Nebius 嵌入¶

本笔记本演示了如何将 Nebius AI Studio 嵌入与 LlamaIndex 结合使用。Nebius AI Studio 实现了所有最先进的嵌入模型，可供商业使用。

首先，让我们安装 LlamaIndex 和 Nebius AI Studio 的依赖项。

输入 [ ]

%pip install llama-index-embeddings-nebius llama-index

%pip install llama-index-embeddings-nebius llama-index

从下面的系统变量上传您的 Nebius AI Studio 密钥，或者直接插入。您可以通过在 Nebius AI Studio 免费注册并在 API 密钥部分颁发密钥来获取。

输入 [ ]

import os

NEBIUS_API_KEY = os.getenv("NEBIUS_API_KEY")  # NEBIUS_API_KEY = ""

import os NEBIUS_API_KEY = os.getenv("NEBIUS_API_KEY") # NEBIUS_API_KEY = ""

现在让我们使用 Nebius AI Studio 获取嵌入

输入 [ ]

from llama_index.embeddings.nebius import NebiusEmbedding

embed_model = NebiusEmbedding(api_key=NEBIUS_API_KEY)

from llama_index.embeddings.nebius import NebiusEmbedding embed_model = NebiusEmbedding(api_key=NEBIUS_API_KEY)

基本用法¶

输入 [ ]

text = "Everyone loves justice at another person's expense"
embeddings = embed_model.get_text_embedding(text)
assert len(embeddings) == 4096
print(len(embeddings), embeddings[:5], sep="\n")

text = "Everyone loves justice at another person's expense" embeddings = embed_model.get_text_embedding(text) assert len(embeddings) == 4096 print(len(embeddings), embeddings[:5], sep="\n")

4096
[-0.0024051666259765625, 0.0083770751953125, -0.005413055419921875, 0.007396697998046875, -0.022247314453125]

异步用法¶

输入 [ ]

text = "Everyone loves justice at another person's expense"
embeddings = await embed_model.aget_text_embedding(text)
assert len(embeddings) == 4096
print(len(embeddings), embeddings[:5], sep="\n")

text = "Everyone loves justice at another person's expense" embeddings = await embed_model.aget_text_embedding(text) assert len(embeddings) == 4096 print(len(embeddings), embeddings[:5], sep="\n")

4096
[-0.0024051666259765625, 0.0083770751953125, -0.005413055419921875, 0.007396697998046875, -0.022247314453125]

批量用法¶

输入 [ ]

texts = [
    "As the hours pass",
    "I will let you know",
    "That I need to ask",
    "Before I'm alone",
]

embeddings = embed_model.get_text_embedding_batch(texts)
assert len(embeddings) == 4
assert len(embeddings[0]) == 4096
print(*[x[:3] for x in embeddings], sep="\n")

texts = [ "As the hours pass", "I will let you know", "That I need to ask", "Before I'm alone", ] embeddings = embed_model.get_text_embedding_batch(texts) assert len(embeddings) == 4 assert len(embeddings[0]) == 4096 print(*[x[:3] for x in embeddings], sep="\n")

[-0.0003848075866699219, 0.0004799365997314453, 0.011199951171875]
[-0.0037078857421875, 0.0114288330078125, 0.00878143310546875]
[0.005924224853515625, 0.005153656005859375, 0.001438140869140625]
[-0.009490966796875, -0.004852294921875, 0.004779815673828125]

异步批量用法¶

输入 [ ]

texts = [
    "As the hours pass",
    "I will let you know",
    "That I need to ask",
    "Before I'm alone",
]

embeddings = await embed_model.aget_text_embedding_batch(texts)
assert len(embeddings) == 4
assert len(embeddings[0]) == 4096
print(*[x[:3] for x in embeddings], sep="\n")

texts = [ "As the hours pass", "I will let you know", "That I need to ask", "Before I'm alone", ] embeddings = await embed_model.aget_text_embedding_batch(texts) assert len(embeddings) == 4 assert len(embeddings[0]) == 4096 print(*[x[:3] for x in embeddings], sep="\n")

[-0.0003848075866699219, 0.0004799365997314453, 0.011199951171875]
[-0.0037078857421875, 0.0114288330078125, 0.00878143310546875]
[0.005924224853515625, 0.005153656005859375, 0.001438140869140625]
[-0.009490966796875, -0.004852294921875, 0.004779815673828125]