Comparative analysis of Weaviate and Milvus vector databases using a dataset of over 1K resumes. Includes build time, search time, and throughput benchmarks with varying data and query loads.
To replicate this analysis or to conduct your own, follow these steps:
- Clone this repository.
- Ensure you have Python installed, along with the libraries listed in
requirements.txt
. - Set up instances of Weaviate and Milvus using weaviate/run.sh and milvus/run.sh respectively.
- Run the benchmarking scripts to perform the tests.