TDA2EXEVM: How could validate the TIDL model accuracy with own test data set?

Part Number: TDA2EXEVM

Hi:

   I had traind a simple classification model with ticaffe,got 90% accuracy on my test data set(about 1000 pictures);

   then I convert ticaffe model to TIDL model;and use the "quantStatsTool" ,I test 10 picture datas manual，the classifacation result is correct;and here is my infer file:  

  rawImage    = 0
 numFrames   = 1
 preProcType  = 3
 inData   = ..\..\test\testvecs\input\1040.jpg
 traceDumpBaseName   = ".\trace_dump_"
 outData   = ".\stats_tool_out.bin"
 updateNetWithStats   = 0

 ......

   But I have 1000 test datas(more later)，how can I automatic test the 1000 datas one time with "quantStatsTool" and get the accuracy ？not only test 1 data one time,so need repeated 1000(or more) times?

  Pls help me!Thank you.