Generative AI Google Brain

translation-intro-translation

2 min read

Getting Started with Translation

Source notebook

Repo path: translation/intro_translation.ipynb · Open on GitHub · intro

Translates English text to Spanish with Cloud Translation API and a Cloud Storage-backed glossary.

Summary

This notebook teaches basic use of the Google Cloud Translation API for translating text from English to Spanish. It sets project, region, and authentication, creates a TranslationServiceClient, defines helpers for text translation and glossary creation, then demonstrates default translation versus glossary-based translation for Google Cloud product names.

Key code patterns

Create Translation client

from google.cloud import translate
 
client = translate.TranslationServiceClient()

Creates the Cloud Translation API client used for translate and glossary requests.

Translate text

response = client.translate_text(
    request=translate.TranslateTextRequest(
        parent=client.common_location_path(project_id, location),
        contents=[text],
        source_language_code="en",
        target_language_code="es",
    )
)

Shows the core request shape for translating English input text to Spanish.

Optional glossary config

glossary_config=(
    translate.TranslateTextGlossaryConfig(glossary=glossary)
    if glossary
    else None
)

Adds a glossary only when supplied, allowing the same helper to handle default and glossary translations.

Create glossary from GCS

glossary = translate.Glossary(
    name=client.glossary_path(project_id, location, glossary_id),
    language_pair=translate.Glossary.LanguageCodePair(
        source_language_code="en", target_language_code="es"
    ),
    input_config=translate.GlossaryInputConfig(
        gcs_source=translate.GcsSource(input_uri=input_uri)
    ),
)
operation = client.create_glossary(
    parent=client.common_location_path(project_id, location), glossary=glossary
)

Defines a unidirectional glossary sourced from a Cloud Storage TSV file.

Models & APIs used

  • APIs / services: Cloud Translation API, Cloud Storage
  • SDKs / libraries: google-cloud-translate

When to use this

Use this pattern when translating text with Cloud Translation API and enforcing domain-specific terminology through glossaries.

Gotchas & caveats

  • Cloud Translation and Cloud Storage are billable components.
  • Translation has documented available regions and optional advanced endpoints.
  • Colab requires user authentication, project configuration, and application-default login.
  • Glossary source data is read from Cloud Storage, such as a TSV file in a bucket.
  • Glossaries are for words or short phrases, usually fewer than five words.

Best practices

  • Use environment variables for project ID and region when notebook parameters are not provided.
  • Set source_language_code and target_language_code explicitly.
  • Use client.common_location_path for regional Translation API resource paths.
  • Use glossaries to consistently translate domain-specific words and phrases.
  • Check supported language codes and supported content formats in the Translation API documentation.

Graph View

Backlinks