Lutetium
@deshipu Absolutely! AI is a very broad term, it can mean lots of things and the implementation doesn't have to be "deep learning". Deep learning is particularly interesting because it's simplicity and performance :)
@Inderpreet Singh Do you mean by digging into a lower level. as in "what's inside the neural network"?
People ask about Arduinos and not microcontrollers hence the term AI or deep learning are more broad.
neurons, mostly...
Could focus on say vision based stuff. It is difficult for people to wade through ALL the information out there
some connections, a bunch of weights heh...
In other news I arrive 1 hour late to the hackchat again. Really looking forward to daylight savings.
I assusmed this hackachat was for vision based
@Dan Maloney That's probably the best way to learn. A lot of the concepts you can learn with very simple networks (the simplest is basically just a matrix multiplication!). When it comes to making projects, the neural networks designed to be good at image processing are particularily useful, which is we provide higher level tutorials (We think it can kickstart making projects)
@Josh Lloyd Are you on the reminder email list? You get a reminder 30 minutes before the chat.
Types of networks and then types of tools like TF vs pyTorch. It can be very overwhelming. I have taught ANN to engineering but the current stuff on the net is just too much UNLESS you focus on a particular problem to be solved
@Inderpreet Singh Right, it can be a lot to take in. There are many components. I think its important to try to learn the fundamentals (like a single layer ANN) as well as how to really apply to practical problems using existing architectures
@Dan Maloney The issue is that I'm asleep. Timezones :)
In the JetBot project we actually focus on a component of deep learning that can sometimes go under the radar, which is actually collecting a dataset
I spent a lot of time meandering all over th net until I found this: https://github.com/dusty-nv/jetson-inference
Following the link tutorial. Best intro ever.
@Tegwyn☠Twmffat agreed.
Its a good place to start IF you have a jetson board.
@John I really enjoy the idea of running models on low power hardware, because once the model is trained it really is magnitudes less expensive to run inference. Is there a suggested means of training at this point? Is NVIDIA offering cloud based training? Should it just be done on one's own means, on their own computer perhaps, for now?
@John Is it likely that something such as Training as a Service might be offerred in the future? At reasonable cost, that would be competative vs. me just running it on my own GTX ?
@Josh Lloyd use Nvidia container on AWS
@Josh Lloyd I think it depends what stage you are at. As a gamer, I train on my desktop with GPU to allow me to easily iterate, experiment, etc. Once you've got a lot of data and a complex pipeline you might consider a cloud pipeline or something else.
@Josh Lloyd You can even train some smaller datasets on the Jetson Nano itself when you're just getting started :)
@John I was hoping youd' say that
As @Tegwyn☠Twmffat we provide containers that you can launch on a cloud provider that come with deep learning software (like TensorFlow) pre-installed
We're getting to the top of the hour, which is the official end of the chat, but if @John wants to stick around and answer questions, that fine. Of course he may need to get back to work, so we'll leave it up to him.
Either way, I want to say a huge thanks to John for taking time out of his busy day for us. This was really helpful, both to AI noobs like me and the more seasoned vets.
I did training on Jetson TX2, which was 'OK'.
@John I've built a smart doll house that recognizes IMU gesture patterns to activate items in the doll house and hoping to port it over to Jetson Nano / Tensorflow, it uses really simple architecture: https://maxoffsky.com/research-progress/project-myhouse-a-smart-dollhouse-with-gesture-recognition/
Thanks @John
@John I would assume that anyone with a gaming PC has a far more capable piece of hardware in their desktop than the Jetson Nano. I have a very outdated (at this point) GTX 760, and that has about 3 times as many CUDA cores.
@John Should it be expected that newer CUDA cores on newer hardware are more performant? In whatever the GFX equivilent for DMIPS is?
Sadly, I don't have a Hack Chat lined up for next week yet - we had a host lined up but they had to reschedule. So watch for announcements in case I get a host. Thanks everyone!
@Dan Maloney I'd love to keep talking with everyone. I do need to grab lunch soon :) Please feel free to send me direct messages, but I'll also come back and check this log. This has been awesome!
Thanks @Dan Maloney and @John
thanks all
I'll be pulling a transcript of the session and posting it on the event page. I'll throw a link in here later.
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.