add to public

add medium and large public dataset

add medium and large dataset to public

add medium and large to public

Author: Caroline-an777

vectordb_bench/__init__.py

@@ -53,6 +53,9 @@ class config:
     LOAD_TIMEOUT_1536D_500K = 24 * 3600  # 24h
     LOAD_TIMEOUT_1536D_5M = 240 * 3600  # 10d
 
+    LOAD_TIMEOUT_1024D_1M = 24 * 3600  # 24h
+    LOAD_TIMEOUT_1024D_10M = 240 * 3600  # 10d
+
     OPTIMIZE_TIMEOUT_DEFAULT = 24 * 3600  # 24h
     OPTIMIZE_TIMEOUT_768D_100K = 24 * 3600  # 24h
     OPTIMIZE_TIMEOUT_768D_1M = 24 * 3600  # 24h
@@ -62,6 +65,9 @@ class config:
     OPTIMIZE_TIMEOUT_1536D_500K = 24 * 3600  # 24h
     OPTIMIZE_TIMEOUT_1536D_5M = 240 * 3600  # 10d
 
+    OPTIMIZE_TIMEOUT_1024D_1M = 24 * 3600  # 24h
+    OPTIMIZE_TIMEOUT_1024D_10M = 240 * 3600  # 10d
+
     def display(self) -> str:
         return [
             i

vectordb_bench/backend/cases.py

@@ -42,6 +42,9 @@ class CaseType(Enum):
     Performance1536D500K99P = 14
     Performance1536D5M99P = 15
 
+    Performance1024D1M = 17
+    Performance1024D10M = 20
+
     Performance1536D50K = 50
 
     Custom = 100
@@ -309,6 +312,30 @@ class Performance1536D5M99P(IntFilterPerformanceCase):
     optimize_timeout: float | int | None = config.OPTIMIZE_TIMEOUT_1536D_5M
 
 
+class Performance1024D1M(PerformanceCase):
+    case_id: CaseType = CaseType.Performance1024D1M
+    filter_rate: float | int | None = None
+    dataset: DatasetManager = Dataset.BIOASQ.manager(1_000_000)
+    name: str = "Search Performance Test (1M Dataset, 1024 Dim)"
+    description: str = """This case tests the search performance of a vector database with a medium 1M dataset
+    (<b>Bioasq 1M vectors</b>, 1024 dimensions), at varying parallel levels. Results will show index building time,
+    recall, and maximum QPS."""
+    load_timeout: float | int = config.LOAD_TIMEOUT_1024D_1M
+    optimize_timeout: float | int | None = config.OPTIMIZE_TIMEOUT_1024D_1M
+
+
+class Performance1024D10M(PerformanceCase):
+    case_id: CaseType = CaseType.Performance1024D10M
+    filter_rate: float | int | None = None
+    dataset: DatasetManager = Dataset.BIOASQ.manager(10_000_000)
+    name: str = "Search Performance Test (10M Dataset, 1024 Dim)"
+    description: str = """This case tests the search performance of a vector database with a large 10M dataset
+    (<b>Bioasq 10M vectors</b>, 1024 dimensions), at varying parallel levels. Results will show index building time,
+    recall, and maximum QPS."""
+    load_timeout: float | int = config.LOAD_TIMEOUT_1024D_10M
+    optimize_timeout: float | int | None = config.OPTIMIZE_TIMEOUT_1024D_10M
+
+
 class Performance1536D50K(PerformanceCase):
     case_id: CaseType = CaseType.Performance1536D50K
     filter_rate: float | int | None = None
@@ -497,6 +524,8 @@ def filters(self) -> Filter:
     CaseType.Performance1536D5M1P: Performance1536D5M1P,
     CaseType.Performance1536D500K99P: Performance1536D500K99P,
     CaseType.Performance1536D5M99P: Performance1536D5M99P,
+    CaseType.Performance1024D1M: Performance1024D1M,
+    CaseType.Performance1024D10M: Performance1024D10M,
     CaseType.Performance1536D50K: Performance1536D50K,
     CaseType.PerformanceCustomDataset: PerformanceCustomDataset,
     CaseType.StreamingPerformanceCase: StreamingPerformanceCase,

vectordb_bench/backend/dataset.py

@@ -166,6 +166,20 @@ class Cohere(BaseDataset):
     scalar_label_percentages: list[float] = [0.001, 0.002, 0.005, 0.01, 0.02, 0.05, 0.1, 0.2, 0.5]
 
 
+class Bioasq(BaseDataset):
+    name: str = "Bioasq"
+    dim: int = 1024
+    metric_type: MetricType = MetricType.COSINE
+    use_shuffled: bool = config.USE_SHUFFLED_DATA
+    with_gt: bool = True
+    _size_label: dict = {
+        1_000_000: SizeLabel(1_000_000, "MEDIUM", 1),
+        10_000_000: SizeLabel(10_000_000, "LARGE", 10),
+    }
+    with_scalar_labels: bool = True
+    scalar_label_percentages: list[float] = [0.001, 0.002, 0.005, 0.01, 0.02, 0.05, 0.1, 0.2, 0.5]
+
+
 class Glove(BaseDataset):
     name: str = "Glove"
     dim: int = 200
@@ -361,6 +375,7 @@ class Dataset(Enum):
     LAION = LAION
     GIST = GIST
     COHERE = Cohere
+    BIOASQ = Bioasq
     GLOVE = Glove
     SIFT = SIFT
     OPENAI = OpenAI
@@ -376,6 +391,8 @@ class DatasetWithSizeType(Enum):
     CohereSmall = "Small Cohere (768dim, 100K)"
     CohereMedium = "Medium Cohere (768dim, 1M)"
     CohereLarge = "Large Cohere (768dim, 10M)"
+    BioasqMedium = "Medium Bioasq (1024dim, 1M)"
+    BioasqLarge = "Large Bioasq (1024dim, 10M)"
     OpenAISmall = "Small OpenAI (1536dim, 50K)"
     OpenAIMedium = "Medium OpenAI (1536dim, 500K)"
     OpenAILarge = "Large OpenAI (1536dim, 5M)"
@@ -410,6 +427,8 @@ def get_optimize_timeout(self) -> float:
     DatasetWithSizeType.CohereSmall: Dataset.COHERE.manager(100_000),
     DatasetWithSizeType.CohereMedium: Dataset.COHERE.manager(1_000_000),
     DatasetWithSizeType.CohereLarge: Dataset.COHERE.manager(10_000_000),
+    DatasetWithSizeType.BioasqMedium: Dataset.BIOASQ.manager(1_000_000),
+    DatasetWithSizeType.BioasqLarge: Dataset.BIOASQ.manager(10_000_000),
     DatasetWithSizeType.OpenAISmall: Dataset.OPENAI.manager(50_000),
     DatasetWithSizeType.OpenAIMedium: Dataset.OPENAI.manager(500_000),
     DatasetWithSizeType.OpenAILarge: Dataset.OPENAI.manager(5_000_000),

vectordb_bench/frontend/config/dbCaseConfigs.py

@@ -231,6 +231,9 @@ def generate_label_filter_cases(dataset_with_size_type: DatasetWithSizeType) ->
             UICaseItem(cases=generate_normal_cases(CaseType.Performance768D10M)),
             UICaseItem(cases=generate_normal_cases(CaseType.Performance768D1M)),
             UICaseItem(isLine=True),
+            UICaseItem(cases=generate_normal_cases(CaseType.Performance1024D1M)),
+            UICaseItem(cases=generate_normal_cases(CaseType.Performance1024D10M)),
+            UICaseItem(isLine=True),
             UICaseItem(cases=generate_normal_cases(CaseType.Performance1536D5M)),
             UICaseItem(cases=generate_normal_cases(CaseType.Performance1536D500K)),
             UICaseItem(cases=generate_normal_cases(CaseType.Performance1536D50K)),
@@ -301,6 +304,8 @@ def generate_label_filter_cases(dataset_with_size_type: DatasetWithSizeType) ->
     CaseType.Performance1536D5M,
     CaseType.Performance1536D500K,
     CaseType.Performance1536D50K,
+    CaseType.Performance1024D1M,
+    CaseType.Performance1024D10M,
     CaseType.CapacityDim960,
     CaseType.CapacityDim128,
 ]