V1基于Prefect Pony XL作为底模进行训练所得的多个lora融合的产物。The product of multiple Lora fusion based on Prefect Pony XL as the base model.Prefect Pony XL作为底模，使用个人在NAI3和tPony生成图片、以及收集到的AI生成图片进行训练所得的LoRa，进行融合并微调，同时添加了Hyper-SDXL-8steps加速生成速度。Prefect Pony XL is used as the base model, use the images generated in NAI3 and tPony and the collected AI-generated images to train the LoRa, fusion and fine-tuning, and Hyper-SDXL-8steps is added to accelerate the generation speed.当前仅在Comfyui中使用，未在Webui中测试。Currently only used in Comfyui, not tested in Webui.已发现问题：在某些未知提示词下，可能出现画质大幅下降/出现色块的问题。在全身和半身图等人物脸部占比较小的情况下，即使高清修复后，眼部质量仍然较差。Problem found:There may be a significant decrease in image quality/color patches under some unknown cue words.In cases where the face of a person is relatively small, such as full body and half body images, the eye quality is still poor even after high definition repair.推荐/Recommend：postive: score_9, score_8_up, score_7_up, negative: score_6, score_5, score_4, source_furry, source_pony, ugly, bad feet, bad hands, Size: 832*1216CFG: 2 | Dynamic CFG: 8-2, Cosine DownStep: 12生成后进行高清修复，并使用adetailer/detailed fix对脸部、眼部进行修复。After generation, perform high definition repair, and use adetailer/detailed fix to repair the face and eyes.高清修复推荐/HireFix Recommend：CFG:2Step: 5-10本文英文部分为AI生成，如有错误敬请见谅。The English part of this article is generated by AI. Sorry for any errors.以上，祝你愉快。The above. Wish you a pleasant time.
