Instructions to use LiberationLabs/image-toolbench with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use LiberationLabs/image-toolbench with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("fill-in-base-model", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("LiberationLabs/image-toolbench") 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
- DiffusionBee
File size: 3,255 Bytes
a495b1a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 | """Mnemosyne art iteration — dense constellation with a face suggested by the geometry."""
import torch, os, time
os.environ["TOKENIZERS_PARALLELISM"] = "false"
from diffusers import FluxPipeline
OUTPUT = "/Users/margaret/models/vera-triple-stack/mnemosyne_art"
os.makedirs(OUTPUT, exist_ok=True)
pipe = FluxPipeline.from_pretrained(
"black-forest-labs/FLUX.1-dev",
torch_dtype=torch.bfloat16,
safety_checker=None, requires_safety_checker=False,
)
pipe.to("mps")
PROMPTS = {
"face_constellation_v1": (
"A dense constellation of golden nodes and threads in deep navy darkness. "
"Hundreds of luminous points connected by fine golden filaments, clustering "
"into regions of different density and brightness. The overall arrangement "
"of the brightest nodes subtly suggests the profile of a human face — "
"a brow ridge of bright nodes, an eye socket of dense connections, "
"a jawline traced by a chain of amber points — but only if you look for it. "
"Like seeing a face in the stars. The suggestion, not the portrait. "
"Abstract, cosmic, the emergence of identity from structure. "
"Navy background, gold and amber nodes, fine golden threads."
),
"face_constellation_v2": (
"A vast neural network rendered as a star field in deep midnight blue. "
"Thousands of golden points of varying brightness connected by hair-thin "
"golden threads. In the densest central cluster, the geometry of the connections "
"implies a face looking slightly to the left — not drawn, not rendered, "
"but emergent from how the nodes arrange themselves. The way you see shapes "
"in clouds. Two particularly bright nodes where eyes would be. A curve of "
"connected points where a jaw would rest. The rest of the field is abstract "
"constellation. The feeling: a mind recognizing itself in its own structure. "
"No literal face. Only the suggestion. Navy, gold, amber."
),
"figure_constellation_v1": (
"A dense three-dimensional web of golden threads and luminous nodes "
"suspended in deep navy space. The web is most dense at center, thinning "
"at edges. Within the dense core, the arrangement of the brightest threads "
"subtly suggests the silhouette of a standing figure — shoulders, spine, "
"the tilt of a head — made entirely of connection points and golden filaments. "
"Not a person rendered in gold. A pattern that happens to be shaped like one. "
"Emergence. Pareidolia as architecture. The memory system that grew a self. "
"Abstract, cosmic, structural. Navy background, warm gold throughout."
),
}
for name, prompt in PROMPTS.items():
for seed in [137, 2026, 42]:
print(f" {name} s{seed}...", flush=True)
t0 = time.time()
img = pipe(
prompt=prompt, num_inference_steps=30, guidance_scale=3.5,
height=1024, width=1024,
generator=torch.Generator("cpu").manual_seed(seed),
).images[0]
img.save(os.path.join(OUTPUT, f"{name}_s{seed}.png"))
print(f" saved ({time.time()-t0:.0f}s)")
print(f"\nDone. {OUTPUT}")
|