Skip to content

fix es quant & rescore settings #648

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
XuanYang-cn merged 1 commit into from
Nov 27, 2025
Merged

fix es quant & rescore settings #648

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
43 changes: 8 additions & 35 deletions vectordb_bench/backend/clients/elastic_cloud/elastic_cloud.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -180,48 +180,21 @@ def search_embedding(
"""
assert self.client is not None, "should self.init() first"

knn = {
"field": self.vector_col_name,
"k": k,
"num_candidates": self.case_config.num_candidates,
"filter": self.filter,
"query_vector": query,
}
if self.case_config.use_rescore:
oversample_k = int(k * self.case_config.oversample_ratio)
oversample_num_candidates = int(self.case_config.num_candidates * self.case_config.oversample_ratio)
knn = {
"field": self.vector_col_name,
"k": oversample_k,
"num_candidates": oversample_num_candidates,
"filter": self.filter,
"query_vector": query,
}
rescore = {
"window_size": oversample_k,
"query": {
"rescore_query": {
"script_score": {
"query": {"match_all": {}},
"script": {
"source": f"cosineSimilarity(params.queryVector, '{self.vector_col_name}')",
"params": {"queryVector": query},
},
}
},
"query_weight": 0,
"rescore_query_weight": 1,
},
}
else:
knn = {
"field": self.vector_col_name,
"k": k,
"num_candidates": self.case_config.num_candidates,
"filter": self.filter,
"query_vector": query,
}
rescore = None
knn["rescore_vector"] = {"oversample": self.case_config.oversample_ratio}
size = k

res = self.client.search(
index=self.indice,
knn=knn,
routing=self.routing_key,
rescore=rescore,
size=size,
_source=False,
docvalue_fields=[self.id_col_name],
Expand Down
7 changes: 4 additions & 3 deletions vectordb_bench/frontend/config/dbCaseConfigs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -839,7 +839,7 @@ class CaseConfigInput(BaseModel):
)

CaseConfigParamInput_IndexType_ES = CaseConfigInput(
label=CaseConfigParamType.IndexType,
label=CaseConfigParamType.index,
inputType=InputType.Option,
inputConfig={
"options": [
Expand Down Expand Up @@ -881,15 +881,16 @@ class CaseConfigInput(BaseModel):
label=CaseConfigParamType.use_rescore,
inputType=InputType.Bool,
inputConfig={"value": False},
isDisplayed=lambda config: config.get(CaseConfigParamType.IndexType, None) != IndexType.ES_HNSW.value,
isDisplayed=lambda config: config.get(CaseConfigParamType.index, None) != IndexType.ES_HNSW.value,
inputHelp="Recalculating scores using the original (non-quantized) vectors.",
)

CaseConfigParamInput_OversampleRatio_ES = CaseConfigInput(
label=CaseConfigParamType.oversample_ratio,
inputType=InputType.Float,
inputConfig={"min": 1.0, "max": 100.0, "value": 2.0},
isDisplayed=lambda config: config.get(CaseConfigParamType.use_rescore, False),
inputHelp="num_oversample = oversample_ratio * top_k.",
inputHelp="Retrieving more candidates per shard for rescoring.",
)

CaseConfigParamInput_UseRouting_ES = CaseConfigInput(
Expand Down
1 change: 1 addition & 0 deletions vectordb_bench/models.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,6 +44,7 @@ class CaseConfigParamType(Enum):
"""

IndexType = "IndexType"
index = "index"
M = "M"
EFConstruction = "efConstruction"
ef_construction = "ef_construction"
Expand Down