【線上會(huì)議】推測 GPT-4 新的視覺+文本模型結(jié)構(gòu)、訓(xùn)練數(shù)據(jù)和訓(xùn)練成本
https://www.meetup.com/speech-and-language-technology-meetup-group/events/292587691
RSVP and see the zoom link if you want know know what OpenAI people possibly did in the new GPT-4, since all other paper readers (no offense) would have no clue since GPT-4's tech report did not mention a word about the new Vision+NLP model architecure, training data, and cost to train. 9 pm PDT this Friday. [Had been critizied for not sticking to the exact content during paper reading and talked too much about 'I think/guess' -- now I THINK/GUESS my time has finally come ?? ]?#gpt4?#openai?#paperpresentation
請點(diǎn)擊 https://www.meetup.com/speech-and-language-technology-meetup-group/events/292587691，RSVP 并查看 Zoom 鏈接，如果您想知道 OpenAI 的人在新的 GPT-4 中可能做了什么，因?yàn)樗衅渌恼撐淖x者（無意冒犯）都不會(huì)知道 GPT-4 的技術(shù)報(bào)告沒有提到關(guān)于新的 Vision+NLP 模型架構(gòu)、訓(xùn)練數(shù)據(jù)和訓(xùn)練成本的任何字眼。本次講座將于本周五晚上 9 點(diǎn)太平洋夏令時(shí)舉行。[曾因不按照論文閱讀內(nèi)容而被批評，并談?wù)撎嘧约旱牟聹y和想法，現(xiàn)在我想我的時(shí)機(jī)終于到了 ??] #gpt4 #openai #paperpresentation
