Instructions to use quarterturn/id4_nichijou with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use quarterturn/id4_nichijou with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("ideogram-ai/ideogram-4-fp8", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("quarterturn/id4_nichijou") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
metadata
tags:
- lora
- diffusers
- ideogram4
- bf8
base_model: ideogram-ai/ideogram-4-fp8
license: cc-by-nc-4.0
An Ideogram 4 fp8 style LoRA for making Nichijou images.
Trained on my Nichijou dataset https://huggingface.co/datasets/quarterturn/nichijou-hd-json-captioned. Checkpoints to step 20000 are provided, but I think 19500 was closest so for now that is id4_nichijou.safetensors.
The dataset was character tagged (to the best of qwen3.6-35b-a3b's ability) so you should be able to make a scene with any main character just by name, though you'll still need to desrcribe their state of dress, since that varies in the actual anime.
