Documentation for custom scripts?

[pjain]

Using the example below, I'd like to instantiate the model once and generate multiple images. In some cases, the image may require a LoRA or Multiple LoRAs.

# Load the model
flux = Flux1.from_name(
   model_name="schnell",  # "schnell" or "dev"
   quantize=8,            # 4 or 8
)

# Generate an image
image = flux.generate_image(
   seed=2,
   prompt="Luxury food photograph",
   config=Config(
      num_inference_steps=2,  # "schnell" works well with 2-4 steps, "dev" works well with 20-25 steps
      height=1024,
      width=1024,
   )
)

image.save(path="image.png")

Is there a way to do something like this to set and reset LoRAs per image being created rather than having to initialize the model again?

from mflux import Flux1, Config

# Load the model
flux = Flux1.from_name(
   model_name="schnell",  # "schnell" or "dev"
   quantize=8,            # 4 or 8
)


foo = {
    'a': {'prompt': 'foo',
          'lora': '/usr/local/fbar'},
    'b': {'prompt': 'bar'},
    'c': {'prompt': 'fubar',
          'lora': ['/usr/local/fbar', '/usr/lcoal/bbar'],
          'scales': '0.5 0.9'
          }
}

for key, value in foo.items():
    if 'lora' in value and value['lora']:
        flux.lora_paths = value['lora']
    else:
        # reset the lora to nothing in case this image doesn't have one
        flux.lora_paths = None
    if 'scales' in value and value['scales']:
        flux.lora_scales = value['scales']
    else:
        # reset the lora to nothing in case this image doesn't have one
        flux.lora_scales = Nonex
                                   
    # Generate an image
    image = flux.generate_image(
        seed=2,
        prompt=p,
        config=Config(
            num_inference_steps=2,  # "schnell" works well with 2-4 steps, "dev" works well with 20-25 steps
            height=1024,
            width=1024,
        )
    )

    image.save(path="image.png")