kanji_lookup / config.py
etrotta's picture
First Release
be12cc9
raw
history blame
723 Bytes
import os
qdrant_location = os.getenv('QDRANT_URL', "localhost")
qdrant_api_key = os.getenv('QDRANT_API_KEY')
description = """This is a Kanji image search demo. Draw or upload an image of an individual Kanji character."""
article = """
### About this project
You can find the source code as well as more information in https://github.com/etrotta/kanji_lookup
It uses the "kha-white/manga-ocr-base" ViT Encoder model to create embeddings, then uses a vector database (qdrant) to find similar characters.
The vector database has been populated with over 10k characters from [The KANJIDIC project](https://www.edrdg.org/wiki/index.php/KANJIDIC_Project), each rendered in multiple fonts downloaded from Google Fonts
"""