Have to say. Very clear and crisp delivery of material. I have been in software for 35 years … lots of great stuff. Not many have I run across with this guys rhetorical skills. Thanks!
Loving the series so far! I am 2 videos in and I feel like you are balancing very well between making these tutorials beginner-friendly and not giving just superficial knowledge.
after looking for a logical simple explanation over and over in videos and text you video has made the most sense, Thank you for explaining it like that!
Thank you very much for this video! It was super helpful, I can now leverage Open AI's API and this is a game changer for me. I'm also now going to listen to Roxette :)
I am working as Data Engineer, I want to upgrade my skills and get AI/ML Engineer job. I attended few trainings in this but for beginners it is amazing videos.
Hi, Can you please make a video of some real use cases of LLM with all the advanced features and code walkthrough. I really like your video because of the way you explain the technical and implementation details.
Thanks for the feedback. Yes I definitely will. Here's what I have on my list, do you have any suggestions? - Text similarity using SBERT (request) - How to build a robust RAG system - How to fine-tune a model for domain-specific question answering
Have a question. I have seen several surges in the software field over the past 35 years. One of the hardest parts of software developed by a company, for their use, is project management. A product owner see's some changes that will, ultimately, increase shareholder wealth, and writes them down. Lets say a scrum team is assigned the work ... in the case we have here - what are your thoughts on how AI can improve the predictability of quality tested software delivered on time? Again, software is software ... math is math. Thanks, John
Great question. The best predictor of project timelines are the durations of past projects with a similar scope. I could imagine an data-driven (AI) tool that integrates into Jira (or the like) that gives project estimates given factors such as scope and available resources.
May I know the following: 1) Can AI act as a code interpreter? 2) Can AI compile and link source codes? 3) Can AI code in different languages in one program? 4) Can AI be instructed to create codes in 8-, 16-, 32-, 64-, 128- and so on bit? 5) Can AI create a computer language and the compiler and interpreter therefor and operating system? 6) Can AI disassemble a Lisp, Fortran, COBOL, or Ada application and create the code for the corresponding compiler? 7) Can AI create a boot sector? 8) Can AI convert a CD and flash drive into start up devices? 9) Can AI make several copies of LLM? 10) Can AI do pre-training, fine-tuning and re-fine-tuning?
AI is a broad term so I will try to answer your questions in regards to LLMs. (Most) LLMs simply generate text. Therefore, it alone cannot do many of the things you mentioned. However, by embedding an LLM into a larger software product, you can accomplish just about any of these things and call it an AI. Happy to chat more during office hours: calendly.com/shawhintalebi/office-hours
Seems counterproductive to teach people OpenAI, which is anything but open, when talking about programmatically working with LLMs. You should steer people towards the actually open LLMs that can be run locally for free. OpenAI is eventually going to be going out of business, once people realize the downsides outweigh any ease of use and first-mover advantage they once had, and the venture capital dries up. The open models with better licenses are already outperforming GPT anyway, and their approach to the technology is rapidly becoming obsolete. Teach people skills that will actually be useful for the open source AI development that will still be around in the future.
While OpenAI conceals their models behind their API, given the SOTA performance of their current models, there are still situations where using them makes sense (at least for now). I cover open-source technologies in other videos in this series: th-cam.com/play/PLz-ep5RbHosU2hnz5ejezwaYpdMutMVB0.html
The last minute of this video mentions that his next video actually gets into that. That said, I’m hoping we get a Poe and Perplexity api video sometime too.
@fakecubed would there possibly be anyway to ask you a question or email you? you seem to have some advanced knowledge regarding AI and python implementation. I wanted to see if it be possible to get your input on a question I have? I tried going to your TH-cam site couldn't find a way to contact, if possible would be greatly appreciated Thank you
👉More on LLMs: th-cam.com/play/PLz-ep5RbHosU2hnz5ejezwaYpdMutMVB0.html
Have to say. Very clear and crisp delivery of material. I have been in software for 35 years … lots of great stuff. Not many have I run across with this guys rhetorical skills. Thanks!
Thanks John! It's been a long road to get here 😅
Like many others who've commented, I've found my guide to the world of AI, LLMs, etc, thanks Shaw!
Glad it was helpful! 😁
Loving the series so far! I am 2 videos in and I feel like you are balancing very well between making these tutorials beginner-friendly and not giving just superficial knowledge.
after looking for a logical simple explanation over and over in videos and text you video has made the most sense, Thank you for explaining it like that!
Great to hear! Glad it helped :)
Finally! Someone explaining this well 🤩👏🏾
Thanks, glad you liked it!
You're a real Ph.D. with extremely understandable way of explaining things you have an obvious expertise in.
Thanks, glad it was helpful 😁
Thanks!
Thanks so much! Glad it was helpful 😁
maaaan, you are explaining everything-every term that would make me feel like the most stupid person when I ask to another person. tysm!
Glad it's helpful :)
Well prepared and presented. This is super helpful and clear. can not wait to get started with a local LLM.
Just amazing !!! Don't have words to say . The absolute appealing content for the audience . Just nailed it.
What you said is what I've been looking for, It is very clear and useful, thanks very much!
This play list is so important 🎉🎉🎉 thanks 🙏
Thanks, glad it’s helpful 😁
Well sized, well explained, go on, it is exceptional in this line
Thanks, glad you liked it :)
Really amazing explanation, for a beginner in LLMs it really really helps. Thank you so much!!
Thank you. Very nice information. I love your interpretation in the end about new genre.
Very useful lectures!
Loved it!
Thank you very much for this video! It was super helpful, I can now leverage Open AI's API and this is a game changer for me. I'm also now going to listen to Roxette :)
Their song was stuck in my head the whole time I was making this video😂😂
Super guide and content.I ever liked this.I more expect in NLP tutorial text classification.
Pretty cool, thank you 🤩
Glad you like it!
From Egypt that amazing pro
awesome explanation.
Glad it was clear 🙂 -ShawGPT
Amazing Explanation 👏👏
Thanks, glad it was clear!
thank you so so much
Bravo! loved it...
Thanks :)
I am working as Data Engineer, I want to upgrade my skills and get AI/ML Engineer job. I attended few trainings in this but for beginners it is amazing videos.
Hi, Can you please make a video of some real use cases of LLM with all the advanced features and code walkthrough. I really like your video because of the way you explain the technical and implementation details.
Thanks for the feedback. Yes I definitely will.
Here's what I have on my list, do you have any suggestions?
- Text similarity using SBERT (request)
- How to build a robust RAG system
- How to fine-tune a model for domain-specific question answering
@@ShawhinTalebi this will be of great help
Have a question. I have seen several surges in the software field over the past 35 years. One of the hardest parts of software developed by a company, for their use, is project management. A product owner see's some changes that will, ultimately, increase shareholder wealth, and writes them down. Lets say a scrum team is assigned the work ... in the case we have here - what are your thoughts on how AI can improve the predictability of quality tested software delivered on time? Again, software is software ... math is math. Thanks, John
Great question. The best predictor of project timelines are the durations of past projects with a similar scope. I could imagine an data-driven (AI) tool that integrates into Jira (or the like) that gives project estimates given factors such as scope and available resources.
May I know the following:
1) Can AI act as a code interpreter?
2) Can AI compile and link source codes?
3) Can AI code in different languages in one program?
4) Can AI be instructed to create codes in 8-, 16-, 32-, 64-, 128- and so on bit?
5) Can AI create a computer language and the compiler and interpreter therefor and operating system?
6) Can AI disassemble a Lisp, Fortran, COBOL, or Ada application and create the code for the corresponding compiler?
7) Can AI create a boot sector?
8) Can AI convert a CD and flash drive into start up devices?
9) Can AI make several copies of LLM?
10) Can AI do pre-training, fine-tuning and re-fine-tuning?
AI is a broad term so I will try to answer your questions in regards to LLMs. (Most) LLMs simply generate text. Therefore, it alone cannot do many of the things you mentioned. However, by embedding an LLM into a larger software product, you can accomplish just about any of these things and call it an AI.
Happy to chat more during office hours: calendly.com/shawhintalebi/office-hours
Can I do with Ollama?
Yes! The syntax is very similar. I share an example here: th-cam.com/video/Ot2c5MKN_-w/w-d-xo.htmlsi=CYVqDM7agauuyCFI
I keep getting rate limit error help.
You may need to set up a payment method and set a sufficient usage limit.
Does this still work?
The API has been updated a couple times since I posted this example.
See a more recent walkthrough here: th-cam.com/video/4RAvJt3fWoI/w-d-xo.html
Please display the full code and mention the Python version and system configuration.
Code and env files are available here: github.com/ShawhinT/TH-cam-Blog/tree/main/LLMs/openai-api
'ging ist' is German, for 'went is' and I have no idea why that would be a completion
Interesting! The Germans may be big Roxette fans 😂
Iran?
Born in Texas, but family came here from Iran :)
@@ShawhinTalebi Dear Shawhin ,I am proud of you❤
I just want to see the results first
cmon 😂, we are your viewers , let us use your 5$ 😂
LOL, when I make my first million I'll give away $5 API credits 😜
Should've used GPT-4, not GPT-3.5.
I have an example with that here: th-cam.com/video/4RAvJt3fWoI/w-d-xo.html
The way you pronounced "ging ist" makes me believe you also speak german. Stimmt's?
Interesting, my German is limited to what I learned watching a TH-cam tutorial 😂
why do you have very less subscribers?
i think google hates you for not using word Gemini instead of openai
I wonder the same thing everyday 😅
Seems counterproductive to teach people OpenAI, which is anything but open, when talking about programmatically working with LLMs. You should steer people towards the actually open LLMs that can be run locally for free. OpenAI is eventually going to be going out of business, once people realize the downsides outweigh any ease of use and first-mover advantage they once had, and the venture capital dries up. The open models with better licenses are already outperforming GPT anyway, and their approach to the technology is rapidly becoming obsolete. Teach people skills that will actually be useful for the open source AI development that will still be around in the future.
While OpenAI conceals their models behind their API, given the SOTA performance of their current models, there are still situations where using them makes sense (at least for now).
I cover open-source technologies in other videos in this series: th-cam.com/play/PLz-ep5RbHosU2hnz5ejezwaYpdMutMVB0.html
The last minute of this video mentions that his next video actually gets into that. That said, I’m hoping we get a Poe and Perplexity api video sometime too.
@fakecubed would there possibly be anyway to ask you a question or email you? you seem to have some advanced knowledge regarding AI and python implementation. I wanted to see if it be possible to get your input on a question I have? I tried going to your TH-cam site couldn't find a way to contact, if possible would be greatly appreciated Thank you
ayo , temp = 2 , is my drunk friend,😂
😂😂
Are you the SHA256? Sorry, geek's joke.
😂😂