Instructions to use fal/FLUX.2-dev-Turbo with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use fal/FLUX.2-dev-Turbo 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.2-dev", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("fal/FLUX.2-dev-Turbo") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Inference
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- Draw Things
- DiffusionBee
LoRA Key discussion
#2
by ByteZSzn - opened
LOL
You credited me now i shall meet you in Valhalla
ByteZSzn changed discussion status to closed
I downloaded the original lora and your Comfy'd lora and am getting the exact same output in ComfyUI. Am I doing something wrong?
ByteZSzn changed discussion title from At least credit me, you took my converted comfyui lora and didnt even credit me lol to LoRA Key discussion
ByteZSzn changed discussion status to open
ByteZSzn changed discussion status to closed
@eepos it should work, update your comfyui to the latest, make sure you're using the comfyui version. I converted the keys 1-1 with the expected format so it should work.
It does work yes, the lora has effect. My point is both work the same, the original non-comfyui'd version works just as well.
So I'm just not understanding what the comfyuification of the lora is improving upon.
ByteZSzn changed discussion status to open
ByteZSzn changed discussion status to closed