Other Parts Discussed in Thread: INA226
Tool/software: TI-RTOS
Hey,
I have a keras model that I imported to TIDL. Importing seems to have worked fine.
The model basically just takes an .png input file and classifies the picture in 0 and 1.
Now when I run the model in host emulation mode (eve_test_dl_algo.out.exe) and read the quat_stats_tool for generated output I get the following results for the images (Image_*_Class_* is the name of the image and what class it is supposed to be classified to):
Image_0_Class_0 = FE 00
Image_1_Class_0 = FE 00
Image_2_Class_0 = FE 00
....... Output for images of class 0 is always the same.
Image_0_Class_1 = C7 37
Image_1_Class_1 = 07 F7
Image_2_Class_1 = 1E E0
Image_3_Class_1 = E9 15
etc. so for images of class 1 tjhe results is different most of the time but with the sum of first Hex and second Hex always resulting to FE.
Somehow this tells me the model is working in a way, since it´s able to make a classification, but the numbers don´t really make sense to me.
I added a .txt file that contains the output of the console when runing the eve_test_dl_algo_out.exe
Do you have an idea on how to interprete those number?
thank you
Nico
Processing config file ..\..\test\testvecs\config\infer\tidl_config.txt ! 0, TIDL_DataLayer , 0, -1 , 1 , x , x , x , x , x , x , x , x , 0 , 0 , 0 , 0 , 0 , 1 , 3 , 227 , 227 , 1, TIDL_ConvolutionLayer , 1, 1 , 1 , 0 , x , x , x , x , x , x , x , 1 , 1 , 3 , 227 , 227 , 1 , 64 , 113 , 113 , 2, TIDL_PoolingLayer , 1, 1 , 1 , 1 , x , x , x , x , x , x , x , 2 , 1 , 64 , 113 , 113 , 1 , 64 , 56 , 56 , 3, TIDL_ConvolutionLayer , 1, 1 , 1 , 2 , x , x , x , x , x , x , x , 3 , 1 , 64 , 56 , 56 , 1 , 16 , 56 , 56 , 4, TIDL_ConvolutionLayer , 1, 1 , 1 , 3 , x , x , x , x , x , x , x , 4 , 1 , 16 , 56 , 56 , 1 , 64 , 56 , 56 , 5, TIDL_ConvolutionLayer , 1, 1 , 1 , 3 , x , x , x , x , x , x , x , 5 , 1 , 16 , 56 , 56 , 1 , 64 , 56 , 56 , 6, TIDL_ConcatLayer , 1, 2 , 1 , 4 , 5 , x , x , x , x , x , x , 6 , 1 , 64 , 56 , 56 , 1 , 128 , 56 , 56 , 7, TIDL_ConvolutionLayer , 1, 1 , 1 , 6 , x , x , x , x , x , x , x , 7 , 1 , 128 , 56 , 56 , 1 , 16 , 56 , 56 , 8, TIDL_ConvolutionLayer , 1, 1 , 1 , 7 , x , x , x , x , x , x , x , 8 , 1 , 16 , 56 , 56 , 1 , 64 , 56 , 56 , 9, TIDL_ConvolutionLayer , 1, 1 , 1 , 7 , x , x , x , x , x , x , x , 9 , 1 , 16 , 56 , 56 , 1 , 64 , 56 , 56 , 10, TIDL_ConcatLayer , 1, 2 , 1 , 8 , 9 , x , x , x , x , x , x , 10 , 1 , 64 , 56 , 56 , 1 , 128 , 56 , 56 , 11, TIDL_PoolingLayer , 1, 1 , 1 , 10 , x , x , x , x , x , x , x , 11 , 1 , 128 , 56 , 56 , 1 , 128 , 27 , 27 , 12, TIDL_ConvolutionLayer , 1, 1 , 1 , 11 , x , x , x , x , x , x , x , 12 , 1 , 128 , 27 , 27 , 1 , 32 , 27 , 27 , 13, TIDL_ConvolutionLayer , 1, 1 , 1 , 12 , x , x , x , x , x , x , x , 13 , 1 , 32 , 27 , 27 , 1 , 128 , 27 , 27 , 14, TIDL_ConvolutionLayer , 1, 1 , 1 , 12 , x , x , x , x , x , x , x , 14 , 1 , 32 , 27 , 27 , 1 , 128 , 27 , 27 , 15, TIDL_ConcatLayer , 1, 2 , 1 , 13 , 14 , x , x , x , x , x , x , 15 , 1 , 128 , 27 , 27 , 1 , 256 , 27 , 27 , 16, TIDL_ConvolutionLayer , 1, 1 , 1 , 15 , x , x , x , x , x , x , x , 16 , 1 , 256 , 27 , 27 , 1 , 32 , 27 , 27 , 17, TIDL_ConvolutionLayer , 1, 1 , 1 , 16 , x , x , x , x , x , x , x , 17 , 1 , 32 , 27 , 27 , 1 , 128 , 27 , 27 , 18, TIDL_ConvolutionLayer , 1, 1 , 1 , 16 , x , x , x , x , x , x , x , 18 , 1 , 32 , 27 , 27 , 1 , 128 , 27 , 27 , 19, TIDL_ConcatLayer , 1, 2 , 1 , 17 , 18 , x , x , x , x , x , x , 19 , 1 , 128 , 27 , 27 , 1 , 256 , 27 , 27 , 20, TIDL_PoolingLayer , 1, 1 , 1 , 19 , x , x , x , x , x , x , x , 20 , 1 , 256 , 27 , 27 , 1 , 256 , 13 , 13 , 21, TIDL_ConvolutionLayer , 1, 1 , 1 , 20 , x , x , x , x , x , x , x , 21 , 1 , 256 , 13 , 13 , 1 , 48 , 13 , 13 , 22, TIDL_ConvolutionLayer , 1, 1 , 1 , 21 , x , x , x , x , x , x , x , 22 , 1 , 48 , 13 , 13 , 1 , 192 , 13 , 13 , 23, TIDL_ConvolutionLayer , 1, 1 , 1 , 21 , x , x , x , x , x , x , x , 23 , 1 , 48 , 13 , 13 , 1 , 192 , 13 , 13 , 24, TIDL_ConcatLayer , 1, 2 , 1 , 22 , 23 , x , x , x , x , x , x , 24 , 1 , 192 , 13 , 13 , 1 , 384 , 13 , 13 , 25, TIDL_ConvolutionLayer , 1, 1 , 1 , 24 , x , x , x , x , x , x , x , 25 , 1 , 384 , 13 , 13 , 1 , 64 , 13 , 13 , 26, TIDL_ConvolutionLayer , 1, 1 , 1 , 25 , x , x , x , x , x , x , x , 26 , 1 , 64 , 13 , 13 , 1 , 256 , 13 , 13 , 27, TIDL_ConvolutionLayer , 1, 1 , 1 , 25 , x , x , x , x , x , x , x , 27 , 1 , 64 , 13 , 13 , 1 , 256 , 13 , 13 , 28, TIDL_ConcatLayer , 1, 2 , 1 , 26 , 27 , x , x , x , x , x , x , 28 , 1 , 256 , 13 , 13 , 1 , 512 , 13 , 13 , 29, TIDL_ConvolutionLayer , 1, 1 , 1 , 28 , x , x , x , x , x , x , x , 29 , 1 , 512 , 13 , 13 , 1 , 128 , 13 , 13 , 30, TIDL_ConvolutionLayer , 1, 1 , 1 , 29 , x , x , x , x , x , x , x , 30 , 1 , 128 , 13 , 13 , 1 , 512 , 13 , 13 , 31, TIDL_ConvolutionLayer , 1, 1 , 1 , 29 , x , x , x , x , x , x , x , 31 , 1 , 128 , 13 , 13 , 1 , 512 , 13 , 13 , 32, TIDL_ConcatLayer , 1, 2 , 1 , 30 , 31 , x , x , x , x , x , x , 32 , 1 , 512 , 13 , 13 , 1 , 1024 , 13 , 13 , 33, TIDL_ConvolutionLayer , 1, 1 , 1 , 32 , x , x , x , x , x , x , x , 33 , 1 , 1024 , 13 , 13 , 1 , 2 , 13 , 13 , 34, TIDL_PoolingLayer , 1, 1 , 1 , 33 , x , x , x , x , x , x , x , 34 , 1 , 2 , 13 , 13 , 1 , 1 , 1 , 2 , 35, TIDL_SoftMaxLayer , 1, 1 , 1 , 34 , x , x , x , x , x , x , x , 35 , 1 , 1 , 1 , 2 , 1 , 1 , 1 , 2 , 36, TIDL_DataLayer , 0, 1 , -1 , 35 , x , x , x , x , x , x , x , 0 , 1 , 1 , 1 , 2 , 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 72 68 72 32 32 32 3 64 3 1 8 1 3 4 4 4896 1024 1 3 32 28 32 32 28 32 64 16 64 8 8 1 8 2 2 896 896 1 4 32 28 32 32 28 32 16 64 16 8 8 1 2 2 2 896 896 1 5 40 30 40 32 28 32 16 64 16 7 8 1 3 2 2 1200 896 1 7 32 28 32 32 28 32 128 16 128 8 8 1 16 2 2 896 896 1 8 32 28 32 32 28 32 16 64 16 8 8 1 2 2 2 896 896 1 9 40 30 40 32 28 32 16 64 16 7 8 1 3 2 2 1200 896 1 12 32 27 32 32 27 32 128 32 128 8 8 1 16 1 1 864 864 1 13 32 27 32 32 27 32 32 128 32 8 8 1 4 1 1 864 864 1 14 40 29 40 32 27 32 32 128 32 8 8 1 4 1 1 1160 864 1 16 32 27 32 32 27 32 256 32 256 8 8 1 32 1 1 864 864 1 17 32 27 32 32 27 32 32 128 32 8 8 1 4 1 1 864 864 1 18 40 29 40 32 27 32 32 128 32 8 8 1 4 1 1 1160 864 1 21 16 13 16 16 13 16 256 48 256 8 8 1 32 1 1 208 208 1 22 16 13 16 16 13 16 48 192 48 8 8 1 6 1 1 208 208 1 23 24 15 24 16 13 16 48 192 48 8 8 1 6 1 1 360 208 1 25 16 13 16 16 13 16 384 64 384 8 8 1 48 1 1 208 208 1 26 16 13 16 16 13 16 64 256 64 8 8 1 8 1 1 208 208 1 27 24 15 24 16 13 16 64 256 64 8 8 1 8 1 1 360 208 1 29 16 13 16 16 13 16 512 128 512 8 8 1 64 1 1 208 208 1 30 16 13 16 16 13 16 128 512 128 8 8 1 16 1 1 208 208 1 31 24 15 24 16 13 16 128 512 128 8 8 1 16 1 1 360 208 1 33 16 13 16 16 13 16 1024 2 1024 8 2 1 128 1 1 208 208 1 Processing Frame Number : 0 Layer 1 : Out Q : 13559 , TIDL_ConvolutionLayer, PASSED #MMACs = 22.06, 27.07, Sparsity : -22.69 Layer 2 :TIDL_PoolingLayer, PASSED #MMACs = 0.20, 0.20, Sparsity : 0.00 Layer 3 : Out Q : 10538 , TIDL_ConvolutionLayer, PASSED #MMACs = 3.21, 3.21, Sparsity : 0.00 Layer 4 : Out Q : 16444 , TIDL_ConvolutionLayer, PASSED #MMACs = 3.21, 3.17, Sparsity : 1.17 Layer 5 : Out Q : 10998 , TIDL_ConvolutionLayer, PASSED #MMACs = 28.90, 28.15, Sparsity : 2.60 Layer 6 : Out Q : 11041 , TIDL_ConcatLayer, PASSED #MMACs = 0.00, 0.00, Sparsity : -1.#J Layer 7 : Out Q : 10030 , TIDL_ConvolutionLayer, PASSED #MMACs = 6.42, 6.42, Sparsity : 0.00 Layer 8 : Out Q : 14115 , TIDL_ConvolutionLayer, PASSED #MMACs = 3.21, 3.21, Sparsity : 0.00 Layer 9 : Out Q : 11907 , TIDL_ConvolutionLayer, PASSED #MMACs = 28.90, 28.51, Sparsity : 1.35 Layer 10 : Out Q : 11954 , TIDL_ConcatLayer, PASSED #MMACs = 0.00, 0.00, Sparsity : -1.#J Layer 11 :TIDL_PoolingLayer, PASSED #MMACs = 0.09, 0.09, Sparsity : 0.00 Layer 12 : Out Q : 7088 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.99, 2.98, Sparsity : 0.10 Layer 13 : Out Q : 14662 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.99, 2.99, Sparsity : 0.00 Layer 14 : Out Q : 6430 , TIDL_ConvolutionLayer, PASSED #MMACs = 26.87, 25.91, Sparsity : 3.59 Layer 15 : Out Q : 6455 , TIDL_ConcatLayer, PASSED #MMACs = 0.00, 0.00, Sparsity : -1.#J Layer 16 : Out Q : 7759 , TIDL_ConvolutionLayer, PASSED #MMACs = 5.97, 5.97, Sparsity : 0.00 Layer 17 : Out Q : 15378 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.99, 2.96, Sparsity : 0.78 Layer 18 : Out Q : 7969 , TIDL_ConvolutionLayer, PASSED #MMACs = 26.87, 25.88, Sparsity : 3.70 Layer 19 : Out Q : 8000 , TIDL_ConcatLayer, PASSED #MMACs = 0.00, 0.00, Sparsity : -1.#J Layer 20 :TIDL_PoolingLayer, PASSED #MMACs = 0.04, 0.04, Sparsity : 0.00 Layer 21 : Out Q : 7431 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.08, 2.08, Sparsity : 0.03 Layer 22 : Out Q : 13602 , TIDL_ConvolutionLayer, PASSED #MMACs = 1.56, 1.56, Sparsity : 0.00 Layer 23 : Out Q : 8083 , TIDL_ConvolutionLayer, PASSED #MMACs = 14.02, 13.47, Sparsity : 3.93 Layer 24 : Out Q : 8115 , TIDL_ConcatLayer, PASSED #MMACs = 0.00, 0.00, Sparsity : -1.#J Layer 25 : Out Q : 21971 , TIDL_ConvolutionLayer, PASSED #MMACs = 4.15, 4.15, Sparsity : 0.00 Layer 26 : Out Q : 35565 , TIDL_ConvolutionLayer, PASSED #MMACs = 2.77, 2.77, Sparsity : 0.00 Layer 27 : Out Q : 15754 , TIDL_ConvolutionLayer, PASSED #MMACs = 24.92, 24.90, Sparsity : 0.08 Layer 28 : Out Q : 15816 , TIDL_ConcatLayer, PASSED #MMACs = 0.00, 0.00, Sparsity : -1.#J Layer 29 : Out Q : 9723 , TIDL_ConvolutionLayer, PASSED #MMACs = 11.08, 11.08, Sparsity : 0.00 Layer 30 : Out Q : 17323 , TIDL_ConvolutionLayer, PASSED #MMACs = 11.08, 11.08, Sparsity : 0.00 Layer 31 : Out Q : 4192 , TIDL_ConvolutionLayer, PASSED #MMACs = 99.68, 99.54, Sparsity : 0.14 Layer 32 : Out Q : 4209 , TIDL_ConcatLayer, PASSED #MMACs = 0.00, 0.00, Sparsity : -1.#J Layer 33 : Out Q : 4120 , TIDL_ConvolutionLayer, PASSED #MMACs = 0.35, 0.25, Sparsity : 27.93 Layer 34 : Out Q : 8816 , TIDL_PoolingLayer, PASSED #MMACs = 0.00, 0.00, Sparsity : 0.00 Layer 35 :-------Max Index 0 : 254 ------- #MMACs = 0.00, 0.00, Sparsity : 0.00 End of config list found !