Issues

• Is there any way to compile with cuda from source code ?

  Tiiiger/QPyTorch#73 · Filter by author HiIcy opened on Dec 24, 2025

• RuntimeError: Ninja is required to load C++ extensions

  Tiiiger/QPyTorch#70 · Filter by author pin-xue-lou opened on Jun 14, 2024
  2

• float_quantize at multi-gpu works wrong.

  Tiiiger/QPyTorch#67 · Filter by author jinsol-neubla opened on Mar 16, 2023

• is there any way to control exponent bias?

  Tiiiger/QPyTorch#65 · Filter by author Moreh-LeeJunhyeok opened on Jan 2, 2023

• quant_cuda does not compile.

  Tiiiger/QPyTorch#64 · Filter by author 4rrw opened on Nov 21, 2022

• 'gbk' codec can't decode

  Tiiiger/QPyTorch#62 · Filter by author ckinsung85 opened on Sep 21, 2022

• Code Freezes When Using QPyTorch

  Tiiiger/QPyTorch#60 · Filter by author hasnainnaeem opened on Sep 4, 2022
  9

• Why the gradient scaling factor is multiplied before quantization?

  Tiiiger/QPyTorch#59 · Filter by author Guangxuan-Xiao opened on Aug 17, 2022

• Model Export to ONNX dose not work due to quant functions

  Tiiiger/QPyTorch#56 · Filter by author JagdishKolhe opened on Jan 27, 2022

• Floatpoint(8,23)flips the input values

  Tiiiger/QPyTorch#48 · Filter by author ASHWIN2605 opened on Jul 8, 2021
  4