How to use the ChatGPT API with Python!!
ฝัง
- เผยแพร่เมื่อ 5 มี.ค. 2023
- Get notified of the free Python course on the home page at www.coursesfromnick.com
Sign up for the Full Stack course here and use TH-cam50 to get 50% off:
www.coursesfromnick.com/bundl...
Code: github.com/nicknochnack/ChatG...
Hopefully you enjoyed this video.
💼 Find AWESOME ML Jobs: www.jobsfromnick.com
Oh, and don't forget to connect with me!
LinkedIn: bit.ly/324Epgo
Facebook: bit.ly/3mB1sZD
GitHub: bit.ly/3mDJllD
Patreon: bit.ly/2OCn3UW
Join the Discussion on Discord: bit.ly/3dQiZsV
Happy coding!
Nick
P.s. Let me know how you go and drop a comment if you need a hand!
#machinelearning #python #datascience - วิทยาศาสตร์และเทคโนโลยี
One thing that's nice to add is streaming, so that the words show up as they are being generated. This avoids the longer wait times for lengthy responses. Also, I've noticed that openai is still working out the kinks with their api's and some days you'll get a pretty high failure rate and have to retry a response. With streaming, you can set a reasonable timeout to try again ( usually if the first word hasn't arrived within 3 seconds, it's going to fail and hang indefinitely). If you don't have streaming, you have to wait as long as your longest possible response to retry which could be more than 10 seconds in some cases. Getting into the minutiae a little bit, but just some things I've discovered in my travels.
Hi there! Can you tell me how to add streaming as you said above?!
does this include getting responses like "permissions denied"?
Awesome!
ChatGPT has been AMAZING the experience for me was priceless
You re a lifesaver, I’ll be using it for my school project thanks 🙌
What do you mean recently? The API has been available for years...
I'm scraping all of the chat and teams log data from my company and training gpt with it. I will build a UI for my team members so that our private technical business questions that are asked frequently can be answered quickly.
I wish you would add more data..
how u implemented
usecase: training using private data for private AI requires an instance of a model
workflow: prompt + private data >chatbot >model (GPT LLM) >completion
unless u have a local instance of model (GPT LLM) then your data is still going to the internet (to OpenAi 's GPT instance)...
feb2024: your post is 10 months old (2023) I don't think running local model instances (llama, Gemini, GPT) were available to download & install locally
Thank's for letting us know that openAI have released API for chatGPT. I wonder how many of my students already have a account
Anytime @Yrgen! Got some plans to do more content around it!
Turnitin will catch the content generated by AI.
I love how these videos - so short and direct it makes it so much simpler to follow thank you 🙏
Is called dumbing it down for the masses.
everyone thinks they are an expert till they hit a basic problem...
Awesome 😎 Thank you for sharing with us❤
Thank's for that. Can you make a video on explainability in ML models, I want to know why and how the models work.
You got it!
Can you make a video on explainability in ML models such as visualisation techniques in CNN and more
Hey Nicolas, enjoying your content! You are an inspiration! Just curious which final cut effect plugins you use? They make videos quiet enjoyable.
No plugins anymore, I do all the special effects in Adobe After Effects. Occasionally I use DeepGlow but all the custom stuff you see here is from AE
Damn 😅 Gotta work out better placement for the keyboard cam.
Also, lmk if any of you are coming to the forum I mentioned on LinkedIn! Would love to grab coffee with y'all!!
i dont get the .ChatCompletion method
i get .completion
I’ve been doing a lot of cool stuff with the API. It’s incredibly easy to implement and find new ways to work with other libraries
Do you think it's worth the price?
@@BigBoss-sm9xj a better question to ask, does it pay for itself and the answer is yes.
Can you share how to get just the output, and not the rest of the object?
how much the pricing?
Nice, thanks.
Cheers Terry!
Any full screen video for this please... Or any other source for this...
How does the results compaire to use gui connected to API or using the web app or in browser
How does this work with continuity between messages?
You need to store it, you can store it temporarily in a list and that way you can maintain continuity
Thanks mate 🤍
for the content, to make it more dynamic, have the user supply the prompt as a command line argument.
is a programmatic example to show the open api
not to show a scripting example for other usecases
What dependencies do I have to install before running this code?
Hey nic can you do a stock market (of your choice) data analysis
So I've made a tkinter app that had a prompt box and returns in a canvas with a save as need to clean it up need to learn to style more
Nice!! Awesome Josh!
What is the benefit of using this instead of the default chat gpt webiste?
Do we need a chatgpt acount
How many such requests can be made? Is there some limits in your plus subscription?
It's using GPT3.5 turbo, how do you use GPT4 instead?
what is the purpose of asking chatgpt questions in vscode
Nice vlog ❣️Kannayya while seeing in mobile be like Enti nenu intha cute ga vunna 😊😅
I'd like to see a simple example with PHP. I tried a month ago it never worked
How can I feed it data to build a custom trained model ?
That's a good question!
you can litterly make an ai app that just runs chatgpt
@noobody. no no, I mean an interface that works via ChatGPT-API
is this script working without openai api key?
Would have to pay tho?
Can you pls do the same for llama 2
Wait, I think this is actually pretty useful... * looks at Alexa echo dot * hehehehehe
For those who didn't get it, I want to try to make an Alexa skill which let's you communicate with chat gpt through Alexa
Is there a way to use ChatGPT 4 on this?
hey nicholas please made video on kaggle
Can I make chatgpt helps me in prediction and classification with me model and links it with opencv
I mean can I let the mission for chatgpt
Awesome ❤👍💪👏😄
I am trying to install openai to my file in visual studio code but its not working what do i do
Is it free to use open ai APIs?
Website or app name??
How about image detection for onscreen images and algo? 😅😂
AWESOME
Cheers Omar!
Bro! I wanna learn Reinforcement Learning, are your old videos still working? I heard that the libraries changed and the code doesn’t work from the comments.
this scares me a lot Nick
Curious? Why so @Mi Name?
Will it give me the gpt4 answers?
This good content
Thanks a mil @jasser!!
Nice one Nicholas. One question, there's the same max characters in the output as in the web version? In this format of video seems longer.
What's the limit in the web version? It generated a whole page for me this morning!
@@NicholasRenotte It was like 4k characters. But I guess that now with the paid version it depends of the budget :')
That’s has to do with the mode 4k token is the gpt4 limit (I heard they will release a 32k token limit )
chatgpt helped me try random generation of api keys until it works and now i have free chatgpt
LOL fr?
@@NicholasRenotte Yea it tries many combinations over and over everytime it returns an error, sometimes it works, sometimes it doesn't. So sorry for whoevers api key was used
Cool man
Public API?
barack obama?!
Hi there. Is it possible you could tell me if paying for the api key gives faster response times... for chatgpt api? Would appreciate your response.
I think you cannot use this without paying, not sure tho
Can we use this to create chatbot? I am a noob to this field. Please answer anyone... 😊
Can you make videos about fine tunning, and semantic search using gpt?🥺
You got it!
hey Nick, I’ve been trying to build a chatbot using the official API but the engine don’t have memory. Is there a way to have such thing with the official api? Right now i m using the un official chatgpt wrapper but I need to login every time
Hmmm not too sure where the memory issue is? On ChatGPTs side or for your app? I'm doing a walkthrough of how to integrate it into Watson Assistant today. Started working on the Openapi spec yesterday. Should be out soon, I've been getting a ton of questions on it.
@@NicholasRenotte Like: you can use the engine to generate answers as you showed but you can't generate a conversation instance through the API. Or can you? Looking foward for the video!
Nice
Eerie how time and time again that we align our projects lmfao
I posted on LinkedIn about an integration I built with it. I am getting slammmed with work now 😂
hey, im continuously getting this error
openai.error.RateLimitError: You exceeded your current quota, please check your plan and billing details
even though its a brand new key and i used the old key way back in feb
can someone please help, thanks :D
I am getting the same problem
it says no moduled name openai
Why would I do this over using the playground
ChatCompletion line is not working properly
Hello. I installed that module but apikey methods not included. This way of codes gives me error. ModuleNotFoundError: No module named 'apikey'
Do you have any suggestion?
It's a local module he created himself to store his APIKEY I believe.
were is the dicreption
Do we need to generate a key from the open Ai website?
yes
where's code
I need help
Ты не пробовал из какой нибудь локальной модели которая полностью доступна в общем доступе, сделать чат?
Not yet most of the stuff has been with paid APIs unfortunately. Closest was probably the live stream I did with BLOOM.
@@NicholasRenotte Sorry, I'm from Russia and I meant this. Would you like to use a public local model to create a dialog neural network similar to GPTchat. Are you planning this??
only 5-6 month ago.. it was that recently🤔🤓
it says You exceeded your current quota, please check your plan and billing details
you are so lucky you got it for free man
Me too
What did you do next?
BOOOM
Is the API free?
hello me to developer and ı have an question; what is gpt-4 engine name
please help me
Why not just use chatgpt promptline?
Is there any method to take user input as message?
U can create a tkinter and use an Entry
Yeah easily so put this content: input("What is your proompt?")
would you need internet connection before it generates the output or it does it like that
0:00
Bro how to buy api keys in open ai
You can get the free credit and can generate one without buying
"Cannot import name APIKEY from apikey"
O________O
I did pip install apikey and it went through, but the code still doesn't run. What gives?
bruh...
@@luischronos Help a newb out?
Not working
Is there anyway to get an api for free use for learning ?
That api is already free
There will be a leck of programming job's in the future better start with electrical engineering.
That'll be affected as well, not that much further from programming jobs
You don’t
How to get API key ?
you hvae to pay
I'm keep getting error that api is invalid or exhausted quota... Is it paid?
It used to be free but it's paid now
@@tusharrao74 that's sad, you know any api / model / lib to make a simple chat bot
I had the same issue. Bit of a bummer it's no longer free.
is this Api is free
Is this the pay to use API?
you have three months free
Yup!
I would rather use ChatGPT to ask it how to use ChatpGPT APIs 🤭
Sure buddy, if you want the less advanced text-davinci-002 model
This man have Something wrong in his Neck 😢😢😢😢😅
how to make Minecraft in python import Minecraft run yay
Is this really chat gpt, or is this gpt? Big difference. The one is a fixed price monthly, the other costs credits. Its misleading if its gpt and not chatgpt
Is this free?
does this guy get on your nerves or is it just me
nah, gets on my nerves as well
Nowadays they need computers to say absolutely nothing?🤯
Is it free or paid?
This one is paid @Ogunkoya! It seems pretty reasonable though $2 per million words.
@@NicholasRenotte how do I start this? And where can I get an API key?
I fail to see how this is anything other than pointless. Why waste time writing all that code when you could just go to chatgpt and ask there?
Can we maybe get a practical use for this?
wtf. This can be used to automatize a lot of processes.
What’s the point of this, why not just use the GUI?
Dont know anything about it
This is not a ChatGPT