MORE MACHINE LEARNING | 100 Days of Code 2

Daniel Bourke documents days 6-10 of his 100-day coding challenge, mixing Python, Stanford's machine learning course, and Udacity's deep learning nanodegree. He's experimenting with single-tasking each day, building Messenger bots, and scored 100/100 on an ML assignment. An 80,000-hours career test pointed him toward AI risk research.
Chapters
Day 6, 100 Days of Code. What's going on, guys? I've been working on a whole bunch of study today. Just look how messy my desk is. Got cups, got tea, got water bottles, got study. What I do? Started out Python, still updating the 100 Days of Code on Medium, Anki. I'm about to go work out because I've been studying all day. But this morning, finished my machine learning assignment, first programming assignment for machine learning. On the Coursera Udacity, I mean Coursera machine learning course. I then did some work on Udacity and then I did some Python study on Treehouse. So it's been a whole mixture of things. Now I'm gonna go work out so I don't get stuck at a computer all day. But I did this cool quiz on 80000hours.org and I'll link it in the description as well so you can do it too. But essentially it was, what is it? So based on your answers and the problems we reviewed so far, here are the top 3 global problems that we recommend you check out first. So, what is 80,000 hours? Well, the average person will spend 80,000 hours of their life working, and right now I'm in a period of trying to decide what I should do with those 80,000 hours. So, I used this website and the top 3 issues it gave for me was Global Priorities Research, risk posed by artificial intelligence— I'm really liking that one actually, considering I've been studying that lately— and promoting effective altruism. So, I'll look more into them, find out what they're about. They have a really good blog post and article on, on each one of them once you do it, but I'll link the test in in the description so you can do it. But otherwise, I'm going to my friend's house later tonight to practice building some messenger bots, like Facebook Messenger. And then tomorrow I'll be going, I think I'll go to the library and do some machine learning or Udacity study, whatever it is. Be day 7. Catch you then. What's going on everyone? Day 7 of 100 Days of Code. I gotta be a bit quiet, it's early in the morning. I'm at my friend's place. I'm heading off to the library to do some study. On today's agenda is Python, machine learning, deep learning, and then I'll come back here this afternoon and me and my friend, we're trying to build a website to like a directory for all the fitness facilities in Brisbane and other areas. So we'll see how that goes. I think we're— oh, stop saying so, I gotta work on that. I think we're just gonna use WordPress or something like that to build the basic template and then work it up from there. But I'm excited. It's raining outside, so it's going to be a good day to sit in, sit in a warm library and do some coding. Catch you in the next clip. I just spent the last 6-7 hours at the library. My brain is spent as well. I've been fasting all day. It's about— what is it? 2:36. So I'm going to have something to eat. And then maybe do some more study later. What was I learning this morning? Learning some recurrent neural networks through Udacity. So recurrent neural networks are used, what Google uses to make Google Translate so awesome. And in machine learning on Coursera, I was learning about logistic regression as well as gradient descent with logistic regression. Can't even speak right now. And so So, I keep saying so. I need to take it out of my vocabulary. Or is it like a transition word? Is it okay as a transition word? So, I got a whole bunch of notes from that. I did about 6 Pomodoros, so 3 on each, as well as some writing, as well as some uploading of videos. So, the last vlog will be up on YouTube now, so you can see that. Thank you so much for watching it. I love your advice, I love your feedback. Leave a comment, I will gladly reply to you all. But whatever you're up to, have an awesome day. I'll see you in the next clip. I'm actually extremely proud of this feed effort right here. Check it out. Salsa, sweet potato, avocado. Actually, that's got some beans in it as well. Mushrooms, pistachio nuts, broccolini, a little bit of chicken breast. Look at the colors on this bad boy. 10 out of 10 keen for this. I'm gonna refuel the brain, have a nap after this, and then get back into some study, website building, whatever. Day 8. Yeah, so today it's gonna be like a normal day. Um, I spent a few hours in bed this morning reading because it was raining and it was comfortable, and why not? I should have got out of bed earlier, yada yada. Oh well, we'll move on. And then I cleaned up my room It's still a bit messy, but there was lots of dog hair. Nonetheless, I'm going to be studying machine learning, deep learning, and Python today as usual. I'm also going to do some writing, answer a question on Quora. It's the 19th of May, so pretend there's a 19 there. Uh, Python book, workout— I gotta work out this afternoon. I want to do a Medium article, don't know what on yet. So yeah, machine learning, 3 Pomodoros, deep learning, 3 Pomodoros, And then it'll be a good day. Chillax. I find that after about 4 hours, 4 to 6 hours of studying, my brain just turns to mush. I can't really take in anymore. So that's what I'm going to go for today. That's what I try to go for every day. And we'll see how it goes. Just had a little breakfast. First study break of the day. It's about 1:30 PM. And I've got I've got a little slow cooker going. Check it out. All that vegetable goodness, a lot of beans and stuff like that. It's gonna take about 8 hours though. I've set myself a challenge for the rest of May. No meat. So, what does that look like? I'm still gonna eat animal products like eggs and egg whites and whatnot, but no other sort of meats. No chicken, no beef, no whatever. No meat, actual meat. Why? I wanna try it out. I haven't done— I've eaten probably about 90, 80, I'd say 80% vegetarian over the past few months. So I'm going to do 100% vegetarian for the next 10 days. We'll see how it goes. I'll show you this when it's done. You just smile, you're on the vlog. Just smile away, boys. Smile away, boys. Just finished gym. Look at this protein pancake apiece. 1,000 calorie beast. I found you, brother. How you doing? Still got the slow cook going on. I think we've got like 3 or so hours left of it. Easy. Camera's fogging up. I'm just gonna make it worse. I'm gonna eat this, get back into some study before bed. It's Friday night, but I'm not doing anything fun. Actually, yeah, studying is fun for me. Very much succeed. Check it out! 8 hours slow cooked beans, sweet potato, capsicum, greens, corn, tomato, salsa, everything, chili. I'm gonna add some little cup of rice to it, and that'll be my dinner for this evening. Then yeah, I just finished some Python study on Treehouse. I'm trying to do at least a little bit of machine learning, a little bit of deep learning, a little bit of Python on Treehouse every day. Maybe I might mix it up to make like one day deep learning, one day machine learning, one day Python. So who knows, I'm still trying to work out this schedule, and I don't know when I'm gonna upload blogs or vlogs or whatever like that. I've been writing more on Quora, which I'm really enjoying, and I also want to do more vlogs, so I've got to work it out. I've got to work out a schedule. And I still have to do podcast number 101. No excuses, I'll get it done next week. Actually, what's tomorrow? Yeah, by next week, so within the next 3 days. We'll see you in the next clip. Good morning! It's day 9, 100 Days of Code, or maybe day 8. I can't remember. I think it's day 9. Today's gonna be a little quiet day, 'cause it's a Saturday and I'm spending some time with some friends. So I'm gonna— in the morning I'll do a little bit. I'll do some Python, some Treehouse, and Anki. They're the things I usually get done really early in the morning, but I won't be doing any machine learning or deep learning today. I'll let my brain have a rest for tomorrow. Friends come around, we're going to chill out for today, cook some food, have a good time, and yeah, that'll be it for the day. But I think I'm going to upload this tomorrow. Sunday and start to work out an actual schedule for these videos and maybe do some other type of videos like cooking and workout and tutorials and whatnot. But it's a beautiful day. It's extremely sunny where I am right now, hence, uh, hence why my eyes are squinting and I just woke up. I like to get some sun in my eyes when I first wake up. There's a molecule in your eyes that help— that, uh, circulates your circadian rhythm. That is triggered by really bright light. So it helps you wake up in the morning if you look at really bright light. And bright light from inside doesn't measure anything of what the sun's capable of. So, do you want a tip to wake up? Get some bright light on your eyes, get some sun on your skin, do some star jumps or something, get the blood flowing, and then have some caffeine if you really want to, but We've evolved to not have caffeine. I mean, I love caffeine, but you don't really need it. Sometimes you do. So when I say I'm doing Anki, this is what I mean. I've got a few decks here of flashcards— Japanese, HTML, CSS. Right now I'm working on Python. So we go into Python, study now. So what are the Boolean results of AND? So this will come up. True and false equals false. True and true equals true. False and true equals false. False and false equals false. Awesome. What are the Boolean results of not or? Uh, the only one that's true is false or false. That's how I remember that one. How to start a block in Python? With a semicolon and indentation. With a colon indentation. Yes, just like that. So I'll do this. until it's finished. And then I'll come over to Treehouse. As you can see, Dates and Time in Python. I'm on the Python track. I got 2 more videos before I finish this module. So I'll finish that, and then I'll do 1 exercise in Learn to Py— Learn Python the Hard Way book. Um, it's free online, so I think I'm up to chapter 32. I do 1 chapter a day. So we'll see whatever, whatever comes up here. They're the 3 things I'll do today, as well as some writing. I'll also update my 100 Days of Code on Medium, and that's all I'm going to do today. I'm going to take a rest, let the brain, let the brain, uh, recuperate all this knowledge I've been learning over the last few days and get back into it tomorrow. But, uh, keeping the streak going of at least a little bit of code every day to keep my mind in that that mindset. 100 Days of Code. Woo! What's up, y'all? I think it's day 10 of the 100 Day Code series. Yesterday was a quiet day. I took a bit of day off technology. My brain needed to recuperate. I was looking at a screen for far too long the last week, but that's all right. You gotta have those peaks and drops, you know what I mean? So I spent the day, I spent the morning just doing some basic coding stuff like Anki, Python, so I still got a bit of practice in, a little bit, and then spent the rest of the day just relaxing with friends. But today's Sunday, the 21st of May. I'm back into some machine learning. We're learning about gradient descent and what's on the agenda for today. So we've got regularized linear regression, regularized logistic regression, and then, um, I believe I have an assignment due later today for machine learning. So then I'll be fully up to date in the course rather than a week behind. So I'm excited for that. That's what— that's— and for the rest of this week, I'm going to be doing one single task a day. That's what I'm going to try and practice doing rather than switching between machine learning, deep learning, Python all throughout the day. I know they're all interrelated, but task switching is very mentally straining. So I'm going to try instead today focus on one thing, which is machine learning course, do that all day. Tomorrow will be Udacity, the day after that will be Python. So we'll see how it goes. We'll practice one thing at a time and I don't know, it's all an experiment, right? See how we go. Oh, and I'm going to upload this, this vlog later tonight. So Sunday might be an upload day. I don't know. I'm still working this out. I'm happy because I just submitted my second machine learning assignment. 100 out of 100 points. Lots of documentation I had to read for that, but all in all, first day of single tasking on one thing. I've spent— so we got machine learning Coursera week 3 classes— about 3 hours total of deep work on this actual assignment without any distractions whatsoever. So put 3 work— 3 hours of work into finishing probably 2 lectures, so that was about half an hour, and then 2.5 hours straight on the assignment. Single-tasking— day 1 of single-tasking going much better than multitasking because now I feel accomplished. I've actually done a major piece of this assessment. So, I'm gonna continue this for the next few days, at least a week, and see how it goes. Tomorrow is all about Udacity, but until then, I'll see you in the next video because this is gonna be the last clip Finished Machine Learning Week 3 on Coursera Stanford course. You can all do it. It's all free. Check it out. Next tomorrow actually starts Week 4, so I'm up to date now instead of being a week behind, but I'm excited. I'm gonna go work out, go to Costco, get some almond butter, my favorite food in the world, and put my Mac on charge because it's about to die. Catch you later.
More from Daniel

Small Language Models (SLMs) Are the Future: Fine-Tuning AI That Runs on Your iPhone

Introducing Sunny - A MedGemma-Powered Skin Health Tracking App (Kaggle Competition Entry)

Benchmarking: NVIDIA DGX Spark & NVIDIA RTX 4090 | Part 4 - Comparing the results

Benchmarking: NVIDIA DGX Spark & NVIDIA RTX 4090 | Part 3 - Running LLMs on the RTX 4090

Benchmarking: NVIDIA DGX Spark & NVIDIA RTX 4090 | Part 2 - Formalizing the tests

Benchmarking: NVIDIA DGX Spark & NVIDIA RTX 4090 | Part 1 - Creating a series of tests

Local Multimodal RAG on the NVIDIA DGX Spark | Part 4 - Publishing our demo to Hugging Face

Local Multimodal RAG on the NVIDIA DGX Spark | Part 3 - Putting it all together


