This is an automated email from the ASF dual-hosted git repository.
jin pushed a commit to branch graph-query
in repository https://gitbox.apache.org/repos/asf/incubator-hugegraph-ai.git
The following commit(s) were added to refs/heads/graph-query by this push:
new f3c27aa refactor(llm): enhance a string of graph query method
f3c27aa is described below
commit f3c27aaa5cfe2318f863afd534c2e3b2b5076df6
Author: imbajin <[email protected]>
AuthorDate: Fri Sep 27 23:30:26 2024 +0800
refactor(llm): enhance a string of graph query method
---
.../src/hugegraph_llm/demo/gremlin_generate_web_demo.py | 4 ++--
.../src/hugegraph_llm/demo/rag_demo/configs_block.py | 8 ++++----
.../hugegraph_llm/operators/hugegraph_op/graph_rag_query.py | 12 ++++++++----
3 files changed, 14 insertions(+), 10 deletions(-)
diff --git a/hugegraph-llm/src/hugegraph_llm/demo/gremlin_generate_web_demo.py
b/hugegraph-llm/src/hugegraph_llm/demo/gremlin_generate_web_demo.py
index 1ba7aba..6166321 100644
--- a/hugegraph-llm/src/hugegraph_llm/demo/gremlin_generate_web_demo.py
+++ b/hugegraph-llm/src/hugegraph_llm/demo/gremlin_generate_web_demo.py
@@ -91,7 +91,7 @@ if __name__ == '__main__':
]
else:
llm_config_input = []
- llm_config_button = gr.Button("apply configuration")
+ llm_config_button = gr.Button("Apply Configuration")
def apply_configuration(arg1, arg2, arg3, arg4):
llm_option = settings.llm_type
@@ -139,7 +139,7 @@ if __name__ == '__main__':
]
else:
embedding_config_input = []
- embedding_config_button = gr.Button("apply configuration")
+ embedding_config_button = gr.Button("Apply Configuration")
def apply_configuration(arg1, arg2, arg3):
embedding_option = settings.embedding_type
diff --git a/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/configs_block.py
b/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/configs_block.py
index c3ef73c..b289482 100644
--- a/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/configs_block.py
+++ b/hugegraph-llm/src/hugegraph_llm/demo/rag_demo/configs_block.py
@@ -192,7 +192,7 @@ def create_configs_block():
gr.Textbox(value=settings.graph_pwd, label="pwd",
type="password"),
gr.Textbox(value=settings.graph_space,
label="graphspace(Optional)"),
]
- graph_config_button = gr.Button("Apply config")
+ graph_config_button = gr.Button("Apply Configuration")
graph_config_button.click(apply_graph_config, inputs=graph_config_input)
# pylint: disable=no-member
with gr.Accordion("2. Set up the LLM.", open=False):
@@ -227,7 +227,7 @@ def create_configs_block():
]
else:
llm_config_input = []
- llm_config_button = gr.Button("apply configuration")
+ llm_config_button = gr.Button("Apply Configuration")
llm_config_button.click(apply_llm_config, inputs=llm_config_input)
# pylint: disable=no-member
with gr.Accordion("3. Set up the Embedding.", open=False):
@@ -262,7 +262,7 @@ def create_configs_block():
else:
embedding_config_input = []
- embedding_config_button = gr.Button("apply configuration")
+ embedding_config_button = gr.Button("Apply Configuration")
# Call the separate apply_embedding_configuration function here
embedding_config_button.click( # pylint: disable=no-member
@@ -299,7 +299,7 @@ def create_configs_block():
]
else:
reranker_config_input = []
- reranker_config_button = gr.Button("apply configuration")
+ reranker_config_button = gr.Button("Apply Configuration")
# TODO: use "gr.update()" or other way to update the config in
time (refactor the click event)
# Call the separate apply_reranker_configuration function here
diff --git
a/hugegraph-llm/src/hugegraph_llm/operators/hugegraph_op/graph_rag_query.py
b/hugegraph-llm/src/hugegraph_llm/operators/hugegraph_op/graph_rag_query.py
index eb90684..9feb48b 100644
--- a/hugegraph-llm/src/hugegraph_llm/operators/hugegraph_op/graph_rag_query.py
+++ b/hugegraph-llm/src/hugegraph_llm/operators/hugegraph_op/graph_rag_query.py
@@ -30,8 +30,9 @@ VERTEX_QUERY_TPL = "g.V({keywords}).as('subj').toList()"
ID_QUERY_NEIGHBOR_TPL = """
g.V({keywords}).as('subj')
.repeat(
- bothE({edge_labels}).as('rel').otherV().as('obj')
-).times({max_deep})
+
bothE({edge_labels}).limit({edge_limit}).as('rel').otherV().dedup().as('obj')
+).times({max_deep}).emit()
+.simplePath()
.path()
.by(project('label', 'id', 'props')
.by(label())
@@ -71,7 +72,7 @@ g.V().has('{prop}', within({keywords})).as('subj')
class GraphRAGQuery:
- def __init__(self, max_deep: int = 2, max_items: int = 30, prop_to_match:
Optional[str] = None):
+ def __init__(self, max_deep: int = 2, max_items: int = 20, prop_to_match:
Optional[str] = None):
self._client = PyHugeClient(
settings.graph_ip,
settings.graph_port,
@@ -112,6 +113,8 @@ class GraphRAGQuery:
_, edge_labels = self._extract_labels_from_schema()
edge_labels_str = ",".join("'" + label + "'" for label in edge_labels)
+ # TODO: enhance the limit logic later
+ edge_limit_amount = len(edge_labels) * 10
use_id_to_match = self._prop_to_match is None
if use_id_to_match:
@@ -126,8 +129,9 @@ class GraphRAGQuery:
gremlin_query = ID_QUERY_NEIGHBOR_TPL.format(
keywords=match_vids,
max_deep=self._max_deep,
- max_items=self._max_items,
edge_labels=edge_labels_str,
+ edge_limit=edge_limit_amount,
+ max_items=self._max_items,
)
log.debug("Kneighbor query: %s", gremlin_query)
