FastEmbedProvider

​

Parameters

Parameter	Type	Default	Description
model_name	str	"BAAI/bge-small-en-v1.5"	FastEmbed model name
cache_dir	Optional[str]	None	Model cache directory
threads	Optional[int]	None	Number of threads (auto-detected if None)
providers	List[str]	["CPUExecutionProvider"]	ONNX execution providers
enable_gpu	bool	False	Enable GPU acceleration if available
enable_parallel_processing	bool	True	Enable parallel text processing
doc_embed_type	str	"default"	Document embedding type (default, passage)
max_memory_mb	Optional[int]	None	Maximum memory usage in MB
model_warmup	bool	True	Warm up model on initialization
enable_sparse_embeddings	bool	False	Use sparse embeddings for better performance
sparse_model_name	Optional[str]	None	Sparse model name if different from dense

​

Functions

​

__init__

Initialize the FastEmbedProvider. Parameters:

• config (Optional[FastEmbedConfig]): Configuration object
• **kwargs: Additional configuration options

​

_setup_providers

Setup ONNX execution providers based on configuration.

​

_initialize_models

Initialize FastEmbed models.

​

_warmup_models

Warm up models with sample data.

​

supported_modes

Get supported embedding modes. Returns:

• List[EmbeddingMode]: List of supported embedding modes

​

pricing_info

Get FastEmbed pricing info (local execution is free). Returns:

• Dict[str, float]: Pricing information

​

get_model_info

Get information about the current FastEmbed model. Returns:

• Dict[str, Any]: Model information

​

_process_embeddings

Process embeddings iterator into list format. Parameters:

• embeddings_iterator (Iterator[np.ndarray]): Embeddings iterator

Returns:

• List[List[float]]: Processed embeddings

​

_embed_batch

Embed a batch of texts using FastEmbed. Parameters:

• texts (List[str]): List of text strings to embed
• mode (EmbeddingMode): Embedding mode (affects processing strategy)

Returns:

• List[List[float]]: List of embedding vectors

​

_normalize_embeddings

Normalize embeddings to unit length. Parameters:

• embeddings (List[List[float]]): List of embedding vectors

Returns:

• List[List[float]]: Normalized embeddings

​

validate_connection

Validate FastEmbed model is working. Returns:

• bool: True if model is working

​

get_performance_info

Get performance and resource usage information. Returns:

• Dict[str, Any]: Performance information

​

list_available_models

List available FastEmbed models. Returns:

• List[Dict[str, Any]]: List of available models

​

get_cache_info

Get information about model caching. Returns:

• Dict[str, Any]: Cache information

​

close

Clean up FastEmbed models and clear memory.

​

create_bge_small_embedding

Create BGE-small FastEmbed provider (fast and efficient). Parameters:

• **kwargs: Additional configuration options

Returns:

• FastEmbedProvider: Configured FastEmbedProvider instance

​

create_bge_large_embedding

Create BGE-large FastEmbed provider (high quality). Parameters:

• **kwargs: Additional configuration options

Returns:

• FastEmbedProvider: Configured FastEmbedProvider instance

​

create_e5_embedding

Create E5 FastEmbed provider (multilingual). Parameters:

• **kwargs: Additional configuration options

Returns:

• FastEmbedProvider: Configured FastEmbedProvider instance

​

create_sparse_embedding

Create sparse embedding provider for efficiency. Parameters:

• **kwargs: Additional configuration options

Returns:

• FastEmbedProvider: Configured FastEmbedProvider instance

​

create_gpu_accelerated_embedding

Create GPU-accelerated FastEmbed provider. Parameters:

• model_name (str): Model name to use
• **kwargs: Additional configuration options

Returns:

• FastEmbedProvider: Configured FastEmbedProvider instance