INT8 Instruct Fix (5 bugs):
- blocks_to_swap no longer forced to 0 for INT8 in generate()
- Added _load_int8_block_swap() method in loader with CB/SCB guard hooks
- Fixed INT8 model size from 40GB to 80GB in memory budget
- CB/SCB now fixed after .to(device) in _move_non_block_components_to_gpu()
- Quant-specific gb_per_block in _calculate_optimal_config()
VAE Decode Fix:
- Fixed super() crash from closure cell corruption in cache clearing
- Added isinstance(val, type) guard to prevent __class__ cell nuke
- Cleaned up instance-level forward attrs after hook removal (4 files)
New Features:
- Experimental latent control nodes (Empty Latent, Noise Shaping, Generate with Latent)
- Unified Generate V2 node (auto-detects NF4/INT8/BF16, block swap, memory mgmt)
- HuggingFace upload script for v2 models (quantization/upload_v2_to_hf.py)
README:
- Updated to v1.3.0 with v2 HuggingFace model links
- INT8 Instruct status changed to Working across all tables
- Added Latent Control Nodes documentation section
- Added v2 download links for base and Instruct models"
- Added HunyuanImage3GenerateLowVRAM node for 24GB VRAM cards
- Fixed loader model selection defaults (NF4/INT8 prioritize correct model type)
- Added tooltips to Low VRAM node
- Documented INT8 loader as non-functional due to bitsandbytes validation
- Added TODO list to README
Features:
- Multiple loader nodes (BF16, NF4, Multi-GPU, optimized variants)
- Advanced generation with official HunyuanImage-3.0 prompt rewriting
- Supports any OpenAI-compatible LLM API (DeepSeek, OpenAI, Claude, local LLMs)
- Organized resolution controls with MP indicators
- Large image generation with CPU offload support
- Comprehensive VRAM management and error handling
- NF4 quantization tools included
License: CC BY-NC 4.0 (Non-Commercial) / Commercial License Available
Author: Eric Hiss (github.com/ericRollei)
Model: HunyuanImage-3.0 by Tencent (Apache 2.0)
Eric Rollei