Tiiiger / QPyTorch

Low Precision Arithmetic Simulation in PyTorch

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Nearest Rounding Kernel

Tiiiger opened this issue · comments

Right now I have only implemented stochastic rounding CUDA kernel. We need to add nearest rounding CUDA kernel.

Done. Need testing

Need to modify quant_module. Please do.

Need to Implement nearest rounding for float kernel. See CUDA implementation. You can ignore clip_exponent for now. @linzhiqiu