List your company

Question Answering on TQ (F1)

80.6F1 ScoreIGPO
66.97670.51374.0577.587Oct 16, 2025

Evaluation results

MethodLinks
IGPOBackbone=Qwen2.5-7B-Instruct2025.1080.6
GSPOBackbone=Qwen2.5-7B-Instruct2025.1077.7
GRPOBackbone=Qwen2.5-7B-Instruct2025.1077
PPOBackbone=Qwen2.5-7B-Instruct2025.1075.4
RLOOBackbone=Qwen2.5-7B-Instruct2025.1072.5
Reinforce++Backbone=Qwen2.5-7B-Instruct2025.1067.5