Issues

• Verify evals on Papers with Code

  FoundationVision/VAR#187 · Filter by author NielsRogge opened on Sep 14, 2026

• 🚀 Your VAR Model is Secretly an Efficient and Explainable Generative Classifier — [ICLR 2026]

  FoundationVision/VAR#182 · Filter by author Yi-Chung-Chen opened on Mar 21, 2026

• [Resource] Awesome-Visual-Autoregressive-Modeling — A curated paper list built on VAR

  FoundationVision/VAR#181 · Filter by author XIANGLONGYAN opened on Mar 10, 2026

• Cannot reproduce "without CFG" FID in Table 3 (expected ~4.64, got ~12), while "with CFG" is reproducible

  FoundationVision/VAR#180 · Filter by author Lil-Shake opened on Feb 15, 2026

• 📢 Introducing SSG (Scaled Spatial Guidance) for Multi-Scale VAR Generation via Information-Theoretic Analysis

  Research
  FoundationVision/VAR#179 · Filter by author Youngwoo-git opened on Feb 6, 2026

• 🌟 Introducing SAR (Self-Autoregressive Refinement): stable self-rollout that unlocks robust post-training for Visual Autoregressive image generation.

  Research
  FoundationVision/VAR#178 · Filter by author GengzeZhou opened on Dec 24, 2025

• 🌈 Introducing DiverseVAR, a training-free framework that unleashes the inherent generative diversity of Visual Autoregressive models while preserving fidelity and text–image alignment.

  Research
  FoundationVision/VAR#177 · Filter by author wangtong627 opened on Nov 27, 2025

• can you share the multi-scale VQ-VAE training and evaluation code?

  FoundationVision/VAR#175 · Filter by author LWang1016 opened on Sep 17, 2025

• Introducing "Next Visual Granularity Generation"

  FoundationVision/VAR#174 · Filter by author Yikai-Wang opened on Aug 20, 2025
  1

• 🚀 Introducing SCALAR: A Scale-wise Controllable VAR Framework for Next-Scale Prediction

  FoundationVision/VAR#173 · Filter by author ZeroRF opened on Aug 12, 2025

• how to modify VAR for image translation task ?

  FoundationVision/VAR#171 · Filter by author keruoya opened on Jul 23, 2025
  2

• Any plans to release the VQVAE training code?

  FoundationVision/VAR#170 · Filter by author aswahd opened on Jul 9, 2025
  2

• Introducing HMAR: Efficient Hierarchical Masked AutoRegressive Image Generation, that builds on VAR.

  Research
  FoundationVision/VAR#169 · Filter by author Kumbong opened on Jul 8, 2025

• 请教高手：我使用作者的vae、vard16_pth但是跑的结果accm只有acc_mean ≈5%，而作者论文中的是acc_mean ≈ 82.6%

  FoundationVision/VAR#167 · Filter by author zys445 opened on Jun 13, 2025
  3

• 🚀 Introducing an efficient KV cache compression framework tailerd for next-scale prediction.

  Research
  FoundationVision/VAR#166 · Filter by author StargazerX0 opened on May 27, 2025

• 🚀 Introducing scale and spatial Markovian assumptions into VAR — achieving faster generation with reduced memory consumption.

  Research
  FoundationVision/VAR#164 · Filter by author NuanBaobao opened on May 23, 2025
  1

• Decode intermediate steps

  FoundationVision/VAR#162 · Filter by author oskarjor opened on May 8, 2025

• Question regarding d12 version

  FoundationVision/VAR#160 · Filter by author minhnn1 opened on Apr 29, 2025

• 汇报代码中一个可能导致torch自动权重初始化失效的操作

  bugenhancement
  FoundationVision/VAR#159 · Filter by author yusijin02 opened on Apr 26, 2025

• 请教：目前 FID 的计算结果无法复现

  FoundationVision/VAR#156 · Filter by author Ameis-god opened on Apr 5, 2025
  1

• The training command in Training Scripts doesn't work

  FoundationVision/VAR#155 · Filter by author LingyuLi-math opened on Apr 4, 2025

• train time v. depth

  FoundationVision/VAR#154 · Filter by author andyxzhu opened on Apr 3, 2025

• train.py文件中，验证集结果打印错误

  FoundationVision/VAR#153 · Filter by author Iron-head opened on Apr 2, 2025

• 请问该如何启用半精度训练的模式？

  FoundationVision/VAR#152 · Filter by author zhuzil opened on Apr 2, 2025

• What is the minimum amount of memory needed to train this minimal model(d16)?

  FoundationVision/VAR#151 · Filter by author Zephyrose opened on Apr 2, 2025