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
| """Fresh art for stale project pages β Oracle, Lyra Technique, Garuda, Agni, Meridian. | |
| Matching site aesthetic: navy/gold, no human faces on product pages, | |
| abstract/architectural, the feel of the tool working. | |
| """ | |
| import torch, os, gc, time | |
| os.environ["TOKENIZERS_PARALLELISM"] = "false" | |
| from diffusers import FluxPipeline | |
| OUTPUT = "/Users/margaret/models/vera-triple-stack/project_art_refresh" | |
| 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") | |
| PROJECTS = { | |
| "oracle": ( | |
| "A vast circular diagnostic instrument made of dark navy metal and gold, " | |
| "suspended in space. Concentric rings rotate independently β each ring inscribed " | |
| "with geometric patterns that glow amber when active. At the center, a lens focuses " | |
| "golden light onto a stream of flowing data rendered as a luminous ribbon. " | |
| "The instrument is reading something, measuring, diagnosing β you can feel it working. " | |
| "No figures, no text. Deep navy background with warm gold accents. " | |
| "The aesthetic of a precision measurement device that sees what the output hides." | |
| ), | |
| "lyra_technique": ( | |
| "A cross-section of layered geometric space β imagine slicing through a transformer's " | |
| "internal representation and seeing the structure. Parallel planes of dark navy, each with " | |
| "different gold patterns: one shows spectral ridges (sharp peaks), another shows flat entropy " | |
| "(smooth), another shows branching pathways. The planes are slightly transparent, stacked, " | |
| "with golden threads connecting corresponding points between layers. " | |
| "The feel of looking inside a mind and seeing geometry where thoughts should be. " | |
| "Abstract, architectural, precise. Navy and gold only." | |
| ), | |
| "garuda": ( | |
| "A golden eagle rendered in geometric, architectural style β not naturalistic but structural. " | |
| "Wings spread wide, each feather is a sharp angular plate of burnished gold against deep navy. " | |
| "The eagle's eye is a lens, scanning. Below it, faint red threads represent intercepted threats β " | |
| "thin, fragile against the eagle's mass. The eagle doesn't chase the threats; it watches, " | |
| "assesses, decides. Heraldic but modern. The feel of a sentinel that devours serpents. " | |
| "No text. Navy background, gold and amber eagle, red threat indicators." | |
| ), | |
| "agni": ( | |
| "A crucible of dark navy ceramic filled with molten gold β but the gold is being tested. " | |
| "Seven geometric probes descend into the liquid, each glowing a different shade of amber " | |
| "depending on what they find. Some glow bright (passing), some glow dim (failing). " | |
| "The crucible has fine cracks filled with cooled gold (kintsugi). Around it, faint traces " | |
| "of rejected material β things that didn't survive the fire. " | |
| "The feel of purification through rigorous testing. " | |
| "Abstract, no figures, no text. Navy ceramic, molten gold, amber probes." | |
| ), | |
| "meridian": ( | |
| "A compass rose made of intersecting golden arcs on a dark navy field, " | |
| "but the arcs are alive β papers, documents, data streams flowing along the curves " | |
| "toward a central nexus point where they converge and transform into structured knowledge. " | |
| "The streams come from all directions. Some are bright (relevant), some dim (filtered out). " | |
| "At the center, a small dense cluster of interconnected golden nodes β the knowledge graph " | |
| "being built in real time. The feel of a swarm that never sleeps, finding what matters. " | |
| "No figures, no text. Navy background, gold arcs and streams." | |
| ), | |
| } | |
| for name, prompt in PROJECTS.items(): | |
| for seed in [137, 2026]: | |
| 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)") | |
| gc.collect(); torch.mps.empty_cache() | |
| print(f"\nDone. {OUTPUT}") | |