PyTorch Project: Handwritten Digit Recognition

Today we train a neural network to classify handwritten digits using PyTorch.
◾◾◾◾◾◾◾◾◾◾◾◾◾◾◾◾◾
📚 Programming Books & Merch 📚
🐍 The Python Bible Book: www.neuralnine...
💻 The Algorithm Bible Book: www.neuralnine...
👕 Programming Merch: www.neuralnine...
💼 Services 💼
💻 Freelancing & Tutoring: www.neuralnine...
🌐 Social Media & Contact 🌐
📱 Website: www.neuralnine...
📷 Instagram: / neuralnine
🐦 Twitter: / neuralnine
🤵 LinkedIn: / neuralnine
📁 GitHub: github.com/Neu...
🎙 Discord: / discord

[@moondevonyt]

jumping into pi torch after tensorflow and keras is like diving into the deep end, but i can see why some might vibe with it
it feels more hands-on and gives you more control
but for some, especially beginners, it might be too low-level and intimidating
that being said, mad respect for the creator for diving deep and walking us through the process
even if it's a hello world example with handwritten digit recognition, the way it's broken down is legit
not everyone's cup of tea, but it's dope for those who want to really get their hands dirty with the nitty-gritty

[@Patrick-wn6uj]

for real bro, I was shocked when he said it is more manual, I thought the reason it popular is because alot of stuff are automated

[@esser50k]

I used your architecture basically and yes, it performs great on the testset. But when inputing my self drawn numbers it sucks, can't get a single one right... I don't understand this..

[@lanreuzamere4994]

Nice one bro! GOD Bless you greatly 🙏

[@xxnotmuchxx]

is the code posted online?

[@tyronebiggumz69]

10/10 video bro! I love your expositions, do you think you could show us how to do spatial recognition with a neural network to extract information from specific entities on forms? For example, extracting information from a tax form or medical record

[@safyantariq99]

I have just started learning PyTorch and setting up a neural network was looking really complex to me. Luckily, your video proved to be greatly helpful. It might have looked like rocket science in the beginning but after running few iterations I am starting to understand the basics of PyTorch involved. Do keep making these videos as I am sure it would prove beneficial for other learners as well.

[@paulthomas1052]

Great project as usual - thanks so much !🤓

[@vladmirbc8712]

What for to use shuffle=True on a test data?

[@Lucas-rz3vl]

I wish you explained some of your steps or reasoning more, I followed along but I feel like I just copied code without truly understanding what is doing what. Still useful, but as a beginner it leaves me wanting!

[@scarysticks66]

please more pytorch videos, do you think of doing lives about kaggle comp?

[@yeyintaung9837]

Awesome ❤

[@pascal831]

Awesome!!🎉🎉

[@matt_milack]

Can you please recommend me sources from which I can study Python for system and network administration? Thank you very much!

[@noerdesamo3858]

How can one modify the system,so that one can see the predictions of the network instead of the end result?

[@candykoko90]

why nobody is sharing videos on Pycharm? 🤔 why is everybody using different versions

[@ErnestoConfused]

Having that massive f string in one line was truly quite not very cash money of you.

[@VitalyEpifanov]

Real

[@Kaustubh56]

I tried to add a code for selecting my own image from the laptop. It's not working. Anybody who has tried it?

[@BonBonEdits-de3xx]

Given groups=1, weight of size [20, 10, 5, 5], expected input[100, 1, 28, 28] to have 10 channels, but got 1 channels instead

[@Kaustubh56]

Similar error for me. What's the fix?
RuntimeError: Given groups=1, weight of size [20, 1, 5, 5], expected input[100, 10, 12, 12] to have 1 channels, but got 10 channels instead

[@чертополох-д4ж]

had similar error, check the CNN function.

[@andiglazkov4915]

😊thanks

[@tarikfarouji8912]

Nice pro❤

[@johnro]

Hi. It’s me again.

[@philtoa334]

Thx_.