Inference with input precision FP16

TCE Options

TCE Open Date: 

Tuesday, November 26, 2019 - 21:00

Inference with input precision FP16


I tried to inference with model precision FP16, input precision FP16 and output precision FP32 (FP16 also) on Movidius Myriad X and Intel HD Graphics.

Inference has completed, but output value is incorrect. What should I do?

 

Thank you.

3 posts / 0 new

Best Reply

Hi, Atsushi,

Did you use public model? if so, which one did you use? Also please provide the way to reproduce your result, required document, file is needed, thank you.


Hi Cary P.

Thank you for replying.

I found bug in my source code.

After debug my code, output value became correct.

 

Thank you.

Leave a Comment

Please sign in to add a comment. Not a member? Join today