haive.core.engine.embedding.providers.AzureOpenAIEmbeddingConfig¶
Azure OpenAI embedding configuration.
Classes¶
Configuration for Azure OpenAI embeddings. |
Module Contents¶
- class haive.core.engine.embedding.providers.AzureOpenAIEmbeddingConfig.AzureOpenAIEmbeddingConfig[source]¶
Bases:
haive.core.engine.embedding.base.BaseEmbeddingConfigConfiguration for Azure OpenAI embeddings.
This configuration provides access to OpenAI embedding models deployed on Azure OpenAI Service. It supports both standard and data zone deployments.
Examples
Basic usage:
config = AzureOpenAIEmbeddingConfig( name="azure_embeddings", model="text-embedding-3-large", deployment_name="text-embedding-3-large", azure_endpoint="https://your-resource.openai.azure.com/", api_key="your-api-key" ) embeddings = config.instantiate()
Using environment variables:
# Set AZURE_OPENAI_API_KEY, AZURE_OPENAI_ENDPOINT, etc. config = AzureOpenAIEmbeddingConfig( name="azure_embeddings", model="text-embedding-3-large", deployment_name="text-embedding-3-large" )
With custom API version:
config = AzureOpenAIEmbeddingConfig( name="azure_embeddings", model="text-embedding-3-large", deployment_name="text-embedding-3-large", api_version="2024-02-15-preview" )
- embedding_type¶
Always EmbeddingType.AZURE_OPENAI
- deployment_name¶
Azure deployment name for the model
- azure_endpoint¶
Azure OpenAI service endpoint URL
- api_version¶
Azure OpenAI API version
- api_key¶
Azure OpenAI API key
- dimensions¶
Output dimensions (optional, model-dependent)
- instantiate()[source]¶
Create an Azure OpenAI embeddings instance.
- Returns:
AzureOpenAIEmbeddings instance configured with the provided parameters
- Raises:
ImportError – If langchain-openai is not installed
ValueError – If configuration is invalid
- Return type:
Any
- classmethod validate_azure_endpoint(v)[source]¶
Validate Azure OpenAI endpoint format.
- Return type:
Any