Have a ComfyUI workflow? Get a readable Workflow Doc in your browser.Make a workflow readable.Try it now ↗
license: other language: en zh libraryname: diffusers pipelinetag: text-to-image tasks: text-to-image-synthesis basemodel: black-forest-labs/FLUX.2-klein-9B frameworks: PyTorch basemodelrelation: finetune
Model source
Source description
针对低显存的用户,推荐 smthemex/ComfyUI_UniBlockSwap,只需 4-6GB 显存,即可跑 bf16 全量模型。 Recommend smthemex/ComfyUI_UniBlockSwap plugin for LOW VRAM users, it only requires 4-6GB of VRAM to run the full bf16 model.
Sources
1 sourceVerified Jul 18
Model artifacts
1 artifactFlux2-Klein-9B-True-v2-bf16.safetensors
safetensors · 16.9 GB · SHA-256 ad0da30a2efc…abad · Hugging Face
DownloadSource excerpts
3 excerpts您可以尝试一下全新的 ComfyUI GGUF 模型加载插件 smthemex/ComfyUI_Dif_GGUF,它能适配更多的 GGUF 文件格式,并且将很快集成低显存(4-8GB)显卡的 GGUF 模型块卸载管理能力。 Recommend to try the new GGUF ComfyUI loader plugin, smthemex/ComfyUI_Dif_GGUF, it compatible with more GGUF format, and will add low VRAM (4-8GB) management for GGUF soon.
V2 版本模型在 FLUX.2-Klein-9B-True-V1 的基础上再次进行了全量微调,相比 V1 模型,在如下方面得到了改善和提升:
具体效果参见如下对比图片。
The V2 version of this model has undergone a full fine-tuning based on FLUX.2-Klein-9B-True-V1. Compared to the V1 version, it has some improvements and enhancements as below:
The realism and texture of images have been further improved, basically reaching the level of Nano Banana.
Prompt adherence and restoration capability, parameter adaptability, and LoRA compatibility have been further enhanced.
The image structure is more organized and reasonable, with some aesthetic improvements as well.
In image editing ability, prompt adherence is better, pixel shift and color bias have also seen some improvements.
For specific results, please refer to the comparison images as below:
与版本 V1 使用方法相同。
The same with V1 vesrion.
可以使用 Turbo / Distill LoRA 加速,但加速会影响画质,尽可能使用较小的 LoRA Weight 适配。
for Turbo Lora : anyMODE's rank128 / rank64 Civitai or Huggingface, weight 0.2 - 0.3 for steps 4 - 8.
for Distill LoRA : vafipas663's rank128 / rank256 Huggingface, weight 0.4 - 0.6 for steps: 4 - 8.
针对编辑后图像会有偏色等问题,推荐使用 ComfyUI-LCS 插件进行 Lentent 修正。
for Color and Shape correct of Image, you can use plugin of : ComfyUI-LCS
Also on: Modelscope.cn
本模型请在遵循 black-forest-labs 公司非商业许可 (FLUX Non-Commercial License) 下使用。
This model falls under the FLUX Non-Commercial License.
--- license: other language: - en - zh library_name: diffusers pipeline_tag: text-to-image tasks: - text-to-image-synthesis base_model: - black-forest-labs/FLUX.2-klein-9B frameworks: PyTorch base_model_relation: finetune --- ## Flux2-Klein-9B-True-V2 -------------------------- 针对低显存的用户，推荐 [**smthemex/ComfyUI_UniBlockSwap**](https://github.com/smthemex/ComfyUI_UniBlockSwap)，只需 **4-6GB** 显存，即可跑 bf16 全量模型。 Recommend [**smthemex/ComfyUI_UniBlockSwap**](https://github.com/smthemex/ComfyUI_UniBlockSwap) plugin for LOW VRAM users, it only requires **4-6GB** of VRAM to run the full bf16 model. 您可以尝试一下全新的 ComfyUI GGUF 模型加载插件 [**smthemex/ComfyUI_Dif_GGUF**](https://github.com/smthemex/ComfyUI_Dif_GGUF)，它能适配更多的 GGUF 文件格式，并且将很快集成低显存(4-8GB)显卡的 GGUF 模型块卸载管理能力。 Recommend to try the new GGUF ComfyUI loader plugin, [**smthemex/ComfyUI_Dif_GGUF**](https://github.com/smthemex/ComfyUI_Dif_GGUF), it compatible with more GGUF format, and will add low VRAM (4-8GB) management for GGUF soon. -------------------------- V2 版本模型在 [FLUX.2-Klein-9B-True-V1](https://huggingface.co/wikeeyang/Flux2-Klein-9B-True-V1) 的基础上再次进行了全量微调，相比 V1 模型，在如下方面得到了改善和提升： 1. 图像的**真实感**和**质感**进一步改善，基本接近**香蕉**(Nano Banana)的水平。 2. 提示词**遵循**和**还原**能力，参数适配性和**LoRA兼容性**进一步改善，。 3. 成图结构更归整、合理，美学方面也有一些改善。 4. 图像编辑能方面，提示词遵循更好，偏移和偏色也有一些改善。 具体效果参见如下对比图片。 The V2 version of this model has undergone a full fine-tuning based on [FLUX.2-Klein-9B-True-V1](https://huggingface.co/wikeeyang/Flux2-Klein-9B-True-V1). Compared to the V1 version, it has some improvements and enhancements as below: 1. The **realism** and **texture** of images have been further improved, basically reaching the level of **Nano Banana**. 2. Prompt **adherence** and **restoration capability**, parameter adaptability, and **LoRA compatibility** have been further enhanced. 3. The image structure is more organized and reasonable, with some aesthetic improvements as well. 4. In image editing ability, prompt adherence is better, pixel shift and color bias have also seen some improvements. For specific results, please refer to the comparison images as below: <p align="center"> <img src="Sample-V2.jpg" width="1200"/> <p> ### 使用方法： 与版本 V1 使用方法相同。 The same with V1 vesrion. 可以使用 Turbo / Distill LoRA 加速，但加速会影响画质，尽可能使用较小的 LoRA Weight 适配。 for Turbo Lora : [anyMODE's](https://civitai.com/user/anyMODE) rank128 / rank64 [Civitai](https://civitai....
--- license: other language: - en - zh library_name: diffusers pipeline_tag: text-to-image tasks: - text-to-image-synthesis base_model: - black-forest-labs/FLUX.2-klein-9B frameworks: PyTorch base_model_relation: finetune --- ## Flux2-Klein-9B-True-V2 -------------------------- 您可以尝试一下全新的 ComfyUI GGUF 模型加载插件 [**smthemex/ComfyUI_Dif_GGUF**](https://github.com/smthemex/ComfyUI_Dif_GGUF)，它能适配更多的 GGUF 文件格式，并且将很快集成低显存(4-8GB)显卡的 GGUF 模型块卸载管理能力。 Recommend to try the new GGUF ComfyUI loader plugin, [**smthemex/ComfyUI_Dif_GGUF**](https://github.com/smthemex/ComfyUI_Dif_GGUF), it compatible with more GGUF format, and will add low VRAM (4-8GB) management for GGUF soon. -------------------------- V2 版本模型在 [FLUX.2-Klein-9B-True-V1](https://huggingface.co/wikeeyang/Flux2-Klein-9B-True-V1) 的基础上再次进行了全量微调，相比 V1 模型，在如下方面得到了改善和提升： 1. 图像的**真实感**和**质感**进一步改善，基本接近**香蕉**(Nano Banana)的水平。 2. 提示词**遵循**和**还原**能力，参数适配性和**LoRA兼容性**进一步改善，。 3. 成图结构更归整、合理，美学方面也有一些改善。 4. 图像编辑能方面，提示词遵循更好，偏移和偏色也有一些改善。 具体效果参见如下对比图片。 The V2 version of this model has undergone a full fine-tuning based on [FLUX.2-Klein-9B-True-V1](https://huggingface.co/wikeeyang/Flux2-Klein-9B-True-V1). Compared to the V1 version, it has some improvements and enhancements as below: 1. The **realism** and **texture** of images have been further improved, basically reaching the level of **Nano Banana**. 2. Prompt **adherence** and **restoration capability**, parameter adaptability, and **LoRA compatibility** have been further enhanced. 3. The image structure is more organized and reasonable, with some aesthetic improvements as well. 4. In image editing ability, prompt adherence is better, pixel shift and color bias have also seen some improvements. For specific results, please refer to the comparison images as below: <p align="center"> <img src="Sample-V2.jpg" width="1200"/> <p> ### 使用方法： 与版本 V1 使用方法相同。 The same with V1 vesrion. 可以使用 Turbo / Distill LoRA 加速，但加速会影响画质，尽可能使用较小的 LoRA Weight 适配。 for Turbo Lora : [anyMODE's](https://civitai.com/user/anyMODE) rank128 / rank64 [Civitai](https://civitai.com/models/2324315/klein-4b9b-base-to-turbo-lora) or [Huggingface](https://huggingface.co/anyMODE/Flux-2-Klein-Base-9B-to-turbo-lora), weight 0.2 - 0.3 for steps 4 - 8. for Distill LoRA : [vafipas663's](https://huggingface.co/vafipas663) rank128 / rank256 [Huggingface](https://huggingface.co/vafipas663/flux2-klein...
Source context: 286759 downloads · 172 likes · Pipeline text-to-image · Library diffusers · Repo wikeeyang/Flux2-Klein-9B-True-V2
