[Feature] Pinecone Vector DB support (#723)
This commit is contained in:
@@ -339,7 +339,6 @@ class EmbedChain(JSONSerializable):
|
||||
metadatas = embeddings_data["metadatas"]
|
||||
ids = embeddings_data["ids"]
|
||||
new_doc_id = embeddings_data["doc_id"]
|
||||
|
||||
if existing_doc_id and existing_doc_id == new_doc_id:
|
||||
print("Doc content has not changed. Skipping creating chunks and embeddings")
|
||||
return [], [], [], 0
|
||||
@@ -404,7 +403,6 @@ class EmbedChain(JSONSerializable):
|
||||
skip_embedding=(chunker.data_type == DataType.IMAGES),
|
||||
)
|
||||
count_new_chunks = self.db.count() - chunks_before_addition
|
||||
print((f"Successfully saved {src} ({chunker.data_type}). New chunks count: {count_new_chunks}"))
|
||||
return list(documents), metadatas, ids, count_new_chunks
|
||||
|
||||
def _format_result(self, results):
|
||||
|
||||
Reference in New Issue
Block a user