Issues

• Output size ignores dilation in cutlassConv2d_withPadding_smallAlignment

  thu-nics/MixDQ#22 · Filter by author Pushazf opened on Oct 29, 2025

• How to get act_ratio_config and weight_ratio_config in tensor_ratio?

  thu-nics/MixDQ#21 · Filter by author Godzilla0517 opened on Sep 22, 2025

• Is it possible to publish LCM-lora quantization recipes?

  thu-nics/MixDQ#20 · Filter by author china-mashina opened on Sep 8, 2025

• Why save the active scale and zero_point?

  thu-nics/MixDQ#19 · Filter by author songh11 opened on Feb 25, 2025

• Missing nodes on ComfyUI & ComfyUI Manager can't find the nodes LoadPipeline and MixdqIntegral

  thu-nics/MixDQ#18 · Filter by author alvarest-valentino opened on Feb 16, 2025

• How reproduce the FID

  thu-nics/MixDQ#13 · Filter by author suppwilly opened on Aug 14, 2024
  1

• libpeng error for mixed-precision process

  thu-nics/MixDQ#12 · Filter by author Harahan opened on Jul 25, 2024
  1

• is MixDQ a PTQ or QAT method？

  thu-nics/MixDQ#11 · Filter by author LiMa-cas opened on Jul 20, 2024
  6

• how many gpu hours it costs during your The Mixed Precision Search Process?

  thu-nics/MixDQ#10 · Filter by author greasebig opened on Jun 28, 2024
  1

• got bad results: quant with 1024 image using sdxl-tubo, using 1.3.1 Normal infer, follow your steps

  thu-nics/MixDQ#9 · Filter by author greasebig opened on Jun 28, 2024
  17

• Can it be used in Controlnet and IP adapter? Why am I using an error message? Is there a special method for this

  thu-nics/MixDQ#4 · Filter by author libai-lab opened on Jun 8, 2024
  1

• does onnx can support the model an inference with onnxruntime?

  thu-nics/MixDQ#3 · Filter by author dragen1860 opened on Jun 7, 2024
  2

• Can it be used in sd1.5 and can it be combined with other acceleration methods such as ByteDance/Hyper SD

  thu-nics/MixDQ#1 · Filter by author libai-lab opened on Jun 4, 2024
  6