topic-benchmark 0.2.2

Benchmarking topic models for a paper

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"