Amazon S3 Vectors¶
Amazon S3 Vectors provides native vector storage in Amazon S3 for AI/ML applications. Use the operators below to manage S3 Vectors resources.
Create a Vector Bucket¶
To create an Amazon S3 Vectors vector bucket, use
S3VectorsCreateVectorBucketOperator.
create_vector_bucket = S3VectorsCreateVectorBucketOperator(
task_id="create_vector_bucket",
vector_bucket_name=bucket_name,
)
Create an Index¶
To create an index in an Amazon S3 Vectors vector bucket, use
S3VectorsCreateIndexOperator.
create_index = S3VectorsCreateIndexOperator(
task_id="create_index",
vector_bucket_name=bucket_name,
index_name=index_name,
data_type="float32",
dimension=4,
distance_metric="cosine",
)
Delete an Index¶
To delete an index from an Amazon S3 Vectors vector bucket, use
S3VectorsDeleteIndexOperator.
delete_index = S3VectorsDeleteIndexOperator(
task_id="delete_index",
vector_bucket_name=bucket_name,
index_name=index_name,
trigger_rule=TriggerRule.ALL_DONE,
)
Delete a Vector Bucket¶
To delete an Amazon S3 Vectors vector bucket, use
S3VectorsDeleteVectorBucketOperator.
delete_vector_bucket = S3VectorsDeleteVectorBucketOperator(
task_id="delete_vector_bucket",
vector_bucket_name=bucket_name,
trigger_rule=TriggerRule.ALL_DONE,
)
Put Vectors¶
To insert vectors into an Amazon S3 Vectors index, use
S3VectorsPutVectorsOperator.
put_vectors = S3VectorsPutVectorsOperator(
task_id="put_vectors",
vector_bucket_name=bucket_name,
index_name=index_name,
vectors=[{"key": "test-vec-1", "data": {"float32": [0.1, 0.2, 0.3, 0.4]}}],
)
Query Vectors¶
To query vectors by similarity in an Amazon S3 Vectors index, use
S3VectorsQueryVectorsOperator.
query_vectors = S3VectorsQueryVectorsOperator(
task_id="query_vectors",
vector_bucket_name=bucket_name,
index_name=index_name,
top_k=3,
query_vector={"float32": [0.1, 0.2, 0.3, 0.4]},
)