<aside> 📢 Here listed will be all the models that I have personally experimented on, and are provided in the previous Google Colab lists. Most of these models are available and compatible with the Diffusion UI Google Colab, however in the case of newer models, a newly created Google Colab is provided with the model ready to be processed. Once again, I did not create the models below, full credit is given in the Resources page. Model sources will be placed under each heading.
I have not written full blown entries for each model since there are quite a lot that I have chosen. Entries for models are listed below the testing summary.I update this section frequently as I research so enjoy! 👍
</aside>
Models are designed for the purpose of producing certain images (Different models, different images generated)
In this document, I will be focusing on mostly anime style models and selected quite a lot which I believe are pretty good
Be careful when downloading .ckpt files as they may contain viruses, try to get a .safetensors version
There are sometimes different versions of the same model (fp16, fp32, pruned) -> fp16 will be the smallest file. Often they produce very similar images, but there still may be some differences.
AnythingV3Pruned
AnythingV3 (fp32)
Normally when I do testing I choose the smallest file as it takes the least time to load
There are quite a lot of models that I will be covering after extending the document. Since it is too much to explain at the moment, I will place down the hugging face links on those models since it is explained better than I probably will.(Needs to be updated, but for now, I have done a lot of testing with new models, please check Overall Model Comparison for that 👍)
Personal favorite models are listed with a ⭐