Created
July 21, 2024 16:33
-
-
Save idontcalculate/f28a225dbb6749e6e5f1c9a175c327f1 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| import os | |
| import csv | |
| from concurrent.futures import ThreadPoolExecutor | |
| from qdrant_client import QdrantClient | |
| from qdrant_client.http.models import PointStruct | |
| from fastembed import TextEmbedding | |
| from dotenv import load_dotenv | |
| # Load environment variables from a .env file | |
| load_dotenv() | |
| # Access the environment variables | |
| QDRANT_URL = os.getenv('QDRANT_URL') | |
| QDRANT_API_KEY = os.getenv('QDRANT_API_KEY') | |
| CSV_FILE_PATH = '/mnt/c/Users/SUPERADMIN/book-hunt-engine/data/books.csv' | |
| COLLECTION_NAME = 'book-lib' # Set the new collection name | |
| # Print out the environment variables to debug | |
| print(f"QDRANT_URL: {QDRANT_URL}") | |
| print(f"QDRANT_API_KEY: {QDRANT_API_KEY}") | |
| print(f"CSV_FILE_PATH: {CSV_FILE_PATH}") | |
| # Initialize Qdrant client | |
| qdrant_client = QdrantClient( | |
| url=QDRANT_URL, | |
| api_key=QDRANT_API_KEY | |
| ) | |
| # Initialize the FastEmbed model | |
| embedding_model = TextEmbedding(model_name="BAAI/bge-base-en") | |
| # Function to generate embedding | |
| def generate_embedding(book_data): | |
| text = f"{book_data['title']} {book_data['description']}" | |
| embeddings = list(embedding_model.embed([text])) | |
| return embeddings[0] | |
| # Function to prepare point data | |
| def prepare_point(row): | |
| try: | |
| # Validate and convert data | |
| doc_id = int(row["isbn13"]) | |
| average_rating = float(row["average_rating"]) if row["average_rating"] else None | |
| published_year = int(row["published_year"]) if row["published_year"] else None | |
| authors = row["authors"].split(", ") if row["authors"] else [] | |
| vector = generate_embedding(row) | |
| payload = { | |
| "title": row["title"], | |
| "categories": row["categories"], | |
| "thumbnail": row["thumbnail"], | |
| "description": row["description"], | |
| "average_rating": average_rating, | |
| "published_year": published_year, | |
| "authors": authors | |
| } | |
| return PointStruct( | |
| id=doc_id, | |
| vector=vector, | |
| payload=payload | |
| ) | |
| except (ValueError, KeyError) as e: | |
| # Skip rows with invalid data | |
| print(f"Skipping row due to error: {e}") | |
| return None | |
| # Function to upsert points in batch | |
| def upsert_points_batch(points_batch): | |
| valid_points = [point for point in points_batch if point is not None] | |
| if valid_points: | |
| qdrant_client.upsert( | |
| collection_name=COLLECTION_NAME, | |
| points=valid_points | |
| ) | |
| # Main function to process CSV and upsert data | |
| def main(): | |
| points_batch = [] | |
| batch_size = 100 # Adjust the batch size as needed | |
| if CSV_FILE_PATH is None: | |
| print("CSV_FILE_PATH is not set. Exiting.") | |
| return | |
| if not os.path.isfile(CSV_FILE_PATH): | |
| print(f"CSV file does not exist at path: {CSV_FILE_PATH}") | |
| return | |
| with open(CSV_FILE_PATH, newline='', encoding='utf-8') as csvfile: | |
| reader = csv.DictReader(csvfile) | |
| with ThreadPoolExecutor(max_workers=4) as executor: # Adjust the number of workers as needed | |
| for row in reader: | |
| point = prepare_point(row) | |
| points_batch.append(point) | |
| if len(points_batch) >= batch_size: | |
| executor.submit(upsert_points_batch, points_batch) | |
| points_batch = [] | |
| if points_batch: | |
| executor.submit(upsert_points_batch, points_batch) | |
| if __name__ == "__main__": | |
| main() | |
| print("Data ingestion completed successfully.") |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment