Hugging Face's logo

6DammK9
/
AstolfoMix

Text-to-Image
Diffusers
Safetensors
English
StableDiffusionPipeline
stable-diffusion
stable-diffusion-diffusers
Merge
Inference Endpoints
Model card Files Community
AstolfoMix / README.md
6DammK9's picture
6DammK9
Update README.md
5df519a
preview code
|
raw
history blame
4.68 kB
metadata
language:
  - en
license: creativeml-openrail-m
tags:
  - stable-diffusion
  - stable-diffusion-diffusers
  - text-to-image
  - safetensors
library_name: diffusers

AstolfoMix (Baseline)

  • A (baseline) merge model focusing on absurdres, and let me wait for a big anime SDXL finetune.
  • Behind the "absurdres", the model should be very robust and capable for most LoRAs / embeddings / addons you can imagine.
  • The image below is 2688x1536 without upscaler. With upscaler, it reaches 8K already.
  • The image below is 10752x6143, and it is a 3.25MB JPEG. "upscaler 4x". See PNG info below.

230725-2.jpg

parameters
(aesthetic:0), (quality:0), (solo:0), (boy:0), (ushanka:0.98), [[braid]], [astolfo], [[moscow, russia]]
Negative prompt: (worst:0), (low:0), (bad:0), (exceptional:0), (masterpiece:0), (comic:0), (extra:0), (lowres:0), (breasts:0.5)
Steps: 48, Sampler: Euler, CFG scale: 4.5, Seed: 3179120067, Size: 1344x768, Model hash: f52ee1e6b3, Model: vcbp_mtd8_cwl-sd, VAE hash: 551eac7037, VAE: vae-ft-mse-840000-ema-pruned.ckpt, Denoising strength: 0.7, Clip skip: 2, Hires upscale: 2, Hires steps: 48, Hires upscaler: Latent, Dynamic thresholding enabled: True, Mimic scale: 1, Separate Feature Channels: False, Scaling Startpoint: MEAN, Variability Measure: AD, Interpolate Phi: 0.3, Threshold percentile: 100, Version: v1.6.0
postprocessing
Postprocess upscale by: 4, Postprocess upscaler: SwinIR_4x
extras
Postprocess upscale by: 4, Postprocess upscaler: SwinIR_4x
  • Current version: 08-vcbpmt_d8cwlbd_aweb5-sd.safetensors (merge of 8 models)
  • Recommended version: "06a" or "08"
  • Receipe Models: Merging UNETs into SD V1.4
  • "Roadmap" / "Theory" in my Github.
  • Recommended prompt: "SD 1.4's Text Encoder"
  • Recommended resolution: 1024x1024 (native T2I), HiRes 1.75x (RTX 2080Ti 11GB)
  • It can generate images up to 1280x1280 with HiRes 2.0x (Tesla M40 24GB), but the yield will be very low and time consuming to generate a nice image.
  • Recommended CFG: 4.5 (also tested on all base models), 6.0 (1280 mode)

Receipe

  • Uniform merge. M = 1 / "number of models in total".
  • M=0.5 (02-vbp23-cbp2-sd)
  • M=0.33 (03-vcbp-mzpikas_tmnd-sd)
  • M=0.25 (04-vcbp_mzpt_d8-sd)
  • M=0.2 (05-vcbp_mtd8_cwl-sd)
  • M=0.167 (06-vcbp_mtd8cwl_bd-sd)
  • M=0.143 (07-vcbp_mtd8cwl_bdaw-sd)
  • M=0.125 (08-vcbpmt_d8cwlbd_aweb5-sd)

xyz_grid-0083-342433587-10240-1656-4.5-48-20230916031640.jpg

Extra: Comparing with merges with original Text Encoders

  • Uniform merge. M = 1 / "number of models in total".
  • M=0.5 (02a-vbp23-cbp2)
  • M=0.33 (03a-vcbp-mzpikas_tmnd)
  • M=0.25 (04a-vcbp_mzpt_d8)
  • M=0.2 (05a-vcbp_mtd8_cwl)
  • M=0.167 (06a-vcbp_mtd8cwl_bd)
  • M=0.143 (07a-vcbp_mtd8cwl_bdaw)
  • M=0.125 (08a-vcbpmt_d8cwlbd_aweb5)

Image coming soon. Suprisingly, they looks similar, with only minor difference in background and unnamed details (semantic relationships).

License

This model is open access and available to all, with a CreativeML OpenRAIL-M license further specifying rights and usage. The CreativeML OpenRAIL License specifies:

  1. You can't use the model to deliberately produce nor share illegal or harmful outputs or content
  2. The authors claims no rights on the outputs you generate, you are free to use them and are accountable for their use which must not go against the provisions set in the license
  3. You may re-distribute the weights and use the model commercially and/or as a service. If you do, please be aware you have to include the same use restrictions as the ones in the license and share a copy of the CreativeML OpenRAIL-M to all your users (please read the license entirely and carefully) Please read the full license here