DeepInfra

DeepInfra 是一個 serverless 推論即服務，提供對 各種 LLM 和 嵌入模型的存取。 本筆記本介紹如何將 LangChain 與 DeepInfra 一起使用以進行文字嵌入。

# sign up for an account: https://deepinfra.com/login?utm_source=langchain

from getpass import getpass

DEEPINFRA_API_TOKEN = getpass()

 ········

import os

os.environ["DEEPINFRA_API_TOKEN"] = DEEPINFRA_API_TOKEN

from langchain_community.embeddings import DeepInfraEmbeddings

API 參考：DeepInfraEmbeddings

embeddings = DeepInfraEmbeddings(
    model_id="sentence-transformers/clip-ViT-B-32",
    query_instruction="",
    embed_instruction="",
)

docs = ["Dog is not a cat", "Beta is the second letter of Greek alphabet"]
document_result = embeddings.embed_documents(docs)

query = "What is the first letter of Greek alphabet"
query_result = embeddings.embed_query(query)

import numpy as np

query_numpy = np.array(query_result)
for doc_res, doc in zip(document_result, docs):
    document_numpy = np.array(doc_res)
    similarity = np.dot(query_numpy, document_numpy) / (
        np.linalg.norm(query_numpy) * np.linalg.norm(document_numpy)
    )
    print(f'Cosine similarity between "{doc}" and query: {similarity}')

Cosine similarity between "Dog is not a cat" and query: 0.7489097144129355
Cosine similarity between "Beta is the second letter of Greek alphabet" and query: 0.9519380640702013

相關

• 嵌入模型 概念指南
• 嵌入模型 操作指南