FLUX.1-dev leaf LoRA
According to the model card, these are LoRA adaptation weights for black-forest-labs/FLUX.1-dev, adding a leaf-image fine-tune on top of the base model.
Open Source Model Profile · lamm-mit
leaf-L-FLUX.1-dev
leaf-L-FLUX.1-dev is a lamm-mit LoRA adapter for FLUX.1-dev. According to the model card, it adds a leaf-microstructure style triggered with <leaf microstructure>.
- Publisher
- lamm-mit
- Task
- text-to-image
- Model type
- Unknown
- License
- apache-2.0
- Library
- diffusers
- Publication status
- Accepted · not indexed
Model overview
leaf-L-FLUX.1-dev is published by lamm-mit as a diffusers LoRA for text-to-image generation. According to the model card, it adapts black-forest-labs/FLUX.1-dev with a rank-64 adapter trained for 4,000 steps on leaf imagery. Prompts use the <leaf microstructure> trigger, and card data records apache-2.0 licensing.
Recorded capabilities
Rank-64, 4,000-step adapter
According to the model card, the adapter uses rank 64 and alpha 64 trained for 4,000 steps, and earlier checkpoints such as steps 3000 and 3500 can be selected through the weight-name parameter.
Leaf-microstructure trigger
According to the model card, prompts should include <leaf microstructure> to trigger the fine-tuned feature, which was trained on lamm-mit/leaf-flux-images-and-captions.
Documented diffusers workflow
According to the model card, loading runs through FluxPipeline with torch bfloat16, a 512 sequence length in the example, and load_lora_weights for the chosen checkpoint.
Use cases in the source record
- Leaf-microstructure text-to-image generation using the documented <leaf microstructure> trigger phrase.
- Biomimetic design-concept rendering following the card's example prompts, such as eco-friendly architectural concepts and staged object scenes.
Limitations and unknowns
- No parameter count was extracted for the adapter itself; the card describes the FLUX.1-dev base as a 12-billion-parameter transformer, which is a publisher claim rather than measured adapter data.
- The publisher states that access to the base model must be obtained first, so the adapter cannot be used standalone.
- No evaluation results were extracted from this record.
- Provider state is historical snapshot data, not independently refreshed current availability.
Source and provenance
Source: lamm-mit/leaf-L-FLUX.1-dev
Captured: Unknown. Processed: 2026-09-07T19:34:49.376874+00:00.
FLUX.1 [dev] Fine-tuned with Leaf Images FLUX.1 [dev] is a 12 billion parameter rectified flow transformer capable of generating images from text descriptions. Install diffusers pip install -U diffusers Model description These are LoRA adaption weights for the FLUX.1 [dev] model ( black-forest-labs/FLUX.1-dev ). The base model is, and you must first get access to it before loading this LoRA adapter. This LoRA adapter has rank=64 and alpha=64, trained for 4,000 steps. Earlier checkpoints are available in this repository as well (you can load these via the adapter parameter, see example below). Trigger keywords The following images we…
F001F002F003F004F005F006F007F010F011F012F013F014F016F019