diff --git a/pyproject.toml b/pyproject.toml
index aafe70750..03349ebdf 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -35,10 +35,11 @@ dependencies = [
     "psutil",
     "polars",
     "plotly",
-    "environs",
+    "environs<14.1.0",
     "pydantic<v2",
     "scikit-learn",
     "pymilvus", # with pandas, numpy, ujson
+    "ujson"
 ]
 dynamic = ["version"]
 
diff --git a/vectordb_bench/backend/clients/api.py b/vectordb_bench/backend/clients/api.py
index a86849e96..28dae441f 100644
--- a/vectordb_bench/backend/clients/api.py
+++ b/vectordb_bench/backend/clients/api.py
@@ -22,7 +22,7 @@ class IndexType(str, Enum):
     IVFSQ8 = "IVF_SQ8"
     Flat = "FLAT"
     AUTOINDEX = "AUTOINDEX"
-    ES_HNSW = "hnsw"
+    ES_HNSW = "ybhnsw"
     ES_IVFFlat = "ivfflat"
     GPU_IVF_FLAT = "GPU_IVF_FLAT"
     GPU_IVF_PQ = "GPU_IVF_PQ"
diff --git a/vectordb_bench/backend/clients/pgvector/cli.py b/vectordb_bench/backend/clients/pgvector/cli.py
index 55a462055..fba329d2d 100644
--- a/vectordb_bench/backend/clients/pgvector/cli.py
+++ b/vectordb_bench/backend/clients/pgvector/cli.py
@@ -118,6 +118,28 @@ class PgVectorTypedDict(CommonTypedDict):
             callback=set_default_quantized_fetch_limit,
         ),
     ]
+    create_index_before_load: Annotated[
+        bool | None,
+        click.option(
+            "--create-index-before-load",
+            type=bool,
+            help="Enable create index before loading vectors",
+            default=True,
+            required=False,
+            show_default=True,
+        ),
+    ]
+    create_index_after_load: Annotated[
+        bool | None,
+        click.option(
+            "--create-index-after-load",
+            type=bool,
+            help="Enable create index after loading vectors",
+            default=False,
+            required=False,
+            show_default=True,
+        ),
+    ]
 
 
 class PgVectorIVFFlatTypedDict(PgVectorTypedDict, IVFFlatTypedDict): ...
@@ -149,6 +171,8 @@ def PgVectorIVFFlat(
             reranking=parameters["reranking"],
             reranking_metric=parameters["reranking_metric"],
             quantized_fetch_limit=parameters["quantized_fetch_limit"],
+            create_index_before_load=parameters["create_index_before_load"],
+            create_index_after_load=parameters["create_index_after_load"],
         ),
         **parameters,
     )
@@ -185,6 +209,8 @@ def PgVectorHNSW(
             reranking=parameters["reranking"],
             reranking_metric=parameters["reranking_metric"],
             quantized_fetch_limit=parameters["quantized_fetch_limit"],
+            create_index_before_load=parameters["create_index_before_load"],
+            create_index_after_load=parameters["create_index_after_load"],
         ),
         **parameters,
     )
diff --git a/vectordb_bench/backend/clients/pgvector/config.py b/vectordb_bench/backend/clients/pgvector/config.py
index c386d75ef..ea7263e8d 100644
--- a/vectordb_bench/backend/clients/pgvector/config.py
+++ b/vectordb_bench/backend/clients/pgvector/config.py
@@ -171,6 +171,8 @@ class PgVectorIVFFlatConfig(PgVectorIndexConfig):
     reranking: bool | None = None
     quantized_fetch_limit: int | None = None
     reranking_metric: str | None = None
+    create_index_before_load: bool | None = True
+    create_index_after_load: bool | None = False
 
     def index_param(self) -> PgVectorIndexParam:
         index_parameters = {"lists": self.lists}
@@ -215,6 +217,8 @@ class PgVectorHNSWConfig(PgVectorIndexConfig):
     reranking: bool | None = None
     quantized_fetch_limit: int | None = None
     reranking_metric: str | None = None
+    create_index_before_load: bool | None = True
+    create_index_after_load: bool | None = False
 
     def index_param(self) -> PgVectorIndexParam:
         index_parameters = {"m": self.m, "ef_construction": self.ef_construction}
diff --git a/vectordb_bench/backend/clients/pgvector/pgvector.py b/vectordb_bench/backend/clients/pgvector/pgvector.py
index 4164461fb..61f030cde 100644
--- a/vectordb_bench/backend/clients/pgvector/pgvector.py
+++ b/vectordb_bench/backend/clients/pgvector/pgvector.py
@@ -374,11 +374,11 @@ def _create_table(self, dim: int):
                     "CREATE TABLE IF NOT EXISTS public.{table_name} (id BIGINT PRIMARY KEY, embedding vector({dim}));",
                 ).format(table_name=sql.Identifier(self.table_name), dim=dim),
             )
-            self.cursor.execute(
-                sql.SQL(
-                    "ALTER TABLE public.{table_name} ALTER COLUMN embedding SET STORAGE PLAIN;",
-                ).format(table_name=sql.Identifier(self.table_name)),
-            )
+            # self.cursor.execute(
+            #     sql.SQL(
+            #         "ALTER TABLE public.{table_name} ALTER COLUMN embedding SET STORAGE PLAIN;",
+            #     ).format(table_name=sql.Identifier(self.table_name)),
+            # )
             self.conn.commit()
         except Exception as e:
             log.warning(f"Failed to create pgvector table: {self.table_name} error: {e}")
diff --git a/vectordb_bench/backend/dataset.py b/vectordb_bench/backend/dataset.py
index 62700b0fa..b53dabfa9 100644
--- a/vectordb_bench/backend/dataset.py
+++ b/vectordb_bench/backend/dataset.py
@@ -202,6 +202,7 @@ def prepare(
         self,
         source: DatasetSource = DatasetSource.S3,
         filters: float | str | None = None,
+        load_train_data: bool = True,
     ) -> bool:
         """Download the dataset from DatasetSource
          url = f"{source}/{self.data.dir_name}"
@@ -210,6 +211,7 @@ def prepare(
             source(DatasetSource): S3 or AliyunOSS, default as S3
             filters(Optional[int | float | str]): combined with dataset's with_gt to
               compose the correct ground_truth file
+            load_train_data(bool): whether to download train files, default True
 
         Returns:
             bool: whether the dataset is successfully prepared
@@ -217,15 +219,19 @@ def prepare(
         """
         file_count, use_shuffled = self.data.file_count, self.data.use_shuffled
 
-        train_files = utils.compose_train_files(file_count, use_shuffled)
-        all_files = train_files
+        all_files = []
+        
+        # Only include train files if load_train_data is True
+        if load_train_data:
+            train_files = utils.compose_train_files(file_count, use_shuffled)
+            all_files.extend(train_files)
 
         gt_file, test_file = None, None
         if self.data.with_gt:
             gt_file, test_file = utils.compose_gt_file(filters), "test.parquet"
             all_files.extend([gt_file, test_file])
 
-        if not self.data.is_custom:
+        if not self.data.is_custom and all_files:
             source.reader().read(
                 dataset=self.data.dir_name.lower(),
                 files=all_files,
diff --git a/vectordb_bench/backend/task_runner.py b/vectordb_bench/backend/task_runner.py
index 2a583b4f5..de7228614 100644
--- a/vectordb_bench/backend/task_runner.py
+++ b/vectordb_bench/backend/task_runner.py
@@ -96,7 +96,9 @@ def init_db(self, drop_old: bool = True) -> None:
     def _pre_run(self, drop_old: bool = True):
         try:
             self.init_db(drop_old)
-            self.ca.dataset.prepare(self.dataset_source, filters=self.ca.filter_rate)
+            # Only download train data if LOAD stage is enabled
+            load_train_data = TaskStage.LOAD in self.config.stages
+            self.ca.dataset.prepare(self.dataset_source, filters=self.ca.filter_rate, load_train_data=load_train_data)
         except ModuleNotFoundError as e:
             log.warning(f"pre run case error: please install client for db: {self.config.db}, error={e}")
             raise e from None