I invested a ton of time into preparing this training so I hope you find it valuable! Let me know what you like and what could be improved in the future, please :)
I commend the hard work you put into this. You synthesized a lot of complexity down to something that is easy to consume and build off for our own projects. Thanks again for this!
Exceptionally clear and concise, despite being 2+ hours long. Excellent introduction to LLMs holding enough advanced bits to keep more seasoned viewers hooked as well!
Hi Jon, I haven't gone through this video yet, but from other comments I learn that this is a great tutorial. I have seen your other videos on the math behind ML and it is very impressive. You are doing this as a service to educate people on concepts that is not easy to understand sometimes. This inspires me to be of service to others in whichever way I can.
Thank you so much for your kind words, John! It's truly rewarding to hear that the content is not only helpful but also inspiring. I'm glad it motivates you to be of service to others as well. Enjoy the video when you get a chance to watch it, and I hope it provides value to your learning journey in ML. Cheers!
Extremely well done Jon, I was able to immediately get to work and learn thanks to your concise explainations of everything. I'm blown away by how much you covered with out any fluff in a manner that was very approachable.
Hey John, I wanted to express my sincere gratitude for the incredible training session on generative AI with LLMs you shared with us. The depth of knowledge and insights you provided in so short time was truly amazing.
Regarding your question about what could be improved, based on my experience and the SDSP episode with Cole, I would say: reduce the amount of content while diving deeper into details. Again, thank you for sharing this enlightening material with us.
This is the best instructions for getting started with LLM for laymen. A true lifesaver. Appreciate your generosity to unlocking the treasure trove. Can't wait to dig deep on this tut. Difficult to suggest improvements but I do want to know, say how to allow the custom inputs of 10 different type of restaurants to have stronger influence on the output. Since these 10 restaurants speak the same way but differ slightly. Their outputs should be similar. Use cases, custom chatbots for each industry and save time for manual inputs for the new restaurant. Subscribed and liked. Thanks a lot!
Thank you so much for your thoughtful comment! I'm glad to hear that you found the tutorial helpful, and I appreciate your support :) As for your question about customizing the model's output, if you're trying to ensure that ten different types of restaurants have a stronger influence on the model's output, and those restaurants speak in similar ways but with slight differences, you'll want to focus on fine-tuning the model using data that reflects these specificities. Fine-tuning is a process in which the language model is further trained on a specific dataset after its initial training. In this case, you would gather and use textual data from the ten different types of restaurants. This data could be reviews, menus, or any other text that encapsulates the unique language and nuances of each restaurant type. Once you've collected and prepared your data, you'll need to fine-tune the model on it. There's no one-size-fits-all approach to this, as the best fine-tuning process will depend on factors like your specific use case, the amount and quality of your data, and the capabilities of the model. Episode #674 of my podcast (Super Data Science) might be useful to you. Also, don't forget about adjusting the model's parameters when generating outputs. For instance, the "temperature" parameter can control the randomness of the output - a lower temperature will produce more focused and deterministic outputs, while a higher temperature will yield more diverse and creative results.
Hi Jon, Many thanks for your efforts and time put to prepare this content. As always your sessions are simple, crisp, complex concepts explains very nicely along with hands on example.
Hi Jon! I tried to run the example in Colab but the GPT4All-inference file read that there is no module named: ‘nomic.gpt4all’. So it won’t take me past the second step of the download. I’m sure it’s user error, but could you point me in the right direction please? Thanks!
It's really a great training. It covers all different aspects in graceful manner. Greatly appreciated. Thanks you so much! Just curious to know if any plan to make tutorial on "Deep Learning with PyTorch and TensorFlow” available to community.
Thanks! Glad you liked it! I had a technical issue that prevented me from being able to record my "Deep Learning with PyTorch and TensorFlow” training at ODSC East. Hopefully I'll have a chance to be able to record it for RUclips soon. In the meantime, you can access this training via O'Reilly. I have a training there online called "Deep Learning with TensorFlow, Keras, and PyTorch". If you don't already have access to the platform, you can use my code "SDSPOD23" to get a free 30-day trial.
Hi Jon, thank you for all your awesome content. Trying to follow along but get a module not found error when first importing the model in google colab. Any ideas what's changed?
Sir, can I start ML parallely with this ML foundation series or after completing this maths from algebra then I start ML . my bigest problam 😞 . please answer
This lecture is relatively advanced if you're just getting started in ML, but you could watch it first if you'd like inspiration or what your skills will enable you to do! Then, I'd dig into my ML Foundations series, starting with Linear Algebra.
I will get back to releasing Probability for ML and then Statistics for ML videos on RUclips as soon as I can. In the meantime you can access all my "Probability and Statistics for ML" content in the O'Reilly platform. If you don't already have access, you can use my special code "SDSPOD23" to get a free 30-day trial.
Hey sir,i have a lot of questions😅like, Is AI and ML is good or bed (without any degree or certificate) ? I can learn AI and ML without a laptop or computer ? And most important and stupid question is What is the best and free skill i learn without any laptop or computer and certificate or degree (i am 18yo boy who needs to become successful,but how) ? Please sir,give me a straight way to be a 1% in this addicted world. I need to become a successful man,but how i don't know. Please give me a straight path. Thanks 😊
Hi there! Sign up for my email newsletter on jonkrohn.com for updates. I hope to be able to record more of my ML Foundations content for RUclips soon. In the meantime, you can get the entire series online via O'Reilly. You can use my code "SDSPOD23" to get a free 30-day trial, which is plenty of time to cover the stats content and beyond.
Hi EclypH! If you're asking about the Statistics content from my ML Foundations curriculum, the status on various platforms is provided in the "Where and When" section of my ML Foundations GitHub repo's README file (I would put a link but it seems like RUclips sometimes automatically removes my comments when I do). Essentially, the full studio-recorded version of my ML Foundations curriculum is available today within the O'Reilly online platform (link also provided in the GitHub repo as well as a code to get a free 30-day trial of the O'Reilly if you don't already have access). I haven't had bandwidth recently to record the home-recorded version of my ML Foundations curriculum (which I publish to RUclips and Udemy) but hope to have the bandwidth soon. You can sign up for my email newsletter on my website to be sure not to miss when I start publishing ML Foundations videos on RUclips again.
I invested a ton of time into preparing this training so I hope you find it valuable! Let me know what you like and what could be improved in the future, please :)
Thanks Jon for sharing 🤩
@@square007tube most welcome!
Thank you , undoubtedly one of the best tutorials i have ever seen
@@rezabagherian3331 cheers, Reza! That's a huge compliment, thank you :)
I commend the hard work you put into this. You synthesized a lot of complexity down to something that is easy to consume and build off for our own projects. Thanks again for this!
Exceptionally clear and concise, despite being 2+ hours long. Excellent introduction to LLMs holding enough advanced bits to keep more seasoned viewers hooked as well!
Thanks, Jord! And glad to hear it hit the mark for a range of levels!
Hi Jon, I haven't gone through this video yet, but from other comments I learn that this is a great tutorial. I have seen your other videos on the math behind ML and it is very impressive. You are doing this as a service to educate people on concepts that is not easy to understand sometimes. This inspires me to be of service to others in whichever way I can.
Thank you so much for your kind words, John! It's truly rewarding to hear that the content is not only helpful but also inspiring. I'm glad it motivates you to be of service to others as well. Enjoy the video when you get a chance to watch it, and I hope it provides value to your learning journey in ML. Cheers!
Extremely well done Jon, I was able to immediately get to work and learn thanks to your concise explainations of everything. I'm blown away by how much you covered with out any fluff in a manner that was very approachable.
Wow, thanks! Super stoked you could dive right in, Dave! Aim is to keep it clear and fluff-free... More coming your way :)
Hey John, I wanted to express my sincere gratitude for the incredible training session on generative AI with LLMs you shared with us. The depth of knowledge and insights you provided in so short time was truly amazing.
Regarding your question about what could be improved, based on my experience and the SDSP episode with Cole, I would say: reduce the amount of content while diving deeper into details.
Again, thank you for sharing this enlightening material with us.
Awesome Jon, This is like A-Z in one session you have covered. I really appreciate your time and efforts to grow the community.
Pure 🔥 John! Just like your Deep Learning Illustrated!🎉🏆
Thank you so much Jon. I appreciate your time and efforts. Please keep uploading videos on NLP.
Thank you, I will!
This is the best instructions for getting started with LLM for laymen. A true lifesaver. Appreciate your generosity to unlocking the treasure trove. Can't wait to dig deep on this tut. Difficult to suggest improvements but I do want to know, say how to allow the custom inputs of 10 different type of restaurants to have stronger influence on the output. Since these 10 restaurants speak the same way but differ slightly. Their outputs should be similar. Use cases, custom chatbots for each industry and save time for manual inputs for the new restaurant. Subscribed and liked. Thanks a lot!
Thank you so much for your thoughtful comment! I'm glad to hear that you found the tutorial helpful, and I appreciate your support :)
As for your question about customizing the model's output, if you're trying to ensure that ten different types of restaurants have a stronger influence on the model's output, and those restaurants speak in similar ways but with slight differences, you'll want to focus on fine-tuning the model using data that reflects these specificities.
Fine-tuning is a process in which the language model is further trained on a specific dataset after its initial training. In this case, you would gather and use textual data from the ten different types of restaurants. This data could be reviews, menus, or any other text that encapsulates the unique language and nuances of each restaurant type.
Once you've collected and prepared your data, you'll need to fine-tune the model on it. There's no one-size-fits-all approach to this, as the best fine-tuning process will depend on factors like your specific use case, the amount and quality of your data, and the capabilities of the model. Episode #674 of my podcast (Super Data Science) might be useful to you.
Also, don't forget about adjusting the model's parameters when generating outputs. For instance, the "temperature" parameter can control the randomness of the output - a lower temperature will produce more focused and deterministic outputs, while a higher temperature will yield more diverse and creative results.
Hi Jon, Many thanks for your efforts and time put to prepare this content. As always your sessions are simple, crisp, complex concepts explains very nicely along with hands on example.
Hey Abhijit, thanks a ton for the props! Super happy to hear you're finding my videos helpful.
Dude you still got it!!! Great video!!!
Haha, thanks Tad :D
Came after listening to your podcast
As usual Awesome John !
Thanks :D
Thank you, Jon. It was fascinating.
You're welcome! Glad you found it fascinating :)
Loved this, really insightful and helpful. Thank you! And did anyone tell you you sound a lot like Sean Carroll, which is also very cool :-)
That is a new one, but a very flattering comparison, thank you :)
Jon Krohn! Thanks my teacher! I'm going to enjoy and eat this content! ❤
haha I hope you find it to be a delicious treat!
@@JonKrohnLearns thanks man!
Thank you sir,your Deep learning illustrated(all 3 videos ) are interested ,now iam foĺlowing all your lectures
Nice! Glad you're enjoying my content :)
Hi Jon! I tried to run the example in Colab but the GPT4All-inference file read that there is no module named: ‘nomic.gpt4all’. So it won’t take me past the second step of the download. I’m sure it’s user error, but could you point me in the right direction please? Thanks!
I am encountering the same issue
Great session Jon! This was very helpful 😁
Glad you enjoyed it!
It's really a great training. It covers all different aspects in graceful manner. Greatly appreciated. Thanks you so much!
Just curious to know if any plan to make tutorial on "Deep Learning with PyTorch and TensorFlow” available to community.
Thanks! Glad you liked it!
I had a technical issue that prevented me from being able to record my "Deep Learning with PyTorch and TensorFlow” training at ODSC East. Hopefully I'll have a chance to be able to record it for RUclips soon.
In the meantime, you can access this training via O'Reilly. I have a training there online called "Deep Learning with TensorFlow, Keras, and PyTorch". If you don't already have access to the platform, you can use my code "SDSPOD23" to get a free 30-day trial.
@@JonKrohnLearns Thanks!
God bless you for putting this up. I have no other means to train
Hi Jon, thank you for all your awesome content. Trying to follow along but get a module not found error when first importing the model in google colab. Any ideas what's changed?
great work
Sir, can I start ML parallely with this ML foundation series or after completing this maths from algebra then I start ML . my bigest problam 😞 . please answer
This lecture is relatively advanced if you're just getting started in ML, but you could watch it first if you'd like inspiration or what your skills will enable you to do! Then, I'd dig into my ML Foundations series, starting with Linear Algebra.
It was awesome! A lot but I'll review the rest later. I need a Udemy course though : - )
Well, then make one - I am unlikely to have the bandwidth myself! Glad this training was a good headstart :)
@@JonKrohnLearns thank you Dr. Jon!
Sir make videos on statistics for machine learning
I will get back to releasing Probability for ML and then Statistics for ML videos on RUclips as soon as I can. In the meantime you can access all my "Probability and Statistics for ML" content in the O'Reilly platform. If you don't already have access, you can use my special code "SDSPOD23" to get a free 30-day trial.
Hey sir,i have a lot of questions😅like,
Is AI and ML is good or bed (without any degree or certificate) ?
I can learn AI and ML without a laptop or computer ?
And most important and stupid question is
What is the best and free skill i learn without any laptop or computer and certificate or degree (i am 18yo boy who needs to become successful,but how) ?
Please sir,give me a straight way to be a 1% in this addicted world.
I need to become a successful man,but how i don't know.
Please give me a straight path.
Thanks 😊
sir what about statistics for ml Foundation series please answer 😀
Hi there! Sign up for my email newsletter on jonkrohn.com for updates. I hope to be able to record more of my ML Foundations content for RUclips soon. In the meantime, you can get the entire series online via O'Reilly. You can use my code "SDSPOD23" to get a free 30-day trial, which is plenty of time to cover the stats content and beyond.
Sir, can I start ML parallelly with this or after completing this math from algebra then I start ML ? @@JonKrohnLearns
whats gonna happen with the statistics course?
Hi EclypH! If you're asking about the Statistics content from my ML Foundations curriculum, the status on various platforms is provided in the "Where and When" section of my ML Foundations GitHub repo's README file (I would put a link but it seems like RUclips sometimes automatically removes my comments when I do).
Essentially, the full studio-recorded version of my ML Foundations curriculum is available today within the O'Reilly online platform (link also provided in the GitHub repo as well as a code to get a free 30-day trial of the O'Reilly if you don't already have access). I haven't had bandwidth recently to record the home-recorded version of my ML Foundations curriculum (which I publish to RUclips and Udemy) but hope to have the bandwidth soon. You can sign up for my email newsletter on my website to be sure not to miss when I start publishing ML Foundations videos on RUclips again.