search

AUTOMATIC1111 / stable-diffusion-webui

Stable Diffusion web UI
GNU Affero General Public License v3.0
139.6k stars 26.47k forks source link

[Bug]: Use built-in refiner breaks checkpoint model infotext #12459

Closed rltgjqmcpgjadyd closed 1 year ago

rltgjqmcpgjadyd commented 1 year ago

Is there an existing issue for this?

What happened?

Using the built-in refiner, checkpoint model infotext is fixed to the refiner model. It should say checkpoint model information, not refiner model information

here`s infotext

conquerer, Merchant, Doppelganger, digital cinematic color grading natural lighting cool shadows warm highlights soft focus actor directed cinematography dolbyvision Gil Elvgren Negative prompt: cropped-frame, imbalance, poor image quality, limited video, specialized creators, polymorphic, washed-out low-contrast (deep fried) watermark, cropped, out-of-frame, low quality, low res, poorly drawn Steps: 20, Sampler: DPM++ 2M SDE Karras, CFG scale: 4.5, Seed: 3568500540, Size: 832x1216, Model hash: 8d0ce6c016, Model: sd_xl_refiner_1.0_0.9vae, Refiner: sd_xl_refiner_1.0_0.9vae [8d0ce6c016], Refiner switch at: 0.7, Version: v1.5.1-366-g458eda13 Template: conquerer, Merchant, Doppelganger, digital cinematic color grading natural lighting cool shadows warm highlights soft focus actor directed cinematography dolbyvision Gil Elvgren Negative Template: cropped-frame, imbalance, poor image quality, limited video, specialized creators, polymorphic, washed-out low-contrast (deep fried) watermark, cropped, out-of-frame, low quality, low res, poorly drawn

Steps to reproduce the problem

  1. Set Refiner
  2. Generate Image
  3. See infotext

What should have happened?

It should say checkpoint model information

Version or Commit where the problem happens

458eda13211ac3498485f1e5154d90808fbcfb60

What Python version are you running on ?

Python 3.11.x (above, no supported yet)

What platforms do you use to access the UI ?

Windows

What device are you running WebUI on?

Nvidia GPUs (RTX 20 above)

Cross attention optimization

Automatic

What browsers do you use to access the UI ?

Google Chrome

Command Line Arguments

--skip-python-version-check --xformers --xformers-flash-attention

List of extensions

None

Console logs

conquerer, Merchant, Doppelganger,
digital cinematic color grading natural lighting cool shadows warm highlights soft focus actor directed cinematography dolbyvision Gil Elvgren
Negative prompt: cropped-frame, imbalance, poor image quality, limited video, specialized creators, polymorphic, washed-out low-contrast (deep fried) watermark, cropped, out-of-frame, low quality, low res, poorly drawn
Steps: 20, Sampler: DPM++ 2M SDE Karras, CFG scale: 4.5, Seed: 3568500540, Size: 832x1216, Model hash: 8d0ce6c016, Model: sd_xl_refiner_1.0_0.9vae, Refiner: sd_xl_refiner_1.0_0.9vae [8d0ce6c016], Refiner switch at: 0.7, Version: v1.5.1-366-g458eda13
Template: conquerer, Merchant, Doppelganger,
digital cinematic color grading natural lighting cool shadows warm highlights soft focus actor directed cinematography dolbyvision Gil Elvgren
Negative Template: cropped-frame, imbalance, poor image quality, limited video, specialized creators, polymorphic, washed-out low-contrast (deep fried) watermark, cropped, out-of-frame, low quality, low res, poorly drawn

Additional information

No response

altoiddealer commented 1 year ago

Are you using the Refiner extension? sd-webui-refiner (https://github.com/wcde/sd-webui-refiner.git)

I don't think A1111 has a native Refiner support? You could use it in img2img tab but that's not the intended usage for the refiner model.

If you are indeed using the Refiner extension, then your issue is probably better suited to be reported there...

I am using that extension which was recently updated to include refiner model and steps in the infotext. The images I've been generating include all the typical information but now also includes the refiner model and refiner steps, tacked onto the end of the infotext.

Panchovix commented 1 year ago

Are you using the Refiner extension? sd-webui-refiner (https://github.com/wcde/sd-webui-refiner.git)

I don't think A1111 has a native Refiner support? You could use it in img2img tab but that's not the intended usage for the refiner model.

If you are indeed using the Refiner extension, then your issue is probably better suited to be reported there...

I am using that extension which was recently updated to include refiner model and steps in the infotext. The images I've been generating include all the typical information but now also includes the refiner model and refiner steps, tacked onto the end of the infotext.

It got support natively supported since some hours ago on dev https://github.com/AUTOMATIC1111/stable-diffusion-webui/commit/36762f0eaf04c270dde23849cb198446ecdc4100

tusing commented 1 year ago

Are you using the Refiner extension? sd-webui-refiner (https://github.com/wcde/sd-webui-refiner.git)

I don't think A1111 has a native Refiner support? You could use it in img2img tab but that's not the intended usage for the refiner model.

If you are indeed using the Refiner extension, then your issue is probably better suited to be reported there...

I am using that extension which was recently updated to include refiner model and steps in the infotext. The images I've been generating include all the typical information but now also includes the refiner model and refiner steps, tacked onto the end of the infotext.

btw, that extension is bugged, this one's the one to use - https://github.com/lisanet/sdxl-webui-refiner-fixed

rltgjqmcpgjadyd commented 1 year ago

after 64311faa6848d641cc452115e4e1eb47d2a7b519

no refiner

Female Mystic Steps: 20, Sampler: DPM++ 2M SDE Karras, CFG scale: 7.5, Seed: 35246188, Size: 832x1216, Model hash: 8df8177fd2, Model: protovisionXLHighFidelity3D_beta0411Bakedvae, VAE hash: c2af720eca, VAE: sdxl_vae.safetensors, Version: v1.5.1-385-gb293ed30 Template: Female Mystic

use refiner

Female Mystic Steps: 20, Sampler: DPM++ 2M SDE Karras, CFG scale: 7.5, Seed: 35246188, Size: 832x1216, Model hash: 8d0ce6c016, Model: sd_xl_refiner_1.0_0.9vae, VAE hash: c2af720eca, VAE: sdxl_vae.safetensors, Refiner: sd_xl_refiner_1.0_0.9vae [8d0ce6c016], Refiner switch at: 0.8, Version: v1.5.1-385-gb293ed30

catboxanon commented 1 year ago

Looks like https://github.com/AUTOMATIC1111/stable-diffusion-webui/commit/fa9370b7411166c19e8e386400dc4e6082f47b2d fixes this. I can't test currently so I'll let you close this issue at your own leisure.

rltgjqmcpgjadyd commented 1 year ago

Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 1458703055, Size: 1195x896, Model hash: 8df8177fd2, Model: protovisionXLHighFidelity3D_beta0411Bakedvae, VAE hash: c2af720eca, VAE: sdxl_vae.safetensors, Clip skip: 2, ENSD: 31337, Refiner: sd_xl_refiner_1.0_0.9vae [8d0ce6c016], Refiner switch at: 0.8, Version: v1.5.1-407-gda80d649

ok it`s fixed

i`ll close issue