search

hollance / Forge

A neural network toolkit for Metal
MIT License
1.27k stars 173 forks source link

Quantized model from Tensorflow #17

Closed dongwu92 closed 7 years ago

dongwu92 commented 7 years ago

Did Forge support 8-bit float value for quantized model from tensorflow?

hollance commented 7 years ago

It currently does not. But you can add this yourself by writing a ParameterLoader class that un-quantizes the weights back to floats.

dongwu92 commented 7 years ago

Good idea! Thank you!