what next now tho? i co-incidentally completed watching his last vid of training up gpt-2 today :-) .
Neural Networks: Zero to Hero
21–30 of 88 posts
Re: Neural Networks: Zero to Hero
#22Earlier quoted context omitted.
https://xkcd.com/1053/
Is it weird that I now know exactly which xkcd it will be just with conversational context? Granted I'm a bit of a Randall Munroe content addict, but it's become second nature now.
I think, in the spirit of the xkcd, you were supposed to pretend you have never heard of it
Re: Neural Networks: Zero to Hero
#23should have a (2022) label
it's an ongoing project, the last lecture is about i year old
Re: Neural Networks: Zero to Hero
#24Saving this
Re: Neural Networks: Zero to Hero
#25I’ve gone through this series of videos earlier this year.
In the past I’ve gone through many “educational resources” about deep neural networks - books, coursera courses (yeah, that one), a university class, the fastai course - but I don’t work with them at all in my day to day.
This series of videos was by far the best, most “intuition building”, highest signal-to-noise ratio, and least “annoying” content to get through. Could of course be that his way of teaching just clicks with me, but in general - very strong recommend. It’s the primary resource I now recommend when someone wants to get into lower level details of DNNs.
Re: Neural Networks: Zero to Hero
#26[flagged]
Re: Neural Networks: Zero to Hero
#27Earlier quoted context omitted.
Is it weird that I now know exactly which xkcd it will be just with conversational context? Granted I'm a bit of a Randall Munroe content addict, but it's become second nature now.
You're not alone. At this point I'm starting to recognise some by number as well.
Communicating the number of XKCD comics, especially in binary, is a very efficient and energy-preserving way to get a laugh.
A: 10000011101 !
B: ACK. LOL !
Re: Neural Networks: Zero to Hero
#28[flagged]
Re: Neural Networks: Zero to Hero
#29I'm not sure how it compares, but another option is the Hugging Face learning portal [0]. I'm doing the Deep RL Course and so far it's pretty straight forward (although when it gets math heavy I'm going to suffer). [0] - https://huggingface.co/learn
I found the Karpathy videos very approachable. While I did study CS, I never went deep into ML. My main knowledge about matrices is for graphic development, so vectors and matrices up to 4x4 in size only. But following the videos, starting to learn about backprob and building the tiny GPT was understandable to me.
Karpathy's lessons are great to really grog the background and underlying basics. They do not go into the many libraries available, the course you link might be more practically applicable.
Re: Neural Networks: Zero to Hero
#30Earlier quoted context omitted.
https://xkcd.com/1053/
Is it weird that I now know exactly which xkcd it will be just with conversational context? Granted I'm a bit of a Randall Munroe content addict, but it's become second nature now.
I know exactly what you mean. It broke my workflow too.