<li><a href="http://cs/f:devtools/kokoro/config/prod/$KOKORO_JOB_NAME">Codesearch - job definition</a></li>
<li><a href="http://cs/f:learning/brain/testing/kokoro/$(echo "$KOKORO_JOB_NAME" | sed 's!tensorflow/!!g')">Codesearch - build definition & scripts</a></li>
<li><a href="http://cs/$KOKORO_JOB_NAME">Codesearch - All references to this job</a></li>
</ul>

* **更簡單**：
    * 不需要學習新的 micro-language 來定義結構。
    * 如果你知道 Python 型別，你就知道如何使用 Pydantic。
* 和你的 **<abbr title="Integrated Development Environment，和程式碼編輯器類似">IDE</abbr>/<abbr title="一個檢查程式碼錯誤的工具">linter</abbr>/brain** 都能好好配合：
    * 因為 Pydantic 的資料結構其實就是你自己定義的類別實例，所以自動補齊、linting、mypy 以及你的直覺都能很好地在經過驗證的資料上發揮作用。
* 驗證**複雜結構**：
    * 使用 Pydantic 模型時，你可以把資料結構分層設計，並且用 Python 的 `List` 和 `Dict` 等型別來定義。

    * If you know Python types you know how to use Pydantic.
* Plays nicely with your **<abbr title="Integrated Development Environment, similar to a code editor">IDE</abbr>/<abbr title="A program that checks for code errors">linter</abbr>/brain**:
    * Because pydantic data structures are just instances of classes you define; auto-completion, linting, mypy and your intuition should all work properly with your validated data.
* Validate **complex structures**:

    /**
     * Logger entries returned by {@link #drain()} method.
     * @param level The logging level, never {@code null}.
     * @param message The logging message, never {@code null}.
     * @param error The error, if applicable.
     */
    record Entry(@Nonnull Level level, @Nonnull String message, @Nullable Throwable error) {}

    /**

		result := drain(input)
		if result != test.output {
			t.Errorf("%s: got %q expected %q", test.name, result, test.output)
		}
	}
}

// lines joins the arguments together as complete lines.
func lines(a ...string) string {
	return strings.Join(a, "\n") + "\n"
}

// drain returns a single string representing the processed input tokens.

from anywhere, and executes the graphs it is sent without performing any checks.
Therefore, if you run a `tf.train.Server` in your network, anybody with access
to the network can execute arbitrary code with the privileges of the user
running the `tf.train.Server`.

## Untrusted inputs during training and prediction

TensorFlow supports a wide range of input data formats. For example it can

= vendor/github.com/cpuguy83/go-md2man/v2 licensed under: =

The MIT License (MIT)

Copyright (c) 2014 Brian Goff

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell

l'application efficacement et sans interruption ni problème.

Ceci contraste avec les étapes de **développement**, où vous êtes constamment en train de modifier le code, de le casser
et de le réparer, d'arrêter et de redémarrer le serveur de développement, _etc._

## Stratégies de déploiement

        requireNonNull(level, "level");
        requireNonNull(message, "message");
        entries.get().add(new Entry(level, message, error));
    }

    @Override
    public List<Entry> drain() {
        return entries.getAndSet(new CopyOnWriteArrayList<>());
    }

Le **Guide d'utilisation avancé**, qui s'appuie sur cette base, utilise les mêmes concepts et vous apprend quelques fonctionnalités supplémentaires.

Mais vous devez d'abord lire le **Tutoriel - Guide d'utilisation** (ce que vous êtes en train de lire en ce moment).