VQModel (Vector Quantized Model)

The VQModel (Vector Quantized Model) is often used in diffusion models like Kandinsky V2.2. It functions similarly to a Variational Autoencoder (VAE) but uses vector quantization instead of a continuous latent space. It's responsible for encoding images into discrete latent representations and decoding those representations back into images. RBLN NPUs can accelerate VQModel inference using Optimum RBLN.

Key Classes

• RBLNVQModel: The main model class for running VQModel on RBLN NPU
• RBLNVQModelConfig: Configuration class for VQModel models

API Reference

Classes

RBLNVQModel

Bases: RBLNModel

RBLN implementation of VQModel used in Kandinsky models.

This model accelerates the VQModel (Vector Quantized Model), which serves a similar purpose to the VAE in other diffusion models (encoding images to latents and decoding latents to images), but utilizes vector quantization.

It can be configured to include both encoder and decoder, or just the decoder part.

This class inherits from [RBLNModel].

Functions

encode(x, return_dict=True, **kwargs)

Encode an input image into a quantized latent representation.

Parameters:

Name	Type	Description	Default
x	FloatTensor	The input image to encode.	required
return_dict	bool	Whether to return output as a dictionary. Defaults to True.	True
**kwargs	Dict[str, Any]	Additional arguments to pass to the encoder/quantizer.	{}

Returns:

Type	Description
Union[FloatTensor, VQEncoderOutput]	Union[FloatTensor, object]: The quantized latent representation or a specific output object.

decode(z, return_dict=True, **kwargs)

Decode a quantized latent representation back into an image.

Parameters:

Name	Type	Description	Default
z	FloatTensor	The quantized latent representation to decode.	required
return_dict	bool	Whether to return output as a dictionary. Defaults to True.	True
**kwargs	Dict[str, Any]	Additional arguments to pass to the decoder.	{}

Returns:

Type	Description
Union[FloatTensor, DecoderOutput]	Union[FloatTensor, DecoderOutput]: The decoded image or a DecoderOutput object.

from_pretrained(model_id, export=False, rbln_config=None, **kwargs) classmethod

The from_pretrained() function is utilized in its standard form as in the HuggingFace transformers library. User can use this function to load a pre-trained model from the HuggingFace library and convert it to a RBLN model to be run on RBLN NPUs.

Parameters:

Name	Type	Description	Default
model_id	Union[str, Path]	The model id of the pre-trained model to be loaded. It can be downloaded from the HuggingFace model hub or a local path, or a model id of a compiled model using the RBLN Compiler.	required
export	bool	A boolean flag to indicate whether the model should be compiled.	False
rbln_config	Optional[Union[Dict, RBLNModelConfig]]	Configuration for RBLN model compilation and runtime. This can be provided as a dictionary or an instance of the model's configuration class (e.g., RBLNLlamaForCausalLMConfig for Llama models). For detailed configuration options, see the specific model's configuration class documentation.	None
kwargs	Dict[str, Any]	Additional keyword arguments. Arguments with the prefix 'rbln_' are passed to rbln_config, while the remaining arguments are passed to the HuggingFace library.	{}

Returns:

Type	Description
Self	A RBLN model instance ready for inference on RBLN NPU devices.

from_model(model, *, rbln_config=None, **kwargs) classmethod

Converts and compiles a pre-trained HuggingFace library model into a RBLN model. This method performs the actual model conversion and compilation process.

Parameters:

Name	Type	Description	Default
model	PreTrainedModel	The PyTorch model to be compiled. The object must be an instance of the HuggingFace transformers PreTrainedModel class.	required
rbln_config	Optional[Union[Dict, RBLNModelConfig]]	Configuration for RBLN model compilation and runtime. This can be provided as a dictionary or an instance of the model's configuration class (e.g., RBLNLlamaForCausalLMConfig for Llama models). For detailed configuration options, see the specific model's configuration class documentation.	None
kwargs	Dict[str, Any]	Additional keyword arguments. Arguments with the prefix 'rbln_' are passed to rbln_config, while the remaining arguments are passed to the HuggingFace library.	{}

The method performs the following steps:

1. Compiles the PyTorch model into an optimized RBLN graph
2. Configures the model for the specified NPU device
3. Creates the necessary runtime objects if requested
4. Saves the compiled model and configurations

Returns:

Type	Description
Self	A RBLN model instance ready for inference on RBLN NPU devices.

save_pretrained(save_directory)

Saves a model and its configuration file to a directory, so that it can be re-loaded using the [from_pretrained] class method.

Parameters:

Name	Type	Description	Default
save_directory	Union[str, PathLike]	The directory to save the model and its configuration files. Will be created if it doesn't exist.	required

Classes

RBLNVQModelConfig

Bases: RBLNModelConfig

Configuration class for RBLN VQModel models, used in Kandinsky.

This class inherits from RBLNModelConfig and provides specific configuration options for VQModel, which acts similarly to a VAE but uses vector quantization.

Functions

__init__(batch_size=None, sample_size=None, uses_encoder=None, scaling_factor=None, in_channels=None, latent_channels=None, num_vq_embeddings=None, **kwargs)

Parameters:

Name	Type	Description	Default
batch_size	Optional[int]	The batch size for inference. Defaults to 1.	None
sample_size	Optional[Tuple[int, int]]	The spatial dimensions (height, width) of the input/output images. If an integer is provided, it's used for both height and width.	None
uses_encoder	Optional[bool]	Whether to include the encoder part of the VQModel in the model. When False, only the decoder is used (for latent-to-image conversion).	None
scaling_factor	Optional[int]	The integer downsampling factor between pixel space and latent space.	None
in_channels	Optional[int]	Number of input channels for the model.	None
latent_channels	Optional[int]	Number of channels in the latent space.	None
num_vq_embeddings	Optional[int]	Number of embeddings in the VQ codebook.	None
**kwargs	Dict[str, Any]	Additional arguments passed to the parent RBLNModelConfig.	{}