PikkieWyn commited on
Commit
8594037
1 Parent(s): a30bf63

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -21,7 +21,7 @@ I say this, because - while I found a merge that I'm happy enough with to share
21
 
22
  The models are licensed under [openrail++ (DPO)](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/LICENSE.md) and [GPL3 (Proteus)](https://choosealicense.com/licenses/gpl-3.0/), both allowing for free use and redistribution within the original model use-case. The merge is published under GPL3.
23
 
24
- The two variants of this model, and the original base models (converted to transformers-base for use with stable diffusion, etc) can be found in the HuggingFace Repository.
25
 
26
  ## The Goal
27
 
@@ -112,7 +112,7 @@ I tested multiple sampler/scheduler/CFG/Step combinations before deciding a foun
112
 
113
  `Tests were run in ComfyUI with the model's CLIP and VAE, no Loras and with no pre/post-processing. Worth noting is that DataPulse recommends a CLIP skip of -2. I did not make any modification to the CLIP skip layer during my merge variant testing. My only adjustment was setting the CLIP scaling to 4 to improve the final quality and clarity of the images. The workdlow was kept as simple as possible, running only base sampling in a single KSampler node. Any errors or small issues were left as is for consisency.`
114
 
115
- The workflow is available in the [Huggingface repository](https://huggingface.co/PikkieWyn/Proteus-RunDiffusion-DPO_Merge/tree/main/comparisons). The comparison images are also the original, unaltered output images, and have their workflows embedded. They're all the same, of course, the only variable being the model used and the prompt text for each test. If you plan on loading the workflow, Use Everywhere (UE Nodes) was used to neaten the flow (CLIP weight and seed distribution), and OneButtonPrompt's preset node was used to generate some random base prompts - so you'll either need to install these or make the necessary tweaks to bypass them.
116
 
117
  **Universal Sampling Parameters:**
118
 
 
21
 
22
  The models are licensed under [openrail++ (DPO)](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/LICENSE.md) and [GPL3 (Proteus)](https://choosealicense.com/licenses/gpl-3.0/), both allowing for free use and redistribution within the original model use-case. The merge is published under GPL3.
23
 
24
+ The two variants of this model, and the original base models (converted to transformers-base for use with stable diffusion, etc) can be found in this repo.
25
 
26
  ## The Goal
27
 
 
112
 
113
  `Tests were run in ComfyUI with the model's CLIP and VAE, no Loras and with no pre/post-processing. Worth noting is that DataPulse recommends a CLIP skip of -2. I did not make any modification to the CLIP skip layer during my merge variant testing. My only adjustment was setting the CLIP scaling to 4 to improve the final quality and clarity of the images. The workdlow was kept as simple as possible, running only base sampling in a single KSampler node. Any errors or small issues were left as is for consisency.`
114
 
115
+ The workflow is available in this repository. The comparison images are also the original, unaltered output images, and have their workflows embedded. They're all the same, of course, the only variable being the model used and the prompt text for each test. If you plan on loading the workflow, Use Everywhere (UE Nodes) was used to neaten the flow (CLIP weight and seed distribution), and OneButtonPrompt's preset node was used to generate some random base prompts - so you'll either need to install these or make the necessary tweaks to bypass them.
116
 
117
  **Universal Sampling Parameters:**
118