Bubble io Clone Chat GPT-3.5 turbo | Full tutorial (Design + Open AI API)

Подписаться 83

Просмотров 2,2 тыс.

50% 1

Discover Chat GPT Clone on Bubble.io! 🚀
Join in creating an AI-powered chatbot using the no-code platform Bubble.io. Learn the process, harness the technology, and be part of the AI revolution!
#ChatGPT #Bubble.io #AIChatbot #NoCode #TechRevolution 🏆🔝

Наука

Опубликовано:

17 мар 2023

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 29

@user-gf2qh4ip6i Год назад

Cool

@MuhammadArslanishtiaq Год назад

Hey Nakita, Good Video. A few suggestions though: - Make the Design first then Implement the Logic. I think it'd be easier to follow. - Create meaningful names of the elements in the design tree. Keep up the good work!

@nikitanocode Год назад

Thank you for the suggestions! Will try to improve the quality of the video.

@plugk3557 Год назад

can we have the ediotr link please it was difficult to follow along

@jvanh8926 Год назад

great video, hopefully you ll keep sharing!

@nikitanocode Год назад

Thank you! I have a couple more videos that I want to record. :)

@yoshieeee Год назад

@@nikitanocode awesome 👍

@ojiemekemichaelolusegun8095 11 месяцев назад

God bless you. This is so helpful :)

@nikitanocode 10 месяцев назад

Glad it was helpful!

@BERATTMEL 20 дней назад

Hello Nikita, first of all, thank you for this video. I chose to use the Claude language model instead of GPT. Just like you did in the video, I organized and structured the JSON and workflows layout based on existing databases. When I send test messages, I receive up to 5-6 replies, but after the 7th or 8th test message, the response from Claude stops. When I check the answer and answer JSON fields in the App data section of the "Data" tab, I see that the answer is created but it is empty. I reset all data and send test messages again and the same problem occurs again after the 6th or 7th message. I use a paid Bubble account and there are no restrictions on my Claude API token. I can't see where I made a mistake. Can you help with this? Thank you.

@BERATTMEL 17 дней назад

I noticed this in the test messages I sent; When you send the same message a second time, the artificial intelligence does not respond. It gives more and more errors, and this problem only occurs when you send the same message twice. When he sends different messages, he responds to each one correctly. Can you help me how to solve this problem, friend?

@user-dw1ne2op7z 9 месяцев назад

Great video! How do you create real time streaming of content like chatgpt? any ideas?

@nikitanocode 9 месяцев назад

Hi, thank you, you would need to get respond of chat GPT firstly (you can show some animation in the meantime) and then use typewriter css animation

@monthewilfriedtutorials5423 11 месяцев назад

Great tutorial. Keep up

@nikitanocode 10 месяцев назад

Thanks!

@YNTube2 Год назад

Hi Nikita, Thank you for the great video. I created the same chat for my daughter, how would you setup the "system" in the api call to act as "a kids friendly teacher name AVA"? I want the chat to answer a certain way and have a name. Do I need to do this in the workflow or change the api call? Thanks!

@nikitanocode Год назад

Hi, thank you for the question. Simply, I can suggest using "role": "system", so basically the system message helps set the behavior of the assistant, it would like like this: [ {"role": "system", "content": "You are an AI assistant designed to provide friendly, age-appropriate, and educational answers for children."}, {"role": "user", "content": "Who won the world series in 2020?"}, {"role": "assistant", "content": "The Los Angeles Dodgers won the World Series in 2020."}, {"role": "user", "content": "Where was it played?"} ] just create logic: when someone creates chat this first system message is in the beginning by default, let me know if I explained it well.

@YNTube2 Год назад

@@nikitanocode Thank you Nikita! It worked :)

$@goodfractalspoker7179$

@goodfractalspoker7179 Год назад

Hello, do you ever receive a 404 error when you reach the max token length? I am unsure of how to reset the conversation and token count to avoid this. Thanks and great video!

@nikitanocode Год назад

Hi, could you copy the error text and send it here. Maybe the problem will be solved with the paid account if you don't have it yet (the limit will be 65k tokens). Or as an option you could create a workflow that counts the number of characters in the entire conversation that you put in the API prompt and then based on the condition that the length is less or more than the maximum take all the latest items from the dialog that are less than the maximum length. Let me know if you understood :)

@MikeWarnus 6 месяцев назад

dont stop making video bro..peace

@tobiasgr2277 9 месяцев назад

Bro thanks can you put the code un the comments that you use in the body Because i can t find them

@nikitanocode 9 месяцев назад

{ "model": "gpt-3.5-turbo", "messages": [] } to initialise API call put this part into variable: {"role": "user", "content": "Hello!"} Also make variable not private to be able to use variable "messages" dynamically

@tobiasgr2277 9 месяцев назад

@@nikitanocode thanks bro, love you