创建和编辑提示

2025-06-11

重要

此功能在 Beta 版中。

本指南介绍如何在 MLflow 提示注册表中创建新提示和管理其版本。你将了解如何使用 MLflow Python SDK 执行这些作。

先决条件

安装 MLflow 和所需包

pip install --upgrade "mlflow[databricks]>=3.1.0" openai

请按照设置环境快速指南创建 MLflow 试验。
访问 Unity 目录中的 CREATE FUNCTION
- 为什么？提示以函数的形式存储在 UC 中

步骤 1. 创建新提示

可以使用 Python SDK 以编程方式创建提示。

使用 mlflow.genai.register_prompt(). 以编程方式创建提示。提示对模板变量使用双大括号语法（{{variable}}）。

import mlflow

# Replace with a Unity Catalog schema where you have CREATE FUNCTION permission
uc_schema = "workspace.default"
# This table will be created in the above UC schema
prompt_name = "summarization_prompt"

# Define the prompt template with variables
initial_template = """\
Summarize content you are provided with in {{num_sentences}} sentences.

Content: {{content}}
"""

# Register a new prompt
prompt = mlflow.genai.register_prompt(
    name=f"{uc_schema}.{prompt_name}",
    template=initial_template,
    # all parameters below are optional
    commit_message="Initial version of summarization prompt",
    tags={
        "author": "data-science-team@company.com",
        "use_case": "document_summarization",
        "task": "summarization",
        "language": "en",
        "model_compatibility": "gpt-4"
    }
)

print(f"Created prompt '{prompt.name}' (version {prompt.version})")

步骤 2：在应用程序中使用提示

下面是使用上述提示模板的简单应用程序。

从注册表加载提示

# Load a specific version using URI syntax
prompt = mlflow.genai.load_prompt(name_or_uri=f"prompts:/{uc_schema}.{prompt_name}/1")

# Alternative syntax without URI
prompt = mlflow.genai.load_prompt(name_or_uri=f"{uc_schema}.{prompt_name}")

在应用程序中使用提示

import mlflow
from openai import OpenAI

# Enable MLflow's autologging to instrument your application with Tracing
mlflow.openai.autolog()

# Connect to a Databricks LLM via OpenAI using the same credentials as MLflow
# Alternatively, you can use your own OpenAI credentials here
mlflow_creds = mlflow.utils.databricks_utils.get_databricks_host_creds()
client = OpenAI(
    api_key=mlflow_creds.token,
    base_url=f"{mlflow_creds.host}/serving-endpoints"
)

# Use the trace decorator to capture the application's entry point
@mlflow.trace
def my_app(content: str, num_sentences: int):
    # Format with variables
    formatted_prompt = prompt.format(
        content=content,
        num_sentences=num_sentences
    )

    response = client.chat.completions.create(
    model="databricks-claude-sonnet-4",  # This example uses a Databricks hosted LLM - you can replace this with any AI Gateway or Model Serving endpoint. If you provide your own OpenAI credentials, replace with a valid OpenAI model e.g., gpt-4o, etc.
    messages=[
        {
        "role": "system",
        "content": "You are a helpful assistant.",
        },
        {
        "role": "user",
        "content": formatted_prompt,
        },
    ],
    )
    return response.choices[0].message.content

result = my_app(content="This guide shows you how to integrate prompts from the MLflow Prompt Registry into your GenAI applications. You'll learn to load prompts, format them with dynamic data, and ensure complete lineage by linking prompt versions to your MLflow Models.", num_sentences=1)
print(result)

步骤 3. 编辑指令

创建的提示版本一旦生成后就无法更改。若要编辑提示内容，请创建一个包含您的更改的新版本。类似 Git 的这种版本控制确保完整的历史记录，并支持回滚功能。

使用现有提示名称调用 mlflow.genai.register_prompt() 创建新版本：

import mlflow

# Define the improved template
new_template = """\
You are an expert summarizer. Condense the following content into exactly {{ num_sentences }} clear and informative sentences that capture the key points.

Content: {{content}}

Your summary should:
- Contain exactly {{num_sentences}} sentences
- Include only the most important information
- Be written in a neutral, objective tone
- Maintain the same level of formality as the original text
"""

# Register a new version
updated_prompt = mlflow.genai.register_prompt(
    name=f"{uc_schema}.{prompt_name}",
    template=new_template,
    commit_message="Added detailed instructions for better output quality",
    tags={
        "author": "data-science-team@company.com",
        "improvement": "Added specific guidelines for summary quality"
    }
)

print(f"Created version {updated_prompt.version} of '{updated_prompt.name}'")

步骤 4. 使用新提示

# Load a specific version using URI syntax
prompt = mlflow.genai.load_prompt(name_or_uri=f"prompts:/{uc_schema}.{prompt_name}/2")

# Or load from specific version
prompt = mlflow.genai.load_prompt(name_or_uri=f"{uc_schema}.{prompt_name}", version="2")

步骤 5. 搜索和探索提示

在 Unity 目录架构中查找提示：

# REQUIRED format for Unity Catalog - specify catalog and schema
results = mlflow.genai.search_prompts("catalog = 'workspace' AND schema = 'default'")

# Using variables for your schema
catalog_name = uc_schema.split('.')[0]  # 'workspace'
schema_name = uc_schema.split('.')[1]   # 'default'
results = mlflow.genai.search_prompts(f"catalog = '{catalog_name}' AND schema = '{schema_name}'")

# Limit results
results = mlflow.genai.search_prompts(
    filter_string=f"catalog = '{catalog_name}' AND schema = '{schema_name}'",
    max_results=50
)

注意： Unity 目录搜索仅支持目录 + 架构筛选。即将推出对名称模式和标记筛选的支持。

后续步骤

评估提示版本 - 比较不同的提示版本以识别最佳性能
使用应用版本跟踪提示 - 将提示版本链接到应用程序版本
使用在已部署应用中的提示——利用别名将提示部署到生产环境

通过