模型介绍：

Model Introduction:

此LoRA模型采用Flux1-Dev为底模训练，推荐使用Flux1-Dev. safetensors模型， 如果显存不足，剪枝类型可以用fp8，或者用Flux1-dev-fp8模型也可以，经测试效果差不多。

This LoRA model is trained using Flux 1-Dev as the base model. It is recommended to use the Flux 1-Dev. safetensors model. If there is insufficient video memory, the pruning type can be fp8, or the Flux 1-dev fp8 model can also be used. The tested results are similar.

此Lora主要由室外自然环境数据集素材进行训练，对生成森系自然环境比较友好，生成的森林植被偏黄调，真实、自然。也有一些室内和其他室外环境数据集。

This Lora is mainly trained on outdoor natural environment dataset materials, which are relatively friendly to the generated forest natural environment. The generated forest vegetation has a yellowish tone and is authentic and natural. There are also some indoor and other outdoor environmental datasets.

用该此Lora生成的画面不论从色彩、清晰度、美感等方面都有比较明显的提升。

The images generated using this Lora show significant improvements in color, clarity, and aesthetics.

运行方式：工作流

Operation mode: ComfyUI

模型：Flux1-Dev. safetensors

推荐尺寸：1024x1536

采样器：默认采样器Euler

采样步数：30

触发词为：changjing

The trigger word : changjing

LORA权重：0.6-1，都可以试试

LORA weight: 0.6-1, you can give it a try

文生图示例：

Example of text2image:

changjing, In the outdoor forest, a small bottle of perfume filled with orange liquid is surrounded by green plants around the stone, blooming flowers and buds, dotted with green branches and leaves. There is a pressure pump head on the top of the perfume bottle. The brand name "LOUIS" is written on the bottle of perfume. Natural light, high-quality masterpiece. Photography Master, 8k

更详细的提示词可以获得更丰富的元素和细节。

More detailed prompts can provide richer elements and details.

此Lora生成的画面不论从色彩、清晰度、美感等方面都有比较明显的提升。见下图使用Lora和不使用Lora的效果对比：

The images generated by Lora show significant improvements in color, clarity, and aesthetics. Compare the effects of using Lora and not using Lora as shown in the following figure: