Vous pourriez aussi l'utiliser pour générer du code automatiquement, pour les clients qui communiquent avec votre API. Comme par exemple, des applications frontend, mobiles ou IOT.

## Récapitulatif, étape par étape

### Étape 1 : import `FastAPI`

```Python hl_lines="1"
{!../../../docs_src/first_steps/tutorial001.py!}
```

Disons que vous voulez ajouter des traductions pour le Créole, et que ce n'est pas encore dans les documents.

En vérifiant le lien ci-dessus, le code pour "Créole" est `ht`.

L'étape suivante consiste à exécuter le script pour générer un nouveau répertoire de traduction :

<div class="termy">

```console
// Use the command new-lang, pass the language code as a CLI argument

    ],
    visibility = [
        "//tensorflow:internal",
    ],
    deps = [
        ":abstract_operation",
        ":c_api_unified_internal",
        "//tensorflow/c/experimental/gradients/tape:tape_operation",
        "//tensorflow/core/lib/llvm_rtti",
        "//tensorflow/core/platform:errors",
    ],
)

cc_library(
    name = "gradients_internal",
    srcs = [
        "gradients.cc",

running models or processes in the same system. GPUs can also have
implementation bugs that might allow attackers to leave malicious code running
and leak or tamper with applications from other users. Please report
vulnerabilities to the vendor of the affected hardware accelerator.

## Reporting vulnerabilities

### Vulnerabilities in TensorFlow

  optional k8s.io.apimachinery.pkg.apis.meta.v1.LabelSelector selector = 2;

  // template is the object that describes the pod that will be created if
  // insufficient replicas are detected. Each pod stamped out by the StatefulSet
  // will fulfill this Template, but have a unique identity from the rest
  // of the StatefulSet. Each pod will be named with the format

pkg debug/elf, const DT_MIPS_SYMBOL_LIB DynTag
pkg debug/elf, const DT_MIPS_SYMTABNO = 1879048209
pkg debug/elf, const DT_MIPS_SYMTABNO DynTag
pkg debug/elf, const DT_MIPS_TIME_STAMP = 1879048194
pkg debug/elf, const DT_MIPS_TIME_STAMP DynTag
pkg debug/elf, const DT_MIPS_UNREFEXTNO = 1879048210
pkg debug/elf, const DT_MIPS_UNREFEXTNO DynTag
pkg debug/elf, const DT_MOVEENT = 1879047674
pkg debug/elf, const DT_MOVEENT DynTag

  optional k8s.io.apimachinery.pkg.apis.meta.v1.LabelSelector selector = 2;

  // template is the object that describes the pod that will be created if
  // insufficient replicas are detected. Each pod stamped out by the StatefulSet
  // will fulfill this Template, but have a unique identity from the rest
  // of the StatefulSet. Each pod will be named with the format

        "//tensorflow/c/eager:dlpack",
        "//tensorflow/c/eager:tape",
        "//tensorflow/c/eager:tfe_context_internal",
        "//tensorflow/c/eager:tfe_op_internal",
        "//tensorflow/c/eager:tfe_tensorhandle_internal",
        "//tensorflow/c/experimental/gradients",
        "//tensorflow/c/experimental/gradients/tape",
        "//tensorflow/c/experimental/next_pluggable_device:c_api",

Mais c'est beaucoup plus complexe que cela.

!!! tip
    Si vous êtes pressé ou si cela ne vous intéresse pas, passez aux sections suivantes pour obtenir des instructions étape par étape afin de tout configurer avec différentes techniques.

Pour apprendre les bases du HTTPS, du point de vue d'un utilisateur, consultez <a href="https://howhttps.works/"

    deps = [
        "//tensorflow/c/eager:gradient_checker",
        "//tensorflow/c/eager:gradients_internal",
        "//tensorflow/c/eager:unified_api_testutil",
        "//tensorflow/c/experimental/gradients/tape:tape_context",
        "//tensorflow/core:test",
        "//tensorflow/core:test_main",
    ],
)

tf_cuda_cc_test(
    name = "nn_grad_test",
    size = "small",
    srcs = [