wehub/mlflow--mlflow
1
0
Fork 0
Code Issues Pull Requests Actions 15 Deployments Models Agent Notes Packages Projects Releases Wiki Activity
Files
main
mlflow--mlflow/README.md
T
wehub-resource-sync 0a86ca5e2a docs: make Chinese README the default
2026-07-13 10:33:55 +00:00

358 lines
32 KiB
Markdown
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
<!-- WEHUB_ZH_README -->
> [!NOTE]
> 本文档由 WeHub 基于上游 README 翻译整理，属于社区翻译，非官方中文文档。
> [English](./README.en.md) · [原始项目](https://github.com/mlflow/mlflow) · [上游 README](https://github.com/mlflow/mlflow/blob/HEAD/README.md)
> 原作者、版权与许可证归属以原始项目及本仓库 LICENSE 文件为准。
<h1 align="center" style="border-bottom: none">
<a href="https://mlflow.org/">
<img alt="MLflow logo" src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/assets/logo.svg" width="200" />
</a>
</h1>
<h2 align="center" style="border-bottom: none">面向智能体、LLM 与模型的开源 AI 工程平台</h2>
MLflow 是规模最大的开源**面向智能体、LLM 和 ML 模型的 AI 工程平台**。MLflow 使各种规模的团队能够[调试](https://mlflow.org/llm-tracing),
[评估](https://mlflow.org/llm-evaluation), [监控](https://mlflow.org/ai-monitoring), 并[优化](https://mlflow.org/prompt-optimization) 生产级 AI 应用，同时
控制成本并管理对模型和数据的访问。凭借每月超过 **6000 万次下载**，
每天有数千家组织依赖 MLflow 充满信心地将 AI 交付到生产环境。
MLflow 面向智能体和 LLM 应用的全面功能集包括生产级[可观测性](https://mlflow.org/docs/latest/genai/tracing), [评估](https://mlflow.org/docs/latest/genai/eval-monitor),
[提示词管理](https://mlflow.org/docs/latest/genai/prompt-registry), [提示词优化](https://mlflow.org/prompt-optimization) 以及用于管理成本和模型访问的 [AI 网关](https://mlflow.org/docs/latest/genai/governance/ai-gateway)。
了解更多请参阅 [面向 LLM 与智能体的 MLflow](https://mlflow.org/docs/latest/genai).
<div align="center">
[![Python SDK](https://img.shields.io/pypi/v/mlflow)](https://pypi.org/project/mlflow/)
[![PyPI Downloads](https://img.shields.io/pypi/dm/mlflow)](https://pepy.tech/projects/mlflow)
[![License](https://img.shields.io/github/license/mlflow/mlflow)](https://github.com/mlflow/mlflow/blob/master/LICENSE.txt)
<a href="https://twitter.com/intent/follow?screen_name=mlflow" target="_blank">
<img src="https://img.shields.io/twitter/follow/mlflow?logo=X&color=%20%23f5f5f5"
alt="在 X（Twitter）上关注"></a>
<a href="https://www.linkedin.com/company/mlflow-org/" target="_blank">
<img src="https://custom-icon-badges.demolab.com/badge/LinkedIn-0A66C2?logo=linkedin-white&logoColor=fff"
alt="在 LinkedIn 上关注"></a>
[![Ask DeepWiki](https://deepwiki.com/badge.svg)](https://deepwiki.com/mlflow/mlflow)
</div>
<div align="center">
<div>
<a href="https://mlflow.org/"><strong>官网</strong></a> ·
<a href="https://demo.mlflow.org/"><strong>试用演示</strong></a> ·
<a href="https://mlflow.org/docs/latest"><strong>文档</strong></a> ·
<a href="https://mlflow.org/blog"><strong>新闻</strong></a> ·
<a href="https://lu.ma/mlflow?k=c"><strong>活动</strong></a>
</div>
</div>
<br>
## 三步即可上手
几分钟内从零搭建完整 LLMOps 技术栈。无需复杂配置或大规模代码改动。[立即上手 →](https://mlflow.org/docs/latest/genai/tracing/quickstart/)
> **最快上手 — 使用 CLI 设置追踪（tracing）**
>
> ```bash
> uvx mlflow@latest agent setup
> ```
>
> 一条命令即可安装 MLflow 技能，并启动你选择的编程智能体，为应用添加追踪。想自己动手接入？请按照以下三个步骤操作。
**1. 启动 MLflow Server**
```bash
uvx mlflow server
```
**2. 启用日志记录**
```python
import mlflow
mlflow.set_tracking_uri("http://localhost:5000")
mlflow.openai.autolog()
```
**3. 运行你的代码**
```python
from openai import OpenAI
client = OpenAI()
client.responses.create(
model="gpt-5.4-mini",
input="Hello!",
)
```
在 `http://localhost:5000` 的 MLflow UI 中探索追踪和指标。
## LLM 与智能体
MLflow 提供构建、调试、评估和部署生产级 LLM 应用与 AI 智能体所需的一切。支持 Python、TypeScript/JavaScript、Java 及任何其他编程语言。MLflow 还与 [OpenTelemetry](https://opentelemetry.io/) 和 MCP 原生集成。
<table>
<tr>
<td width="50%">
<img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/assets/readme-tracing.png" alt="Observability" width=100%>
<div align="center">
<br>
<a href="https://mlflow.org/docs/latest/genai/tracing/"><strong>Observability</strong></a>
<br><br>
<div>捕获 LLM 应用和智能体的完整追踪，以深入洞察行为。基于 OpenTelemetry 构建，支持任何 LLM 提供商和智能体框架。监控生产质量、成本与安全性。</div><br>
<a href="https://mlflow.org/docs/latest/genai/tracing/quickstart/">入门指南 →</a>
<br>
<a href="https://demo.mlflow.org/#/experiments/1/traces?startTimeLabel=CUSTOM&startTime=2026-04-17T16%3A47%3A15.258Z&endTime=2026-04-24T21%3A20%3A50.781Z">试用演示 →</a>
<br><br>
</div>
</td>
<td width="50%">
<img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/assets/readme-llm-eval.png" alt="Evaluation" width=100%>
<div align="center">
<br>
<a href="https://mlflow.org/docs/latest/genai/eval-monitor/"><strong>Evaluation</strong></a>
<br><br>
<div>运行系统化评估，随时间跟踪质量指标，在回归进入生产环境之前及时发现。可从 50 多种内置指标和 LLM 评判器中选择，或自定义指标。</div><br>
<a href="https://mlflow.org/docs/latest/genai/eval-monitor/">入门指南 →</a>
<br>
<a href="https://demo.mlflow.org/#/experiments/1/runs/d690ad8bb7a546c5a74b79691bb32b27/evaluations">试用演示 →</a>
<br><br>
</div>
</td>
</tr>
<tr>
<td width="50%">
<img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/assets/readme-prompt.png" alt="Prompts & Optimization" width=100%>
<div align="center">
<br>
<a href="https://mlflow.org/docs/latest/genai/prompt-registry/"><strong>Prompts & Optimization</strong></a>
<br><br>
<div>对提示词进行版本管理、测试和部署，并完整追踪血缘关系。<a href="https://mlflow.org/prompt-optimization">自动优化提示词</a>，采用前沿算法提升性能。</div><br>
<a href="https://mlflow.org/docs/latest/genai/prompt-registry/create-and-edit-prompts/">入门指南 →</a>
<br>
<a href="https://demo.mlflow.org/#/experiments/1/prompts/mlflow-demo.prompts.code-reviewer?promptVersion=4">试用演示 →</a>
<br><br>
</div>
</td>
<td width="50%">
<img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/assets/readme-gateway.png" alt="AI Gateway" width=100%>
<div align="center">
<br>
<a href="https://mlflow.org/docs/latest/genai/governance/ai-gateway/"><strong>AI Gateway</strong></a>
<br><br>
<div>面向所有 LLM 提供商的统一 API 网关。通过兼容 OpenAI 的接口路由请求、管理速率限制、处理故障转移并控制成本，内置凭据管理、防护栏（guardrails）以及用于 A/B 测试的流量分流。</div><br>
<a href="https://mlflow.org/docs/latest/genai/governance/ai-gateway/quickstart/">入门指南 →</a>
<br><br>
</div>
</td>
</tr>
</table>
## 模型训练
对于机器学习和深度学习模型开发，MLflow 提供全套工具来管理 ML 生命周期：
- [**实验跟踪（Experiment Tracking）**](https://mlflow.org/docs/latest/ml/tracking/) — 跨实验跟踪模型、参数、指标和评估结果
- [**模型评估（Model Evaluation）**](https://mlflow.org/docs/latest/ml/evaluation/) — 与实验跟踪集成的自动化评估工具
- [**模型注册表（Model Registry）**](https://mlflow.org/docs/latest/ml/model-registry/) — 协作管理 ML 模型的完整生命周期
- [**部署（Deployment）**](https://mlflow.org/docs/latest/ml/deployment/) — 将模型部署到 Docker、Kubernetes、Azure ML、AWS SageMaker 等平台，支持批处理和实时评分
了解更多请参阅 [面向模型训练的 MLflow](https://mlflow.org/docs/latest/ml).
## 集成
MLflow 支持所有智能体（agent）框架、大语言模型（LLM）提供商、工具和编程语言。我们为 60 多个框架提供一行代码即可启用的自动追踪（tracing）。请参阅[完整集成列表](https://mlflow.org/docs/latest/genai/tracing/integrations/).
### OpenTelemetry
<table>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/app-instrumentation/opentelemetry"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/opentelemetry-logo-only.png" height="40"><br><sub><b>OpenTelemetry</b></sub></a></td>
</tr>
</table>
### 智能体框架（Python）
<table>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/langchain"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/langchain-logo-only.png" height="40"><br><sub><b>LangChain</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/langgraph"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/langgraph-logo-only.png" height="40"><br><sub><b>LangGraph</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/openai-agent"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/openai-logo-only.png" height="40"><br><sub><b>OpenAI Agent</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/dspy"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/dspy-logo.png" height="40"><br><sub><b>DSPy</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/pydantic_ai"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/pydantic-ai-logo-only.png" height="40"><br><sub><b>PydanticAI</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/google-adk"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/google-adk-logo.png" height="40"><br><sub><b>Google ADK</b></sub></a></td>
</tr>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/microsoft-agent-framework"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/microsoft-agent-framework-logo.png" height="40"><br><sub><b>Microsoft Agent</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/crewai"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/crewai-logo.svg" height="40"><br><sub><b>CrewAI</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/llama_index"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/llamaindex-logo.svg" height="40"><br><sub><b>LlamaIndex</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/autogen"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/autogen-logo.png" height="40"><br><sub><b>AutoGen</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/strands"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/strands-logo.png" height="40"><br><sub><b>Strands</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/livekit"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/livekit-logo.png" height="40"><br><sub><b>LiveKit Agents</b></sub></a></td>
</tr>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/agno"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/agno-logo.png" height="40"><br><sub><b>Agno</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/bedrock-agentcore"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/bedrock-logo.png" height="40"><br><sub><b>Bedrock AgentCore</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/smolagents"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/smolagents-logo.png" height="40"><br><sub><b>Smolagents</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/semantic_kernel"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/semantic-kernel-logo.png" height="40"><br><sub><b>Semantic Kernel</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/deepagent"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/deepagent-logo.svg" height="40"><br><sub><b>DeepAgent</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/ag2"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/ag2-logo.png" height="40"><br><sub><b>AG2</b></sub></a></td>
</tr>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/haystack"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/haystack-logo.png" height="40"><br><sub><b>Haystack</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/koog"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/koog.png" height="40"><br><sub><b>Koog</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/txtai"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/txtai-logo.png" height="40"><br><sub><b>txtai</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/pipecat"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/pipecat.png" height="40"><br><sub><b>Pipecat</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/watsonx-orchestrate"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/watsonx-orchestrate.png" height="40"><br><sub><b>Watsonx</b></sub></a></td>
</tr>
</table>
### 智能体框架（TypeScript）
<table>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/langchain"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/langchain-logo-only.png" height="40"><br><sub><b>LangChain</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/langgraph"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/langgraph-logo-only.png" height="40"><br><sub><b>LangGraph</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/vercelai"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/vercel-logo.svg" height="40"><br><sub><b>Vercel AI SDK</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/mastra"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/mastra-logo.png" height="40"><br><sub><b>Mastra</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/voltagent"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/voltagent-logo.png" height="40"><br><sub><b>VoltAgent</b></sub></a></td>
</tr>
</table>
### 智能体框架（Java）
<table>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/spring-ai"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/spring-ai-logo.png" height="40"><br><sub><b>Spring AI</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/quarkus-langchain4j"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/langchain4j.svg" height="40"><br><sub><b>Quarkus LangChain4j</b></sub></a></td>
</tr>
</table>
### 模型提供商
<table>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/openai"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/openai-logo-only.png" height="40"><br><sub><b>OpenAI</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/anthropic"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/anthropic-logo.png" height="40"><br><sub><b>Anthropic</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/databricks"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/databricks-logo.png" height="40"><br><sub><b>Databricks</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/gemini"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/google-gemini-logo.svg" height="40"><br><sub><b>Gemini</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/bedrock"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/bedrock-logo.png" height="40"><br><sub><b>Amazon Bedrock</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/litellm"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/litellm-logo.png" height="40"><br><sub><b>LiteLLM</b></sub></a></td>
</tr>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/mistral"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/mistral-ai-logo.svg" height="40"><br><sub><b>Mistral</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/xai-grok"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/grok-logo.png" height="40"><br><sub><b>xAI / Grok</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/ollama"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/ollama-logo.png" height="40"><br><sub><b>Ollama</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/groq"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/groq-logo.svg" height="40"><br><sub><b>Groq</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/deepseek"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/deepseek-logo.png" height="40"><br><sub><b>DeepSeek</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/qwen"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/qwen-logo.jpg" height="40"><br><sub><b>Qwen</b></sub></a></td>
</tr>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/moonshot"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/kimi-logo.png" height="40"><br><sub><b>Moonshot AI</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/cohere"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/cohere-logo.png" height="40"><br><sub><b>Cohere</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/byteplus"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/byteplus-logo.png" height="40"><br><sub><b>BytePlus</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/novitaai"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/novitaai-logo.jpg" height="40"><br><sub><b>Novita AI</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/fireworksai"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/fireworks-ai-logo.png" height="40"><br><sub><b>FireworksAI</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/togetherai"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/together-ai-logo.png" height="40"><br><sub><b>Together AI</b></sub></a></td>
</tr>
</table>
### 网关
<table>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/databricks-ai-gateway"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/databricks-logo.png" height="40"><br><sub><b>Databricks</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/litellm-proxy"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/litellm-logo.png" height="40"><br><sub><b>LiteLLM Proxy</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/vercel-ai-gateway"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/vercel-logo.svg" height="40"><br><sub><b>Vercel AI Gateway</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/openrouter"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/openrouter-logo.png" height="40"><br><sub><b>OpenRouter</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/portkey"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/portkey-logo.png" height="40"><br><sub><b>Portkey</b></sub></a></td>
</tr>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/helicone"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/helicone-logo.png" height="40"><br><sub><b>Helicone</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/kong"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/kong-logo.png" height="40"><br><sub><b>Kong AI Gateway</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/pydantic-ai-gateway"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/pydantic-ai-logo-only.png" height="40"><br><sub><b>PydanticAI Gateway</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/truefoundry"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/truefoundry-logo.png" height="40"><br><sub><b>TrueFoundry</b></sub></a></td>
</tr>
</table>
### 工具与无代码（No-Code）
<table>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/instructor"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/instructor-logo.svg" height="40"><br><sub><b>Instructor</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/claude_code"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/claude-code-logo.png" height="40"><br><sub><b>Claude Code</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/opencode"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/opencode-logo.png" height="40"><br><sub><b>Opencode</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/langfuse"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/langfuse-logo.png" height="40"><br><sub><b>Langfuse</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/arize"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/arize-phoenix-logo.png" height="40"><br><sub><b>Arize / Phoenix</b></sub></a></td>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/goose"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/goose-logo.png" height="40"><br><sub><b>Goose</b></sub></a></td>
</tr>
<tr>
<td align="center" width="110"><a href="https://mlflow.org/docs/latest/genai/tracing/integrations/listing/langflow"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/langflow.svg" height="40"><br><sub><b>Langflow</b></sub></a></td>
</tr>
</table>
## 托管 MLflow
MLflow 可用于多种环境，包括本地环境、本地部署（on-premises）集群、云平台以及托管服务。作为一款开源平台，MLflow **厂商中立（vendor-neutral）**——无论你是在构建 AI 智能体、LLM 应用还是 ML 模型，都能使用 MLflow 的核心能力。
<table>
<tr>
<td align="center" width="130"><a href="https://docs.databricks.com/aws/en/mlflow3/genai/"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/databricks-logo.png" height="40"><br><sub><b>Databricks</b></sub></a></td>
<td align="center" width="130"><a href="https://aws.amazon.com/sagemaker-ai/experiments/"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/amazon-sagemaker-logo.png" height="40"><br><sub><b>Amazon SageMaker</b></sub></a></td>
<td align="center" width="130"><a href="https://learn.microsoft.com/en-us/azure/machine-learning/concept-mlflow?view=azureml-api-2"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/azure-ml-logo.png" height="40"><br><sub><b>Azure ML</b></sub></a></td>
<td align="center" width="130"><a href="https://nebius.com/services/managed-mlflow"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/nebius-logo.png" height="40"><br><sub><b>Nebius</b></sub></a></td>
<td align="center" width="130"><a href="https://mlflow.org/docs/latest/ml/tracking/"><img src="https://raw.githubusercontent.com/mlflow/mlflow/refs/heads/master/docs/static/images/logos/kubernetes-logo.png" height="40"><br><sub><b>Self-Hosted</b></sub></a></td>
</tr>
</table>
## 💭 支持
- 如需 MLflow 使用方面的帮助或提问（例如「如何做 X？」），请访问[文档](https://mlflow.org/docs/latest).
- 在文档中，你可以向我们的 AI 聊天机器人提问。点击右下角的 **"Ask AI"** 按钮。
- 参加[线上活动](https://lu.ma/mlflow?k=c)，例如 Office Hours 和聚会（meetup）。
- 如需报告 bug、提交文档问题或提出功能请求，请[在 GitHub 上提交 issue](https://github.com/mlflow/mlflow/issues/new/choose).
- 如需接收版本发布公告并参与其他讨论，请订阅我们的邮件列表（mlflow-users@googlegroups.com），或在 [Slack](https://mlflow.org/slack). 上加入我们。
## 🤝 贡献
我们诚挚欢迎为 MLflow 做出贡献！
- 提交 [bug 报告](https://github.com/mlflow/mlflow/issues/new?template=bug_report_template.yaml) 和 [功能请求](https://github.com/mlflow/mlflow/issues/new?template=feature_request_template.yaml)
- 参与 [good-first-issues](https://github.com/mlflow/mlflow/issues?q=is%3Aissue+is%3Aopen+label%3A%22good+first+issue%22) 和 [help-wanted](https://github.com/mlflow/mlflow/issues?q=is%3Aissue+is%3Aopen+label%3A%22help+wanted%22)
- 撰写关于 MLflow 的文章并分享你的使用经验
请参阅我们的[贡献指南](CONTRIBUTING.md)，了解更多关于为 MLflow 做出贡献的信息。
## ⭐️ Star 历史
<a href="https://star-history.com/#mlflow/mlflow&Date">
<picture>
<source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=mlflow/mlflow&type=Date&theme=dark" />
<source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=mlflow/mlflow&type=Date" />
<img alt="Star 历史图表" src="https://api.star-history.com/svg?repos=mlflow/mlflow&type=Date" />
</picture>
</a>
## ✏️ 引用
如果你在研究中使用 MLflow，请通过 [GitHub 仓库页面](https://github.com/mlflow/mlflow), 顶部的 "Cite this repository" 按钮进行引用，该按钮会提供包括 APA 和 BibTeX 在内的引用格式。
## 👥 核心成员
MLflow 目前由以下核心成员维护，并得到了数百位才华横溢的社区成员的重要贡献。
- [Ben Wilson](https://github.com/BenWilson2)
- [Corey Zumar](https://github.com/dbczumar)
- [Daniel Lok](https://github.com/daniellok-db)
- [Gabriel Fu](https://github.com/gabrielfu)
- [Harutaka Kawamura](https://github.com/harupy)
- [Humair Khan](https://github.com/HumairAK)
- [Joel Robin P](https://github.com/joelrobin18)
- [Matthew Prahl](https://github.com/mprahl)
- [Pat Sukprasert](https://github.com/PattaraS)
- [Serena Ruan](https://github.com/serena-ruan)
- [Tomu Hirata](https://github.com/TomeHirata)
- [Weichen Xu](https://github.com/WeichenXu123)
- [Yuki Watanabe](https://github.com/B-Step62)
Reference in New Issue View Git Blame Copy Permalink