https://github.com/egorsmkv/qdrant-ukrainian

Some experiments with Qdrant for Ukrainian search using embeddings
https://github.com/egorsmkv/qdrant-ukrainian

qdrant ukrainian vector-search

Last synced: 25 days ago
JSON representation

Some experiments with Qdrant for Ukrainian search using embeddings

• Host: GitHub
• URL: https://github.com/egorsmkv/qdrant-ukrainian
• Owner: egorsmkv
• Created: 2024-07-04T11:57:22.000Z (4 months ago)
• Default Branch: master
• Last Pushed: 2024-07-09T18:36:57.000Z (4 months ago)
• Last Synced: 2024-07-09T23:17:07.358Z (4 months ago)
• Topics: qdrant, ukrainian, vector-search
• Language: Python
• Homepage:
• Size: 401 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Test Qdrant vector similarity search engine for Ukrainian

Dataset:



Search result:



---

Tested with Qdrant 1.10.0

## Install deps

### Create an env

```

uv venv --python 3.12

```

### Activate the env

```

source .venv/bin/activate

```

### Install libraries

```

uv pip install qdrant-client sentence-transformers polars sentencepiece

```

optional:

```

uv pip install ruff

```

## Run

### Start qdrant server

```

docker run -p 6333:6333 qdrant/qdrant

```

### Download dataset and convert from Parquet to JSON

```

python download_dataset.py

```

### Embedd docs

```

python embedd_docs.py

```

Also, you can see them in the UI: http://localhost:6333/dashboard

### Search docs

```

python search_docs.py

```

## Lint and format the code

```

ruff check

ruff format

```