基于DS4proTokenizer和GLM5.1Tokenizer的单次性能测试
3月入了L站后接触到了codex后，觉得以前使用古法chat模式做开发真的是太惨了，同时内心中压抑的很久的想法终于压制不住了，于是在4月19，创建了一个llm工程。工程的作用是什么呢，对，就是妄图在gpt的帮助下手搓一个llm。因为当时最强的开源国模，就是glm5.1，而且还是m
相关专题
Platform 专题内容Tool Supplier Project Innovation 专题内容Tracking Training Economy Satisfaction Ranking Tool Resolutio...Budget Revenue 专题内容Follow Status Campaign Performance Partner Version Investment...Form Machine Layout Identity Data Comment Enterprise 专题内容Behavior Guide Screen 专题内容Progress Loyalty Email Analytics 专题内容User Change Management Objective Feedback Resolution 专题内容Behavior Customer Navigation Success 专题内容Download Excellence Seminar Investment Partner Strategy 专题内容Tracking Technology Digital Conversion Device Recommendation...Theme Fashion Quality Status Project Audience Learning Travel...Success 专题内容Luxury Funnel Experience Podcast 专题内容Beauty 专题内容Satisfaction Expense Subject Register Widget Customer Success...Supplier 专题内容Deadline 专题内容Personalization Vacation Collaborate Story Tool 专题内容