Thank you so much for watching, Everone. I hope you find this course valuable and if you have any question don’t hesitate to reach out to me :) again, thank you for the opportunity 🎉
@@joyaljose3666 you have to have the API subscription, however, if you open an account (for the first time) you'll have a few Free credits that you can use :)
I am casually watching this course, and I like the teacher. I am casually considering trying to follow along. I appreciate the relaxed speed of the course. I have seen some other hotshot young TH-camrs, talking about Assistants API, but they are so fast, so I don't understand anything. This teacher's speed is just right for me :)
Thank you very much for putting this video together. Just as an FYI for anyone trying to get the Studdy Buddy to work. You should pin your openai python package to 1.14.0 ( later versions might work, but the latest will not ) because the API changed in the latest version of the OpenAI Python API.
Randomly came across this tutorial and the instructor has such a calming voice and so so easy to grasp the lesson. This is something I haven’t even considered learning before and I am so focused on the subject that I’m excited to give it a try. Thank you!!
I was looking for this many weeks, most of the other information is out of date and code is not working. Thank you so much! if you can make another video about how to connect this to simple webinterface - would be great!
In the Study Buddy you need to fix upload_to_openai so that the file actually gets a name when uploaded: def upload_to_openai(filepath): with open(filepath, "rb") as file: response = client.files.create(file=(os.path.basename(filepath), file), purpose="assistants") return response.id
it really gets to me that error prevented the assistant from working and only you mentioned it. im lucky i saw your comment and connected the problem to the code,. not familiar with python i would have never figured that out. i cant thank you enough. made all the difference. i have a big question for you: do you think my sql tables with extensive business calculation knowledge examples can be read in and recognized as the models for building any workflow? organized by access, functions and sequence, this methodology can define any business process. if an assistant can read in my proven formulas that reside in 15 sql tables and let users select from them and build new ones anything can be defined, matched and executed.
I wanted to see the final results with the user interface. I used to work for Dow Jones (publishers of the WSJ) and my job was programming that news compilation, once at the end of each month, but it was made in C++. Running that executable took more than 12 hours. We left the program running overnight. In any other language, Visual Basic, or any other, the program took almost 24 hours to run. It was even slower for scripting languages. So the only solution was C++.
Thanks a lot for this tutorial. I've learned a lot from this. 2:14:48 I wonder why we need to create the assistant everytime users click the button? This creates a lot of assistant in the OpenAI account. Should we just create it one time when the app is built?
I just finished this course and I must say I thoroughly enjoyed it! The content was well-explained and beginner-friendly. I would be really excited to see more videos on this topic in the future. Please continue with the great work, and I'm looking forward to the next videos!
thank you so much for this video. saved us all a ton of time. the api from the application is generating inconsistent responses and behaves as if the files don't event exist. However, from the playground, it works as expected.
This was really excellent! For the study buddy, I think things are different now since the code doesn't work, so I converted the related functions to work like it does in the openai documentation for file searching. Works great! Thanks so much!
i watched the first 10 minutes but i'm still confused regarding what assistant api actually does, i even asked chatgpt but it doesn't know either.. is it just chatgpt api or i can use this to teach the chat's api to learn only my specific dataset and answer questions regarding that data without answering other topics? thanks! p.s. i've subscribed to your youtube channel btw, i saw few interesting python projects there i hope you keep it up and not give up after a year or. so like most youtubers@@vincibits
Thank you for such a indepth tutorial and it has help me alot and I think you have alot to offer.You can really provide alot of value if you just brush up a little in your teaching skills as I am totally lost sometimes on which part I am coding when your writing a monolith and only seeing the final result in the end. I dont know which part is broken or which part does what. Again I dont mean this to come in a bad way.
This was a nice video, the only issue I have is that it was very difficult to follow you creating the class, as there were no comments on the github nor any sort of flowchart showing how the functions were working with each other. Other than that, loved your work @vincibits!
Amazing tutorial! One question about the cost. Is cheaper to create a new thread for each message or it's the same we use the same thread for each messages ? How to calculate cost of usage ?
Thank you for the tutorial, it's truly awesome. I need to find time to thoroughly review it. I have a question: I'm currently looking to analyze simulated data from interviews between psychologists and clients. Each interview is about 20,000 tokens long, and I want to summarize the Q&A content to create a case report for each client. However, loading all the data at once is a significant challenge. I've considered dividing the data into batches and feeding it to the assistant for processing, but this often leads to incoherent outputs, as if the assistant is hallucinating. My current plan is to split the data into multiple files, then gradually submit these to the assistant to generate summaries, and finally integrate these into a comprehensive case report. The issue is that my current programming skills are limited to simple interactions-submitting a question and receiving an answer. Although I can use a for loop to process batches of data for summaries, these seem to be treated as separate conversations without shared context, which means the AI might not remember past interactions. How can I program it so that my prompts are processed in a single thread, similar to how the assistant interface maintains context throughout a conversation? Thank you for your help.
This is a wonderful video @vincibits This may be premature since I haven’t tried it yet, but I’ve been doing a lot of work so I’m not token hogging and this assistants API feels like it may eat up tokens. I’ll benchmark each for comparison. The convo history compression alone would be something I’d rather managed in a thread, but if it’s using up tokens, it’s gonna be expensive
Interesting how about 70% of the video just shows things you can already do with the normal chat completions API (function calling), but with way less overhead and cost. Anyway, good video 👍
I used assistants api a week after it released. I'm just a small text file it was super slow I'm saying 10 seconds for a simple question. It's a really unoptimized rag system.
How did you get intellisense on the openai parameters (enums, etc) in VS Code? Which extension? (there are dozens, but I like what I'm seeing on your video)
Does anyone know how to enable the widget that pops up with the function arguments when the parentheses are created--just as in the video? I've been trying to figure out how to enable that but couldn't figure it out.
Now see the thing is, they have mentioned multiple forms of pricing in OpenAI, it's confusing, if you could help. You see, if wer'e using assistants api with 4 or 4.5 turbo models, also fine tuning a bit(if this means training our assistant to some extent), then they are going to charge separately on the three fronts they mentioned there? model + assistant api(session thing) + fine tuning($8/M tokens or so)?
Paulo, wonderful course, and thank you. Is the OpenAI Assistant python class code you demonstrated available anywhere? I looked at the links provided (two were the same) and didn't see anything regarding the class development. Again, thank you so much for your time.
Can I create a single PDF chatbot assistant and use it for multiple users or I have to create a new assistant for each user ? I am talking about android application.
why didn't you get directly with google colab? No need to install coder or python libraries as is all in one. Makes sense ora do you see any issue in using google colab?
I would like to ask how to activate, (source myenv/bin/activate), I am stuck here, I am a windows user and I followed the steps you instructed but I am stuck at this early hahahaha, could you instruct me?
sorry im new OAI asssistant API..i have already had 'credentials' from my custom GPT in GPT store. how can i connect it to my separate app with assistant API? not recreating the instruction on assistant API but just connect it to custom GPT in GPT store that i have already created earlier
Does anyone face very slow api calls? I even tried with latest model '4o' but I'm getting many failures and in case of success, it could take up to 10 seconds or more to get a response.
thanks alot for the awesome tutorial, very informative and easy to follow. I have a question, if i send a copy of the session history "Thread" at each API call, is the thread content 'tokens' calculated in my usage again each time?
Please How to connect to an existing assistant API using streamlit, assistant'id and openai api Key? All the video i see just learn How to create a New one. I created my assistant API in playgroung openai and i want ton connect to it by using streamlit. How Can i do it please??
hi where you supply the 3 links for your course code, there is an error the 2nd project link, the first persona-trainer code is repeated so i have been trying to type in every word as you say it and its very hard to comprehend at the same time. could you please fix that link to go to NEWS-SUMMERIZER code instead?
Thank you so much for watching, Everone. I hope you find this course valuable and if you have any question don’t hesitate to reach out to me :) again, thank you for the opportunity 🎉
Love your style. Subscribed. Waiting for more quality videos.
@@noNullMoments Thank you so much for your support. It means the world to me... :)
Discord server was the early birds? 😀@@vincibits
Without having the API
subscription can we able to do the python code
@@joyaljose3666 you have to have the API subscription, however, if you open an account (for the first time) you'll have a few Free credits that you can use :)
This dude's voice is so soothing I might learn this lesson in my sleep.
Hahh.. Is that a good thing? I mean I want you to learn, and not to fall asleep :)
Also liked it. Soft and clear. It's a good thing for sure.
for real, it feels like I'm being taught by Kendrick Lamar
@@rolandcucicea6006 lol! I hope that's a good thing?!
@@vincibits, To me I feel like you teaching me like sitting next to me. 😄
I am casually watching this course, and I like the teacher. I am casually considering trying to follow along. I appreciate the relaxed speed of the course. I have seen some other hotshot young TH-camrs, talking about Assistants API, but they are so fast, so I don't understand anything. This teacher's speed is just right for me :)
Thank you!!
Thank you very much for putting this video together. Just as an FYI for anyone trying to get the Studdy Buddy to work. You should pin your openai python package to 1.14.0 ( later versions might work, but the latest will not ) because the API changed in the latest version of the OpenAI Python API.
thank you! that helped me very much
Thanks
Paulo was the first to do Flutter Courses. Love to see he’s still making great videos.
Thank you so much for your kind words :)
Mr.bo this gentleman is an excellent teacher take care of him
Randomly came across this tutorial and the instructor has such a calming voice and so so easy to grasp the lesson. This is something I haven’t even considered learning before and I am so focused on the subject that I’m excited to give it a try. Thank you!!
I was looking for this many weeks, most of the other information is out of date and code is not working. Thank you so much! if you can make another video about how to connect this to simple webinterface - would be great!
Спасибо, наверное, это для меня лично самый полезный видос за пару лет, русскоязычному сегменту не хватает гайдов/обзоров по теме п2п
In the Study Buddy you need to fix upload_to_openai so that the file actually gets a name when uploaded:
def upload_to_openai(filepath):
with open(filepath, "rb") as file:
response = client.files.create(file=(os.path.basename(filepath), file),
purpose="assistants")
return response.id
Correct.
thanks very much,
Thanks a million. i had nothing working until you provided this. very kind of you!!!
it really gets to me that error prevented the assistant from working and only you mentioned it. im lucky i saw your comment and connected the problem to the code,. not familiar with python i would have never figured that out. i cant thank you enough. made all the difference.
i have a big question for you: do you think my sql tables with extensive business calculation knowledge examples can be read in and recognized as the models for building any workflow? organized by access, functions and sequence, this methodology can define any business process. if an assistant can read in my proven formulas that reside in 15 sql tables and let users select from them and build new ones anything can be defined, matched and executed.
@@aeharrison1able JUST ASK CHAT GPT :)
I wanted to see the final results with the user interface. I used to work for Dow Jones (publishers of the WSJ) and my job was programming that news compilation, once at the end of each month, but it was made in C++. Running that executable took more than 12 hours. We left the program running overnight. In any other language, Visual Basic, or any other, the program took almost 24 hours to run. It was even slower for scripting languages. So the only solution was C++.
Smooth, masterful, and timely walk-though of the OpenAI Assistants API. Very clear. Much more pragmatic than reading the docs. Fantastic job!
i was just saying the exact opposite to myself 😅
Anyway, thank authors for the effort and for the ideas included 😌
@@kalamatej You don't like the course? I am just curious :)
Thanks a lot for this tutorial. I've learned a lot from this.
2:14:48 I wonder why we need to create the assistant everytime users click the button? This creates a lot of assistant in the OpenAI account. Should we just create it one time when the app is built?
I just finished this course and I must say I thoroughly enjoyed it! The content was well-explained and beginner-friendly. I would be really excited to see more videos on this topic in the future. Please continue with the great work, and I'm looking forward to the next videos!
Thank you so. much!
thank you so much for this video. saved us all a ton of time. the api from the application is generating inconsistent responses and behaves as if the files don't event exist. However, from the playground, it works as expected.
Thank you so much!
Great content! Unlike speedy, marketer, shallow content that is often seen, this man provides in depth, nicely paced, original content. 👏👏
Thank you so much! I really appreciate it!
This was really excellent! For the study buddy, I think things are different now since the code doesn't work, so I converted the related functions to work like it does in the openai documentation for file searching. Works great!
Thanks so much!
Paulo is great. I did one of his courses on another plataform. Happy to see him here
Thank you so much for your kind work! I am humbled!
Thank you for your kind words. It means a lot!
i watched the first 10 minutes but i'm still confused regarding what assistant api actually does, i even asked chatgpt but it doesn't know either.. is it just chatgpt api or i can use this to teach the chat's api to learn only my specific dataset and answer questions regarding that data without answering other topics? thanks! p.s. i've subscribed to your youtube channel btw, i saw few interesting python projects there i hope you keep it up and not give up after a year or. so like most youtubers@@vincibits
Amazing Teacher, Very detailed Tutorial.
Thank you so much! I really appreciate you!
Great instructor, calm and methodical. Thank you 🙏
Thanks for this tutorial, this solves my tasks for this sprint.
Very detailed and good pace. This is something I would bookmark and watch over and over again. Thanks.
Thanks for taking the time to make this, great course! Followed it to the end and have a real handle on the API now, thanks again!
Mr paulo really your style is excellent and that was very helpfull,hope to see more frequently in this channel,hope mr bo will accept
Thank you so much!
I think there is a little mistake in the video description. The personal trainer github repo shows up twice and the newsSummarizer is missing!
Paulo 's flutter course was good. I took it in udemy.
Thank you!
Thank you for such a indepth tutorial and it has help me alot and I think you have alot to offer.You can really provide alot of value if you just brush up a little in your teaching skills as I am totally lost sometimes on which part I am coding when your writing a monolith and only seeing the final result in the end. I dont know which part is broken or which part does what. Again I dont mean this to come in a bad way.
Please share the slides / presentation for notes. Thanks 🙏
he didnt specify that we have to add the credit card credentials???
someone please clarify what to do about the card details
This is so well explained. Just perfect, thanks
As always Steve Jobs giving great advice thanks steve.
Love Paulo! Always a pleasure to see him.
Thank you so much! It means a lot to me. Thank you!
Thank you so much!
This was a nice video, the only issue I have is that it was very difficult to follow you creating the class, as there were no comments on the github nor any sort of flowchart showing how the functions were working with each other.
Other than that, loved your work @vincibits!
Amazing tutorial! One question about the cost. Is cheaper to create a new thread for each message or it's the same we use the same thread for each messages ? How to calculate cost of usage ?
same question I have
Thank you for the tutorial, it's truly awesome. I need to find time to thoroughly review it. I
have a question: I'm currently looking to analyze simulated data from interviews between psychologists and clients. Each interview is about 20,000 tokens long, and I want to summarize the Q&A content to create a case report for each client.
However, loading all the data at once is a significant challenge. I've considered dividing the data into batches and feeding it to the assistant for processing, but this often leads to incoherent outputs, as if the assistant is hallucinating.
My current plan is to split the data into multiple files, then gradually submit these to the assistant to generate summaries, and finally integrate these into a comprehensive case report.
The issue is that my current programming skills are limited to simple interactions-submitting a question and receiving an answer. Although I can use a for loop to process batches of data for summaries, these seem to be treated as separate conversations without shared context, which means the AI might not remember past interactions.
How can I program it so that my prompts are processed in a single thread, similar to how the assistant interface maintains context throughout a conversation?
Thank you for your help.
Thank you so much.
This is what i need 🎉
Great content, thanks! Do you have a preference when building a RAG app to work with the OpenAI Assistants API, Llama Index, or Langchain?
Thank you very much. Easy to follow. I'll try it in NodeJS ❤
Thank you!
This guy is AMAZINGLY GOOODDDD !
This course is very nice and useful for us thanks all of that
Thank you! I am glad you find it helpful!
Thank you so much for this tutorial!! Very helpful
This is a wonderful video @vincibits
This may be premature since I haven’t tried it yet, but I’ve been doing a lot of work so I’m not token hogging and this assistants API feels like it may eat up tokens. I’ll benchmark each for comparison. The convo history compression alone would be something I’d rather managed in a thread, but if it’s using up tokens, it’s gonna be expensive
Amazing teacher!
Great Course!
watched from Brazil
Interesting how about 70% of the video just shows things you can already do with the normal chat completions API (function calling), but with way less overhead and cost. Anyway, good video 👍
I used assistants api a week after it released. I'm just a small text file it was super slow I'm saying 10 seconds for a simple question. It's a really unoptimized rag system.
Great course, thank you for making available this high quality content 😃
Exactly what I needed, thank you
This isn't possible without credit card info, right? Even the $5 credits on the new account aren't a lot?
thanks a million for this content
This is really awesome.
thanks for sharing great knowledge sir, can we have assistant api implementation in php
Great job. It would be great if you can teach us how to implement stream responses when it release for tge Asistant api
Will do!
Please do a short follow up with the streaming responses 😊😊😊
How did you get intellisense on the openai parameters (enums, etc) in VS Code?
Which extension? (there are dozens, but I like what I'm seeing on your video)
Good but perhaps a bit slow? Speed x10
Does anyone know how to enable the widget that pops up with the function arguments when the parentheses are created--just as in the video? I've been trying to figure out how to enable that but couldn't figure it out.
Thank you very much
This was excellent. Do you do consultations as well? Thanks.
Is there a reason that we have to wait for the run to complete? I see that OpenAI allows streaming - client.beta.threads.runs.stream
Good explanation. 👍👍
But what if i want to read from more than one file for last project of study buddy then? how can I do it?
This is a goldmine
What are the benefits of using this over say the custom GPT’s?
Using someone else's configuration versus making your own. More power and more control to you at the expense of having to do the work...
But you make your own Custom GPT's anyway
Love this video! By the way, is there the code of function calling provided? I do not find it in the profile.
What is the difference between the assistant and fine tuning?
As a knowledge base can’t you just get it to retrieve large files and then the chat gpt will incorporate it into its answer without all the coding?
would assistant API follow the same response with custom GPT if we give them both the same instruction?
Now see the thing is, they have mentioned multiple forms of pricing in OpenAI, it's confusing, if you could help. You see, if wer'e using assistants api with 4 or 4.5 turbo models, also fine tuning a bit(if this means training our assistant to some extent), then they are going to charge separately on the three fronts they mentioned there? model + assistant api(session thing) + fine tuning($8/M tokens or so)?
Fantastic job Paulo!
Thank you so much!
Very useful, Thank you Paulo
Thank you!
Paulo, wonderful course, and thank you. Is the OpenAI Assistant python class code you demonstrated available anywhere? I looked at the links provided (two were the same) and didn't see anything regarding the class development. Again, thank you so much for your time.
Can I create a single PDF chatbot assistant and use it for multiple users or I have to create a new assistant for each user ? I am talking about android application.
@vincibits
Thank you very much for your explanations.
I can't find the link to get code about News Summarizer.
Could-you publish it again?
awesome video!
why didn't you get directly with google colab? No need to install coder or python libraries as is all in one. Makes sense ora do you see any issue in using google colab?
если вам дать в управление USDT за какое время сможете иксануть?
I would like to ask how to activate, (source myenv/bin/activate), I am stuck here, I am a windows user and I followed the steps you instructed but I am stuck at this early hahahaha, could you instruct me?
where will we get these notes ?? plz help
Thanks!
nice. so assistance api is something which is doing what langchain does??
in the function call, how can we connect to the database?
Oh my god finally 😭I've been waiting for this
Please make an AI app which is tracking movements of maga, project 2025 staffing, legal actions or policies
sorry im new OAI asssistant API..i have already had 'credentials' from my custom GPT in GPT store. how can i connect it to my separate app with assistant API? not recreating the instruction on assistant API but just connect it to custom GPT in GPT store that i have already created earlier
thanks for the video, but I a getting this error:"TypeError: OpenAI.__init__() takes 1 positional argument but 2 were given"
Who thought that Open AI was intelligent? God help anyone who has to go through this to get it to do something useful.
Plz tell me which version of dell laptop is required for OpenAI project
Its awesome.
Why I cant see is no save button, when I am creating assistant?
Does anyone face very slow api calls? I even tried with latest model '4o' but I'm getting many failures and in case of success, it could take up to 10 seconds or more to get a response.
awesome
anyone knows how to use an existing assistant and existing thread from vscode, and how to catch the assistant responses?
The best!
This video needs a warning that it uses the older OpenAI Assistant codebase and Example #2 (News-summarizer) won't run as shown in the video.
thanks alot for the awesome tutorial, very informative and easy to follow. I have a question, if i send a copy of the session history "Thread" at each API call, is the thread content 'tokens' calculated in my usage again each time?
(58:58) why the response when call assistant was so slow ? Do u have a solution or something explain that ?
Please How to connect to an existing assistant API using streamlit, assistant'id and openai api Key?
All the video i see just learn How to create a New one.
I created my assistant API in playgroung openai and i want ton connect to it by using streamlit. How Can i do it please??
Thx bro
hi where you supply the 3 links for your course code, there is an error the 2nd project link, the first persona-trainer code is repeated so i have been trying to type in every word as you say it and its very hard to comprehend at the same time. could you please fix that link to go to NEWS-SUMMERIZER code instead?
Yes, I can't find the news summarizer code as well @vincibits
That's a lot of money for one extra day in Paris 😮
Damn, Paris is expensive