Instructions to use noema-art/threadguyflux with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use noema-art/threadguyflux with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("noema-art/threadguyflux") prompt = "This is a photograph of a young threadguy sitting in a black director's chair in what appears to be a studio or warehouse setting with a polished concrete floor. The threadguy, who has light skin and a well-groomed beard, is smiling and making a peace sign with his right hand. He is wearing a dark grey T-shirt with the text \"KILLER G\" printed in white on the back, along with a yellow graphic design. His pants are light brown, and he is wearing white sneakers." image = pipe(prompt).images[0] - Inference
- Notebooks
- Google Colab
- Kaggle
| job: extension | |
| config: | |
| name: threadguyflux | |
| process: | |
| - type: sd_trainer | |
| training_folder: output | |
| device: cuda:0 | |
| trigger_word: threadguy | |
| network: | |
| type: lora | |
| linear: 32 | |
| linear_alpha: 32 | |
| save: | |
| dtype: float16 | |
| save_every: 100 | |
| max_step_saves_to_keep: 16 | |
| datasets: | |
| - folder_path: /workspace/dataset | |
| caption_ext: txt | |
| caption_dropout_rate: 0.05 | |
| shuffle_tokens: false | |
| cache_latents_to_disk: true | |
| resolution: | |
| - 512 | |
| - 768 | |
| - 1024 | |
| train: | |
| batch_size: 1 | |
| steps: 4000 | |
| gradient_accumulation_steps: 1 | |
| train_unet: true | |
| train_text_encoder: false | |
| gradient_checkpointing: true | |
| noise_scheduler: flowmatch | |
| optimizer: adamw8bit | |
| lr: 0.0001 | |
| skip_first_sample: true | |
| disable_sampling: true | |
| ema_config: | |
| use_ema: true | |
| ema_decay: 0.99 | |
| dtype: bf16 | |
| model: | |
| name_or_path: black-forest-labs/FLUX.1-dev | |
| is_flux: true | |
| quantize: true | |
| low_vram: true | |
| sample: | |
| sampler: flowmatch | |
| sample_every: 500 | |
| width: 1024 | |
| height: 1024 | |
| prompts: null | |
| neg: '' | |
| seed: 42 | |
| walk_seed: true | |
| guidance_scale: 4 | |
| sample_steps: 20 | |
| meta: | |
| name: threadguyflux | |
| version: '1.0' | |