open-finllm-reasoning-leaderboard / model_performance.csv
Xueqing's picture
fix: bugs
ac1edfa
raw
history blame
902 Bytes
Model,Type,finqa,dm-simplong,xbrl-math
4o,instruction-tuned,72.49,60.0,72.22
o1,instruction-tuned,49.07,56.0,74.44
o3-mini,instruction-tuned,60.87,59.0,76.67
v3,instruction-tuned,73.2,53.0,76.67
r1,instruction-tuned,65.13,53.0,86.67
deepseek-70b,instruction-tuned,66.73,53.0,86.67
llama3-70B-instruct,instruction-tuned,58.92,41.0,56.67
llama31-70B-instruct,instruction-tuned,63.18,48.0,63.33
llama33-70B-instruct,instruction-tuned,68.15,54.0,70.0
deepseek-32b,instruction-tuned,65.48,55.0,84.44
deepseek-14b,instruction-tuned,63.27,44.0,84.44
deepseek-8b,instruction-tuned,45.96,33.0,81.11
llama3 8b-instruct,instruction-tuned,41.97,29.0,48.89
llama31 8b-instruct,instruction-tuned,54.13,34.0,62.22
Qwen2.5-32B-Instruct,instruction-tuned,,,
Qwen2.5-72B-Instruct,instruction-tuned,73.38,59.0,67.78
Qwen2.5-72B-Instruct-math,instruction-tuned,69.74,42.0,83.33
Fino1-8B,instruction-tuned,60.87,40.0,82.22