CivArchive
    Joey Potter (Dawson's Creek) - 1.5 v1.0
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined

    A character from TV series Dawson's Creek, portrayed by American actress Katie Holmes. Recommended to use together with the pair embedding (LINK), in such case adjust the LoRA weight as necessary.

    Description

    FAQ

    Comments (6)

    jacklaughedFeb 13, 2024
    CivitAI

    What is the recommended weights for combining the TI and Lora? Is there a preferred checkpoint to get the best results? Thanks.

    dolirama126
    Author
    Feb 13, 2024· 1 reaction

    Hi, all images are made in ComfyUI and have full workflows embedded so you can just copy Nodes to ComfyUI canvas and get all the prompts and settings. I am using TIs at full weight and LoRAs in the range of 0.5-1.0, in this case it worked best at 0.7 for me. My choice of checkpoints are Photon and ICBINP Relapse, but you can get good results with other photorealistic checkpoints as well.

    johnehly18309Feb 13, 2024

    @dolirama126 @dolirama126 Thanks for the info, I'm not using Comfy UI. I'd bet a lot of people here aren't so you might consider publishing that data for those who don't.

    dolirama126
    Author
    Feb 13, 2024

    @johnehly18309 Sorry, but the data are published. It's Civitai who you need to urge to find ways how to display them in more user friendly way - I can't do it on my side.

    jacklaughedFeb 13, 2024

    @dolirama126 I'm not on ComfyUI either, I'm on Automatic1111. You can simply add the basic data into your description or manually enter prompt data for your posted images, everyone else does that. But thanks for the info!

    dolirama126
    Author
    Feb 13, 2024

    @jacklaughed Sorry, that's too much extra work I'd rather spend on training new models. The main difference between A1111 and Comfy data is that Civitai can parse and display the first one while it struggles with the second one, so this is really just Civitai problem (you couldn't read A1111 data either if they could not parse them as they do). But there is still another solution you can use to read my prompts (which are very basic, boring and repetitive) without installing ComfyUI - just copy them from image data (it's the icon at the end of Workflow row) and paste them into Notepad. The prompt will be at the end of that file, looks like this:

    "photon_v1.safetensors","Baked VAE",-1,"!new\\JYP.safetensors",0.7000000000000001,1,"RAW portrait photo, embedding:JYP2 wearing collared sweater in busy LA street, embedding:tech\\nvjobfc","embedding:tech\\fcNeg-neg","none","comfy",512,768,1]

    LORA
    SD 1.5

    Details

    Downloads
    494
    Platform
    CivitAI
    Platform Status
    Deleted
    Created
    2/12/2024
    Updated
    5/13/2026
    Deleted
    5/23/2025

    Files

    JoeyPotter.safetensors

    Mirrors

    CivitAI (1 mirrors)
    TensorFiles (1 mirrors)