Updated results for Polish tasks

#22
No description provided.
rafalposwiata changed pull request status to open
Massive Text Embedding Benchmark org

Great! I made a few changes:

  • Readded some other language scores that were accidentally removed
  • Named the MSMARCO-PL scores back to validation. The conversion to test is handled in the results.py file of this repository. It is better to have the correct split names in the results json file.
Muennighoff changed pull request status to merged

Thank you for the corrections and my apologies.
Unfortunately these errors still occur in multilingual-e5 models. I will correct it and make a commit.

Massive Text Embedding Benchmark org

Oh sorry I had missed them - let me fix it!

Massive Text Embedding Benchmark org

Should be fixed now! Will update the LB!

Thank you for everything! I will pay more attention to these multilingual tasks in the future.

Massive Text Embedding Benchmark org

No worries! Really amazing job on PL-MTEB and the models 😊 Happy to help in case you plan to write a paper on your work.

Sign up or log in to comment