Issues

• DeepCompressor2: a fork built so an agent can bring up a new model — multi-GPU calibration, disk-backed weights, faster search

  nunchux-ai/deepcompressor#125 · Filter by author lantudou opened on Aug 19, 2026
  1

• 如何设置rank为128正常出图

  nunchux-ai/deepcompressor#123 · Filter by author lzcchl opened on Jun 12, 2026

• Quantization of Diffusion Model with W4A16 Setting

  nunchux-ai/deepcompressor#117 · Filter by author Ahmaddimran opened on Jan 14, 2026

• convert svdq failed

  nunchux-ai/deepcompressor#116 · Filter by author gonggp opened on Jan 12, 2026

• convert llama model: TypeError: LlamaAttention.forward() missing 1 required positional argument: 'position_embeddings'

  nunchux-ai/deepcompressor#114 · Filter by author lisuying214 opened on Dec 5, 2025

• [bug] of convert llama-30b using lmquant.llm.run

  nunchux-ai/deepcompressor#113 · Filter by author lisuying214 opened on Dec 5, 2025

• How to convert model using multi-gpus?

  nunchux-ai/deepcompressor#112 · Filter by author lisuying214 opened on Dec 4, 2025

• 【BUG】 convert flux quant result to nunchaku safetensors question

  nunchux-ai/deepcompressor#110 · Filter by author dingdingdong-sun opened on Nov 24, 2025

• Running on A6000 Gives OOM

  nunchux-ai/deepcompressor#109 · Filter by author protyayofficial opened on Nov 14, 2025

• sdxl-turbo colab inside + questions

  nunchux-ai/deepcompressor#108 · Filter by author eyaler opened on Nov 10, 2025
  4

• [BUG]in MJHQ dataset, the format of feature is different in function _info and _generate_examples

  nunchux-ai/deepcompressor#107 · Filter by author dingdingdong-sun opened on Nov 7, 2025
  2

• Regarding quantization parameter adjustment

  nunchux-ai/deepcompressor#106 · Filter by author QQQ27 opened on Oct 17, 2025
  1

• 如何自定义量化qwen-edit模型

  nunchux-ai/deepcompressor#105 · Filter by author bank010 opened on Oct 14, 2025

• Qwen-Image/Edit quantization code open source

  nunchux-ai/deepcompressor#104 · Filter by author xuhongtao001 opened on Sep 19, 2025
  11

• SVDQuant for WAN 2.2

  nunchux-ai/deepcompressor#103 · Filter by author zjq0455 opened on Sep 12, 2025
  1

• How to get Flux-Fill INT8 ?

  nunchux-ai/deepcompressor#102 · Filter by author 631068264 opened on Sep 9, 2025
  1

• Quantize Flux.1-kontext

  nunchux-ai/deepcompressor#100 · Filter by author Leo-A11A opened on Sep 1, 2025
  4

• Bug with latest diffusers

  nunchux-ai/deepcompressor#99 · Filter by author pingzhili opened on Aug 21, 2025
  3

• nvfp4's double scaling

  nunchux-ai/deepcompressor#98 · Filter by author arkhodamoradi opened on Aug 21, 2025

• Is there a guide / script for quantizing flux kontext?

  nunchux-ai/deepcompressor#97 · Filter by author mehamednews opened on Aug 19, 2025

• WAN 2.2 support

  nunchux-ai/deepcompressor#96 · Filter by author aslyansky-m opened on Aug 17, 2025

• quantizing a custom flux kontext model

  nunchux-ai/deepcompressor#95 · Filter by author comartlab opened on Aug 14, 2025
  3

• FLUX Dev Quantization fails on an A100 with 40 GB VRAM

  nunchux-ai/deepcompressor#94 · Filter by author andrsatalia opened on Aug 4, 2025
  6

• I looked at the usage examples, but there were a lot of details that I didn't understand.

  nunchux-ai/deepcompressor#92 · Filter by author dernuiJK opened on Jul 28, 2025

• Missing FluxAttentionStruct? KeyError: <class 'diffusers.models.transformers.transformer_flux.FluxAttention'>

  nunchux-ai/deepcompressor#91 · Filter by author lym000000 opened on Jul 22, 2025
  8