Machine Learning🤖🧠 can be scary for developers!😱
Let me help you with this!💪🏽
If you are a Mobile📱, Web🕸️ or Backend🖥️ developer, you can use ML from the comfort of your preferred IDE without needing the heavy math part of it!
🤩🤯
Imagine you need to load a very large (eg: 5.7GB) csv file to train your model!🤔
This is a very common problem in real world situations and also in many Kaggle competitions!
How can we use Pandas 🐼 effectively to do that?
Let's dive in…
[2 effective min]
1/10🧵
Today it's my 6 year Google anniversary!
I started as an android developer 📱🤖 and today I'm an ML developer 🧠🤖
I learned a lot of technical stuff (A LOT!), but there's much more than that.
Let me tell you some of the things I learned...
1/11🧵
My goals for this year are to help you learn:
- Python 🐍
- Machine Learning 🤖🧠
- TensorFlow 🤖🛠️
If any of these interest you, follow me so you don't miss the content!
And share with your friends so they can learn too!🤝🏾
"How do I learn Python?"
🤔
3 tips:
• Do one basic tutorial 🤓
• Practice, practice, practice 💪🏾
• Start/Find a project to apply what you learned 🧐
"Ok Gus, how about some links?"
👇
Oi
@cauemoura
e
@FeCastanhari
Sou muito fã de seus vídeos!
Vi alguns cortes de vcs no
@Podpah
e entendo sua preocupação com AI.
Não concordo, mas entendo
Se quiserem conversar e entender o ponto de vista de alguém que trabalha nessa área todo dia, minha DM está aberta.
Hi👋🏾
I'm Gus and work as a Developer Advocate on the Google AI team
I share content about:
🤖🧠Machine Learning
🤖🧡TensorFlow
🐍🧑🏾💻Python
🥑🤝DevRel
I'm almost at 16K followers!
Can you help me get there?
leave your intro in the comments and share to help me reach it!
🥺🥺
Machine learning goes beyond Deep Learning and Neural Networks
Sometimes a simpler technique might give you better results and be easier to understand
A very versatile algorithm is the Decision Forest
🌴🌲🌳?
What is it and how does it work?
Let me tell you..
[7 min]
1/10🧵
How can we change a 3 minute load time to 1 second?
⚡️⚡️⚡️🤯
As a Pandas🐼 user, the read_csv method might be very dear 💕to you.
But even with a lot of tuning, it will still be slow.
Let's make it faster!!!
[1 ⚡️ min]
1/7🧵
Do you need a Python course to start 2023?
So, here is a tip💡 for you:
➡️Learn Python with the Kaggle course:
Pros:
• Kaggle Kernels allows you to try the code in the browser
• Very good pace of content
• Fun and challenging puzzles
If you code in Python🐍 for more than an hour, you probably have seen the keyword lambda
In essence lambda enables us to create an anonymous function or function without a name
When and why should you use it?
[1 (⚡️) min]
1/5🧵
I like Python🐍 because:
-> Very easy to read. It looks like pseudo code
-> Fast to create scripts for automating work
-> Huge amount of libraries for all kinds of specific tasks (NumPy, Beautiful Soup, Pandas, TF, Pillow, Librosa)
Why do (or don't 👀) you like Python?
🤔
What are Python🐍 decorators🎀?
Decorator is an Object Oriented pattern that allows behavior to be added to individual objects
They can be more efficient than subclassing and in some cases it can make your code 1000 faster!👀🤯
[1 min]
1/5🧵
Everyone that does some Data Analysis or Machine Learning knows the Pandas library 🐼
One thing that not everyone is aware of is how to use it efficiently!
Have you thought about how much memory your dataframe is using? 🤔
How to use less? 🗜️
Let me show you…
[2 min]
1/8🧵
Excellent opportunity to learn Python basics and maybe charge your career!
The curriculum is very interesting
- Python basics
- git
- cloud
And you get a certificate!
Very good!
Today is my birthday!
As my gift to you, I created this thread with all my NLP posts of this week to give you some technical content for your weekend!
[2 minutes]
1/11🧵
Hi everyone, I don't like asking but can you help me reach the magic 10K? I'm almost there!
In return, I'll post daily quality content about:
- Python🐍
- Machine Learning🧠🤖
- Career Development👨🏾🔬
Share to increase the reach!
I prefer working with a good programmer that is a nice person overall than with an amazing 10x 🧑🏽🎤 programmer that is an a**hole
People skills matter!
I was telling a friend that one cool feature from Python is list slice notation
So instead of just posting the link I decided to do a brief explanation.
[5 min]
Python's regular array indexing as in multiple languages is: a[index]
a = [0, 1, 2, 3, 4]
a[0] == 0
1/12🧵
New year and you want to start your Machine Learning 🤖🧠 journey
Here are the 3 things you'll need to start:
1⃣ - A good ML Tutorial/Course
2⃣ - An environment to code
3⃣ - The Secret sauce
Let me give you some directions🎯
[1 minute]
1/7🧵
Effective Pandas🐼 tip [4]:
When you start to work on a real dataset with more data (millions of records) and want to run a transformation on the data, what should you do?
Let me tell you how to make your execution more than 19000 times faster!!
🤯🤯🤯
[1 effective min]
1/7🧵
💡Machine Learning🤖🧠 models are designed to make predictions🎱🔮 and not giving you insights about the data!
If you have a bunch of data and want to find out "interesting" things about it, use Statistical Approaches!📊📈
🧵
Are you interested in:
- Machine Learning🤖🧠?
- have fun?👍🏾🥳
- help the community?🤝👥
If so you should definitely look into the Google summer of code!👀
I'm one of the mentors for TensorFlow so there's a chance we can collaborate on cool projects 😉
🟩 Are you a
#Developer
new to open source?
We've compiled a list of ways to increase your chances of being selected as a 2022 Google Summer of code Contributor.
🟨 Get application tips and learn more about the program.
👉
How deep do you need to understand Machine Learning to start using it?🤔
Let's use an analogy to explain this
Imagine 3 personas:
• 🧑🏾✈️Driver
• 👩🏾🔧Mechanic
• 👩🏾🔬Car Manufacturer
[1 minute]
1/5🧵
When you start reading about Machine Learning one term comes up frequently: NLP
What is NLP?
• Natural Language Processing
• NLP = Linguistics + Computer Science + AI
• The field that studies how can computers understand text
Some history about it📖
[30 sec]
1/5🧵
If you are looking for something to learn during the weekend,
How about on-device Machine Learning?
You'll only need some understanding of ML and some of Mobile development.
Let me give you all the pointers in FAQ style:
[reading: 5.84 min]
1/12🧵
The main reason why Python🐍 is so popular with Machine Learning🤖🧠 developers is because it's so easy to read.👓
But another feature that really helps is the collections manipulation tools, specifically Slicing!🔪
Let me show you how it works
👇🏾👇🏾👇🏾
[2 min]
1/11🧵
@EmmaBostian
Sorry Emma but I disagree.
One example is that big-o is part of problem solving, otherwise you'd use brute force to solve every problem and that's not always the case.
My Christmas gift arrived and right on time for my eoy break! 🤩
Thanks
@fchollet
for the great work and sharing your knowledge with the community!
The book looks great!
💡Machine Learning🤖🧠 models are designed to make predictions and not giving you insights.
If you have a bunch of data and want to find out "interesting" things about it, use Statistical approaches!
[20 seconds]👀
1/4🧵
Machine Learning🤖🧠 models can be classified regarding how much human supervision they need. 🧐
This affects the algorithms used and the types of tasks that it can solve.
You can categorize Machine Learning models in 4 major categories:
👇🏽
🤖🧠🧮
To start with Machine Learning you don't need to understand the Math part in all the details
But if you want to have a better grasp of why and how the "magic" works, here are the topics you'll need to learn:
👇
Why do we need the yield keyword in Python🐍?
To understand it, you need to know what is a:
• Generator
• Iterables
• Lazy evaluation
Quick thread to start the week...
[2 minutes]
1/10🧵
Let's start with some theory
I've been working with ML on the Audio domain and at first I couldn't understand much but as I kept reading I managed to figure out some things.
Let me share some of the basic theory with you:
[10 minutes read]
1/n🧵
Machine learning goes beyond Neural Networks
Sometimes a simpler technique might give you way better results and much easier to understand
A very versatile family of algorithms are Decision Forests
🌴🌲🌳?
Let me tell you how they work…
1/10🧵
Today I learned that you can unpack assignments using a starred expression! 🤯
It's called catch-all unpacking ()
This is better than using slicing and indexing in some situations because it's easier to read and maintain.
Imagine you need to load a very large (eg: 5.7GB) csv file to train your model!🤔
This is a very common problem in real world situations and also in many Kaggle competitions!
How can we use Pandas 🐼 effectively to do that?
Let's dive in…
[2 effective min]
1/10🧵
This is one reason why people are afraid of contributing to the community
-Divam did a great job! spent their time creating something super cool and shared with everyone
-Just to have someone come and shi* on their head for no reason!
This is very sad!
Don't be like that!
@nilmoretto
Meu chute sem ser biologo mas sendo cientista da computacao é:
-> Leptoglossus occidentalis tambem conhecido como western conifer seed bug
👈🏾🌱🪲
Meu metodo pra descobri:
- baixa a image (ou da um screenshot)
- usa esta na pagina aqui:
📱+🤖🧠
Pixel 7 phones were announced this week and if you like Machine Learning, specifically on-device ML, this release brings a lot of cool features!
These are my top 5 best ML features:
👇
When you use environments like Google Colab, Jupyter Notebooks or Kaggle Kernels, you have access to Magic Commands🪄
Let me explain how they work⚙️,
why they matter🤔
and how to create one yourself🏗️
[1 min]
1/12🧵
What is JAX?
JAX is Autograd and XLA, brought together for high-performance numerical computing and ML research. It provides composable transformations of Python+NumPy programs: differentiate, vectorize, parallelize, JIT compile to GPU/TPU, and more.
🤔🧐
#30DaysOfJAX
1/11🧵
Why do you need a train, validation and a test dataset?
The train is obvious, it's what your model learns from.
To avoid overfitting (memorizing) or underfitting you need to try your model with unseen data and that's why you need validation/test datasets.
But why both?
1/4🧵
Today is my daughter's birthday!
🎂🎉🎉🎉
She wants to be a scientist so we created her future lego set for when she's very famous!
Leave a like for her! I'll print it later as a gift!
When I was studying for my technical interviews I used a couple of different resources
Here is a list of the 4 most important ones..
[And some bonus ones! 🎁🎁]
[1 minute of investment]
1/8🧵
One concept that helped me learn Machine Learning is:
-> Transfer Learning 🧐🧠➡️🤓
Let's make an analogy: learning a new programming language.
[30 sec]
1/5🧵
@svpino
Since I skipped the ones for learning the language, my main reference is Effective Python.
It's a great book if you already know Python and want to write better code.
I failed my first Google interview process 😭
Sometimes we fail and that's fine!👍🏽
I've failed many times and in crucial life changing moments 😱
The tip I can give you is:
✅Learn from your failures and grow with them💪🏽
⬆️Use what you've learned and do better next time!
@msoares
Cara quite legal! Excelente trabalho!
Em 2013 eu e o
@jay_santos
criamos um site pra analisar gastos de deputados e também tivemos diversos problemas mas vimos vários gastos estranhos com combustível e tudo que vc imaginar
Infelizmente por falta de tempo matamos o projeto...
Imagine you want to pair two sequences into only one:
a = [💤,🦠,🧮]
b = [🥱,🤒,🤓]
c = [[🥱,💤],[🦠,🤒],[🧮,🤓]]
Python🐍 has the Zip function to do just that!
c = zip(a,b)
But what if you want to go from c back to a and b?
Are you afraid of the code interviews?😱😱😱
You're not alone!🫂
I've done many on both sides (candidate and interviewer)!
Here are some tips that helped me succeed:
👍🏾👍🏾👍🏾
[1.5 minutes]
1/10🧵
Have you ever wanted to add a math description for your
#Python
function but found it time-consuming to do so? latexify_py allows you to add LaTex math description with only one decorator.
Link to latexify_py:
It's balloons-on-my-profile-page day!!!
🎈🎈🎈🥳🎉🎊🪅
The only 🎁 I want is to be able to keep sharing good Machine Learning 🤖🧠 and Python🐍 content that helps you grow everyday!
We’re excited to announce 𝗚𝗲𝗺𝗶𝗻𝗶:
@Google
’s largest and most capable AI model.
Built to be natively multimodal, it can understand and operate across text, code, audio, image and video - and achieves state-of-the-art performance across many tasks. 🧵
To train a ML model, you need a train, validation and a test dataset
The train split is clear, it's what your model learns from
To avoid overfitting or underfitting you need to try your model with unseen data and that's why you need validation/test datasets
But why both?🤔
🧵
Hoje conheci ao vivo uma das pessoas mais incríveis aqui do Twitter:
@import_robs
Aprendi tanto sobre física e astronomia que estou quase achando Interestelar fácil de entender!
🤯
Obrigado pelo papo Roberta!
Sua paixão pelo seu trabalho é inspiradora!
🤖🧠+🌌
OMG!!!😱
I'm 10 followers away from 5K!
Do you think you can help me get there?
Please share this with your friends that are interested in Machine Learning, Python and tech career tips!
Do you wanna learn a cool Python🐍 feature today?
➡️ list slice notation🔪
Python's regular array indexing as in multiple languages is: a[index]
a = [0, 1, 2, 3, 4]
a[0] == 0
👇
This week so far:
- Presented 7+ hours of TensorFlow content to universities and big companies 🤖🧠🥑🗣️
- Joined an internal hackathon, with two projects 👨🏾💻 (14 teams participated)
- Got first and second place in the internal hackathon! 🥇🥈
🥳🎉🎊
Not bad, not bad at all!
day 8 of
#30daysofml
on Kaggle!
After a good basic Python week, today we start with Machine Learning!
The two topics that are mentioned are Decision Trees 🌳🌲🎄🌴 and Pandas 🐼
1/4🧵
Today, after 6 years of deciding to leave my country, family and friends to chase a better life, I can say that I achieved adulthood:
I own a mortgage!
I'm very happy!
Here's a TPU inference example on Colab:
The Colab TPU is actually slower than GPU for single-image generation, but you do get a speedup when generating a large batch (8+ images).
In two hours (11:59AM BST) I'll open my DM
Feel free to ask questions about:
• Python 🐍
• Machine Learning 🤖🧠
• TensorFlow 🦾
• DevRel 🥑🗣️
Talk to you soon!!
Python has one very good built-in module to deal with collections called itertools
Its functions form an "iterator algebra" that is fast and memory efficient. They are inspired by the ones in APL, SML and Haskel
Let's take a look on some of them: count, cycle and repeat
1/5🧵
Working with smarter people than you is sometimes very hard for your ego!
There's a lot of impostor syndrome in the beginning.
But that's a big chance to learn what these smarter people do that you can adapt and grow as bigger fish 🦈.
3/11🧵
Given you have your data loaded on a Pandas 🐼Dataframe.
How can you understand more about its characteristics?🤔🧐
Let's see a couple of options...👇🏾👇🏾👇🏾
[30 quick sec]
1/5🧵
Today I decided to study music generation 🎹🎵🎶 🇮🇪using Machine Learning 🧠🤖.
I followed the great MiT course
more specifically the third class exercise (lab 1)
and I got some nice tunes! Check it out! what do you think
@lmoroney
?
1/5🧵
@atchimolajuwon
@William_Castro
o que piora ainda mais é ele ter a audiencia que tem e normalizando isso do: "eu acho a, b, c pq é minha percepcao sem embasamento nenhum pq é chato."
olha que exemplão legal....
por isso eu marquei o canal dele pra nunca ser uma recomendacao pra mim no youtube!