INT8

--- tags: - comfyui - int8 - qwen2_5_vl - qwen3 - qwen3_vl - Quantizations --- For INT4 convrot W4A4 models, go to https://huggingface.co/ariaotp/int4-ConvRot-W4A4-ComfyUI instead. These models are quantized INT8(both of simple and convrot) for ComfyUI 0.27.0 or later. They are fast in Ampere(RTX30x0)/Ada Lovelace(RTX40x0)/Blackwell(RTX50x0). Original models that I used: * Qwen2.5-VL 7B (Qwen2.5-VL-7B-Instruct-abliterated): https://huggingface.co/huihui-ai/Qwen2.5-VL-7B-Instruct-abliterated (Needed for Qwen-Image and FireRed-Image-Edit) * Qwen3 0.6B (Huihui-Qwen3-0.6B-abliterated-v2) : https://huggingface.co/huihui-ai/Huihui-Qwen3-0.6B-abliterated-v2 (Needed for Anima) * Qwen3 4B (Huihui-Qwen3-4B-abliterated-v2): https://huggingface.co/chinmankokumin/Qwen3-4B-abliterated-v2 (Needed for Z-Image and FLUX.2 Klein 4B) * Qwen3 4B Instruct-2507: (Huihui-Qwen3-4B-Instruct-2507-abliterated): https://huggingface.co/huihui-ai/Huihui-Qwen3-4B-Instruct-2507-abliterated (Alternate text encoder for Z-Image or FLUX.2 Klein 4B) * Qwen3 8B (Huihui-Qwen3-8B-abliterated-v2): https://huggingface.co/huihui-ai/Huihui-Qwen3-8B-abliterated-v2 (Needed for FLUX.2 Klein 9B) * Qwen3-VL 4B (Huihui-Qwen3-VL-4B-Instruct-abliterated): https://huggingface.co/ahmed22xa/Huihui-Qwen3-VL-4B-Instruct-abliterated-comfy (Needed for Krea2) * Qwen3-VL 8B (Huihui-Qwen3-VL-8B-Instruct-abliterated): https://huggingface.co/huihui-ai/Huihui-Qwen3-VL-8B-Instruct-abliterated (Needed for Boogu-Image and Ideogram4) They used Starnodes Model Converter, made by Starnodes2024, quantized from BF16 models https://github.com/Starnodes2024/comfyui-starnodes-modelconverter Or quant_int8_convrot.py of ComfyUI's official ComfyUI model tools https://github.com/Comfy-Org/comfy-model-tools License Qwen2.5-VL, Qwen3, Qwen3-VL: Apache 2.0 * https://www.apache.org/licenses/LICENSE-2.0.txt

--- tags: - comfyui - int8 - booguimage - Quantizations --- For INT4 convrot W4A4 models, go to https://huggingface.co/ariaotp/int4-ConvRot-W4A4-ComfyUI instead. These models are quantized INT8(both of simple and convrot) for ComfyUI 0.27.0 or later. They are fast in Ampere(RTX30x0)/Ada Lovelace(RTX40x0)/Blackwell(RTX50x0). Original models that I used: * Boogu-Image-Edit-Turbo 0708 hotfix: https://huggingface.co/smthem/comfy-org-gguf-int4 (for INT8 version of other types of Boogu-Image, go to https://huggingface.co/Comfy-Org/Boogu-Image instead) For Boogu-Image-Edit-Turbo, Don't used convrot because boogu's attention/feed-forward layer dimensions (840, 3360) are not compatible with ConvRot's group-size requirements (group size must be a power of 4 and evenly divide in_features) License Boogu-IMage: Apache 2.0 * https://www.apache.org/licenses/LICENSE-2.0.txt

Source context: 0 downloads · 4 likes · Repo ariaotp/INT8-ComfyUI