Jevgpt 的 token 探索与质量提升难点
问题概述
The author built an autoregressive chatbot using a small token vocabulary and spec decoding to expand it, but struggles with limited token exploration and wants to know if the verifier pass truly improves output quality; users are curious about potential gains.
痛点标签
- 扩展 token 词汇表以实现更好探索
- 验证器 pass 对输出质量的提升有限或不确定
待验证问题
- 该项目在扩展 token 词汇表时遇到的主要技术障碍是什么？
- 验证器 pass 是否真的能显著提升输出质量？
- 用户是否愿意为此功能付费或更换现有工具？
证据
事实
felt extra autistic today and made jev an autoregressive chatbot by classifier small token vocabularies.https://github.com/giga-james/jevgpthad some trouble getting it to explore enough token options so i did a small spec decoding pass to expand token vocab (jev as drafter and verifier lol)it aint much and it aint honest work either
原话 · @Anonymous
Really interesting approach. I’m curious how much the verifier pass improves the output quality compared to just increasing the token vocabulary.
免责声明：本卡片由 AI 从社区讨论中提炼，属于「痛点证据」，是需求分析与验证的起点，不等同于「已验证的市场需求」。AI 存在分析局限，关键决策前请打开上方原始讨论人工复核。