See also the older blog post about the OpenGL/Metal backend at https://blog.tensorflow.org/2019/01/tensorflow-lite-now-faster-with-mobile.html https://blog.tens

See also the older blog post about the OpenGL/Metal backend at https://blog.tensorflow.org/2019/01/tensorflow-lite-now-faster-with-mobile.html https://blog.tensorflow.org/2019/01/tensorflow-lite-now-fast... Would be interesting to see updated graphs about iOS vs Android that the OpenGL post had.