๐ค Hugging Face | ๐ค ModelScope | ๐ป GitHub | ๐ฌ WeChat (ๅพฎไฟก) | ๐ฎ Discord
# Qwen-Image-Layered-Series > โ ๏ธ **Config-only repository โ no model weights.** > This repo contains **only** QuantFunc per-layer **precision configs** for **Qwen-Image-Layered** (RGBA layer decomposition). It does **not** contain, mirror, or redistribute any model weights. **You bring your own** official [`Qwen/Qwen-Image-Layered`](https://huggingface.co/Qwen/Qwen-Image-Layered); these configs only tell the QuantFunc engine how to quantize it **at load time, on your own machine**. **Powered by the [QuantFunc ComfyUI plugin](https://github.com/RealJonathanYip/ComfyUI-QuantFunc) โ the fastest diffusion inference engine:** - ๐ **2xโ11x speedup** over standard BF16/FP16 Python pipelines. - โ๏ธ **Native C++/CUDA** (`libquantfunc.so` / `quantfunc.dll`), **zero Python model dependencies**. - ๐งฉ **Universal format adapter** โ loads **diffusers / BFL / HF** layouts directly, no manual conversion. - ๐ข **Full GPU coverage** โ RTX 20/30/40/50 ยท A100/H100/H200/B100/B200 (CUDA 12 & 13); native **FP4** on Blackwell. ๐ **Install the plugin:** **https://github.com/RealJonathanYip/ComfyUI-QuantFunc** ## What this repository provides Just the precision configs โ **no weights**: ``` Qwen-Image-Layered-Series/ โโโ config.json # = 50x-below INT4 map (HF download-counter query file) โโโ precision-config/ โโโ 50x-above-fp4-sample.json # NVFP4 (FP4 weights, af8wf4 MLP) โ RTX 50 / SM120+ โโโ 50x-below-int4-sample.json # INT4 per-group-128 โ all SMs (robust fallback) ``` We deliberately **do not host Qwen-Image-Layered weights**. The QuantFunc **Lighting** backend does **runtime** quantization: you load the *official* weights and they are quantized **in-memory at load**, so no pre-quantized checkpoint is ever distributed. ## How to use 1. **Obtain the official model yourself** โ [`Qwen/Qwen-Image-Layered`](https://huggingface.co/Qwen/Qwen-Image-Layered) (diffusers layout). Follow Qwen's distribution channels and license. 2. **Install the QuantFunc ComfyUI plugin:** https://github.com/RealJonathanYip/ComfyUI-QuantFunc 3. **Load the official model** through the **Build Pipeline** node (universal format adapter). 4. **Precision config** โ leave the node on **`auto detect`** (it recognizes Qwen-Image-Layered and applies the right map automatically: **NVFP4** on RTX 50 / SM120+, **INT4** otherwise), or point it at a file manually. ## Precision configs Two GPU tiers (the auto-detect picks by SM): | File | Target GPU | Scheme | |------|-----------|--------| | `50x-above-fp4-sample.json` | RTX 50 / SM120+ | **NVFP4** (FP4 e2m1 weights); **FP8 activations on the MLP only** (`af8wf4`), attention stays W4A4 | | `50x-below-int4-sample.json` | RTX 20/30/40 + datacenter | **INT4** per-group-128 (AUTO_4 โ INT4 on all SMs); robust, fully coherent at any SM | **Why the MLP is `af8wf4` on the NVFP4 map:** `use_additional_t_cond` + layer3d modulation make the MLP input activations large enough to saturate the FP4-activation per-16 FP8 (e4m3 max 448) microscale โ green-noise background. FP8 activation (per-token FP16 act-scale) on the MLP removes it; attention tolerates FP4 activation and stays on the fast W4A4 path. This differs from the base Qwen-Image NVFP4 map by **exactly one layer** (the MLP up-projection `net.0.proj`). In both maps the `img_mod`/`txt_mod` modulation GEMMs stay **INT8**. ### โ ๏ธ Companion settings REQUIRED for coherence (not part of the precision map) - base scheduler (`configs/qwen-image-base-scheduler.json`) - `num_inference_steps = 50` - `true_cfg_scale = 4.0` - **non-empty** `negative_prompt` - a **real RGBA composite** input image - resolution **640** > NVFP4 (`50x-above`) is **SM120+ only** (FP4 is native `sm_120a`, never PTX-JIT). On older GPUs use the INT4 map. ## Legal / Attribution - This repository distributes **only** the QuantFunc precision-config JSON โ our own work, Apache-2.0. - It contains **no Qwen weights** and is **not affiliated with, nor endorsed by, the Qwen team**. - You are solely responsible for obtaining the official model and complying with its license and terms of use. ## Community - ๐ฎ [Discord server](https://discord.gg/jCp9TpFWcn) - ๐ฌ Scan the QR code below to join our WeChat group: