config.py

# Embeddings Configuration

EMBEDDING_MODEL_NAME = "sentence-transformers/all-MiniLM-L6-v2"
# Document Processing Configuration
CHUNK_SIZE = 8192
CHUNK_OVERLAP = 200
BATCH_SIZE = 1000

# Retrieval Configuration
RETRIEVER_K = 3

# File Paths
JSON_FILE_PATH = "index_training.json"
VECTORSTORE_PATH = "vectorstore"