Stable Diffusion

Without paywall: https://archive.ph/QD9v1

Excerpt from the relevant “ComfyUI dev” Matrix room:

matt3o
and what is it then?

comfyanonymous
"safety training"

matt3o
why does it trigger on certain keywords and it's like it's scrambling the image?

comfyanonymous
the 2B wasn't the one I had been working on so I don't really know the specifics

matt3o
I was even able to trick it by sending certain negatives

comfyanonymous
I was working on a T5 only 4B model which would ironically had been safer without breaking everything
because T5 doesn't know any image data so it was only able to generate images in the distribution of the filtered training data

comfyanonymous
but they canned my 4B and I wasn't really following the 2B that closely

[…]

comfyanonymous
yeah they did something with the weights
the model arch of the 2B was never changed at all

BVH
weights directly?
oh boy, abliteration, the worst kind

comfyanonymous
also they apparently messed up the pretraining on the 2B so it was never supposed to actually be released

[…]

comfyanonymous
yeah the 2B apparently was a bit of a failed experiment by the researchers that left
but there was a strong push by the top of the company to release to 2B instead of the 4B and 8B

Additional excerpt (after the Reddit post) from Stable Diffusion Discord “#sd3”:

comfy
Yes I resigned over 2 weeks ago and Friday was my last day at stability

Basic ComfyUI workflows without 10k custom nodes or impossible to follow workflows.

https://archive.is/FkzdE

Without paywall: https://archive.ph/8QkSl

I created a custom SDXL Lora using my dataset. I created the dataset using a previous generative art tool I build to visualize factorio blueprints: https://github.com/piebro/factorio-blueprint-visualizer. I like the lora to create interesting patterns.

I don't know why its cracking me up so much, I dont know what I was expecting.. but also, I can't find any easy reason why this is happening, maybe an artist name i cant find now? A product name translation?

~~Stable Diffusion XL doesn't seem to replicate the same results~~ (actually kinda @ 768px instead of 512) any?? Stable diffusion ~~1.x~~ model with only positive prompt:

turbotits

Settings	Image
Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 912261969, Size: 512x512, Model hash: f968fc436a, Model: analogMadness_v50, VAE hash: 735e4c3a44, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Version: v1.6.0-2-g4afaaf8a
Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 1212726575, Size: 512x512, Model hash: 9aba26abdf, Model: deliberate_v2_sd1.5, VAE hash: 735e4c3a44, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Version: v1.6.0-2-g4afaaf8a
Steps: 20, Sampler: DPM++ 2M, CFG scale: 7, Seed: 1020297828, Size: 512x512, Model hash: af9ab7cc05, Model:realcartoonPixar_v2_sd1.5, VAE hash: 735e4c3a44, VAE: animevae.pt, ADetailer model: yolov8n.pt, ADetailer confidence: 0.3, ADetailer dilate erode: 4, ADetailer mask blur: 4, ADetailer denoising strength: 0.4, ADetailer inpaint only masked: True, ADetailer inpaint padding: 32, ADetailer ControlNet model: control_v11p_sd15_inpaint [ebff9138], ADetailer ControlNet module: inpaint_global_harmonious, ADetailer version: 23.11.1, ControlNet 0: "Module: inpaint_global_harmonious, Model: control_v11p_sd15_inpaint [ebff9138], Weight: 1.0, Resize Mode: ResizeMode.INNER_FIT, Low Vram: False, Guidance Start: 0.0, Guidance End: 1.0, Pixel Perfect: True, Control Mode: ControlMode.BALANCED, Save Detected Map: True", Version: v1.6.0-2-g4afaaf8a
Steps: 20, Sampler: DPM++ 2M, CFG scale: 7, Seed: 2311419033, Size: 512x512, Model hash: cc6cb27103, Model: v1-5-pruned-emaonly, VAE hash: 735e4c3a44, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Version: v1.6.0-2-g4afaaf8a
Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 2505533199, Size: 512x512, Model hash: fc2511737a, Model: chilloutmix_NiPrunedFp32Fix.sd1.5, VAE hash: 735e4c3a44, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Clip skip: 2, Version: v1.6.0-2-g4afaaf8a
Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 3851799520, Size: 512x512, Model hash: fe4efff1e1, Model: HuggingFace_SD1.4, Clip skip: 2, Version: v1.6.0-2-g4afaaf8a
Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 2906017088, Size: 768x768, Model hash: e6bb9ea85b, Model: sdXL_v10VAEFix, Version: v1.6.0-2-g4afaaf8a

some models seem to show better results using clip skip 1 (vs. 2). Anyway, I got a chuckle from it. On a more utilitarian note, i suppose things like this could be used to gauge how "far" from an unknown merge point you are from a base model perhaps? (like DNA evolution i mean, wording it poorly)

settings for post/linked image:

Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 366961196, Size: 512x704, Model hash: 8c4042921a, Model: aZovyaRPGArtistTools_v3VAE_sd1.5, VAE hash: 735e4c3a44, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Denoising strength: 0.52, Hires upscale: 1.5, Hires upscaler: Latent, Refiner: centerflex_v28.sd1.5_safetensors [02e248bf1b], Refiner switch at: 0.8, Version: v1.6.0-2-g4afaaf8a

Using ParchArtXL civitai.com/models/141471/parchartxl