DDIMScheduler
Last updated
Last updated
(DDIM) by Jiaming Song, Chenlin Meng and Stefano Ermon.
The abstract from the paper is:
Denoising diffusion probabilistic models (DDPMs) have achieved high quality image generation without adversarial training, yet they require simulating a Markov chain for many steps to produce a sample. To accelerate sampling, we present denoising diffusion implicit models (DDIMs), a more efficient class of iterative implicit probabilistic models with the same training procedure as DDPMs. In DDPMs, the generative process is defined as the reverse of a Markovian diffusion process. We construct a class of non-Markovian diffusion processes that lead to the same training objective, but whose reverse process can be much faster to sample from. We empirically demonstrate that DDIMs can produce high quality samples 10× to 50× faster in terms of wall-clock time compared to DDPMs, allow us to trade off computation for sample quality, and can perform semantically meaningful image interpolation directly in the latent space.
The original codebase of this paper can be found at , and you can contact the author on .
The paper claims that a mismatch between the training and inference settings leads to suboptimal inference generation results for Stable Diffusion. To fix this, the authors propose:
🧪 This is an experimental feature!
rescale the noise schedule to enforce zero terminal signal-to-noise ratio (SNR)
Copied
train a model with v_prediction (add the following argument to the or scripts)
Copied
change the sampler to always start from the last timestep
Copied
rescale classifier-free guidance to prevent over-exposure
Copied
For example:
Copied
( num_train_timesteps: int = 1000beta_start: float = 0.0001beta_end: float = 0.02beta_schedule: str = 'linear'trained_betas: typing.Union[numpy.ndarray, typing.List[float], NoneType] = Noneclip_sample: bool = Trueset_alpha_to_one: bool = Truesteps_offset: int = 0prediction_type: str = 'epsilon'thresholding: bool = Falsedynamic_thresholding_ratio: float = 0.995clip_sample_range: float = 1.0sample_max_value: float = 1.0timestep_spacing: str = 'leading'rescale_betas_zero_snr: bool = False )
Parameters
num_train_timesteps (int, defaults to 1000) — The number of diffusion steps to train the model.
beta_start (float, defaults to 0.0001) — The starting beta value of inference.
beta_end (float, defaults to 0.02) — The final beta value.
beta_schedule (str, defaults to "linear") — The beta schedule, a mapping from a beta range to a sequence of betas for stepping the model. Choose from linear, scaled_linear, or squaredcos_cap_v2.
trained_betas (np.ndarray, optional) — Pass an array of betas directly to the constructor to bypass beta_start and beta_end.
clip_sample (bool, defaults to True) — Clip the predicted sample for numerical stability.
clip_sample_range (float, defaults to 1.0) — The maximum magnitude for sample clipping. Valid only when clip_sample=True.
set_alpha_to_one (bool, defaults to True) — Each diffusion step uses the alphas product value at that step and at the previous one. For the final step there is no previous alpha. When this option is True the previous alpha product is fixed to 1, otherwise it uses the alpha value at step 0.
steps_offset (int, defaults to 0) — An offset added to the inference steps. You can use a combination of offset=1 and set_alpha_to_one=False to make the last step use step 0 for the previous alpha product like in Stable Diffusion.
thresholding (bool, defaults to False) — Whether to use the “dynamic thresholding” method. This is unsuitable for latent-space diffusion models such as Stable Diffusion.
dynamic_thresholding_ratio (float, defaults to 0.995) — The ratio for the dynamic thresholding method. Valid only when thresholding=True.
sample_max_value (float, defaults to 1.0) — The threshold value for dynamic thresholding. Valid only when thresholding=True.
DDIMScheduler extends the denoising procedure introduced in denoising diffusion probabilistic models (DDPMs) with non-Markovian guidance.
scale_model_input
( sample: FloatTensortimestep: typing.Optional[int] = None ) → torch.FloatTensor
Parameters
sample (torch.FloatTensor) — The input sample.
timestep (int, optional) — The current timestep in the diffusion chain.
Returns
torch.FloatTensor
A scaled input sample.
Ensures interchangeability with schedulers that need to scale the denoising model input depending on the current timestep.
set_timesteps
( num_inference_steps: intdevice: typing.Union[str, torch.device] = None )
Parameters
num_inference_steps (int) — The number of diffusion steps used when generating samples with a pre-trained model.
Sets the discrete timesteps used for the diffusion chain (to be run before inference).
step
( model_output: FloatTensortimestep: intsample: FloatTensoreta: float = 0.0use_clipped_model_output: bool = Falsegenerator = Nonevariance_noise: typing.Optional[torch.FloatTensor] = Nonereturn_dict: bool = True ) → ~schedulers.scheduling_utils.DDIMSchedulerOutput or tuple
Parameters
model_output (torch.FloatTensor) — The direct output from learned diffusion model.
timestep (float) — The current discrete timestep in the diffusion chain.
sample (torch.FloatTensor) — A current instance of a sample created by the diffusion process.
eta (float) — The weight of noise for added noise in diffusion step.
use_clipped_model_output (bool, defaults to False) — If True, computes “corrected” model_output from the clipped predicted original sample. Necessary because predicted original sample is clipped to [-1, 1] when self.config.clip_sample is True. If no clipping has happened, “corrected” model_output would coincide with the one provided as input and use_clipped_model_output has no effect.
generator (torch.Generator, optional) — A random number generator.
variance_noise (torch.FloatTensor) — Alternative to generating noise with generator by directly providing the noise for the variance itself. Useful for methods such as CycleDiffusion.
Returns
~schedulers.scheduling_utils.DDIMSchedulerOutput or tuple
Predict the sample from the previous timestep by reversing the SDE. This function propagates the diffusion process from the learned model outputs (most often the predicted noise).
( prev_sample: FloatTensorpred_original_sample: typing.Optional[torch.FloatTensor] = None )
Parameters
prev_sample (torch.FloatTensor of shape (batch_size, num_channels, height, width) for images) — Computed sample (x_{t-1}) of previous timestep. prev_sample should be used as next model input in the denoising loop.
pred_original_sample (torch.FloatTensor of shape (batch_size, num_channels, height, width) for images) — The predicted denoised sample (x_{0}) based on the model output from the current timestep. pred_original_sample can be used to preview progress or for guidance.
Output class for the scheduler’s step function output.
prediction_type (str, defaults to epsilon, optional) — Prediction type of the scheduler function; can be epsilon (predicts the noise of the diffusion process), sample (directly predicts the noisy sample) or v_prediction` (see section 2.4 of paper).
timestep_spacing (str, defaults to "leading") — The way the timesteps should be scaled. Refer to Table 2 of the for more information.
rescale_betas_zero_snr (bool, defaults to False) — Whether to rescale the betas to have zero terminal SNR. This enables the model to generate very bright and dark samples instead of limiting it to samples with medium brightness. Loosely related to .
This model inherits from and . Check the superclass documentation for the generic methods the library implements for all schedulers such as loading and saving.
return_dict (bool, optional, defaults to True) — Whether or not to return a or tuple.
If return_dict is True, is returned, otherwise a tuple is returned where the first element is the sample tensor.