Download
1 variant available
bf16 SafeTensor
DC Obsidian Monochrome [Krea2Raw][GPT5.2Captioned][Trigger=DCOBMON.][Rank32] v1.safetensors
BF16, good balance • 218 MB
Verified: 4 days ago
Trained @ 10,000 Steps, Rank32, Captioned with GPT5.2, with a dataset of 88 Images
I had trouble training this one "I believe i am probably captioning too much" Infortunately it isn’t as dark and grainy as I would like, will try better for v2.
Done two failed training runs so now you have to add "DCOBMON. Black monochrome image of" or "DCOBMON. blue monochrome image of" etc. at the start of each prompt it can do different colors, like orange, green, blue etc.
Show more

6600 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9
1.6K0 1 2 3 4 5 6 7 8 9.0 1 2 3 4 5 6 7 8 9K
. Krea2 Raw v1 Trigger Word = "DCOBMON. black monochrome image of" you can change the colour. to blue, orange, green etc.
. FluxKlein9B AND Z-ImageTurbo v1 Trigger Word = "DCOBMON."
. SDXL Aspect Ratio's work best with this lora
. Some other good triggers to add at the end of your prompts "some highlight clipping. Heavy film grain and rough texture cover the frame, giving a tense, empty atmosphere."
. This Lora was to the style of @mertzv2 on Instagram he makes amazing black and white/ Monochrome grainy Images and videos check his page out its really good, Trained on my midjourney images with his style, this lora can get close to his style of images with the correct prompting but its hard to get the correct grain which i have added in the workflow to add extra grain if needed. [Only Basic workflow used for Z-Image Turbo] just add a film grain node to decode node and play with settings.
. This lora was trained with Ostris Ai-Toolkit with Runpod for 8,000 steps Rank 32 Tagged with detailed captions consisting of 80-100 words with GPT5.2 (89 Images Total)
. All images have Prompts, generation details Workflow for ComfyUi just click the image you want, then save, then drop into ComfyUI or Open the image with notepad on pc and you can search all the metadata there. My workflow has multiple Upscalers to choose from Seedvr2, Flash VSR, SDXL TILED CONTROLNET, Ultimate SD Upscale and a DetailDaemon Upscaler and an Qwen 3 llm to describe images if needed
. Check out my profile gallery for more examples
If you use this in a merge, to train a model or upload to another site please credit me by showing my civitai page
https://civitai.com/user/dkpc69


