Performance of fine-tuned GPT-4o mini on various tasks compared to previous SOTA.
收藏数据链接：
官方服务：
资源简介：
For sentiment analysis and news categorization, the metric used is macro-F1. For sarcasm detection, F1-score is used. ‘Ft.’ in the table refers to fine-tuned.
创建时间：
2025-10-08
