TBNilles/LEANN
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix some para

Browse Source
This commit is contained in:
yichuan520030910320
2025-07-23 01:46:34 -07:00
parent ae16d9d888
commit 851f0f04c3
2 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/wechat_history_reader_leann.py
View File

@@ -78,7 +78,7 @@ def create_leann_index_from_multiple_wechat_exports(
) )
# Create text splitter with 256 chunk size # Create text splitter with 256 chunk size
text_splitter = SentenceSplitter(chunk_size=256, chunk_overlap=128) text_splitter = SentenceSplitter(chunk_size=192, chunk_overlap=64)
# Convert Documents to text strings and chunk them # Convert Documents to text strings and chunk them
all_texts = [] all_texts = []

2
packages/leann-core/src/leann/embedding_compute.py
View File

@@ -101,7 +101,7 @@ def compute_embeddings_sentence_transformers(
if device == "mps": if device == "mps":
batch_size = 128 # MPS optimal batch size from benchmark batch_size = 128 # MPS optimal batch size from benchmark
if model_name == "Qwen/Qwen3-Embedding-0.6B": if model_name == "Qwen/Qwen3-Embedding-0.6B":
batch_size = 64 batch_size = 32
elif device == "cuda": elif device == "cuda":
batch_size = 256 # CUDA optimal batch size batch_size = 256 # CUDA optimal batch size
# Keep original batch_size for CPU # Keep original batch_size for CPU