This is an automated email from the ASF dual-hosted git repository.
ming pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/incubator-hugegraph-ai.git
The following commit(s) were added to refs/heads/main by this push:
new deced7e Add progress bars and print information for building vector
indexes (#74)
deced7e is described below
commit deced7e627dd4aecf36ed0e9c6f44a989c28f2cf
Author: vichayturen <[email protected]>
AuthorDate: Tue Aug 27 12:39:28 2024 +0800
Add progress bars and print information for building vector indexes (#74)
Co-authored-by: imbajin <[email protected]>
---
.../src/hugegraph_llm/operators/index_op/build_semantic_index.py | 3 ++-
.../src/hugegraph_llm/operators/index_op/build_vector_index.py | 5 ++++-
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git
a/hugegraph-llm/src/hugegraph_llm/operators/index_op/build_semantic_index.py
b/hugegraph-llm/src/hugegraph_llm/operators/index_op/build_semantic_index.py
index a439d46..2582653 100644
--- a/hugegraph-llm/src/hugegraph_llm/operators/index_op/build_semantic_index.py
+++ b/hugegraph-llm/src/hugegraph_llm/operators/index_op/build_semantic_index.py
@@ -19,6 +19,7 @@
import os
from typing import Any, Dict
+from tqdm import tqdm
from hugegraph_llm.config import resource_path, settings
from hugegraph_llm.models.embeddings.base import BaseEmbedding
from hugegraph_llm.indices.vector_index import VectorIndex
@@ -36,7 +37,7 @@ class BuildSemanticIndex:
log.debug("Building vector index for %s vertices...",
len(context["vertices"]))
vids = []
vids_embedding = []
- for vertex in context["vertices"]:
+ for vertex in tqdm(context["vertices"]):
vertex_text = f"{vertex['label']}\n{vertex['properties']}"
vids_embedding.append(self.embedding.get_text_embedding(vertex_text))
vids.append(vertex["id"])
diff --git
a/hugegraph-llm/src/hugegraph_llm/operators/index_op/build_vector_index.py
b/hugegraph-llm/src/hugegraph_llm/operators/index_op/build_vector_index.py
index 51f7bec..1fccf15 100644
--- a/hugegraph-llm/src/hugegraph_llm/operators/index_op/build_vector_index.py
+++ b/hugegraph-llm/src/hugegraph_llm/operators/index_op/build_vector_index.py
@@ -19,9 +19,11 @@
import os
from typing import Dict, Any
+from tqdm import tqdm
from hugegraph_llm.config import settings, resource_path
from hugegraph_llm.indices.vector_index import VectorIndex
from hugegraph_llm.models.embeddings.base import BaseEmbedding
+from hugegraph_llm.utils.log import log
class BuildVectorIndex:
@@ -37,7 +39,8 @@ class BuildVectorIndex:
raise ValueError("chunks not found in context.")
chunks = context["chunks"]
chunks_embedding = []
- for chunk in chunks:
+ log.debug("Building vector index for %s chunks...",
len(context["chunks"]))
+ for chunk in tqdm(chunks):
chunks_embedding.append(self.embedding.get_text_embedding(chunk))
if len(chunks_embedding) > 0:
if os.path.exists(self.index_file) and
os.path.exists(self.content_file):
