fix: Restore embedding-mode parameter to all examples
- All examples now have --embedding-mode parameter (unified interface benefit) - Default is 'sentence-transformers' (consistent with original behavior) - Users can now use OpenAI or MLX embeddings with any data source - Maintains functional equivalence with original scripts
This commit is contained in:
@@ -1,6 +1,6 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
Debug script to test ZMQ communication with the exact same setup as main_cli_example.py
|
||||
Debug script to test ZMQ communication with embedding models.
|
||||
"""
|
||||
|
||||
import sys
|
||||
@@ -13,9 +13,9 @@ from leann_backend_diskann import embedding_pb2
|
||||
|
||||
|
||||
def test_zmq_with_same_model():
|
||||
print("=== Testing ZMQ with same model as main_cli_example.py ===")
|
||||
print("=== Testing ZMQ with embedding model ===")
|
||||
|
||||
# Test the exact same model that main_cli_example.py uses
|
||||
# Test with a common embedding model
|
||||
model_name = "sentence-transformers/all-mpnet-base-v2"
|
||||
|
||||
# Start server with the same model
|
||||
|
||||
Reference in New Issue
Block a user