[1.2.28]

• fix: Limit opensearch-py to below 3.0.0 to prevent grpcio conflicts with milvus and watsonx

[1.2.27]

• feat: Add AWS IAM authentication and full async support to OpenSearch connector

• feat: add astra_generated_embeddings and enable_lexical_search options to AstraDB stager/uploader

[1.2.25]

• fix: dont submit fields for Elasticsearch/Opensearch if there are none specified

[1.2.24]

• fix: Remove --skip-existing from release.yml

• fix: Improve opensearch with correct connector_type metadata and compatible index validation

• feat: Add option to disable binary encoded vectors in AstraDBUploader

What's Changed

• Fix databricks sdk version by @mateuszkuprowski in #605

Full Changelog: 1.2.20...1.2.21

[1.2.20]

• fix: Prevent weaviate cloud precheck from passing with invalid config

What's new in 1.2.19

** fix: pin aiobotocore**

This release creates a pin to avoid a version of aiobotocore that's incompatible with recent versions of botocore.

For the complete changelog, see CHANGELOG.md.

What's New in 1.2.18

feat: add configurable Bedrock inference profile support

This release adds support for configurable Bedrock inference profiles, allowing users to specify custom inference profiles when using AWS Bedrock for embedding and text processing tasks.

Features

• Configurable Bedrock inference profile support
• Enhanced AWS Bedrock integration capabilities

For the complete changelog, see CHANGELOG.md.

🤖 Generated with Claude Code

1.2.14

• Fix: IBM watsonx.data S3 bucket authentication fix