●Change8

v0.3.58

📦 comfyuiView on GitHub →
✨ 5 features🐛 8 fixes🔧 1 symbols

Summary

This release introduces support for Hunyuan 3D 2.1 and Hunyuan image 2.1 models, alongside several bug fixes related to lowVRAM, AMD compatibility, and progress updates. Several new API nodes for video and audio processing were also added.

Migration Steps

  1. If using AMD hardware, note that PyTorch attention will not be enabled unless Triton is available.

✨ New Features

  • Added support for Hunyuan 3D 2.1.
  • Added support for Hunyuan image 2.1 regular model.
  • Enabled bf16 VAE on RDNA4.
  • Added ByteDance video nodes to API nodes.
  • Added StabilityAI Audio API nodes.

🐛 Bug Fixes

  • Fixed progress update crossover between users.
  • Fixed lowVRAM issues with Hunyuan3D 2.1.
  • Added cache headers for images.
  • Fixed amd_min_version crash when on CPU device.
  • Added ZeroDivisionError catch for EasyCache logging statement.
  • Added explicit casting in apply_rope for Qwen VL.
  • Fixed lowVRAM issue with Hunyuan image VAE.
  • Fixed an issue on old PyTorch versions.

🔧 Affected Symbols

validate_inputs