Thread Reader
Chuangbo Li
@chuangbo
Mar 3
2 tweets
Twitter
Share
Translate
Download
你是否因为 ChatGPT 无法“学习”超过 4096 个 token 的上下文而感到困扰?如果是的话,不妨看看这个开源例子。它的基本原理是,通过 Embedding 模型和数据库在大量物料中搜索可能与用户回答相关的段落,然后从这些段落中生成 prompt,以便 ChatGPT 进行回答。
Mckay Wrigley
@mckaywrigley
Mar 02
View on Twitter
I embedded all of
@Paul Graham
’s essays. All 605,870 tokens worth. Use OpenAI’s new model to search & chat with them at
paul-graham-gpt.vercel.app
. Code & dataset is 100% open-source for anyone to use. GitHub:
github.com/mckaywrigley/p…
Show this thread
另外,还推荐大家参考 OpenAI 官方提供的一系列实用的例子和指南,其中包括了各种场景和功能的演示和代码,让你轻松上手 OpenAI 的强大能力。
github.com/openai/openai-…
github.com/openai/openai-…
GitHub - openai/openai-cookbook: Examples and guides for using the OpenAI API
Chuangbo Li
@chuangbo
🇳🇿 NZ / 💻 Dev / 🐈⬛ Cat / 🏂 Snowboarding / WeChat #802 / V2EX #584 / 祝愿世界和平,生生不息,繁荣昌盛。Live long and prosper. 🖖
Follow on Twitter
twitter-thread.com/t/1631461656151887873
Missing some tweets in this thread? Or failed to load images or videos? You can try to
force a refresh
.