FireworksEmbeddings

This notebook explains how to use Fireworks Embeddings, which is included in the langchain_fireworks package, to embed texts in langchain. We use the default nomic-ai v1.5 model in this example.

%pip install -qU langchain-fireworks

Setup

from langchain_fireworks import FireworksEmbeddings

API Reference:FireworksEmbeddings

import getpass
import os

if "FIREWORKS_API_KEY" not in os.environ:
    os.environ["FIREWORKS_API_KEY"] = getpass.getpass("Fireworks API Key:")

Using the Embedding Model

With FireworksEmbeddings, you can directly use the default model 'nomic-ai/nomic-embed-text-v1.5', or set a different one if available.

embedding = FireworksEmbeddings(model="nomic-ai/nomic-embed-text-v1.5")

res_query = embedding.embed_query("The test information")
res_document = embedding.embed_documents(["test1", "another test"])
print(res_query[:5])
print(res_document[1][:5])

[0.01367950439453125, 0.0103607177734375, -0.157958984375, -0.003070831298828125, 0.05926513671875]
[0.0369873046875, 0.00545501708984375, -0.179931640625, -0.018707275390625, 0.0552978515625]