Download modules/domain_knowledge/vector_store.py from recommender-system/llmproccrs: direct link, hf CLI and curl.
- Browser
- Download file 1.97 kB
-
https://huggingface.co/recommender-system/llmproccrs/resolve/main/modules/domain_knowledge/vector_store.py
- Command line
-
hf download hf://recommender-system/llmproccrs/modules/domain_knowledge/vector_store.py
-
curl -L -o vector_store.py https://huggingface.co/recommender-system/llmproccrs/resolve/main/modules/domain_knowledge/vector_store.py
1.97 kB
| import chromadb | |
| from chromadb.config import Settings | |
| import os | |
| class VectorStore: | |
| def __init__(self, persist_path: str, collection_name: str): | |
| """ | |
| Inisialisasi koneksi ke ChromaDB. | |
| """ | |
| if not os.path.exists(persist_path): | |
| os.makedirs(persist_path, exist_ok=True) | |
| print(f"[D&K] Connecting to ChromaDB at: {persist_path}") | |
| # Setup Client | |
| self.client = chromadb.PersistentClient(path=persist_path) | |
| # Get or Create Collection | |
| # metadata={"hnsw:space": "cosine"} agar similarity search akurat | |
| self.collection = self.client.get_or_create_collection( | |
| name=collection_name, | |
| metadata={"hnsw:space": "cosine"} | |
| ) | |
| print(f"[D&K] Collection '{collection_name}' ready. Total items: {self.collection.count()}") | |
| def add_documents(self, ids, documents, embeddings, metadatas): | |
| """ | |
| Menyimpan batch data ke vector database. | |
| """ | |
| try: | |
| self.collection.add( | |
| ids=ids, | |
| documents=documents, | |
| embeddings=embeddings, | |
| metadatas=metadatas | |
| ) | |
| except Exception as e: | |
| print(f"[D&K] Error adding documents: {e}") | |
| def search(self, query_embedding, n_results=10): | |
| """ | |
| Mencari dokumen berdasarkan vector embedding query. | |
| """ | |
| results = self.collection.query( | |
| query_embeddings=[query_embedding], | |
| n_results=n_results | |
| ) | |
| # Flatten results (Chroma returns lists of lists) | |
| documents = results["documents"][0] if results["documents"] else [] | |
| metadatas = results["metadatas"][0] if results["metadatas"] else [] | |
| distances = results["distances"][0] if results["distances"] else [] | |
| return documents, metadatas, distances | |
| def get_count(self): | |
| return self.collection.count() |