Skip to content

armgit / LEANN

Public Imported from GitHub
0

[MLsys2026 Best Paper]: https://arxiv.org/abs/2506.08276. RAG on Everything with LEANN. Enjoy 97% storage savings while running a fast, accurate, and 100% private RAG application on your personal device.

add progress bar in build

yichuan520030910320530f6e4af520
diff --git a/examples/wechat_history_reader_leann.py b/examples/wechat_history_reader_leann.py
index e15e876..49e04a1 100644
--- a/examples/wechat_history_reader_leann.py
+++ b/examples/wechat_history_reader_leann.py
@@ -224,7 +224,7 @@ async def query_leann_index(index_path: str, query: str):
query,
top_k=20,
recompute_beighbor_embeddings=True,
- complexity=128,
+ complexity=32,
beam_width=1,
llm_config={
"type": "openai",
@@ -252,7 +252,7 @@ async def main():
parser.add_argument(
"--index-dir",
type=str,
- default="./wechat_history_june19_test",
+ default="./wechat_history_magic_test",
help="Directory to store the LEANN index (default: ./wechat_history_index_leann_test)",
)
parser.add_argument(
diff --git a/packages/leann-backend-hnsw/third_party/faiss b/packages/leann-backend-hnsw/third_party/faiss
index ff22e2c..2547df4 160000
--- a/packages/leann-backend-hnsw/third_party/faiss
+++ b/packages/leann-backend-hnsw/third_party/faiss
@@ -1 +1 @@
-Subproject commit ff22e2c86be1784c760265abe146b1ab0db90ebe
+Subproject commit 2547df4377ae097e2eabc9b019c15135b1fea2b4
diff --git a/packages/leann-core/src/leann/api.py b/packages/leann-core/src/leann/api.py
index 6f4c536..75a6d2b 100644
--- a/packages/leann-core/src/leann/api.py
+++ b/packages/leann-core/src/leann/api.py
@@ -20,6 +20,7 @@ def compute_embeddings(
mode: str = "sentence-transformers",
use_server: bool = True,
port: Optional[int] = None,
+ is_build=False,
) -> np.ndarray:
"""
Computes embeddings using different backends.
@@ -51,6 +52,7 @@ def compute_embeddings(
chunks,
model_name,
mode=mode,
+ is_build=is_build,
)
@@ -209,6 +211,7 @@ class LeannBuilder:
self.embedding_model,
self.embedding_mode,
use_server=False,
+ is_build=True,
)
string_ids = [chunk["id"] for chunk in self.chunks]
current_backend_kwargs = {**self.backend_kwargs, "dimensions": self.dimensions}
diff --git a/packages/leann-core/src/leann/embedding_compute.py b/packages/leann-core/src/leann/embedding_compute.py
index 20cef9f..c41f009 100644
--- a/packages/leann-core/src/leann/embedding_compute.py
+++ b/packages/leann-core/src/leann/embedding_compute.py
@@ -13,7 +13,7 @@ logger = logging.getLogger(__name__)
def compute_embeddings(
- texts: List[str], model_name: str, mode: str = "sentence-transformers"
+ texts: List[str], model_name: str, mode: str = "sentence-transformers",is_build: bool = False
) -> np.ndarray:
"""
Unified embedding computation entry point
@@ -27,7 +27,7 @@ def compute_embeddings(
Normalized embeddings array, shape: (len(texts), embedding_dim)
"""
if mode == "sentence-transformers":
- return compute_embeddings_sentence_transformers(texts, model_name)
+ return compute_embeddings_sentence_transformers(texts, model_name, is_build=is_build)
elif mode == "openai":
return compute_embeddings_openai(texts, model_name)
elif mode == "mlx":
@@ -42,6 +42,7 @@ def compute_embeddings_sentence_transformers(
use_fp16: bool = True,
device: str = "auto",
batch_size: int = 32,
+ is_build: bool = False,
) -> np.ndarray:
"""
Compute embeddings using SentenceTransformer
@@ -133,7 +134,7 @@ def compute_embeddings_sentence_transformers(
embeddings = model.encode(
texts,
batch_size=batch_size,
- show_progress_bar=False, # Don't show progress bar in server environment
+ show_progress_bar=is_build, # Don't show progress bar in server environment
convert_to_numpy=True,
normalize_embeddings=False, # Keep consistent with original API behavior
device=device,