Added new benchmark AMC23, corrected an error in AIME-2024, and removed leading zeros from AIME-2024 for improved formatting

Kipok / NeMo-Skills

A pipeline to improve skills of large language models

https://kipok.github.io/NeMo-Skills/

Apache License 2.0

185 stars 41 forks source link

Closed wedu-nvidia closed 1 month ago

wedu-nvidia commented 1 month ago

I added the benchmark amc23. Fix one error for aime-2024 for https://artofproblemsolving.com/wiki/index.php/2024_AIME_I_Problems/Problem_12. The correct answer should be 385 not 384 Remove leading zeros for aime-2024 for better formatting