NomicEmbeddings

本笔记本介绍如何开始使用 Nomic 嵌入模型。

安装

# install package
!pip install -U langchain-nomic

确保设置以下环境变量：

from langchain_nomic.embeddings import NomicEmbeddings

embeddings = NomicEmbeddings(model="nomic-embed-text-v1.5")

embeddings.embed_query("My query to look up")

embeddings.embed_documents(
    ["This is a content of the document", "This is another document"]
)

# 异步嵌入查询
await embeddings.aembed_query("My query to look up")

# 异步嵌入文档
await embeddings.aembed_documents(
    ["This is a content of the document", "This is another document"]
)

Nomic的 nomic-embed-text-v1.5 模型经过 Matryoshka 学习训练，能够使用单一模型实现可变长度的嵌入。这意味着您可以在推理时指定嵌入的维度。该模型支持的维度范围为 64 到 768。

embeddings = NomicEmbeddings(model="nomic-embed-text-v1.5", dimensionality=256)

embeddings.embed_query("My query to look up")