This matches the implementation in Denoising Diffusion Probabilistic Models: From Fairseq. Build sinusoidal embeddings. This matches

Get the information of a token, including its start and end index in the current tokenizer. Args: token (str): The token

Compute the continuous-time label t in [0, T] of a given half-logSNR lambda_t.

Load and preprocess a base64 image, resize if necessary.

Create a wrapper function for the noise prediction model. DPM-Solver needs to solve the continuous-time diffusion ODEs. For DPMs trained on discr

Replace external embedding to the embedding layer. Noted that, in this function we use `torch.cat` to avoid inplace modification. Arg

Replace external input ids to 0. Args: input_ids (torch.Tensor): The input ids to be replaced. Returns: torc

Replace the placeholder tokens in text with the original keywords. This function will be called in `self.decode`. Args: t

Restore the parameters stored with the `store` method. Useful to validate the model with EMA parameters without affecting the

DPM-Solver++(2M) SDE.

DPM-Solver++(3M) SDE.

DPM-Solver++ (stochastic).

r""" Enable sliced attention computation. When this option is enabled, the attention module splits the input tensor in slices to comp

r""" Enable sliced attention computation. When this option is enabled, the attention module splits the input tensor in slices to comp

r""" Sets the attention processor to use to compute attention. Parameters: processor (`dict` of `AttentionProcessor` or o

r""" Sets the attention processor to use to compute attention. Parameters: processor (`dict` of `AttentionProcessor` or o