LLMRails

Let’s load the LLMRails Embeddings class.

To use LLMRails embedding you need to pass api key by argument or set it in environment with LLM_RAILS_API_KEY key. To gey API Key you need to sign up in https://console.llmrails.com/signup and then go to https://console.llmrails.com/api-keys and copy key from there after creating one key in platform.

from langchain.embeddings import LLMRailsEmbeddings

embeddings = LLMRailsEmbeddings(model="embedding-english-v1")  # or embedding-multi-v1

text = "This is a test document."

To generate embeddings, you can either query an invidivual text, or you can query a list of texts.

query_result = embeddings.embed_query(text)
query_result[:5]

[-0.09996652603149414,
 0.015568195842206478,
 0.17670190334320068,
 0.16521021723747253,
 0.21193109452724457]

doc_result = embeddings.embed_documents([text])
doc_result[0][:5]

[-0.04242777079343796,
 0.016536075621843338,
 0.10052520781755447,
 0.18272875249385834,
 0.2079043835401535]