<h3 id="this-is-not-an-sdxl-or-sd-1.5-or-pony-model-eevna5fmd">⚠️ This is not an SDXL or SD 1.5 or Pony model</h3>This is a fine tune of PixArt Sigma 1024 on the training dataset I use for training PixelWave.Version 01 was trained on 22k mostly non AI images for 50 epochs, and then merged with the 00 checkpoint to help improve the overall output of the model.The 00 version is only 5 epochs. I included over 70k images generated with SDXL version of PixelWave in the training.I'm still looking for the best settings. More steps usually gives a better image.<a target="_blank" rel="ugc" href="https://comfyworkflows.com/workflows/d06e7c62-5503-4189-b9fa-e4aa0f4515f2">🍝 Workflow</a>You need the extra models node package for comfy. <a target="_blank" rel="ugc" href="https://github.com/city96/ComfyUI_ExtraModels">https://github.com/city96/ComfyUI_ExtraModels</a>If you have the vram you can set the T5 loader to use gpu instead of cpu ram.T5 files go in the models/t5 folder: <a target="_blank" rel="ugc" href="https://huggingface.co/PixArt-alpha/pixart_sigma_sdxlvae_T5_diffusers/tree/main/text_encoder">https://huggingface.co/PixArt-alpha/pixart_sigma_sdxlvae_T5_diffusers/tree/main/text_encoder</a>

PixelWaveSigma

Trained on 22k images, mostly no AI images included this time.CAME Constant, 1e-05, batch size 8, accumulation steps 128, min snr gamma 5

Files with SHA256: c550f57b9dc908b226e116ed3d21a3cbb5a6d3c8084a88bb7aec4197286a0827