[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.
Python aifaissgpt-osslangchainllama-indexllmlocalstorageoffline-firstollamaprivacypythonragretrieval-augmented-generationvector-databasevector-searchvectors
rm useless
yichuan52003091032095a653993a9a
diff --git a/examples/main_cli_example.py b/examples/main_cli_example.pyindex 262fa94..e89e1ee 100644--- a/examples/main_cli_example.py+++ b/examples/main_cli_example.py@@ -23,7 +23,7 @@ file_extractor: dict[str, BaseReader] = {".xlsx": reader,}node_parser = DoclingNodeParser(- chunker=HybridChunker(tokenizer="Qwen/Qwen3-Embedding-4B", max_tokens=64)+ chunker=HybridChunker(tokenizer="Qwen/Qwen3-Embedding-4B", max_tokens=128))print("Loading documents...")documents = SimpleDirectoryReader(diff --git a/packages/leann-core/src/leann/api.py b/packages/leann-core/src/leann/api.pyindex ed6af41..48478bf 100644--- a/packages/leann-core/src/leann/api.py+++ b/packages/leann-core/src/leann/api.py@@ -32,6 +32,7 @@ def _compute_embeddings(chunks: List[str], model_name: str) -> np.ndarray:else:from sentence_transformers import SentenceTransformermodel = SentenceTransformer(model_name)+ model = model.half()print(f"INFO: Computing embeddings for {len(chunks)} chunks using SentenceTransformer model '{model_name}'...")embeddings = model.encode(chunks, show_progress_bar=True)