Hugging Face's logo

Diffusers documentation

Guiders

You are viewing main version, which requires installation from source. If you'd like regular pip install, checkout the latest stable version (v0.34.0).
Hugging Face's logo
Join the Hugging Face community

and get access to the augmented documentation experience

to get started

Guiders

Guiders are components in Modular Diffusers that control how the diffusion process is guided during generation. They implement various guidance techniques to improve generation quality and control.

BaseGuidance

class diffusers.guiders.guider_utils.BaseGuidance

< >

( start: float = 0.0 stop: float = 1.0 )

Base class providing the skeleton for implementing guidance techniques.

cleanup_models

< >

( denoiser: Module )

Cleans up the models for the guidance technique after a given batch of data. This method should be overridden in subclasses to implement specific model cleanup logic. It is useful for removing any hooks or other stateful modifications made during prepare_models.

from_pretrained

< >

( pretrained_model_name_or_path: typing.Union[str, os.PathLike, NoneType] = None subfolder: typing.Optional[str] = None return_unused_kwargs = False **kwargs )

Parameters

  • pretrained_model_name_or_path (str or os.PathLike, optional) — Can be either:

    • A string, the model id (for example google/ddpm-celebahq-256) of a pretrained model hosted on the Hub.
    • A path to a directory (for example ./my_model_directory) containing the guider configuration saved with ~BaseGuidance.save_pretrained.
  • subfolder (str, optional) — The subfolder location of a model file within a larger model repository on the Hub or locally.
  • return_unused_kwargs (bool, optional, defaults to False) — Whether kwargs that are not consumed by the Python class should be returned or not.
  • cache_dir (Union[str, os.PathLike], optional) — Path to a directory where a downloaded pretrained model configuration is cached if the standard cache is not used.
  • force_download (bool, optional, defaults to False) — Whether or not to force the (re-)download of the model weights and configuration files, overriding the cached versions if they exist.
  • proxies (Dict[str, str], optional) — A dictionary of proxy servers to use by protocol or endpoint, for example, {'http': 'foo.bar:3128', 'http://hostname': 'foo.bar:4012'}. The proxies are used on each request.
  • output_loading_info(bool, optional, defaults to False) — Whether or not to also return a dictionary containing missing keys, unexpected keys and error messages.
  • local_files_only(bool, optional, defaults to False) — Whether to only load local model weights and configuration files or not. If set to True, the model won’t be downloaded from the Hub.
  • token (str or bool, optional) — The token to use as HTTP bearer authorization for remote files. If True, the token generated from diffusers-cli login (stored in ~/.huggingface) is used.
  • revision (str, optional, defaults to "main") — The specific model version to use. It can be a branch name, a tag name, a commit id, or any identifier allowed by Git.

Instantiate a guider from a pre-defined JSON configuration file in a local directory or Hub repository.

To use private or gated models, log-in with hf auth login. You can also activate the special “offline-mode” to use this method in a firewalled environment.

prepare_models

< >

( denoiser: Module )

Prepares the models for the guidance technique on a given batch of data. This method should be overridden in subclasses to implement specific model preparation logic.

save_pretrained

< >

( save_directory: typing.Union[str, os.PathLike] push_to_hub: bool = False **kwargs )

Parameters

  • save_directory (str or os.PathLike) — Directory where the configuration JSON file will be saved (will be created if it does not exist).
  • push_to_hub (bool, optional, defaults to False) — Whether or not to push your model to the Hugging Face Hub after saving it. You can specify the repository you want to push to with repo_id (will default to the name of save_directory in your namespace).
  • kwargs (Dict[str, Any], optional) — Additional keyword arguments passed along to the push_to_hub() method.

Save a guider configuration object to a directory so that it can be reloaded using the ~BaseGuidance.from_pretrained class method.

set_input_fields

< >

( **kwargs: typing.Dict[str, typing.Union[str, typing.Tuple[str, str]]] )

Parameters

  • **kwargs (Dict[str, Union[str, Tuple[str, str]]]) — A dictionary where the keys are the names of the fields that will be used to store the data once it is prepared with prepare_inputs. The values can be either a string or a tuple of length 2, which is used to look up the required data provided for preparation.

    If a string is provided, it will be used as the conditional data (or unconditional if used with a guidance method that requires it). If a tuple of length 2 is provided, the first element must be the conditional data identifier and the second element must be the unconditional data identifier or None.

    Example:

Set the input fields for the guidance technique. The input fields are used to specify the names of the returned attributes containing the prepared data after prepare_inputs is called. The prepared data is obtained from the values of the provided keyword arguments to this method.

ClassifierFreeGuidance

class diffusers.ClassifierFreeGuidance

< >

( guidance_scale: float = 7.5 guidance_rescale: float = 0.0 use_original_formulation: bool = False start: float = 0.0 stop: float = 1.0 )

Parameters

  • guidance_scale (float, defaults to 7.5) — The scale parameter for classifier-free guidance. Higher values result in stronger conditioning on the text prompt, while lower values allow for more freedom in generation. Higher values may lead to saturation and deterioration of image quality.
  • guidance_rescale (float, defaults to 0.0) — The rescale factor applied to the noise predictions. This is used to improve image quality and fix overexposure. Based on Section 3.4 from Common Diffusion Noise Schedules and Sample Steps are Flawed.
  • use_original_formulation (bool, defaults to False) — Whether to use the original formulation of classifier-free guidance as proposed in the paper. By default, we use the diffusers-native implementation that has been in the codebase for a long time. See [~guiders.classifier_free_guidance.ClassifierFreeGuidance] for more details.
  • start (float, defaults to 0.0) — The fraction of the total number of denoising steps after which guidance starts.
  • stop (float, defaults to 1.0) — The fraction of the total number of denoising steps after which guidance stops.

Classifier-free guidance (CFG): https://huggingface.co/papers/2207.12598

CFG is a technique used to improve generation quality and condition-following in diffusion models. It works by jointly training a model on both conditional and unconditional data, and using a weighted sum of the two during inference. This allows the model to tradeoff between generation quality and sample diversity. The original paper proposes scaling and shifting the conditional distribution based on the difference between conditional and unconditional predictions. [x_pred = x_cond + scale * (x_cond - x_uncond)]

Diffusers implemented the scaling and shifting on the unconditional prediction instead based on the Imagen paper, which is equivalent to what the original paper proposed in theory. [x_pred = x_uncond + scale * (x_cond - x_uncond)]

The intution behind the original formulation can be thought of as moving the conditional distribution estimates further away from the unconditional distribution estimates, while the diffusers-native implementation can be thought of as moving the unconditional distribution towards the conditional distribution estimates to get rid of the unconditional predictions (usually negative features like “bad quality, bad anotomy, watermarks”, etc.)

The use_original_formulation argument can be set to True to use the original CFG formulation mentioned in the paper. By default, we use the diffusers-native implementation that has been in the codebase for a long time.

ClassifierFreeZeroStarGuidance

class diffusers.ClassifierFreeZeroStarGuidance

< >

( guidance_scale: float = 7.5 zero_init_steps: int = 1 guidance_rescale: float = 0.0 use_original_formulation: bool = False start: float = 0.0 stop: float = 1.0 )

Parameters

  • guidance_scale (float, defaults to 7.5) — The scale parameter for classifier-free guidance. Higher values result in stronger conditioning on the text prompt, while lower values allow for more freedom in generation. Higher values may lead to saturation and deterioration of image quality.
  • zero_init_steps (int, defaults to 1) — The number of inference steps for which the noise predictions are zeroed out (see Section 4.2).
  • guidance_rescale (float, defaults to 0.0) — The rescale factor applied to the noise predictions. This is used to improve image quality and fix overexposure. Based on Section 3.4 from Common Diffusion Noise Schedules and Sample Steps are Flawed.
  • use_original_formulation (bool, defaults to False) — Whether to use the original formulation of classifier-free guidance as proposed in the paper. By default, we use the diffusers-native implementation that has been in the codebase for a long time. See [~guiders.classifier_free_guidance.ClassifierFreeGuidance] for more details.
  • start (float, defaults to 0.01) — The fraction of the total number of denoising steps after which guidance starts.
  • stop (float, defaults to 0.2) — The fraction of the total number of denoising steps after which guidance stops.

Classifier-free Zero(CFG-Zero): https://huggingface.co/papers/2503.18886

This is an implementation of the Classifier-Free Zero* guidance technique, which is a variant of classifier-free guidance. It proposes zero initialization of the noise predictions for the first few steps of the diffusion process, and also introduces an optimal rescaling factor for the noise predictions, which can help in improving the quality of generated images.

The authors of the paper suggest setting zero initialization in the first 4% of the inference steps.

SkipLayerGuidance

class diffusers.SkipLayerGuidance

< >

( guidance_scale: float = 7.5 skip_layer_guidance_scale: float = 2.8 skip_layer_guidance_start: float = 0.01 skip_layer_guidance_stop: float = 0.2 skip_layer_guidance_layers: typing.Union[int, typing.List[int], NoneType] = None skip_layer_config: typing.Union[diffusers.hooks.layer_skip.LayerSkipConfig, typing.List[diffusers.hooks.layer_skip.LayerSkipConfig], typing.Dict[str, typing.Any]] = None guidance_rescale: float = 0.0 use_original_formulation: bool = False start: float = 0.0 stop: float = 1.0 )

Parameters

  • guidance_scale (float, defaults to 7.5) — The scale parameter for classifier-free guidance. Higher values result in stronger conditioning on the text prompt, while lower values allow for more freedom in generation. Higher values may lead to saturation and deterioration of image quality.
  • skip_layer_guidance_scale (float, defaults to 2.8) — The scale parameter for skip layer guidance. Anatomy and structure coherence may improve with higher values, but it may also lead to overexposure and saturation.
  • skip_layer_guidance_start (float, defaults to 0.01) — The fraction of the total number of denoising steps after which skip layer guidance starts.
  • skip_layer_guidance_stop (float, defaults to 0.2) — The fraction of the total number of denoising steps after which skip layer guidance stops.
  • skip_layer_guidance_layers (int or List[int], optional) — The layer indices to apply skip layer guidance to. Can be a single integer or a list of integers. If not provided, skip_layer_config must be provided. The recommended values are [7, 8, 9] for Stable Diffusion 3.5 Medium.
  • skip_layer_config (LayerSkipConfig or List[LayerSkipConfig], optional) — The configuration for the skip layer guidance. Can be a single LayerSkipConfig or a list of LayerSkipConfig. If not provided, skip_layer_guidance_layers must be provided.
  • guidance_rescale (float, defaults to 0.0) — The rescale factor applied to the noise predictions. This is used to improve image quality and fix overexposure. Based on Section 3.4 from Common Diffusion Noise Schedules and Sample Steps are Flawed.
  • use_original_formulation (bool, defaults to False) — Whether to use the original formulation of classifier-free guidance as proposed in the paper. By default, we use the diffusers-native implementation that has been in the codebase for a long time. See [~guiders.classifier_free_guidance.ClassifierFreeGuidance] for more details.
  • start (float, defaults to 0.01) — The fraction of the total number of denoising steps after which guidance starts.
  • stop (float, defaults to 0.2) — The fraction of the total number of denoising steps after which guidance stops.

Skip Layer Guidance (SLG): https://github.com/Stability-AI/sd3.5

Spatio-Temporal Guidance (STG): https://huggingface.co/papers/2411.18664

SLG was introduced by StabilityAI for improving structure and anotomy coherence in generated images. It works by skipping the forward pass of specified transformer blocks during the denoising process on an additional conditional batch of data, apart from the conditional and unconditional batches already used in CFG ([~guiders.classifier_free_guidance.ClassifierFreeGuidance]), and then scaling and shifting the CFG predictions based on the difference between conditional without skipping and conditional with skipping predictions.

The intution behind SLG can be thought of as moving the CFG predicted distribution estimates further away from worse versions of the conditional distribution estimates (because skipping layers is equivalent to using a worse version of the model for the conditional prediction).

STG is an improvement and follow-up work combining ideas from SLG, PAG and similar techniques for improving generation quality in video diffusion models.

Additional reading:

The values for skip_layer_guidance_scale, skip_layer_guidance_start, and skip_layer_guidance_stop are defaulted to the recommendations by StabilityAI for Stable Diffusion 3.5 Medium.

SmoothedEnergyGuidance

class diffusers.SmoothedEnergyGuidance

< >

( guidance_scale: float = 7.5 seg_guidance_scale: float = 2.8 seg_blur_sigma: float = 9999999.0 seg_blur_threshold_inf: float = 9999.0 seg_guidance_start: float = 0.0 seg_guidance_stop: float = 1.0 seg_guidance_layers: typing.Union[int, typing.List[int], NoneType] = None seg_guidance_config: typing.Union[diffusers.hooks.smoothed_energy_guidance_utils.SmoothedEnergyGuidanceConfig, typing.List[diffusers.hooks.smoothed_energy_guidance_utils.SmoothedEnergyGuidanceConfig]] = None guidance_rescale: float = 0.0 use_original_formulation: bool = False start: float = 0.0 stop: float = 1.0 )

Parameters

  • guidance_scale (float, defaults to 7.5) — The scale parameter for classifier-free guidance. Higher values result in stronger conditioning on the text prompt, while lower values allow for more freedom in generation. Higher values may lead to saturation and deterioration of image quality.
  • seg_guidance_scale (float, defaults to 3.0) — The scale parameter for smoothed energy guidance. Anatomy and structure coherence may improve with higher values, but it may also lead to overexposure and saturation.
  • seg_blur_sigma (float, defaults to 9999999.0) — The amount by which we blur the attention weights. Setting this value greater than 9999.0 results in infinite blur, which means uniform queries. Controlling it exponentially is empirically effective.
  • seg_blur_threshold_inf (float, defaults to 9999.0) — The threshold above which the blur is considered infinite.
  • seg_guidance_start (float, defaults to 0.0) — The fraction of the total number of denoising steps after which smoothed energy guidance starts.
  • seg_guidance_stop (float, defaults to 1.0) — The fraction of the total number of denoising steps after which smoothed energy guidance stops.
  • seg_guidance_layers (int or List[int], optional) — The layer indices to apply smoothed energy guidance to. Can be a single integer or a list of integers. If not provided, seg_guidance_config must be provided. The recommended values are [7, 8, 9] for Stable Diffusion 3.5 Medium.
  • seg_guidance_config (SmoothedEnergyGuidanceConfig or List[SmoothedEnergyGuidanceConfig], optional) — The configuration for the smoothed energy layer guidance. Can be a single SmoothedEnergyGuidanceConfig or a list of SmoothedEnergyGuidanceConfig. If not provided, seg_guidance_layers must be provided.
  • guidance_rescale (float, defaults to 0.0) — The rescale factor applied to the noise predictions. This is used to improve image quality and fix overexposure. Based on Section 3.4 from Common Diffusion Noise Schedules and Sample Steps are Flawed.
  • use_original_formulation (bool, defaults to False) — Whether to use the original formulation of classifier-free guidance as proposed in the paper. By default, we use the diffusers-native implementation that has been in the codebase for a long time. See [~guiders.classifier_free_guidance.ClassifierFreeGuidance] for more details.
  • start (float, defaults to 0.01) — The fraction of the total number of denoising steps after which guidance starts.
  • stop (float, defaults to 0.2) — The fraction of the total number of denoising steps after which guidance stops.

Smoothed Energy Guidance (SEG): https://huggingface.co/papers/2408.00760

SEG is only supported as an experimental prototype feature for now, so the implementation may be modified in the future without warning or guarantee of reproducibility. This implementation assumes:

  • Generated images are square (height == width)
  • The model does not combine different modalities together (e.g., text and image latent streams are not combined together such as Flux)

PerturbedAttentionGuidance

class diffusers.PerturbedAttentionGuidance

< >

( guidance_scale: float = 7.5 perturbed_guidance_scale: float = 2.8 perturbed_guidance_start: float = 0.01 perturbed_guidance_stop: float = 0.2 perturbed_guidance_layers: typing.Union[int, typing.List[int], NoneType] = None perturbed_guidance_config: typing.Union[diffusers.hooks.layer_skip.LayerSkipConfig, typing.List[diffusers.hooks.layer_skip.LayerSkipConfig], typing.Dict[str, typing.Any]] = None guidance_rescale: float = 0.0 use_original_formulation: bool = False start: float = 0.0 stop: float = 1.0 )

Parameters

  • guidance_scale (float, defaults to 7.5) — The scale parameter for classifier-free guidance. Higher values result in stronger conditioning on the text prompt, while lower values allow for more freedom in generation. Higher values may lead to saturation and deterioration of image quality.
  • perturbed_guidance_scale (float, defaults to 2.8) — The scale parameter for perturbed attention guidance.
  • perturbed_guidance_start (float, defaults to 0.01) — The fraction of the total number of denoising steps after which perturbed attention guidance starts.
  • perturbed_guidance_stop (float, defaults to 0.2) — The fraction of the total number of denoising steps after which perturbed attention guidance stops.
  • perturbed_guidance_layers (int or List[int], optional) — The layer indices to apply perturbed attention guidance to. Can be a single integer or a list of integers. If not provided, perturbed_guidance_config must be provided.
  • perturbed_guidance_config (LayerSkipConfig or List[LayerSkipConfig], optional) — The configuration for the perturbed attention guidance. Can be a single LayerSkipConfig or a list of LayerSkipConfig. If not provided, perturbed_guidance_layers must be provided.
  • guidance_rescale (float, defaults to 0.0) — The rescale factor applied to the noise predictions. This is used to improve image quality and fix overexposure. Based on Section 3.4 from Common Diffusion Noise Schedules and Sample Steps are Flawed.
  • use_original_formulation (bool, defaults to False) — Whether to use the original formulation of classifier-free guidance as proposed in the paper. By default, we use the diffusers-native implementation that has been in the codebase for a long time. See [~guiders.classifier_free_guidance.ClassifierFreeGuidance] for more details.
  • start (float, defaults to 0.01) — The fraction of the total number of denoising steps after which guidance starts.
  • stop (float, defaults to 0.2) — The fraction of the total number of denoising steps after which guidance stops.

Perturbed Attention Guidance (PAG): https://huggingface.co/papers/2403.17377

The intution behind PAG can be thought of as moving the CFG predicted distribution estimates further away from worse versions of the conditional distribution estimates. PAG was one of the first techniques to introduce the idea of using a worse version of the trained model for better guiding itself in the denoising process. It perturbs the attention scores of the latent stream by replacing the score matrix with an identity matrix for selectively chosen layers.

Additional reading:

PAG is implemented with similar implementation to SkipLayerGuidance due to overlap in the configuration parameters and implementation details.

AdaptiveProjectedGuidance

class diffusers.AdaptiveProjectedGuidance

< >

( guidance_scale: float = 7.5 adaptive_projected_guidance_momentum: typing.Optional[float] = None adaptive_projected_guidance_rescale: float = 15.0 eta: float = 1.0 guidance_rescale: float = 0.0 use_original_formulation: bool = False start: float = 0.0 stop: float = 1.0 )

Parameters

  • guidance_scale (float, defaults to 7.5) — The scale parameter for classifier-free guidance. Higher values result in stronger conditioning on the text prompt, while lower values allow for more freedom in generation. Higher values may lead to saturation and deterioration of image quality.
  • adaptive_projected_guidance_momentum (float, defaults to None) — The momentum parameter for the adaptive projected guidance. Disabled if set to None.
  • adaptive_projected_guidance_rescale (float, defaults to 15.0) — The rescale factor applied to the noise predictions. This is used to improve image quality and fix
  • guidance_rescale (float, defaults to 0.0) — The rescale factor applied to the noise predictions. This is used to improve image quality and fix overexposure. Based on Section 3.4 from Common Diffusion Noise Schedules and Sample Steps are Flawed.
  • use_original_formulation (bool, defaults to False) — Whether to use the original formulation of classifier-free guidance as proposed in the paper. By default, we use the diffusers-native implementation that has been in the codebase for a long time. See [~guiders.classifier_free_guidance.ClassifierFreeGuidance] for more details.
  • start (float, defaults to 0.0) — The fraction of the total number of denoising steps after which guidance starts.
  • stop (float, defaults to 1.0) — The fraction of the total number of denoising steps after which guidance stops.

Adaptive Projected Guidance (APG): https://huggingface.co/papers/2410.02416

AutoGuidance

class diffusers.AutoGuidance

< >

( guidance_scale: float = 7.5 auto_guidance_layers: typing.Union[int, typing.List[int], NoneType] = None auto_guidance_config: typing.Union[diffusers.hooks.layer_skip.LayerSkipConfig, typing.List[diffusers.hooks.layer_skip.LayerSkipConfig], typing.Dict[str, typing.Any]] = None dropout: typing.Optional[float] = None guidance_rescale: float = 0.0 use_original_formulation: bool = False start: float = 0.0 stop: float = 1.0 )

Parameters

  • guidance_scale (float, defaults to 7.5) — The scale parameter for classifier-free guidance. Higher values result in stronger conditioning on the text prompt, while lower values allow for more freedom in generation. Higher values may lead to saturation and deterioration of image quality.
  • auto_guidance_layers (int or List[int], optional) — The layer indices to apply skip layer guidance to. Can be a single integer or a list of integers. If not provided, skip_layer_config must be provided.
  • auto_guidance_config (LayerSkipConfig or List[LayerSkipConfig], optional) — The configuration for the skip layer guidance. Can be a single LayerSkipConfig or a list of LayerSkipConfig. If not provided, skip_layer_guidance_layers must be provided.
  • dropout (float, optional) — The dropout probability for autoguidance on the enabled skip layers (either with auto_guidance_layers or auto_guidance_config). If not provided, the dropout probability will be set to 1.0.
  • guidance_rescale (float, defaults to 0.0) — The rescale factor applied to the noise predictions. This is used to improve image quality and fix overexposure. Based on Section 3.4 from Common Diffusion Noise Schedules and Sample Steps are Flawed.
  • use_original_formulation (bool, defaults to False) — Whether to use the original formulation of classifier-free guidance as proposed in the paper. By default, we use the diffusers-native implementation that has been in the codebase for a long time. See [~guiders.classifier_free_guidance.ClassifierFreeGuidance] for more details.
  • start (float, defaults to 0.0) — The fraction of the total number of denoising steps after which guidance starts.
  • stop (float, defaults to 1.0) — The fraction of the total number of denoising steps after which guidance stops.

AutoGuidance: https://huggingface.co/papers/2406.02507

TangentialClassifierFreeGuidance

class diffusers.TangentialClassifierFreeGuidance

< >

( guidance_scale: float = 7.5 guidance_rescale: float = 0.0 use_original_formulation: bool = False start: float = 0.0 stop: float = 1.0 )

Parameters

  • guidance_scale (float, defaults to 7.5) — The scale parameter for classifier-free guidance. Higher values result in stronger conditioning on the text prompt, while lower values allow for more freedom in generation. Higher values may lead to saturation and deterioration of image quality.
  • guidance_rescale (float, defaults to 0.0) — The rescale factor applied to the noise predictions. This is used to improve image quality and fix overexposure. Based on Section 3.4 from Common Diffusion Noise Schedules and Sample Steps are Flawed.
  • use_original_formulation (bool, defaults to False) — Whether to use the original formulation of classifier-free guidance as proposed in the paper. By default, we use the diffusers-native implementation that has been in the codebase for a long time. See [~guiders.classifier_free_guidance.ClassifierFreeGuidance] for more details.
  • start (float, defaults to 0.0) — The fraction of the total number of denoising steps after which guidance starts.
  • stop (float, defaults to 1.0) — The fraction of the total number of denoising steps after which guidance stops.

Tangential Classifier Free Guidance (TCFG): https://huggingface.co/papers/2503.18137

< > Update on GitHub

←Components and configs IP-Adapter→