Issues

• Out-of-bounds GPU write in DenseMatVecKernel via unchecked full_row_indices from a checkpoint (CWE-787)

  SqueezeAILab/SqueezeLLM#77 · Filter by author professor-moody opened on Jul 14, 2026

• Request for Script to Reproduce Experiments in Figure 2

  SqueezeAILab/SqueezeLLM#75 · Filter by author weizhepei opened on Dec 2, 2024

• When I use SqueezeLLM to quantize the LLaMA2-13B model and test it, the speed is extremely slow.

  SqueezeAILab/SqueezeLLM#71 · Filter by author zhangfzR opened on Jul 3, 2024

• Can you update the version that can quant OPT family?

  SqueezeAILab/SqueezeLLM#70 · Filter by author Chen-1031 opened on Jun 26, 2024

• how can I get the models of 0.45% sparsity by myself?

  SqueezeAILab/SqueezeLLM#69 · Filter by author LiMa-cas opened on Jun 17, 2024

• Why do LLaMA-2-7B have s0 quantized models, but no s5 and s45 sparsity quantized models?

  SqueezeAILab/SqueezeLLM#68 · Filter by author Evane5cence opened on May 26, 2024

• Further speeding up the quantization process

  SqueezeAILab/SqueezeLLM#67 · Filter by author SyphonArch opened on May 5, 2024

• Installation instructions did not lead to the local transformers version being selected, giving errors

  SqueezeAILab/SqueezeLLM#66 · Filter by author RDouglasSharp opened on Apr 9, 2024

• Support JAIS models

  SqueezeAILab/SqueezeLLM#65 · Filter by author 7ossam81 opened on Mar 24, 2024

• Dense-only quantization bit precision

  SqueezeAILab/SqueezeLLM#63 · Filter by author akarkim opened on Mar 5, 2024
  2

• D+S packing in vLLM seems buggy

  SqueezeAILab/SqueezeLLM#62 · Filter by author MingLin-home opened on Feb 27, 2024

• sample_weight is negative when running kmeans clustering

  SqueezeAILab/SqueezeLLM#61 · Filter by author MingLin-home opened on Feb 23, 2024

• A question about LLaMA-2-7B and Mistral models only provide Dense-only (0%) quantized models

  SqueezeAILab/SqueezeLLM#56 · Filter by author WeiMa01 opened on Feb 4, 2024

• On A100 card, speed-up effect does not show up.

  SqueezeAILab/SqueezeLLM#51 · Filter by author leocnj opened on Nov 30, 2023
  2

• Future plan for this project

  SqueezeAILab/SqueezeLLM#45 · Filter by author tjtanaa opened on Nov 1, 2023

• Vicuna-1.5?

  SqueezeAILab/SqueezeLLM#44 · Filter by author mlinmg opened on Oct 29, 2023

• Minor bug for --include_sparse

  SqueezeAILab/SqueezeLLM#39 · Filter by author vuiseng9 opened on Aug 11, 2023