FLUX PRO NSFW TRAINED - DREAM DIFFUSION BY DICE
MY DISCORD FOR THE WORKFLOWS AND DOWNLOADS : https://discord.gg/7aQNJgj4rT
If you like this checkpoint, then you will like the new versions I have released the Tensor Art links to the other 2 Flux models are below:
Dream Diffusion Flux Pro : https://tensor.art/models/782777409377168908/FLUX-PRO-DREAM-DIFFUSION-V3
Dream Diffusion Flux Ultimate : https://tensor.art/models/772796227503912754/FLUX-ULTIMATE-DREAM-DIFFUSION-V2
FLUX PRO NSFW - By Dream Diffusion
MY DISCORD FOR THE WORKFLOWS AND DOWNLOADS : https://discord.gg/7aQNJgj4rT
CAUTION THIS CHECKPOINT CAN CREATE ***** CONTENT
FLUX PRO NSFW - By Dream Diffusion
WELCOME TO THE NEXT LEVEL OF GENERATIVE ART........
Also Built in my own detailer Module and My UHD Module. so you will see higher res images now with extra details. Incorporating ***** content realistic images with text from just prompt. Its quite the milestone in Flux development . I just hope you all enjoy using it on TA.
Its new High Res outputs and better adherence to prompts makes this version the best of all my checkpoints I have trained so far.
I really hope you enjoy using this version and please feel free to leave a comment for feedback or advice you may need with samplers , schedulers etc etc.
I have used a lot of car images in the gallery checkpoint as I find they are a great base to see the detail and some of the colours that this new checkpoint can produce.
I personally use these settings in TA when generating a render ....
Chose Advanced settings.
Sampler : Euler or DPM++2M
Scheduler : Simple or Beta
Steps : Around 20 to 30
Text Encoder: FP8, FP16 and FP32
CFG : For realism 4, For text in a image 4.5
Upscaler : I like to use Latent with around 11 steps with a denoise of 0.63
Enjoy the checkpoint and look forward to seeing your images below ..... DICE
I have optimized this version to be faster at generating images at 576 x 1024 either as landscape or portrait variants. Will render 30% quicker at that size.
I have an article on here that shows a few more complex prompts to try out if you like
https://tensor.art/articles/760080809324596639
Once you have downloaded my Flux Hyper Checkpoint you need to save it to the unet folder inside of comfy, if you dont see a unet folder create a new folder name it unet inside of the models folder.
If you use forge webui then save it to models/stable diffusion folder. Download the links from below and save them to your comfy or Forge folders...
Comfy Workflow : https://openart.ai/workflows/maitruclam/comfyui-workflow-for-flux-simple/iuRdGnfzmTbOOzONIiVV
Vae download this to your Vae folder inside of your model folder: https://huggingface.co/black-forest-labs/FLUX.1-schnell/tree/main/vae
Clip: download clip_l.safetensors and t5xxl_fp8_e4m3fn.safetensors download these 2 and save them to your clip folder inside of your models folder Download them from : https://huggingface.co/comfyanonymous/flux_text_encoders/tree/main
If you want to install forge on your local system, I have made a simple to follow YouTube Video to install Pinokio 1 click installer for forge that will run FLUX, SD3, SDXL and SD1.5. https://youtu.be/gftE_Z5vdHw?feature=shared
Feel Free to come join my Flux Discord Server to share you models https://discord.gg/8Tz3kceR
If you have any questions or issues feel free to drop a comment below and I will get back to you as soon as I can. Enjoy DICE