PixArt-Sigma-1024px_512px-animetune

<ul><li>Pruned Model fp16 (1.15 GB): Inference checkpoint.</li><li>Training Data (9.11 GB): Fine-tuning diffusers model+OneTrainer config data.</li><li>Config (13.06 KB): ComfyUI workflow. Due to upload requirements, it has been changed to .yml. Please change the extension to .json when using.</li></ul>Q: Is it better than the 512px model?A: No, it is not necessarily better.There are a few tags that are undertrained.The training requirements for Pixart are low and excellent! However, my weak PC is insufficient for training the 1024px model, leading to compromises...However, I believe the 1024px model is necessary for future development, so I trained it to a somewhat satisfactory level. I hope you like it!Honestly, using this model alone might not balance quality and generation time, so it might be better to merge the concepts or styles of this model with other models and enjoy it that way...I have updated the workflow. I am using automatic CFG with extensions; please install them if you encounter errors.Please use the latest version as much as possible since some nodes used do not exist in the older versions of ComfyUI.The settings are configured to produce results close to my dataset, but this is not the definitive solution, so please try various workflows!I recommend the Suggested Resources workflow, as I think it’s better optimized than mine!I will update the samples once I find good settings.For those who don’t want to waste time, I think it’s a good idea to practice with the 512px model first to see which tags work before using the 1024px model.

pixartSigma1024px_1024pxV00.yml

pixartSigma1024px_1024pxV00_trainingData.zip

pixartSigma1024px_1024pxV00.safetensors

1024px_v0.0

1024px_v04

512px_0.7

512px_0.6

512px_v0.5

512px_v0.4

512px_v0.3

512px_v0.2

512px_v0.1

1024px_v0.31

1024px_v0.2

1024px_v0.1

512px_v0.0

Files with SHA256: 90fa93a9c240f5a7edcaa6841708ad32692bd7fd4002bff7882b016c7cc2d137