1 
                    
                    crokily      2023-02-23 19:21:36 +08:00    3 周前传上 github 的项目 
                7 天前开始发了几篇公关稿,话术是“超越 GPT3.5” 现在好像没啥热度,没搜到相关的讨论 不知道表现咋样  | 
            
     2 
                    
                    eason1874      2023-02-23 19:23:39 +08:00 
                    
                    吊打?真的吗?我不信 
                 | 
            
     3 
                    
                    crokily      2023-02-23 19:28:23 +08:00 
                    
                    
                 | 
            
     5 
                    
                    lovestudykid      2023-02-24 00:44:04 +08:00 
                    
                    >fine-tuning a smaller model with task-specific datasets (rather than feeding in-context learning demos to a larger generalist LLM) is exactly what we wanted in this study  
                一句话总结,不要拿去跟 chatgpt 比  |