Hacker News new | past | comments | ask | show | jobs | submit login

Yes - here's a doc about doing quantized inference in TensorFlow, for example: https://www.tensorflow.org/performance/quantization

AFAIK, there's still a bit of a performance gap between just using TF and using the specialized gemmlowp library on Android, but that part's getting cleaned up.

Haven't seen much in generalized results on training using lower precision.




Does that work with Pascal CUDA8 INT8 out of the box?


I'm not sure - I believe it depends on getting cuDNN6 working, and from this bug, I can't quite tell if it works or not (but it's probably not officially supported yet): https://github.com/tensorflow/tensorflow/issues/8828




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: