/qnn

Training neural nets with quantized weights on arbitrarily specified bit-depth

Primary LanguagePython

Watchers