Benchmarking topic models for a paper
Project description
topic-benchmark
Just Benchmarking Topic Models :)
Todo:
- Run benchmark with these models and upload the results:
- [ x ] all-MiniLM-L6-v2
- all-mpnet-base-v2 ⌛
- sentence-transformers/average_word_embeddings_glove.6B.300d ⌛
- intfloat/e5-large-v2 (OR intfloat/multilingual-e5-large-instruct, to my knowledge, they are the same size, but this one performs way better on MTEB)
- [ x ] Implement pretty printing and formatting to Latex and MD tables for results.
- [ x ] (Maybe) Implement speed tracking.
Usage:
pip install topic-benchmark
python3 -m topic_benchmark run -e "embedding_model_name"
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
topic_benchmark-0.2.2.tar.gz
(9.8 kB
view hashes)
Built Distribution
Close
Hashes for topic_benchmark-0.2.2-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 3866250082fddb7215785063db2c2f36340c2e473d90e822c12a0e49e87c5843 |
|
MD5 | 6e42d60295e9cf088bb65ea5334d2b29 |
|
BLAKE2b-256 | 5086333ad68ea8b9ec7b1dcaacf6f22ed1d7e81577a46d40ff55a048551b358b |