Type | |
Stats | 5,760 4,412 4 |
Reviews | |
Published | Aug 29, 2023 |
Base Model | |
Training | Steps: 55,428 Epochs: 10 |
Hash | AutoV2 C53DABC181 |
This is a finetuned (and merge) model of sdxl, based on a mixture of checkpoints such as CounterFeitXL, AIO anime, SDXL 0.9, finetuned and further merged. It's a work in progress, implying that there might be further changes or updates.
The model is trained on 300k anime images with danbooru tags for language understanding, and then applied checkpoint merging for a more aesthetically pleasing result.
There are no required tags or embeddings for the model to work, but a mixture of danbooru tags, natural language, and quality tags could be beneficial. Also the model might require a higher CFG value to work have better results.
Discussion
I just wanted to double check & clarify what you had mentioned in your model description at the end about your model being released under a ‘non-commercial license’ but its use is restricted on image regeneration sites. Maybe it’s just me, but it just confused me a tad bit bc I take that as in this model isn’t allowed to be used on any image regeneration sites, which tbh I thought was sorta the main purpose of this website & the reasoning for uploading models - to be used by others on image regeneration websites. But then says this model has a non commercial license? So as long as your model isn’t being used for monetary gain in any way, shape, or form it’s ok to be used on a image regeneration site that’s completely free? Bc the type of license you have means basically that it’s “not primarily intended for or directed towards commercial advantage or monetary compensation”.
“The model is released under non-commercial license, and explicitly do not allow any use on image generation sites.”
Again, I really like your model & just wanted to double check beforehand so I (& hopefully no other jerks) never go against the agreement of your NC license & never cause anything detrimental to your hardwork or that would cause you any issues :)
这个可以训练Lora么，用的什么脚本训练比较好？
Gallery
