MLOps Academy · 课时

只有超过基线时才自动推广

以冠军模型和挑战模型的比较结果作为推广门槛

只有超过基线时才自动推广 是 CoddyKit 上的免费 MLOps Academy 课时。这是第 3 节课，共 4 节。你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。这是 MLOps Academy 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 MLOps Academy 课程共包含 4 节课。

本课时的部分内容尚未翻译，以英文显示。

Newer Is Not Always Better

A freshly retrained model can be worse than the one in production. Never promote a new model blindly just because it is newer.

Champion vs Challenger

The live model is the champion; the new candidate is the challenger. Promotion is a contest the challenger must win on the metrics.

Define the Baseline

Your baseline is the current champion score on a fixed evaluation set. The challenger has to clearly beat that number to earn promotion.

Evaluate on the Same Data

Score both models on the identical, held-out test set. A fair comparison only counts when nothing else changed between them.

A Promotion Gate

Wrap the decision in a check that returns true only if the challenger wins. This gate compares the two scores directly.

def should_promote(new, baseline):
    return new > baseline

Add a Margin

Tiny gains can be noise, so require a real margin. Demanding at least a 1% lift avoids churning the production model for nothing.

def should_promote(new, baseline):
    return new >= baseline + 0.01

Pick the Right Metric

Compare on the metric that matters for your task, like F1 or AUC, not just raw accuracy. The wrong metric promotes the wrong model.

Wire It Into the Pipeline

Put the gate right after evaluation in your DAG. If the challenger loses, the pipeline stops and the champion stays live.

Promote in the Registry

Winning means moving the model stage to Production in the registry. This transition is the one action that changes what users get served.

client.transition_model_version_stage(
    name="fraud", version=7, stage="Production")

Log the Decision

Record both scores and the outcome every run. That audit trail explains later why a model was or was not promoted.

Guard Against Bad Data

A challenger can look amazing because the test set leaked or shrank. Sanity-check the evaluation before trusting any promotion.

Quick Check

When should the new challenger model be auto-promoted?

Recap

You gated promotion on results: compare challenger to champion on the same data, require a margin, then transition the winner in the registry. ✅

常见问题解答

「只有超过基线时才自动推广」课时是免费的吗？

是的 — 「只有超过基线时才自动推广」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 MLOps Academy 课程的其余内容，请升级到 CoddyKit PRO。 MLOps Academy 课程共包含 4 节课。

「只有超过基线时才自动推广」这节课中我会学到什么？

以冠军模型和挑战模型的比较结果作为推广门槛你通过在浏览器中直接运行的动手代码来练习 MLOps Academy，全天候 AI 导师会在你学习这节课的过程中回答你的问题。

学习 MLOps Academy 需要有经验吗？

无需任何先前经验。CoddyKit 上的 MLOps Academy 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。这是第 3 节课，共 4 节。

「只有超过基线时才自动推广」课时需要多长时间？

大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。

我能在这节 MLOps Academy 课中编写并运行代码吗？

能。每节 MLOps Academy 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。

此课程中的所有课时

← 返回 MLOps Academy