CivArchive
    Envy Flux Handdrawn RPG Map 01 - ACTUALLY USABLE Dungeon maps for D&D and other TTRPGs - v1.0
    Preview 29079958
    Preview 29079942
    Preview 29079969
    Preview 29079973
    Preview 29079975
    Preview 29079985
    Preview 29079999
    Preview 29080004
    Preview 29080005
    Preview 29080009
    Preview 29079924
    Preview 29080026
    Preview 29080028
    Preview 29080029
    Preview 29080049
    Preview 29080066
    Preview 29080074
    Preview 29080085
    Preview 29080097
    Preview 29080101

    Dungeon maps. That you can actually use.

    mic drop

    Description

    FAQ

    Comments (7)

    handyman1986Sep 12, 2024
    CivitAI

    Your loras are great. I've been wanting to know how you feel about Flux when compared to models you've used in the past.

    _Envy_
    Author
    Sep 13, 2024· 2 reactions

    It blows everything else out of the water. Take these maps, for instance. It understands how rooms and hallways work, and that things generally have to be connected (there are a couple of floating rooms here and there, but not many). I haven't seen that level of common sense in any other model. It's also generally pretty easy to train, if you have hefty enough hardware. It also has amazing prompt comprehension and it gets small details right in a way that SDXL and SD1.5 could never do.

    I do miss the amazing SDXL promax controlnet, but I imagine Flux will get there eventually. For now, my workflow tends to use both SDXL and Flux, depending on what I need to do.

    handyman1986Sep 13, 2024

    @_Envy_ Yeah this map lora looks really impressive, and the fact that it's usuable is wild too. The Flux model is definitely easier to train on too, I only really need about 20-30 images of celebs and the output is spot on to how they look. I use Civitai's training however, as I don't have the VRAM at all to train locally. Still, Civitai's training is surprisingly not half-bad. I'm gonna definitely upgrade to a very capable desktop in the near future so I can generate faster and local training.

    @_Envy_ and the haaaaands!!! It knows that humans have five fingers!!! And they can hold things properly! 😁 I mean Pony did kinda sorta, too, but you get zombie eyes there instead ... 😋

    SpacedrinkOct 19, 2024
    CivitAI

    Is there no trigger word? Just using the lora and typing a random subject doesn't seem to do anything until I add like, black and white, dungeon, floor plan and things like that

    _Envy_
    Author
    Oct 20, 2024

    See the prompts with the images. I didn't train it with any specific trigger words, but it does need a prompt that's generally in that format.

    furrycontact1Nov 17, 2025
    CivitAI

    Woah...
    So.. i took this LoRa as a good excuse to set up a Flux workflow for use with ComfyUI (which I also rarely used).
    I took the skull workflow as an example to work through... that was.. quite the headspin of stuff to be installed, downloaded,adjusted.. Got it working!

    But, it also became clear that there's a lot of cruft/loose-ends on the workflow that arent used (anymore).
    Is there an example you recommend that is 'cleaner' ?

    LORA
    Flux.1 D

    Details

    Downloads
    880
    Platform
    CivitAI
    Platform Status
    Available
    Created
    9/12/2024
    Updated
    5/12/2026
    Deleted
    -

    Files

    EnvyFluxDungeonMap01.safetensors

    Mirrors

    Available On (1 platform)

    Same model published on other platforms. May have additional downloads or version variants.