Instructions to use nphSi/Z-Image-Lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use nphSi/Z-Image-Lora with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Tongyi-MAI/Z-Image,Tongyi-MAI/Z-Image-Turbo", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("nphSi/Z-Image-Lora") prompt = "Alexandra Chando (vrtlAlexandraChando)" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
LoRa Feedback / Request Re-Training / Tips for Workflows, Prompting, Interferencing etc.
Please be as descriptive as possible. List your used Tools, Models, Samplers etc. SFW samples allowed but keep it in a sane amount and size.
Wow, I just saw you made loras of Malina Weissman and Ciara Renee and downloaded them like lightning!
Thank you so much for taking the time and effort to answer my request!
I really appreciate all your work and the quality of your loras!
Thank you, and have an amazing weekend!
As for the Joanne Kelly lora I've been loving, here's a sample.
It was made using Chroma Forge, the regular Z-Image Turbo checkpoint, 9 steps. Can give prompt if anyone wants it.
The E-Bike LoRA is awesome, but I just can't get the characters to ride standing up. And any character lora I tried gets kind of brown or green eyes. Can you tell me how you trained it? Did your dataset even include riders?
That would be great, because I've been wanting to make one for my Surron Light Bee, but I don't really know how to go about it. Only the bike, with riders, sitting, standing, jumping, caption the images or not, and so on....
Its only the bike without humans or captions. It has no real use other than having roughly the same bike in every generation. Training skills with bikes could be tricky. Before caption you would have to be sure the text encoder understands what you caption and then use the exact same prompt as caption. If you cant get "standing on bike" or anything else to generate you will have a hard time to train it. Anyway good luck π
Btw i bought that bike last week and had the images sitting in my inbound folder. I just quickly made a Lora before deleting them π€£
And caption was "vrtlebike, bicycle". I have no clue why it would change eye color.
@nphSi Thanks for the newest loras of Tala Ashe and Camren Bicondova. Hopefully they're not your last. π₯²
No worries, still a lot of Loras to do.
No worries, still a lot of Loras to do.
That's really great to hear! Thank you for all your amazing work, and have a great weekend!
