造梦实验室-醇彩粒绘-QWEN

EN：

【Large Model/Checkpoint】Qwen Image 20B v1

【Model Weights】0.8~1 (1 is recommended for single LoRa)

【Trigger Word】cclh

【Sampling Method】Euler

【CFG Scale】4.0

【Sampling Steps】40

【About Training】This model's base model was trained using Qwen Image 20B v1.

【About Usage】This model follows the official Qwen license.

【Training Acknowledgements】This model training received technical support from the Moda Community.

【Model Introduction】This model gives illustrations a delicate grainy/noise texture.

ZH：

【大模型/Checkpoint】Qwen Image 20B v1

【模型权重】0.8~1（单lora建议直接上1）

【触发词】cclh

【采样方法/Sampler Method】Euler

【提示词引导系数/CFG Scale】4.0

【采样步数/Sampling Steps】40

【关于训练】本模型底模使用Qwen Image 20B v1训练

【关于使用】本模型遵循Qwen官方许可证

【训练致谢】本模型训练获得魔搭社区技术支持

【模型介绍】让插画具有细腻的颗粒/噪点肌理质感