Implement OpenSearch ANN #1225
Implement OpenSearch ANN #1225
Conversation
|
OpenSearch HNSW is seemingly implemented. Simple queries can be run. However, I still need to benchmark OpenSearch HNSW to make sure that there is really a HNSW index running, and to make sure that it is actually faster. When running the benchmarking script, an error is coming up when running |
|
Fixes #1079 |
|
|
||
| In addition to native Elasticsearch query & filtering, it provides efficient vector similarity search using | ||
| the KNN plugin that can scale to a large number of documents. | ||
| """ | ||
|
|
||
| def __init__(self, **kwargs): |
There was a problem hiding this comment.
It might be more readable to have explicit arguments(with their corresponding defaults).
| @@ -1053,6 +1100,9 @@ def _create_document_index(self, index_name: str): | |||
| if not self.client.indices.exists(index=index_name): | |||
| raise e | |||
|
|
|||
| def stop_service(self): | |||
There was a problem hiding this comment.
Would it make sense to keep this in the utils instead? It can be misleading as it only works in the case when OpenSearch is running as a Docker Container running locally.
OpenSearch has ANN capabilities. Let's add support for these features!
Only OpenSearch has dot product ANN, not Open Distro. Let's also make this switch.