Can I Make The Next Chat GPT??? (Part 2)

preview_player
Показать описание
How would you feel if you couldn't use chat GPT?

⏱️⏱️VIDEO CHAPTERS⏱️⏱️

CHAT WITH US!
===============================

#neelstechshow #coding #programming
Рекомендации по теме
Комментарии
Автор

Bro is pulling out RNN against an instructGPT trained using PPO, you've really got some balls.

techtam
Автор

Bro are you born genius? You are super talented

SubrataDas-ylpl
Автор

If you’re actually trying to pull this off an RNN will get nowhere you need a transformer architecture and the training time needed for functional NLP models is several months on a few dozen A100s so the best option is to use bloom and offload to system ram or an nvme ssd and do some simple low shot learning to get it functioning for your task specific application

danielletemplin
Автор

You can build chat GPT chatbot for your website in minutes. Forbes recently published an article explaining it

aljosabozic
Автор

RNN's were used for language modelling like 7 years ago. using such network is not the best because it just trains really slow and gives off way worse results than for example a transformer architecture

logicon
Автор

A ban is a recommendation not a rule just know that

Balded
Автор

Its nearly impossible to train an large language model on any consumer grade laptop because of the gpu size constraints you can train small rnn, lstm model but that will underfit because of less data to train this make an model that converge soon that is it will give wrong result, now speaking of chat -gpt ita trained on over billion parameters on very large gpu what you can actually do is take online instance of gpu from services like aws to trian an llm but it will cost you vey much 😅

VedantSingh-qnwz
Автор

“Simple” and “AI” will never go together

vincentfarris
Автор

Bro my school don't even know how to use a computer, they don't no a sh*t about chatGPT 😂😂

SASA_maxillo
Автор

UPDATE: I am now using a transformer model instead of a RNN based GAN

neelbanga
Автор

Honestly you're making really great videos

vd
Автор

Dude just make a web proxy that backends to a replit/glitch/heroku server you can bypass the filters stopping you from using chatgpt

TheRaccoonFPV
Автор

Here is a simple fix which takes less than an hour. Build a simple UI run with any backend language, that then sends requests to openAI. Basically just a custom UI for chat gpt

abhimanyusareen
Автор

Dude awesome at an young age!! Train me if you can !! I wanted to learn AI

venwal-qtff
Автор

I’ve been getting your shorts on recommendations and honestly all of them sounds so fkin fake. Keep up the good work if youre not faking it. If you even managed to create a model something thats even 5% comparable to chatgpt might even cost a billion dollar

feloi
Автор

This man is going to cure cancer one day calling it now

kieki
Автор

Welp I hope you have some amazing compute power. If not then just offload with ALPACA 😅

sarthakjain
Автор

Just run a instance of LLaMA or gpt4all and you good

tocraft
Автор

How would I feel? I wouldn't know how to.😅

elw
Автор

Chat GPT, which is openAI has some of the worlds smartest people working on it and billions of dollars in backing from companies like Microsoft. to think you could build anything remotley close in 30 mins is cringe.

Slaughter