An NN is like a stack of pancakes, with computation flowing up when we make predictions.
How does it all work?
THREAD: How is it possible to train a well-performing, advanced Computer Vision model 𝗼𝗻 𝘁𝗵𝗲 𝗖𝗣𝗨? 🤔
At the heart of this lies the most important technique in modern deep learning - transfer learning.
Let's analyze how it
THREAD: Can you start learning cutting-edge deep learning without specialized hardware? \U0001f916
— Radek Osmulski (@radekosmulski) February 11, 2021
In this thread, we will train an advanced Computer Vision model on a challenging dataset. \U0001f415\U0001f408 Training completes in 25 minutes on my 3yrs old Ryzen 5 CPU.
Let me show you how...
An NN is like a stack of pancakes, with computation flowing up when we make predictions.
How does it all work?
An image is a collection of pixels. Each pixel is just a bunch of numbers describing its color.
Here is what it might look like for a black and white image
Each layer performs computation on the image, transforming it and passing it upwards.
The outputs of a layer are called activations, and the outputs of the last layer have a special meaning... they are the predictions!
We would like for it to have a high value, and for other activations in the last layer to be small...
So far so good! But what about transfer learning?
We know that these layers evolve during training to become feature detectors.
What do we mean by that?
A layer above might have windows that construct shapes from these lines.
We might have a window light up when it sees a square, another when it sees a colorful blob.
Maybe one sliding window will combine lines and detect text... maybe another one will learn to detect faces.
Does all of this sound like a hard task?
But, presumably, once we detect all these lower-level features, we can combine them in a plethora of interesting ways? 🤔
Or trains, planes, and ships. Or blood cell boundaries. Or aneurysms in x-rays. The possibilities are endless!
We let researchers, large corporations, spend millions of dollars to train very complex models.
And then we get to build on top of their work! 😇
But so much for the theory. How does it all work in practice?
The @fastdotai framework downloaded the model for us and removed the top of it (the part responsible for predicting 1 of 1000 classes).
During training, we kept nearly the entire model frozen, and only trained the uppermost part, making use of all the lower level features that were being detected.
Ingenius! 😁
We will hopefully get a chance to explore all of them 🙂
If you enjoyed this thread, let me know please and help me reach others who might also be interested 😊🙏
And the visualizations of what the layers can detect?
Next stop - deciphering how it all works in code and finding ways to further improve our model!
Stay tuned for more 🙂
More from Tech
On Wednesday, The New York Times published a blockbuster report on the failures of Facebook’s management team during the past three years. It's.... not flattering, to say the least. Here are six follow-up questions that merit more investigation. 1/
1) During the past year, most of the anger at Facebook has been directed at Mark Zuckerberg. The question now is whether Sheryl Sandberg, the executive charged with solving Facebook’s hardest problems, has caused a few too many of her own. 2/ https://t.co/DTsc3g0hQf
2) One of the juiciest sentences in @nytimes’ piece involves a research group called Definers Public Affairs, which Facebook hired to look into the funding of the company’s opposition. What other tech company was paying Definers to smear Apple? 3/ https://t.co/DTsc3g0hQf
3) The leadership of the Democratic Party has, generally, supported Facebook over the years. But as public opinion turns against the company, prominent Democrats have started to turn, too. What will that relationship look like now? 4/
4) According to the @nytimes, Facebook worked to paint its critics as anti-Semitic, while simultaneously working to spread the idea that George Soros was supporting its critics—a classic tactic of anti-Semitic conspiracy theorists. What exactly were they trying to do there? 5/
1) During the past year, most of the anger at Facebook has been directed at Mark Zuckerberg. The question now is whether Sheryl Sandberg, the executive charged with solving Facebook’s hardest problems, has caused a few too many of her own. 2/ https://t.co/DTsc3g0hQf
2) One of the juiciest sentences in @nytimes’ piece involves a research group called Definers Public Affairs, which Facebook hired to look into the funding of the company’s opposition. What other tech company was paying Definers to smear Apple? 3/ https://t.co/DTsc3g0hQf
3) The leadership of the Democratic Party has, generally, supported Facebook over the years. But as public opinion turns against the company, prominent Democrats have started to turn, too. What will that relationship look like now? 4/
4) According to the @nytimes, Facebook worked to paint its critics as anti-Semitic, while simultaneously working to spread the idea that George Soros was supporting its critics—a classic tactic of anti-Semitic conspiracy theorists. What exactly were they trying to do there? 5/
"I really want to break into Product Management"
make products.
"If only someone would tell me how I can get a startup to notice me."
Make Products.
"I guess it's impossible and I'll never break into the industry."
MAKE PRODUCTS.
Courtesy of @edbrisson's wonderful thread on breaking into comics – https://t.co/TgNblNSCBj – here is why the same applies to Product Management, too.
There is no better way of learning the craft of product, or proving your potential to employers, than just doing it.
You do not need anybody's permission. We don't have diplomas, nor doctorates. We can barely agree on a single standard of what a Product Manager is supposed to do.
But – there is at least one blindingly obvious industry consensus – a Product Manager makes Products.
And they don't need to be kept at the exact right temperature, given endless resource, or carefully protected in order to do this.
They find their own way.
make products.
"If only someone would tell me how I can get a startup to notice me."
Make Products.
"I guess it's impossible and I'll never break into the industry."
MAKE PRODUCTS.
Courtesy of @edbrisson's wonderful thread on breaking into comics – https://t.co/TgNblNSCBj – here is why the same applies to Product Management, too.
"I really want to break into comics"
— Ed Brisson (@edbrisson) December 4, 2018
make comics.
"If only someone would tell me how I can get an editor to notice me."
Make Comics.
"I guess it's impossible and I'll never break into the industry."
MAKE COMICS.
There is no better way of learning the craft of product, or proving your potential to employers, than just doing it.
You do not need anybody's permission. We don't have diplomas, nor doctorates. We can barely agree on a single standard of what a Product Manager is supposed to do.
But – there is at least one blindingly obvious industry consensus – a Product Manager makes Products.
And they don't need to be kept at the exact right temperature, given endless resource, or carefully protected in order to do this.
They find their own way.
Next.js has taken the web dev world by storm
It’s the @reactjs framework devs rave about praising its power, flexibility, and dev experience
Don't feel like you're missing out!
Here's everything you need to know in 10 tweets
Let’s dive in 🧵
Next.js is a @reactjs framework from @vercel
It couples a great dev experience with an opinionated feature set to make it easy to spin up new performant, dynamic web apps
It's used by many high-profile teams like @hulu, @apple, @Nike, & more
https://t.co/whCdm5ytuk
@vercel @hulu @Apple @Nike The team at @vercel, formerly Zeit, originally and launched v1 of the framework on Oct 26, 2016 in the pursuit of universal JavaScript apps
Since then, the team & community has grown expotentially, including contributions from giants like @Google
https://t.co/xPPTOtHoKW
@vercel @hulu @Apple @Nike @Google In the #jamstack world, Next.js pulled a hefty 58.6% share of framework adoption in 2020
Compared to other popular @reactjs frameworks like Gatsby, which pulled in 12%
*The Next.js stats likely include some SSR, arguably not Jamstack
https://t.co/acNawfcM4z
@vercel @hulu @Apple @Nike @Google The easiest way to get started with a new Next.js app is with Create Next App
Simply run:
yarn create next-app
or
npx create-next-app
You can even start from a git-based template with the -e flag
yarn create next-app -e https://t.co/JMQ87gi1ue
https://t.co/rwKhp7zlys
It’s the @reactjs framework devs rave about praising its power, flexibility, and dev experience
Don't feel like you're missing out!
Here's everything you need to know in 10 tweets
Let’s dive in 🧵
Next.js is a @reactjs framework from @vercel
It couples a great dev experience with an opinionated feature set to make it easy to spin up new performant, dynamic web apps
It's used by many high-profile teams like @hulu, @apple, @Nike, & more
https://t.co/whCdm5ytuk
@vercel @hulu @Apple @Nike The team at @vercel, formerly Zeit, originally and launched v1 of the framework on Oct 26, 2016 in the pursuit of universal JavaScript apps
Since then, the team & community has grown expotentially, including contributions from giants like @Google
https://t.co/xPPTOtHoKW
@vercel @hulu @Apple @Nike @Google In the #jamstack world, Next.js pulled a hefty 58.6% share of framework adoption in 2020
Compared to other popular @reactjs frameworks like Gatsby, which pulled in 12%
*The Next.js stats likely include some SSR, arguably not Jamstack
https://t.co/acNawfcM4z
@vercel @hulu @Apple @Nike @Google The easiest way to get started with a new Next.js app is with Create Next App
Simply run:
yarn create next-app
or
npx create-next-app
You can even start from a git-based template with the -e flag
yarn create next-app -e https://t.co/JMQ87gi1ue
https://t.co/rwKhp7zlys
You May Also Like
Great article from @AsheSchow. I lived thru the 'Satanic Panic' of the 1980's/early 1990's asking myself "Has eveyrbody lost their GODDAMN MINDS?!"
The 3 big things that made the 1980's/early 1990's surreal for me.
1) Satanic Panic - satanism in the day cares ahhhh!
2) "Repressed memory" syndrome
3) Facilitated Communication [FC]
All 3 led to massive abuse.
"Therapists" -and I use the term to describe these quacks loosely - would hypnotize people & convince they they were 'reliving' past memories of Mom & Dad killing babies in Satanic rituals in the basement while they were growing up.
Other 'therapists' would badger kids until they invented stories about watching alligators eat babies dropped into a lake from a hot air balloon. Kids would deny anything happened for hours until the therapist 'broke through' and 'found' the 'truth'.
FC was a movement that started with the claim severely handicapped individuals were able to 'type' legible sentences & communicate if a 'helper' guided their hands over a keyboard.
For three years I have wanted to write an article on moral panics. I have collected anecdotes and similarities between today\u2019s moral panic and those of the past - particularly the Satanic Panic of the 80s.
— Ashe Schow (@AsheSchow) September 29, 2018
This is my finished product: https://t.co/otcM1uuUDk
The 3 big things that made the 1980's/early 1990's surreal for me.
1) Satanic Panic - satanism in the day cares ahhhh!
2) "Repressed memory" syndrome
3) Facilitated Communication [FC]
All 3 led to massive abuse.
"Therapists" -and I use the term to describe these quacks loosely - would hypnotize people & convince they they were 'reliving' past memories of Mom & Dad killing babies in Satanic rituals in the basement while they were growing up.
Other 'therapists' would badger kids until they invented stories about watching alligators eat babies dropped into a lake from a hot air balloon. Kids would deny anything happened for hours until the therapist 'broke through' and 'found' the 'truth'.
FC was a movement that started with the claim severely handicapped individuals were able to 'type' legible sentences & communicate if a 'helper' guided their hands over a keyboard.
