer 39 : Out Q : 175 , TIDL_EltWiseLayer, PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 40 : Out Q : 351 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 41 : Out Q : 1196 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 42 : Out Q : 483 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.42, 0.42, Sparsity : 0.00 Layer 43 : Out Q : 220 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 44 : Out Q : 158 , TIDL_EltWiseLayer, PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 45 : Out Q : 320 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 46 : Out Q : 1105 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 47 : Out Q : 459 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.42, 0.42, Sparsity : 0.00 Layer 48 : Out Q : 137 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 49 : Out Q : 136 , TIDL_EltWiseLayer, PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 50 : Out Q : 287 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 51 : Out Q : 1024 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 52 : Out Q : 631 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.42, 0.42, Sparsity : 0.00 Layer 53 : Out Q : 178 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 54 : Out Q : 109 , TIDL_EltWiseLayer, PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 55 : Out Q : 219 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 56 : Out Q : 1488 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 57 : Out Q : 450 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.42, 0.42, Sparsity : 0.00 Layer 58 : Out Q : 143 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 59 : Out Q : 106 , TIDL_EltWiseLayer, PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 60 : Out Q : 213 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 61 : Out Q : 1198 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 62 : Out Q : 556 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.42, 0.42, Sparsity : 0.00 Layer 63 : Out Q : 150 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 64 : Out Q : 102 , TIDL_EltWiseLayer, PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 65 : Out Q : 205 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 66 : Out Q : 1259 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 67 : Out Q : 749 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.42, 0.42, Sparsity : 0.00 Layer 68 : Out Q : 181 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.17, 2.17, Sparsity : 0.00 Layer 69 : Out Q : 103 , TIDL_EltWiseLayer, PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Layer 70 : Out Q : 207 , TIDL_BatchNormLayer , PASSED #MMACs = 0.01, 0.01, Sparsity : 0.00 Hi Praveen Is there any updates? lluo Hi IIuo, I have looked at the log you shared, it looks okay to me, this configurations is supported on TDA2. Please follow steps mentioned in section 3.8 (Matching TIDL inference result) in the TIDL user guide. BTW, what is the TIDL release version that you are using? Thanks, Praveen Hi Praveen I have been using REL.TIDL.01.01.03.00 . And the model & tools has been uploaded in regnet_import.zip. I will follow user guide to check my model on REL.TIDL.01.02.00.00 later. Hi, Please check with REL.TIDL.01.02.00.00 and let us know. Thanks, Praveen