How to Build a Neural Network in Python from SCRATCH (In-Depth Tutorial)

Schedule a 1-on-1: calendly.com/e...
Discord: / discord
Github Repo: github.com/Inf...
I use Nvidia Shadowplay and OBS studio to record
Record with 60fps + 4k res + 50 bitrate
I edit with Davinci Resolve
Socials
Twitch: / elliotcodes
KZbin: / @elliotarledge
Twitter: @elliot_arledge
Equipment:
Mouse: Glorious Model O
Keyboard: Apex Pro TKL (omnipoint switches)
Headset: HyperX Cloud ll
Webcam/Mousecam: Logitech Brio
Mousepad: Aqua Control Plus
Primary Monitor: Acer v279-144hz-1ms
Secondary Monitor: 144hz dell 24 inch monitor
Router: ROG Rapture GT-AC5300
PC Specs:
Graphics Card: Nvidia Geforce 3070 (FHR)
Processor: Ryzen 5 3600x @ 3.8GHz, 6 cores, 12 threads
Motherboard: Msi Tomahawk b450
Memory: 32GB-3200mhz
Storage: 500GB nvme SSD & 1 TB Western Digital SSD
Power Supply: 650 watts +80 Gold

[@irisheyez1231]

This is exactly what I was looking for! Wanted to get a very rudimentary practical understanding of a neural network. Glad to find this. When I ran in my Git Bash, starting at epoch 6300, it returned at loss of 0.0000.

[@earnstein7607]

As someone who loves mathematics and doesn't know how to apply it in the programming world, having seen some of your videos, I'm inspired to give machine learning my all. Thank you tutor

[@earnstein7607]

Hopefully, you'll post more in-depth tutorials in the future❤🎉

[@elliotarledge]

For sure, I highly recommend you join the discord server in the description. We talk about this stuff all the time

[@elliotarledge]

I forgot to add it so just join here: discord.com/invite/autFyVVZ

[@earnstein7607]

@@elliotarledge I'm in , thank you so much.

[@chrispassante8532]

How did you link your visual studio and command prompt. Also, how can I download anaconda prompt? Great video btw. Keep it up.

[@elliotarledge]

Just download anaconda from the website. It's pretty well documented. If you want to run python files in a separate command prompt you just use the cd command until you are in the same folder. cd'ing is something you typically learn early on.

[@yisyvonne]

Thank you so much for this! Subscribed b/c your channel is the best and your tutorials are so well done. I'm a Mathematics Undergrad and I'm very interested in Machine learning and its applications in topology and linguistics.

[@earnstein7607]

This is the channel I've been looking for.

[@sabaokangan]

Thank you so much for sharing your guidance with us on KZbin ❤‍🔥 from Kyoto

[@roberthill5373]

Thank you Elliot for this concise intro on this topic. Is it possible to rewrite or implement this NN using Mojo programming?

[@elliotarledge]

some people in the community have tried, nothing is certain though. you should stay up to date with the modularml github to get in on it faster.

[@foslight4272]

thank you man. for LLM tutorial. it's lot to me. :)

[@technykgs]

Thank you very much!

[@thtguyfrmdwnthestrt]

Great content!

[@ClayShoaf]

So if I understand this correctly, with this example model, once trained, you should be able to put in a random float32 number and the model will multiply it by 2. Is that correct?

[@pencilgame]

That's what I was wondering!

[@elliotarledge]

that is correct, in this video we specifically go through the computation and learning happening, not the data pipeline.

[@forlanbeckham1504]

If possible, make a video of deep fake detection using face forensics++ dataset(image and video). and how to deploy this model and integrate into a website or mobile app.

[@elliotarledge]

I already have deepfakes on my list of content to create. the research has been done. but not super to compile it into something you can just plug and play with

[@forlanbeckham1504]

@@elliotarledge Within one month, will the video come?

[@elliotarledge]

no guarantees, you might want to join the discord server in the desc to get more updates as to what I'm working on

[@forlanbeckham1504]

@@elliotarledge I want to join the Discord server, but where is the invite link?

[@elliotarledge]

@@forlanbeckham1504 Apologies, you should be able to see it now :)