ParallelData

Source: src/AWS/Translate/ParallelData.ts

Amazon Translate parallel data — segment-aligned translation examples imported from S3 that steer the style, tone, and word choice of batch translation jobs (Active Custom Translation). The import is asynchronous: the resource waits for the parallel data to become ACTIVE.

Managing Parallel Data

Import parallel data from S3

const examples = yield* AWS.Translate.ParallelData("StyleExamples", {
  s3Uri: "s3://my-bucket/style-examples.csv",
  format: "CSV",
});

Use parallel data in a batch translation job

const startJob = yield* AWS.Translate.StartTextTranslationJob(dataAccessRole);
yield* startJob({
  InputDataConfig: { S3Uri: "s3://my-bucket/input/", ContentType: "text/plain" },
  OutputDataConfig: { S3Uri: "s3://my-bucket/output/" },
  SourceLanguageCode: "en",
  TargetLanguageCodes: ["es"],
  ParallelDataNames: [examples.parallelDataName],
});