/home/root/ti/EmbeddedSW-Aptiv-OnnxModelIntegrationTidl/scripts/../models_artifacts/topk_model_artifacts/tempDir/subgraph_0_tidl_net.bin Layer 0: TIDL_DataLayer "X"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   numChannels=0 dataQ=0   actParams:      actType=TIDL_Clip      slopeScale=0.000000 clipMin/Max=(0.755766,4.676167)   Graph compiler Info:       space:DDR baseMem(bytes):0-1049600 size(elem):524800       bufferWidth(elem):524288 activeBufferWidth(elem):524288 bufferHeight(elem):1       padRows:0 padCols:0   Outputs:      [0] numDim=6 dims=[1,1,1,1,64,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[524288,524288,524288,524288,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.755766,4.676167)         tensorScale=4096.000000         tensorZeroPoint=0 Data [0 0] Layer 1: TIDL_TopKLayer "tidl_topKcustom_2"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   K=3, axis=4   largest=1, sorted=1   actParams:      actType=TIDL_NoAct      slopeScale=0.000000 clipMin/Max=(0.000000,0.000000)   Graph compiler Info:       space:MSMC baseMem(bytes):0-147584 size(elem):73792       bufferWidth(elem):24576 activeBufferWidth(elem):24576 bufferHeight(elem):3       padRows:0 padCols:0   Outputs:      [1] numDim=0 dims=[1,1,1,3,3,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[73728,73728,73728,24576,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.774658,4.480469)         tensorScale=4096.000000         tensorZeroPoint=0 TopK [1 1] Layer 0: TIDL_DataLayer "X"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   numChannels=0 dataQ=0   actParams:      actType=TIDL_Clip      slopeScale=0.000000 clipMin/Max=(0.755766,4.676167)   Graph compiler Info:       space:DDR baseMem(bytes):0-1049600 size(elem):524800       bufferWidth(elem):524288 activeBufferWidth(elem):524288 bufferHeight(elem):1       padRows:0 padCols:0   Outputs:      [0] numDim=6 dims=[1,1,1,1,64,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[524288,524288,524288,524288,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.755766,4.676167)         tensorScale=4096.000000         tensorZeroPoint=0 ->Layer 1: TIDL_TopKLayer "tidl_topKcustom_2"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   K=3, axis=4   largest=1, sorted=1   actParams:      actType=TIDL_NoAct      slopeScale=0.000000 clipMin/Max=(0.000000,0.000000)   Graph compiler Info:       space:MSMC baseMem(bytes):0-147584 size(elem):73792       bufferWidth(elem):24576 activeBufferWidth(elem):24576 bufferHeight(elem):3       padRows:0 padCols:0   Outputs:      [1] numDim=0 dims=[1,1,1,3,3,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[73728,73728,73728,24576,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.774658,4.480469)         tensorScale=4096.000000         tensorZeroPoint=0 1x1x1x1x64x8192 Layer 2: TIDL_SliceLayer "Values"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   slicePoints=0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0   axis=3 stride=1   , actParams:      actType=TIDL_NoAct      slopeScale=0.000000 clipMin/Max=(0.000000,0.000000)   Graph compiler Info:       space:DDR baseMem(bytes):0-50176 size(elem):25088       bufferWidth(elem):24576 activeBufferWidth(elem):24576 bufferHeight(elem):1       padRows:0 padCols:0   Outputs:      [2] numDim=0 dims=[1,1,1,1,3,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[24576,24576,24576,24576,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.774658,4.480469)         tensorScale=4096.000000         tensorZeroPoint=0 Slice [2 2] Layer 1: TIDL_TopKLayer "tidl_topKcustom_2"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   K=3, axis=4   largest=1, sorted=1   actParams:      actType=TIDL_NoAct      slopeScale=0.000000 clipMin/Max=(0.000000,0.000000)   Graph compiler Info:       space:MSMC baseMem(bytes):0-147584 size(elem):73792       bufferWidth(elem):24576 activeBufferWidth(elem):24576 bufferHeight(elem):3       padRows:0 padCols:0   Outputs:      [1] numDim=0 dims=[1,1,1,3,3,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[73728,73728,73728,24576,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.774658,4.480469)         tensorScale=4096.000000         tensorZeroPoint=0 ->Layer 2: TIDL_SliceLayer "Values"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   slicePoints=0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0   axis=3 stride=1   , actParams:      actType=TIDL_NoAct      slopeScale=0.000000 clipMin/Max=(0.000000,0.000000)   Graph compiler Info:       space:DDR baseMem(bytes):0-50176 size(elem):25088       bufferWidth(elem):24576 activeBufferWidth(elem):24576 bufferHeight(elem):1       padRows:0 padCols:0   Outputs:      [2] numDim=0 dims=[1,1,1,1,3,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[24576,24576,24576,24576,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.774658,4.480469)         tensorScale=4096.000000         tensorZeroPoint=0 1x1x1x3x3x8192 Layer 3: TIDL_SliceLayer "Indices"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   slicePoints=1,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0   axis=3 stride=1   , actParams:      actType=TIDL_NoAct      slopeScale=0.000000 clipMin/Max=(0.000000,0.000000)   Graph compiler Info:       space:DDR baseMem(bytes):0-99328 size(elem):49664       bufferWidth(elem):24576 activeBufferWidth(elem):24576 bufferHeight(elem):2       padRows:0 padCols:0   Outputs:      [3] numDim=0 dims=[1,1,1,2,3,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[49152,49152,49152,24576,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.000000,0.015381)         tensorScale=4096.000000         tensorZeroPoint=0 Slice [3 3] Layer 1: TIDL_TopKLayer "tidl_topKcustom_2"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   K=3, axis=4   largest=1, sorted=1   actParams:      actType=TIDL_NoAct      slopeScale=0.000000 clipMin/Max=(0.000000,0.000000)   Graph compiler Info:       space:MSMC baseMem(bytes):0-147584 size(elem):73792       bufferWidth(elem):24576 activeBufferWidth(elem):24576 bufferHeight(elem):3       padRows:0 padCols:0   Outputs:      [1] numDim=0 dims=[1,1,1,3,3,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[73728,73728,73728,24576,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.774658,4.480469)         tensorScale=4096.000000         tensorZeroPoint=0 ->Layer 3: TIDL_SliceLayer "Indices"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   slicePoints=1,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0   axis=3 stride=1   , actParams:      actType=TIDL_NoAct      slopeScale=0.000000 clipMin/Max=(0.000000,0.000000)   Graph compiler Info:       space:DDR baseMem(bytes):0-99328 size(elem):49664       bufferWidth(elem):24576 activeBufferWidth(elem):24576 bufferHeight(elem):2       padRows:0 padCols:0   Outputs:      [3] numDim=0 dims=[1,1,1,2,3,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[49152,49152,49152,24576,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.000000,0.015381)         tensorScale=4096.000000         tensorZeroPoint=0 1x1x1x3x3x8192 Layer 4: TIDL_DataLayer "X"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   numChannels=0 dataQ=0   actParams:      actType=TIDL_Clip      slopeScale=0.000000 clipMin/Max=(0.000000,340282346638528859811704183484516925440.000000)   Graph compiler Info:       space:DDR baseMem(bytes):0-1049600 size(elem):1049600       bufferWidth(elem):1048576 activeBufferWidth(elem):0 bufferHeight(elem):0       padRows:0 padCols:0 Data [4 0] Layer 2: TIDL_SliceLayer "Values"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   slicePoints=0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0   axis=3 stride=1   , actParams:      actType=TIDL_NoAct      slopeScale=0.000000 clipMin/Max=(0.000000,0.000000)   Graph compiler Info:       space:DDR baseMem(bytes):0-50176 size(elem):25088       bufferWidth(elem):24576 activeBufferWidth(elem):24576 bufferHeight(elem):1       padRows:0 padCols:0   Outputs:      [2] numDim=0 dims=[1,1,1,1,3,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[24576,24576,24576,24576,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.774658,4.480469)         tensorScale=4096.000000         tensorZeroPoint=0 ->Layer 4: TIDL_DataLayer "X"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   numChannels=0 dataQ=0   actParams:      actType=TIDL_Clip      slopeScale=0.000000 clipMin/Max=(0.000000,340282346638528859811704183484516925440.000000)   Graph compiler Info:       space:DDR baseMem(bytes):0-1049600 size(elem):1049600       bufferWidth(elem):1048576 activeBufferWidth(elem):0 bufferHeight(elem):0       padRows:0 padCols:0 1x1x1x1x3x8192 Layer 5: TIDL_DataLayer "X"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   numChannels=0 dataQ=0   actParams:      actType=TIDL_Clip      slopeScale=0.000000 clipMin/Max=(0.000000,340282346638528859811704183484516925440.000000)   Graph compiler Info:       space:DDR baseMem(bytes):0-1049600 size(elem):1049600       bufferWidth(elem):1048576 activeBufferWidth(elem):0 bufferHeight(elem):0       padRows:0 padCols:0 Data [5 0] Layer 3: TIDL_SliceLayer "Indices"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   slicePoints=1,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0   axis=3 stride=1   , actParams:      actType=TIDL_NoAct      slopeScale=0.000000 clipMin/Max=(0.000000,0.000000)   Graph compiler Info:       space:DDR baseMem(bytes):0-99328 size(elem):49664       bufferWidth(elem):24576 activeBufferWidth(elem):24576 bufferHeight(elem):2       padRows:0 padCols:0   Outputs:      [3] numDim=0 dims=[1,1,1,2,3,8192]  elementType=TIDL_SignedShort         padH/W=[0,0] batchPadH/W=[0,0] numBatchH/W=[1,1] pitch=[49152,49152,49152,24576,8192]         dataQ=0 roundBits=0         min/maxValue=(0,0) min/maxTensorValue=(0.000000,0.015381)         tensorScale=4096.000000         tensorZeroPoint=0 ->Layer 5: TIDL_DataLayer "X"   weightsElementSizeInBits=16   kernelType=1   multiCoreMode=TIDL_NOT_MULTI_CORE   strideOffsetMethod=TIDL_StrideOffsetTopLeft   numChannels=0 dataQ=0   actParams:      actType=TIDL_Clip      slopeScale=0.000000 clipMin/Max=(0.000000,340282346638528859811704183484516925440.000000)   Graph compiler Info:       space:DDR baseMem(bytes):0-1049600 size(elem):1049600       bufferWidth(elem):1048576 activeBufferWidth(elem):0 bufferHeight(elem):0       padRows:0 padCols:0 1x1x1x2x3x8192
×

NODE PROPERTIES

INPUTS

OUTPUTS

ATTRIBUTES