experimental_experiment.torch_models.diffusion_model_helper¶

experimental_experiment.torch_models.diffusion_model_helper.get_stable_diffusion_2_unet(inputs_as_dict: bool = False, overwrite: bool = False, **kwargs) → Tuple[Any, Tuple[Any, ...] | Dict[str, Any]][source]¶

Gets a non initialized model.

Parameters:

inputs_as_dict – returns dummy inputs as a dictionary or not
overwrite – do not consider the config from the true model
kwargs – to overwrite the configuration, example num_hidden_layers=1

Returns:

model, inputs

See SableDiffusion2Unet.