Add Tencent Elasticsearch Results #654
Conversation
# Conflicts: # README.md resolved by upstream-a/main version # vectordb_bench/backend/clients/__init__.py resolved by upstream-a/main version # vectordb_bench/backend/clients/tencent_elasticsearch/tencent_elasticsearch.py resolved by upstream-a/main version # vectordb_bench/cli/vectordbbench.py resolved by upstream-a/main version # vectordb_bench/frontend/config/dbCaseConfigs.py resolved by upstream-a/main version
| "metric_type": "COSINE", | ||
| "efConstruction": 200, | ||
| "M": 16, | ||
| "num_candidates": 4000 |
There was a problem hiding this comment.
Thank you for sharing your test results on Tencent-ES with us. However, the num_candidates parameter appears to be set higher than typical. For tasks targeting top-100 retrieval, it is common to set this value around 100, which can achieve recall=0.9. Here, setting it to 4000 achieves a recall above 0.99, but it also significantly increases computational overhead. Please confirm whether this meets your expectations.
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: alwayslove2013, morning-color The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
4 participants
No description provided.