Share this page!

Radek Osmulski Profile picture
Twitter logo
17 Feb, 4 tweets, 2 min read
✍️ not sure if this can be helpful to you, but here is how I configure my Ubuntu desktop after install:

✅ use scripts to install a bunch of stuff
✅ install i3 wm
✅ make font not super tiny

Here is the doc: notion.so/How-I-configur…

More thoughts (+ a demo!) in tweets below 😉
If you do decide to follow through with this approach, you are opening yourself up to a world of hurt 🙂

But with time, this will allow you to work on your computer faster and in an (IMHO) much more pleasant way.

Here is a teaser (I am not using the mouse for any of this).
Quick howto:

✅ alt-d and start typing to start a program
✅ alt-enter opens terminal
✅ ctrl-f c creates a new tmux pane
✅ ctrl-f x closes current pane
✅ ctrl-f n, ctrl-f p, ctrl-f <pane num> switches between panes
✅ shift-alt q closes current program
✅ alt-<h, j, k or l> moves between open windows
✅ alt-<num> switches workspaces

Do you need this to do deep learning? No!

Proceed at your own risk 🙂

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with Radek Osmulski

Radek Osmulski Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @radekosmulski

Radek Osmulski Profile picture
18 Feb
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 works...

2/ For starters, let's look at what a neural network (NN for short) does.

An NN is like a stack of pancakes, with computation flowing up when we make predictions.

How does it all work? Image
3/ We show an image to our model.

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 Image
Read 17 tweets
Radek Osmulski Profile picture
12 Feb
THREAD: Sometimes code can seem impenetrable. But there are various ways we can make our life easier.

A common pattern is a function or an object accepting a function. Here is an example.

A 𝚜𝚙𝚕𝚒𝚝𝚝𝚎𝚛 can be a thing that can be called.

But what inputs will it receive? Image
2/ This is not a straightforward question to answer.

But we can learn so much more about what is going on right here in our Jupyter notebook!

Enter 𝚜𝚎𝚝_𝚝𝚛𝚊𝚌𝚎.

We can manufacture an anonymous function and have 𝚜𝚎𝚝_𝚝𝚛𝚊𝚌𝚎 called from inside the 𝙳𝚊𝚝𝚊𝙱𝚕𝚘𝚌𝚔! Image
3/ We can now query the actors.

𝚂𝚘𝚖𝚎𝚝𝚑𝚒𝚗𝚐 turns out to be an 𝙻 (a 𝚕𝚒𝚜𝚝 with superpowers).

It also seems to consist of 𝙿𝚘𝚜𝚒𝚡𝙿𝚊𝚝𝚑𝚜.

So we now know what whatever is passed as 𝚜𝚙𝚕𝚒𝚝𝚝𝚎𝚛 should take.

But we can do better. Image
Read 5 tweets
Radek Osmulski Profile picture
12 Feb
THREAD: How does machine learning 🤖 differ from regular programming? 🧑‍💻

In both, we tell the computer 𝗲𝘅𝗮𝗰𝘁𝗹𝘆 what to do.

But there is one important difference...
2/ In regular programming, we describe each step the computer will take.

In machine learning, we write a program where the computer can alter some parameters based on the training examples.

How does this work?
3/ Our model has a lot of tiny knobs, known as weights or parameters, that control the functioning of the program.

We show the computer a lot of examples with correct labels.

Here is how this can play out...
Read 6 tweets
Radek Osmulski Profile picture
11 Feb
THREAD: Can you start learning cutting-edge deep learning without specialized hardware? 🤖

In this thread, we will train an advanced Computer Vision model on a challenging dataset. 🐕🐈 Training completes in 25 minutes on my 3yrs old Ryzen 5 CPU.

Let me show you how...
2/ We will train on the challenging Oxford-IIIT Pet Dataset.

It consists of 37 classes, with very few examples (around 200) per class. These breeds are hard to tell apart for machines and humans alike!

Such problems are called fine-grained image classification tasks.
3/ These are all the lines of code that we will need!

Our model trains to a very good accuracy of 92%! This is across 37 classes!

How many people do you know who would be as good at telling dog and cat breeds apart?

Let's reconstruct how we did it...
Read 17 tweets
Radek Osmulski Profile picture
5 Feb
THREAD: Can you go from being a treasurer to doing cutting edge DL research through the power of the community? ✨

Sarada Lee (@moodymwlee) is the founder of the Perth ML Group and a Scholar @DataInstituteSF and @Uni_Newcastle.

Her amazing journey began with a selfless act...
2/ In 2016 Sarada founded the Perth ML Group to help others learn.

How can the community support you? 🤗

It can...

✅ help you set up your environments 🧑‍💻
✅ provide technically-sound answers to challenging questions 💡
✅ make learning more fun! 🥳
3/ What are some tips for community participation?

✅ explaining things to others will help you learn 🦉
✅ it's okay to be anxious about sharing your answers publicly - DMs are always an option 📨
✅ experiment with various approaches and learn in a way that suits you best 💡
Read 4 tweets
Radek Osmulski Profile picture
21 Dec 18
This is how little code it takes to implement a siamese net using @fastdotai and @pytorch.

I share this because I continue to be amazed.
Here is a refactored version that will be easier to change
The models above were my naive adaptations of the simple siamese network concept from cs.utoronto.ca/~gkoch/files/m… (screenshot on the left below) to a CNN setting.

On the right is the network from the Learning Deep Convolutional Feature Hierarchies section but using pretrained models
Read 4 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!

Like this author's thread?

Get emails when @radekosmulski has new unrolls!

Check out other threads from @radekosmulski!

Read all threads by @radekosmulski