List your company

Question Answering on TriviaQA (F1 score)

39.37F1 ScoreGPT-4o (2024-11-20)
11.726818.903426.0833.2566Aug 8, 2025

Evaluation results

MethodLinks
GPT-4o (2024-11-20)Version=2024-11-202025.0839.37
Llama + LoRRBackbone=Llama3.2-3B, Optimization=LoRR2025.0816.56
Llama + DPOBackbone=Llama3.2-3B, Optimization=DPO2025.0816.23
Qwen (Base)Backbone=Qwen2.5-3B, Mode=Base2025.0814.28
Qwen + LoRRBackbone=Qwen2.5-3B, Optimization=LoRR2025.0814.08
Qwen + DPOBackbone=Qwen2.5-3B, Optimization=DPO2025.0813.67
Llama (Base)Backbone=Llama3.2-3B, Mode=Base2025.0812.79