Getting Started With OpenAI ChatGPT (GPT-3.5) API In Python | Tutorial For Beginners

OpenAI has finally released the most anticipated ChatGPT model today, called "gpt-3.5-turbo". Unlike the GPT 3.0 models that work like a search engine, ChatGPT model tracks the conversation to create more useful outputs. In this tutorial, I will go through every single detail on how to get started with the newest OpenAI ChatGPT API in Python.

► Buy Me a Coffee? Your support is much appreciated!
-------------------------------------------------------------------------------------------
☕ Venmo: @Jie-Jenn

► Support my channel so I can continue making free contents
---------------------------------------------------------------------------------------------------------------

00:00 - Intro to OpenAI gpt 3.5 turbo model
00:34 - OpenAI GPT-3.5 Pricing difference vs GPT-3.0
01:51 - OpenAI GPT 3.5 models summary
03:03 - Sign up for an OpenAI account
03:29 - Create an OpenAI API Key
03:47 - Create Python script
04:15 - Install OpenAI Python package
04:27 - open.ChatCompletion library usage
11:00 - Track ChatGPT conversation
15:03 - Make the Python script runs like ChatGPT

#chatgpt #chatgptapi #openai #ai #python #gpt

[jiejenn]

I will be releasing more chatGPT related videos in the future, so don't forget to subscribe!

PS: I was incorrect regarding you cannot set a token limit with ChatCompletion model. In fact you can, by setting the "max_tokens" parameter.

[JGVRP]

From YouTube comment reply tutorials until now, you have kept up with the times, the generation of teenagers like me need your contribution, sir. 😊❤

[toddwmac]

Much appreciation, I decided to learn VS Code and Python (basics) a couple days ago so I could start playing with this API. Last time I cracked open any code was probably VBA (if you can call that code) Appreciate you taking the time to explain your process and some of the underlying pieces.

Update - your code worked perfectly and I know that: In the 2020 NBA Finals, 575 free throws were attempted by both teams combined over the 6 games. Thanks a ton!

[douglashilton]

I just finished building a python code that totally interactive without typing or reading text. you can use an local wakeup word(stays offline until you want to use it). then wakes up and uses ChatGPT-3.5-turbo. It remembers the previous conversations, you talk to it, and it answers back with speech. I want to add it to an interactive AI avatar if anyone can help with that. uses wireless microphone so I don't even need to be by the computer.

[kontgaming]

we need programmers like this. Thank you Jie Jenn

[HaidenSD]

I have subscribe your channel bcoz Jenn cares about his commenters : )

[GokulBalajiGaming]

thankyou so much bro you explained it very clearly.

[bnglr]

"system" is supposed to instruct the model, send it before "user" message, no need to send it as an individual request.

[DigitalSamTV]

hi, how do u set if the api uses gpt3.5 or gpt 4 ? there is no setting when you generate the key as far as I can see... please help. cheers

[HaidenSD]

Thanks Jenn it works am using Pycharm. Does token have limitation? Bcoz i want to summarize abt 50, 000 words. So how do i increase the token or bypass limitation?

[MilkstacheTV]

Today was my first day of using python and i already made a discord bot that reads messages from a specified channel, sends those messages to chat gpt, and then sends chat gpt's response back to the original channel(i know it isn't much but it's my first day come on).

[carspam_nv]

my code is not working i just copied the script and nothing is happening is there a index.html file

[senjin]

Hi, I'm successfully run your script in terminal!
however, i want to export the output, so can you maybe make another video tutorial how to use this on web browser, or maybe a gui app with parameters setting? thank you!

[toddd.]

Thanks for sharing this. Just a quick note, on some lines the word 'conversation' was misspelled as covnersation.

[leehaen]

HI, thanks your shareing. When I follow your step and block in 9:30 . The fuction show : Invalid URL (POST /v1/chat/completions). I fond the solution to this problem is to change the model to engine, but it still not work. Could you help me ? This block me 2 days.

[dragonsaige]

Hey, how are you running whatever code you select in the currently open python interpreter?

[soumojitguhamajumder]

Do we yet know, if we could retain the existing model?

Can I add or change certain layers?

[yuchengpeng]

Thank you so much! But I got his problem Additional properties are not allowed ('content:' was unexpected) - 'messages.2", could you tell me what would be the problem? Thanks a lot!

[thedarkglovemusic]

Is there anything special you need to do to get this to run just using "run pythin file"? If I run the whole code as a block it works but all hell breaks loose if I try to just click the |> button.

[JuanRamirez-dibl]

So, for a work the size of the KJB that contains about 780, 000 words, the cost would be somewhere around $1500? 🤔