Embedding Models

info

This page may contain outdated information. It will be updated as soon as possible.

Amazon Bedrock Embeddings

Used to load embedding models from Amazon Bedrock.

Parameter	Type	Description
`credentials_profile_name`	`str`	Name of the AWS credentials profile in ~/.aws/credentials or ~/.aws/config, which has access keys or role information.
`model_id`	`str`	ID of the model to call, e.g., `amazon.titan-embed-text-v1`. This is equivalent to the `modelId` property in the `list-foundation-models` API.
`endpoint_url`	`str`	URL to set a specific service endpoint other than the default AWS endpoint.
`region_name`	`str`	AWS region to use, e.g., `us-west-2`. Falls back to `AWS_DEFAULT_REGION` environment variable or region specified in ~/.aws/config if not provided.

Used to generate server-side embeddings using DataStax Astra.

Parameter	Type	Description
`provider`	`str`	The embedding provider to use.
`model_name`	`str`	The embedding model to use.
`authentication`	`dict`	Authentication parameters. Use the Astra Portal to add the embedding provider integration to your Astra organization.
`provider_api_key`	`str`	An alternative to the Astra Authentication that let you use directly the API key of the provider.
`model_parameters`	`dict`	Additional model parameters.

Used to load embedding models from Cohere.

Parameter	Type	Description	Default
`cohere_api_key`	`str`	API key required to authenticate with the Cohere service.
`model`	`str`	Language model used for embedding text documents and performing queries.	`embed-english-v2.0`
`truncate`	`bool`	Whether to truncate the input text to fit within the model's constraints.	`False`

Generate embeddings using Azure OpenAI models.

Parameter	Type	Description
`Azure Endpoint`	`str`	Your Azure endpoint, including the resource. Example: `https://example-resource.azure.openai.com/`
`Deployment Name`	`str`	The name of the deployment.
`API Version`	`str`	The API version to use, options include various dates.
`API Key`	`str`	The API key to access the Azure OpenAI service.

Generate embeddings using Hugging Face Inference API models.

Parameter	Type	Description	Default
`API Key`	`str`	API key for accessing the Hugging Face Inference API.
`API URL`	`str`	URL of the Hugging Face Inference API.	`http://localhost:8080`
`Model Name`	`str`	Name of the model to use for embeddings.	`BAAI/bge-large-en-v1.5`
`Cache Folder`	`str`	Folder path to cache Hugging Face models.
`Encode Kwargs`	`dict`	Additional arguments for the encoding process.
`Model Kwargs`	`dict`	Additional arguments for the model.
`Multi Process`	`bool`	Whether to use multiple processes.	`False`

Used to load embedding models from HuggingFace.

Parameter	Type	Description	Default
`Cache Folder`	`str`	Folder path to cache HuggingFace models.
`Encode Kwargs`	`dict`	Additional arguments for the encoding process.
`Model Kwargs`	`dict`	Additional arguments for the model.
`Model Name`	`str`	Name of the HuggingFace model to use.	`sentence-transformers/all-mpnet-base-v2`
`Multi Process`	`bool`	Whether to use multiple processes.	`False`

Used to load embedding models from OpenAI.

Parameter	Type	Description	Default
`OpenAI API Key`	`str`	The API key to use for accessing the OpenAI API.
`Default Headers`	`Dict[str, str]`	Default headers for the HTTP requests.
`Default Query`	`NestedDict`	Default query parameters for the HTTP requests.
`Allowed Special`	`List[str]`	Special tokens allowed for processing.	`[]`
`Disallowed Special`	`List[str]`	Special tokens disallowed for processing.	`["all"]`
`Chunk Size`	`int`	Chunk size for processing.	`1000`
`Client`	`Any`	HTTP client for making requests.
`Deployment`	`str`	Deployment name for the model.	`text-embedding-3-small`
`Embedding Context Length`	`int`	Length of embedding context.	`8191`
`Max Retries`	`int`	Maximum number of retries for failed requests.	`6`
`Model`	`str`	Name of the model to use.	`text-embedding-3-small`
`Model Kwargs`	`NestedDict`	Additional keyword arguments for the model.
`OpenAI API Base`	`str`	Base URL of the OpenAI API.
`OpenAI API Type`	`str`	Type of the OpenAI API.
`OpenAI API Version`	`str`	Version of the OpenAI API.
`OpenAI Organization`	`str`	Organization associated with the API key.
`OpenAI Proxy`	`str`	Proxy server for the requests.
`Request Timeout`	`float`	Timeout for the HTTP requests.
`Show Progress Bar`	`bool`	Whether to show a progress bar for processing.	`False`
`Skip Empty`	`bool`	Whether to skip empty inputs.	`False`
`TikToken Enable`	`bool`	Whether to enable TikToken.	`True`
`TikToken Model Name`	`str`	Name of the TikToken model.

Generate embeddings using Ollama models.

Parameter	Type	Description	Default
`Ollama Model`	`str`	Name of the Ollama model to use.	`llama2`
`Ollama Base URL`	`str`	Base URL of the Ollama API.	`http://localhost:11434`
`Model Temperature`	`float`	Temperature parameter for the model. Adjusts the randomness in the generated embeddings.

Parameter	Type	Description	Default
`credentials`	`Credentials`	The default custom credentials to use.
`location`	`str`	The default location to use when making API calls.	`us-central1`
`max_output_tokens`	`int`	Token limit determines the maximum amount of text output from one prompt.	`128`
`model_name`	`str`	The name of the Vertex AI large language model.	`text-bison`
`project`	`str`	The default GCP project to use when making Vertex API calls.
`request_parallelism`	`int`	The amount of parallelism allowed for requests issued to VertexAI models.	`5`
`temperature`	`float`	Tunes the degree of randomness in text generations. Should be a non-negative value.	`0`
`top_k`	`int`	How the model selects tokens for output, the next token is selected from the top `k` tokens.	`40`
`top_p`	`float`	Tokens are selected from the most probable to least until the sum of their probabilities exceeds the top `p` value.	`0.95`
`tuned_model_name`	`str`	The name of a tuned model. If provided, `model_name` is ignored.
`verbose`	`bool`	This parameter controls the level of detail in the output. When set to `True`, it prints internal states of the chain to help debug.	`False`