feat(amazon-bedrock): binary embeddings

[anatolzak]

Background

The AI SDK currently does not support generating binary embeddings using Amazon Titan Text Embeddings V2.
Binary embeddings offer significant advantages, maintaining about 95% accuracy while having a footprint that is more than 95% smaller compared to float embeddings.

https://youtu.be/8B1RW6uw8GQ?t=1109

Summary

This PR introduces an option to allow generating binary embeddings using Amazon Titan Text Embeddings V2.

const { embedding } = await embed({
  model: bedrock.embedding('amazon.titan-embed-text-v2:0', {
    embeddingType: 'binary',
  }),
  value: 'sunny day at the beach',
});

console.log(embedding);

[
  0, 1, 0, 0, 1, 1, 0, 0, 0, 1, 1, 1,
  1, 0, 0, 0, 1, 0, 1, 1, 0, 1, 1, 0,
  1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1,
  1, 1, 0, 1, 1, 0, 0, 1, 1, 1, 1, 1,
  1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 1, 0,
  0, 1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 0,
  1, 1, 0, 0, 1, 0, 1, 1, 1, 1, 1, 1,
  1, 0, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1,
  0, 0, 0, 1,
  ... 924 more items
]

See the Amazon Titan Text Embeddings v2 response schema for reference:
https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-titan-embed-text.html#model-parameters-titan-embed-text-request-response

Tasks

• Tests for the changes have been added (for bug fixes / features)
• Docs have been added / updated (for bug fixes / features)
• You've followed the PR contributing guidelines for adding a patch changeset to version packages and fixing prettier issues

Future Work