Awesome tut. Thank you. Have you got any vids, or could you show us how to have chatGPT work from a provided context? For example, a repository of PDFs, like instruction manuals / procedures?
I was looking for something like this for soo long......Dudde, you're awesome...... you explained it so clearly.. definitely gonna make something using this infor
I'd be interested in a similar video to write a program that uses ChatGPT API to generate code and then execute it. Maybe generate unit tests or other validation to cross check it? Catch errors and then ask ChatGPT to debug its generated code?
Does using assistant add to the amount of tokens used per interaction and response? Is it resending past messages to maintain context? Thank you for the video by the way. Very informative.
How to make a chatbot that interacts with potential customers on my website? I am a solar consultant and want the chat to assess, qualify and occasionally close my customers.
Great video! Like others, do you think you can create a video and/or write a Medium article on how to train GPT on custom data (documentation)? Love to learn more on this topic. Thanks in advance!
Some changes are needed for this code to run properly as of 1-02-2024: The code below is modified one. ' ' ' import os from openai import OpenAI client = OpenAI(api_key = "key") while True: content = input("User: ") messages.append({"role": "user", "content": content}) completion = client.chat.completions.create( model="gpt-3.5-turbo", messages=messages ) chat_response = completion.choices[0].message.content print(f'ChatGPT: {chat_response}') messages.append({"role": "assistant", "content": chat_response}) ' ' '
I have a quick question, In the assistant mode, how can we create an assistant that we can come back and continue the chat later on? Can we write the responses in the file ? If I close the notebook i presume the assistant is gone?
What's the most elegant solution to prevent the accumulated messages from eventually exceeding the max token limit in your final example? In my testing with the web version of chatgpt it seems like it "forgets" very old Assistant outputs before it forgets user inputs. I was able to ask chatGPT what was the first thing I asked and it was correct but it would make an educated guess as to what it said that far back. I was able to just scroll up and double check.
Hi, recently I've created account with gpt4, however I didn't get any response, I got an error "You exceeded your current quota, please check your plan and billing details."
🔥 Get my FREE ChatGPT Cheat Sheet: frankandrade.ck.page/08c94cf1c1
Very helpful. Especially assistant rule to save history. This is the most needed feature. Many thanks
Such a great video, super clear explanations as well. Could you do a video on how to train gpt on custom data sets soon?
I am new to this and was looking for tuts like this one to boost my confidence. Thanks
Awesome video! The only video that explains what I was looking for! Thank you!
Awesome tut. Thank you. Have you got any vids, or could you show us how to have chatGPT work from a provided context? For example, a repository of PDFs, like instruction manuals / procedures?
I was looking for something like this for soo long......Dudde, you're awesome...... you explained it so clearly.. definitely gonna make something using this infor
I'd be interested in a similar video to write a program that uses ChatGPT API to generate code and then execute it. Maybe generate unit tests or other validation to cross check it? Catch errors and then ask ChatGPT to debug its generated code?
Great content as always
super clear explanations thank you ❤
Does using assistant add to the amount of tokens used per interaction and response? Is it resending past messages to maintain context?
Thank you for the video by the way. Very informative.
Amzaing video! Thank you
Great work Frank! Thank you! Did you share the notebook somewhere ?
Hi. I just added the link to my GitHub repo. Thanks for letting me know!
Thanks for the insightful video, much appreciate. However, I have a question. Can we create a specific product ML using ChatGPT?
How to make a chatbot that interacts with potential customers on my website? I am a solar consultant and want the chat to assess, qualify and occasionally close my customers.
You can do this by fine tuning the model on your own custom data and build the bot.
@@Xohrabcan you tell me how to do that?
Great video! Like others, do you think you can create a video and/or write a Medium article on how to train GPT on custom data (documentation)? Love to learn more on this topic. Thanks in advance!
Some changes are needed for this code to run properly as of 1-02-2024:
The code below is modified one.
' ' '
import os
from openai import OpenAI
client = OpenAI(api_key = "key")
while True:
content = input("User: ")
messages.append({"role": "user", "content": content})
completion = client.chat.completions.create(
model="gpt-3.5-turbo",
messages=messages
)
chat_response = completion.choices[0].message.content
print(f'ChatGPT: {chat_response}')
messages.append({"role": "assistant", "content": chat_response})
' ' '
Good stuff!!
Hi. Does the 40messages/3hours limit apply in Python API?
This is great!
I have a quick question, In the assistant mode, how can we create an assistant that we can come back and continue the chat later on? Can we write the responses in the file ? If I close the notebook i presume the assistant is gone?
Awesome.. Keep it up.
What's the most elegant solution to prevent the accumulated messages from eventually exceeding the max token limit in your final example?
In my testing with the web version of chatgpt it seems like it "forgets" very old Assistant outputs before it forgets user inputs. I was able to ask chatGPT what was the first thing I asked and it was correct but it would make an educated guess as to what it said that far back. I was able to just scroll up and double check.
Can you compare the api and pro version in terms of value such as pricing, speed, etc? Thanks
Wow! I'm very beginner to Python. I wonder if you could suggest if I could do this under my Telegram Bot? I tried some but I face errors. :/
And thanks for this informative video!
since this is for chatgtp 3.5 can I do the same for chatgtp 4?
Hi, recently I've created account with gpt4, however I didn't get any response, I got an error "You exceeded your current quota, please check your plan and billing details."
can it remember previous chats/questions?
A great video
i used it 0 times and it said that i exceed my current quota
same erros
I am getting error 429 - You exceeded your current quota, please check your plan and billing details
What can actually be done with the ChatGPT API in Python?
I got an error=>module 'openai' has no attribute 'ChatCompletion'. Why?
You are likely using an old version of openai try to uninstall then reinstall
How to install it?How to get Jupyter? ???
is this whisper api?
No this is gpt-3.5-turbo, it is different to whisper api
What is difference between GPT-3.5-turbo and whisper?
Is this api paid or free ?
"messages=messages" is not good practice
i don't understand a thing :(
good morning, great video. getting an error on the api key, it claims it doesnt exist...