高级提示词技巧（变量映射与函数）¶

本笔记本将展示一些高级提示词技术。这些功能允许您定义更具定制性/表现力的提示词、复用现有提示词，同时还能用更少的代码行数实现特定操作。

我们将演示以下功能：

1. 部分格式化
2. 提示模板变量映射
3. 提示函数映射
4. 动态少样本示例

In [ ]:

%pip install llama-index-llms-openai

%pip install llama-index-llms-openai

1. 部分格式化¶

部分格式化 (partial_format) 功能允许您对提示进行部分格式化，填充部分变量同时保留其他变量供后续填充。

这是一个便捷的功能，您无需在传递过程中一直维护所有必需的提示变量直到执行 format 操作，可以在变量可用时进行部分格式化。

此操作将创建提示模板的副本。

In [ ]:

from llama_index.core.prompts import RichPromptTemplate

qa_prompt_tmpl_str = """\
Context information is below.
---------------------
{{ context_str }}
---------------------
Given the context information and not prior knowledge, answer the query.
Please write the answer in the style of {{ tone_name }}
Query: {{ query_str }}
Answer: \
"""

prompt_tmpl = RichPromptTemplate(qa_prompt_tmpl_str)

from llama_index.core.prompts import RichPromptTemplate qa_prompt_tmpl_str = """\ Context information is below. --------------------- {{ context_str }} --------------------- Given the context information and not prior knowledge, answer the query. Please write the answer in the style of {{ tone_name }} Query: {{ query_str }} Answer: \ """ prompt_tmpl = RichPromptTemplate(qa_prompt_tmpl_str)

In [ ]:

partial_prompt_tmpl = prompt_tmpl.partial_format(tone_name="Shakespeare")

partial_prompt_tmpl = prompt_tmpl.partial_format(tone_name="Shakespeare")

In [ ]:

partial_prompt_tmpl.kwargs

partial_prompt_tmpl.kwargs

Out[ ]:

{'tone_name': 'Shakespeare'}

In [ ]:

fmt_prompt = partial_prompt_tmpl.format(
    context_str="In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters",
    query_str="How many params does llama 2 have",
)
print(fmt_prompt)

fmt_prompt = partial_prompt_tmpl.format( context_str="In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters", query_str="How many params does llama 2 have", ) print(fmt_prompt)

Context information is below.
---------------------
In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters
---------------------
Given the context information and not prior knowledge, answer the query.
Please write the answer in the style of Shakespeare
Query: How many params does llama 2 have
Answer: 

我们也可以使用 format_messages 将提示格式化为 ChatMessage 对象。

In [ ]:

fmt_prompt = partial_prompt_tmpl.format_messages(
    context_str="In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters",
    query_str="How many params does llama 2 have",
)
print(fmt_prompt)

fmt_prompt = partial_prompt_tmpl.format_messages( context_str="In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters", query_str="How many params does llama 2 have", ) print(fmt_prompt)

[ChatMessage(role=<MessageRole.USER: 'user'>, additional_kwargs={}, blocks=[TextBlock(block_type='text', text='Context information is below.'), TextBlock(block_type='text', text='---------------------'), TextBlock(block_type='text', text='In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters'), TextBlock(block_type='text', text='---------------------'), TextBlock(block_type='text', text='Given the context information and not prior knowledge, answer the query.'), TextBlock(block_type='text', text='Please write the answer in the style of Shakespeare'), TextBlock(block_type='text', text='Query: How many params does llama 2 have'), TextBlock(block_type='text', text='Answer:')])]

2. 提示模板变量映射¶

模板变量映射功能允许您指定从"预期"提示键（例如响应合成所需的 context_str 和 query_str）到实际模板中键名的映射关系。

通过此功能，您可以复用现有的字符串模板，而无需繁琐地修改模板变量。

In [ ]:

from llama_index.core.prompts import RichPromptTemplate

# NOTE: here notice we use `my_context` and `my_query` as template variables
qa_prompt_tmpl_str = """\
Context information is below.
---------------------
{{ my_context }}
---------------------
Given the context information and not prior knowledge, answer the query.
Query: {{ my_query }}
Answer: \
"""

template_var_mappings = {"context_str": "my_context", "query_str": "my_query"}

prompt_tmpl = RichPromptTemplate(
    qa_prompt_tmpl_str, template_var_mappings=template_var_mappings
)

from llama_index.core.prompts import RichPromptTemplate # NOTE: here notice we use `my_context` and `my_query` as template variables qa_prompt_tmpl_str = """\ Context information is below. --------------------- {{ my_context }} --------------------- Given the context information and not prior knowledge, answer the query. Query: {{ my_query }} Answer: \ """ template_var_mappings = {"context_str": "my_context", "query_str": "my_query"} prompt_tmpl = RichPromptTemplate( qa_prompt_tmpl_str, template_var_mappings=template_var_mappings )

In [ ]:

fmt_prompt = prompt_tmpl.format(
    context_str="In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters",
    query_str="How many params does llama 2 have",
)
print(fmt_prompt)

fmt_prompt = prompt_tmpl.format( context_str="In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters", query_str="How many params does llama 2 have", ) print(fmt_prompt)

Context information is below.
---------------------
In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters
---------------------
Given the context information and not prior knowledge, answer the query.
Query: How many params does llama 2 have
Answer: 

3. 提示词函数映射¶

您还可以将函数作为模板变量传入，而非固定值。

这使您能够在查询时动态注入某些依赖于其他值的变量。

以下是一些基础示例。我们会在《RAG提示工程指南》中展示更多高级示例（例如小样本示例）。

In [ ]:

from llama_index.core.prompts import RichPromptTemplate

qa_prompt_tmpl_str = """\
Context information is below.
---------------------
{{ context_str }}
---------------------
Given the context information and not prior knowledge, answer the query.
Query: {{ query_str }}
Answer: \
"""


def format_context_fn(**kwargs):
    # format context with bullet points
    context_list = kwargs["context_str"].split("\n\n")
    fmtted_context = "\n\n".join([f"- {c}" for c in context_list])
    return fmtted_context


prompt_tmpl = RichPromptTemplate(
    qa_prompt_tmpl_str, function_mappings={"context_str": format_context_fn}
)

from llama_index.core.prompts import RichPromptTemplate qa_prompt_tmpl_str = """\ Context information is below. --------------------- {{ context_str }} --------------------- Given the context information and not prior knowledge, answer the query. Query: {{ query_str }} Answer: \ """ def format_context_fn(**kwargs): # format context with bullet points context_list = kwargs["context_str"].split("\n\n") fmtted_context = "\n\n".join([f"- {c}" for c in context_list]) return fmtted_context prompt_tmpl = RichPromptTemplate( qa_prompt_tmpl_str, function_mappings={"context_str": format_context_fn} )

In [ ]:

context_str = """\
In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters.

Our fine-tuned LLMs, called Llama 2-Chat, are optimized for dialogue use cases.

Our models outperform open-source chat models on most benchmarks we tested, and based on our human evaluations for helpfulness and safety, may be a suitable substitute for closed-source models.
"""

fmt_prompt = prompt_tmpl.format(
    context_str=context_str, query_str="How many params does llama 2 have"
)
print(fmt_prompt)

context_str = """\ In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters. Our fine-tuned LLMs, called Llama 2-Chat, are optimized for dialogue use cases. Our models outperform open-source chat models on most benchmarks we tested, and based on our human evaluations for helpfulness and safety, may be a suitable substitute for closed-source models. """ fmt_prompt = prompt_tmpl.format( context_str=context_str, query_str="How many params does llama 2 have" ) print(fmt_prompt)

Context information is below.
---------------------
- In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters.

- Our fine-tuned LLMs, called Llama 2-Chat, are optimized for dialogue use cases.

- Our models outperform open-source chat models on most benchmarks we tested, and based on our human evaluations for helpfulness and safety, may be a suitable substitute for closed-source models.

---------------------
Given the context information and not prior knowledge, answer the query.
Query: How many params does llama 2 have
Answer: 

4. 动态少样本示例¶

通过函数映射，您还可以基于其他提示变量动态注入少样本示例。

以下示例展示了如何使用向量存储，根据查询动态注入文本到 SQL 的少样本示例。

首先，我们定义一个文本到 SQL 的提示模板。

In [ ]:

text_to_sql_prompt_tmpl_str = """\
You are a SQL expert. You are given a natural language query, and your job is to convert it into a SQL query.

Here are some examples of how you should convert natural language to SQL:
<examples>
{{ examples }}
</examples>

Now it's your turn.

Query: {{ query_str }}
SQL: 
"""

text_to_sql_prompt_tmpl_str = """\ You are a SQL expert. You are given a natural language query, and your job is to convert it into a SQL query. Here are some examples of how you should convert natural language to SQL: {{ examples }} Now it's your turn. Query: {{ query_str }} SQL: """

根据此提示模板，我们将定义并索引一些文本到SQL的少样本示例。

In [ ]:

import os

os.environ["OPENAI_API_KEY"] = "sk-..."

import os os.environ["OPENAI_API_KEY"] = "sk-..."

In [ ]:

from llama_index.core import Settings, VectorStoreIndex
from llama_index.core.schema import TextNode
from llama_index.llms.openai import OpenAI
from llama_index.embeddings.openai import OpenAIEmbedding

# Set global default LLM and embed model
Settings.llm = OpenAI(model="gpt-4o-mini")
Settings.embed_model = OpenAIEmbedding(model="text-embedding-3-small")

# Setup few-shot examples
example_nodes = [
    TextNode(
        text="Query: How many params does llama 2 have?\nSQL: SELECT COUNT(*) FROM llama_2_params;"
    ),
    TextNode(
        text="Query: How many layers does llama 2 have?\nSQL: SELECT COUNT(*) FROM llama_2_layers;"
    ),
]

# Create index
index = VectorStoreIndex(nodes=example_nodes)

# Create retriever
retriever = index.as_retriever(similarity_top_k=1)

from llama_index.core import Settings, VectorStoreIndex from llama_index.core.schema import TextNode from llama_index.llms.openai import OpenAI from llama_index.embeddings.openai import OpenAIEmbedding # Set global default LLM and embed model Settings.llm = OpenAI(model="gpt-4o-mini") Settings.embed_model = OpenAIEmbedding(model="text-embedding-3-small") # Setup few-shot examples example_nodes = [ TextNode( text="Query: How many params does llama 2 have?\nSQL: SELECT COUNT(*) FROM llama_2_params;" ), TextNode( text="Query: How many layers does llama 2 have?\nSQL: SELECT COUNT(*) FROM llama_2_layers;" ), ] # Create index index = VectorStoreIndex(nodes=example_nodes) # Create retriever retriever = index.as_retriever(similarity_top_k=1)

借助我们的检索器，我们可以创建带有函数映射的提示模板，从而根据查询动态注入少量示例。

In [ ]:

from llama_index.core.prompts import RichPromptTemplate


def get_examples_fn(**kwargs):
    query = kwargs["query_str"]
    examples = retriever.retrieve(query)
    return "\n\n".join(node.text for node in examples)


prompt_tmpl = RichPromptTemplate(
    text_to_sql_prompt_tmpl_str,
    function_mappings={"examples": get_examples_fn},
)

from llama_index.core.prompts import RichPromptTemplate def get_examples_fn(**kwargs): query = kwargs["query_str"] examples = retriever.retrieve(query) return "\n\n".join(node.text for node in examples) prompt_tmpl = RichPromptTemplate( text_to_sql_prompt_tmpl_str, function_mappings={"examples": get_examples_fn}, )

In [ ]:

prompt = prompt_tmpl.format(
    query_str="What are the number of parameters in the llama 2 model?"
)
print(prompt)

prompt = prompt_tmpl.format( query_str="What are the number of parameters in the llama 2 model?" ) print(prompt)

You are a SQL expert. You are given a natural language query, and your job is to convert it into a SQL query.

Here are some examples of how you should convert natural language to SQL:
<examples>
Query: How many params does llama 2 have?
SQL: SELECT COUNT(*) FROM llama_2_params;
</examples>

Now it's your turn.

Query: What are the number of parameters in the llama 2 model?
SQL: 

In [ ]:

response = Settings.llm.complete(prompt)
print(response.text)

response = Settings.llm.complete(prompt) print(response.text)

SELECT COUNT(*) FROM llama_2_params;