Issues

• Constant prequantization rounds halfway weights after adding the zero point

  qualcomm/aimet#4124 · Filter by author vera-rubin opened on Oct 12, 2026

• QDQ scale gradients omit zero_point_shift for clipped inputs

  qualcomm/aimet#4117 · Filter by author taking-lying-flat opened on Oct 10, 2026

• Docs snippets for AdaScale/SpinQuant call _prefill_inputs and _set_tensors_to_output_n_bit_symmmetric with non-existent keyword arguments

  qualcomm/aimet#4112 · Filter by author Anai-Guo opened on Aug 30, 2026

• have mismatching datatypes. 0x416 != 0x232

  bug
  Bug

  qualcomm/aimet#4104 · Filter by author ThreeD-H opened on Jun 10, 2026
  1

  

• [BUG] LiteMP FP16 mixed precision cannot be deployed via qnn-onnx-converter on QAIRT 2.33.0

  communityQNN
  qualcomm/aimet#4102 · Filter by author charmway opened on Jun 4, 2026
  5

  

• How to Disable Concat Quantization Nodes, and how to Share the Same Quantization Parameters Across Multiple Layers

  question
  qualcomm/aimet#4101 · Filter by author piantou opened on May 26, 2026
  3

  

• Support GPTQ for LLMs INT4 quantization

  question
  qualcomm/aimet#4100 · Filter by author Yichangfa opened on May 19, 2026
  2

  

• Examples.torch.quantize ERROR

  aimet-torchllm
  qualcomm/aimet#4095 · Filter by author bqFirst opened on Apr 29, 2026
  3

  

• AIMET Quantized Model Experiences Severe Accuracy Drop After QAIRT Conversion

  aimet-onnx
  qualcomm/aimet#4094 · Filter by author slatbox opened on Apr 24, 2026
  3

• run lpbq demo Qwen2.5-0.5B-Instruct convert qnn cpp crash

  aimet-onnxdocumentationllm
  qualcomm/aimet#4091 · Filter by author guxiangtao opened on Apr 14, 2026
  

• QAT with Aimet: Gradient & Training Behavior of param_quantizers min/max under two quant_schemes

  aimet-torchquestion
  qualcomm/aimet#4087 · Filter by author piantou opened on Mar 25, 2026
  4

• Quantitative analysis (qwen-3.5)

  communityenhancementllmMedium
  qualcomm/aimet#4083 · Filter by author lchyys opened on Mar 12, 2026
  2

  

• Asymmetric Quantization Offset Overflow Issue with Large Activation Ranges

  questionwaiting on more information
  qualcomm/aimet#4072 · Filter by author mbdibd opened on Jan 30, 2026
  2

  

• quantize qwen3-vl

  enhancementin-progressllm
  qualcomm/aimet#4063 · Filter by author Arvin-928 opened on Jan 22, 2026
  3

  

• QuantSim accuracy is close to FP32, but inference result is incorrect when running with Genie (QNN 2.30, AIMET 2.0.0)

  llmquestion
  qualcomm/aimet#4057 · Filter by author yuzhou164 opened on Dec 30, 2025
  4

  

• Publish Python 3.12 wheels

  enhancementin-progressinstallation
  qualcomm/aimet#4027 · Filter by author janjongboom opened on Aug 29, 2025
  6

  

• QNN_TARGET_ARCH when running qnn-model-lib-generator and qnn-context-binary-generator

  bugQNNwaiting on more information
  qualcomm/aimet#4022 · Filter by author xiaoxiaosuaxuan opened on Aug 25, 2025
  3

  

• Deekseek quant support

  documentationenhancementllm
  qualcomm/aimet#3815 · Filter by author 1826133674 opened on Feb 17, 2025
  1

  

• Read this FIRST!

  community
  qualcomm/aimet#3730 · Filter by author quic-bharathr opened on Jan 14, 2025

• How to estimate export onnx-model time

  questionwaiting on more information
  qualcomm/aimet#3269 · Filter by author songh11 opened on Aug 26, 2024
  1

  

• How to get a Qoperator format ONNX model after quantization?

  qualcomm/aimet#2825 · Filter by author JiliangNi opened on Mar 18, 2024
  4

• examples for YOLOv5

  qualcomm/aimet#2809 · Filter by author CangHaiQingYue opened on Mar 5, 2024
  4

• CUDA out of memory

  qualcomm/aimet#2807 · Filter by author zkdfbb opened on Mar 5, 2024
  6

• The quantization parameters of encodings are inconsistent with the quantization parameters in embeeding.onnx

  qualcomm/aimet#2680 · Filter by author hcqylymzc opened on Jan 29, 2024
  3

  

• QAT issue of yolox

  qualcomm/aimet#2679 · Filter by author ianWeichengZhang opened on Jan 29, 2024