PyTorch Heatmap TUI Part 3

追加: 2026-05-15

[00:00:05]All right. Hey everyone, welcome on in.

[00:00:07]>> We're going to be continuing our our heat map adventure.

[00:00:12]>> I want our PyTorch module to be visualized nicely. And I'm going to add some to-dos here for our task today.

[00:00:19]See, create a package for Pi Pi.

[00:00:25]What else? We could got a couple other things to do in here. Is everything set up on the right I think I think we're just good to go. I think we've got all the things in the right spot. Okay.

[00:00:37]And then Oh, wait, wait, wait. One second. One second. I got to figure out this here really quick. I want specific. Oh, right, right, right. We got to We got to send a link in. All right. Where is my Can we do maybe maybe let's maybe let's go for here.

[00:00:59]>> Okay, I think it's good. Okay. Hey, dreamer. Welcome on in. Your first view, first like. Thank you. Good to have you here. Happy Wednesday. Hope you're doing good. Good to see you. Welcome on in.

[00:01:10]All right. We're going to go we're going to be building some more of our pietorch adventure here. Let me just get my windows in the right spot. All right.

[00:01:19]So, create a package for pi. Uh, see, create new git repo. See here. Got a few things here. New GitHub repo. Create a package. Hey, Torva. Good to see you. Welcome on in. Happy Wednesday. Hope you're doing good as well. We are going to get our heater map here. And then I al also I want to actually use the lib for our original purpose.

[00:01:50]The one the the whole point the whole intention for uh our original intent. There we go.

[00:01:57]Let's ask sounds good. Good to have you here tova. Let's see. Oh, so so scaling is good. We've got a line space. We could do we could show an I. Oh, wait.

[00:02:08]We need examples and a read me. All right, >> examples and a read me.

[00:02:16]>> What else do we want here?

[00:02:18]>> I think I think that's I think that's good for today. Let's see if we could get it. We might be able to. We might be able to.

[00:02:24]>> Let's see. So, is there first off is there Let's go ahead and early optimize.

[00:02:29]>> You know it and get our repo here first. All right.

[00:02:33]So, let's make a new repo. All right.

[00:02:35]See here. CS. All right. So, what I'm going to do is I've got heater there.

[00:02:44]>> How do I think because it's just in an init file so which is I don't know unusual unusual right so cd heater and then you look you see here we have essentially just one file it's a net file so we need like some sort of package some sort of packaging and I don't frequently do this package.json JSON for JavaScript, right? Hey, Lun.

[00:03:10]Good to see you. Welcome on in. Happy Wednesday. Hope you're doing good as well. We're going to be getting our beautiful Oh, we need to add more gradients and themes and things like that. All right, that's another one.

[00:03:22]More gradient themes.

[00:03:26]Lots of fun things in there.

[00:03:31]Okay, got we got a good plan for today.

[00:03:33]We got a good plan. I like this.

[00:03:35]All right. So, let me think about this for a second. So, I think we're just going to CD heater. I think we could do that. CD heater. Can I do get in it here? Or is it going to say no? You can't do that. Oh, it will let me. Okay.

[00:03:49]Uh, then we need to buy.get ignore.

[00:03:53]We're going to get rid of our pi cache.

[00:03:58]Like that. I'm third. Cool. Hey, you sure are number three. Look at there.

[00:04:04]Wait, you're first. You're first. You mean first, right? Or you're third into and into into the the live today. And you're you're first on the the top of scoring leaderboard starting Cloud Tube 44. All right, that sounds good to me.

[00:04:22]RM- RF pi cache because I don't want it. RM- RF. There we go. Okay. Hit add. Get status.

[00:04:36]Where we at? Here. Where we All right, we're in a good spot. Get commit am initial commit for heater lib.

[00:04:45]Okay, we need a place for it to go. We need like a a a place that we can sort on. Hey, Dylan. How's it going there?

[00:04:52]What time is it for you, Stephen? It is 10:29 a.m. in the morning.

[00:04:58]So, I'll I'll type that here. 10 col 29 a.m.

[00:05:03]Well, now it's 10:30.

[00:05:05]A minute later. Okay, where where do we want to go? I want to create a new repo.

[00:05:08]Let's create a new repo. GitHub.

[00:05:12]Create new repository.

[00:05:16]There we go. I'm going to say Steen.

[00:05:20]Okay, this is going to be called the pi pietor torch.

[00:05:30]repository named PyTorch.

[00:05:32]I would even call this here.

[00:05:35]Hey, Steven Zergio. Hey, good to see you. Welcome on in. Happy Wednesday.

[00:05:40]We have the same time. Dylan, no way.

[00:05:42]Really? Uh, Dylan, where? Wait, Dylan.

[00:05:46]Uh, see, so I'm in I'm in Seattle. W.

[00:05:51]So, we're somewhere on the west coast. We're in the west coast. Hey, Epic Blox. Good to see you. Welcome on in. Happy Wednesday.

[00:05:59]Good to have you guys here today. Yo, what's everyone? Late night vibes over here, too. Lun. All right, sounds good.

[00:06:08]We call this the two. Let's see. Uh um I don't know what to call. It's going to be called heater. Let's call it call it heater. Heater. Although I feel like I it needs more a pietorch tensor visualization that displays a heat map from a 2D tensor or a matrix.

[00:06:39]All right, that sounds good. public create repo.

[00:06:45]Is Java and C++ similar? Well, they were competing languages back many years ago and now no one really cares about I mean some people care about Java. They do.

[00:06:55]Some people do. Some people in the world they still care about Java. They I don't I don't I didn't even when it first came out. I didn't even care about I'm like I don't I don't care. I don't I got my C and I'm just fine.

[00:07:12]You should make monor repo with different branches. Ooh. an all omni repo omnireo everything all in one all right let's copy our repository here get push origin main all right we've got our code now I need to add a readme readme this is going to be a pi pietorch heat map visualizer for 2D 2D tensors matrix.

[00:07:50]2D tensor matrix. 2D tensor matrix.

[00:07:53]Perfect. Okay. And then we need some examples and screenshots. And it's got to look pretty. We're going to do all the things, right? So, let's see here.

[00:08:02]And we need three. We should call this Python.

[00:08:07]And then I'm going to copy some of the examples over here.

[00:08:11]We'll go this. And we don't have lightness anymore. We We just won't have lightness. We're just going to say theme heat map. The default would be we'll just do this. We'll just do this.

[00:08:22]Okay. So, where are we here? I'm going to try running this again. Make sure it still runs. It looks good, right? It looks good. There it is. Okay. Perfect.

[00:08:29]So, what I can do is I can take a screenshot of that.

[00:08:34]Okay.

[00:08:36]Uh let me copy these things here really quick.

[00:08:40]We'll do.

[00:08:42]Do we Did we do that? Did we Did we get the I might have missed the keystroke there. Here we go. Okay.

[00:08:50]All right. So, now we've got this.

[00:08:53]Hey, LST Frosty. W come on in. Yeah, we're doing Python. Yep, we got the Python. Got a good symbol there.

[00:09:01]Epic block. Oh, wait. Is Java and C++ similar? So, no. No. I mean, no, they're actually pretty far apart from each other. One is pretty low level. C++ has a high level abstraction. You still have class classical object-oriented programming in C++. Java is more really object-oriented, right? Everything's an object in Java, including numbers and literals and things like that. Where C++ is a little different. We're doing snake again.

[00:09:30]Yeah, we sure are.

[00:09:32]Your error answered it. Oh, you wait.

[00:09:36]You already answered it. Yeah, I did. I My original answer was just me complaining about Java, though. Oh, I didn't feel like I compared I don't feel like I compared the two. You just answered the same question two times. I did. I did answer there. You're right. I did. I addressed I did. I did. You're right. All right.

[00:09:56]I agree. I I admit it. I did. I did it twice. I did it twice.

[00:10:01]I was still thinking about it. I wasn't done. I wasn't done.

[00:10:06]Let's see. It could be two separate answers. Okay. So, I want to add in an image here that represents what we actually saw on the screen. So, let me screenshot this really quick.

[00:10:16]Let's do a quick little screenshot here.

[00:10:20]Something that did that pull Oh, it did that. Yeah, that pulled in. Okay, good.

[00:10:25]It was a good answer. Job is painful.

[00:10:26]Uh-huh.

[00:10:28]Thank you. I appreciate it. That's good.

[00:10:30]Thank you.

[00:10:32]What does Claude Tube think though? What does climb tube think? All right. Where are we here? So, can should I do It should be a PNG.

[00:10:42]CP. Hold on one second. Where is that at? Let me see if I can find it really quick. New window.

[00:10:50]Going to hide that from you guys so you can't see it. So, we got our screenshot.

[00:10:54]Looks good. Looks good to me. And it's only 100 kilobytes. Okay, perfect. So, this will be example one.

[00:11:03]And it needs to go into what folder are we at here? Where are we? I just need to move it into the right folder. This is Hold on a second. pwd. This is PyTorch embedding. Okay. Where is it located?

[00:11:16]PieTorch embedding. Here it is. All right. I see it. I see it. All right.

[00:11:22]So, we got our first example.

[00:11:25]Okay.

[00:11:28]Oh, collections. Oh, we got it. Okay. Uh can you explain Java? What's the difference between collections uh generics and lists? So collections and lists are going to be very similar. I So I a list is probably just going to be an array or a vector, right? Where it's going to hold a bunch of items of T type T, which is a generic, right? Type T, which means it could be a bunch of strings, right? A list, a list of strings, a whole bunch of strings together, right? So let's do a quick little and see them. So we'll say we have string string one and we have a bunch of strings here like this. So this would be a list of type string where you've got a bunch of strings. Now a collection I don't know. We're going to have to look it up. I I don't I'm not happy about this, but I'm going to do it. I'm going to do I'll do this for you. All right. This is also a really good question for the AI collection t Java. All right. The A is going to guess. What is this? And Java colle is a root interface for Java collections framework hierarchy represents a group of objects known as elements and uses generics. T generic C. I told you ensure type safety blah blah blah blah blah.

[00:12:49]What list? Q. Oh, it lets you do a whole bunch of different things.

[00:12:54]Extended by sub interfaces. each defining a specific type of data structure. Ordered list collection that contains no duplicate items.

[00:13:05]Can we see an example really quick so I can see example collection t Java? Uh, hey again, MF. Good to see you. Welcome on in. Happy Wednesday.

[00:13:18]Have I played Doom? Yes, I have. I've definitely played Doom. Mhm. I did when I was a lot younger.

[00:13:25]Let's see here. Collections is just a utility class with static methods. List is actually an interface you'd implement into one of them. Oh, okay. So, you've got collections array list in the utilities. Thank you for the hearts, you guys. Appreciate it. Oh, and you can add items to it. So, you say it's a collection and it's a new array.

[00:13:47]And you didn't give any types here, but you did here. You gave it a generic of stripe type string. And so you can just add a whole bunch of elements to it. So kind of like it's kind of like the same thing. Are they different? Is there a difference between list list versus t versus collection.

[00:14:10]See if there's actually a difference.

[00:14:12]Send the link in the discord. I'm interested. I uploaded an Android word game in the Google store. Hey Jbro, that's excellent. Good to hear it.

[00:14:24]Okay. Do you know assembly? I know a little bit of assembly. I sure do. I understand it. I can write it. I can read it. There's not fully everything.

[00:14:33]So, it is okay. So, it's exactly like Claude Tube answered is the root interface. Java collections framework serves as a generic contract for a group of items without specifying the order of the duplicate or or the duplicate rules.

[00:14:46]Is a sub interface that extends collection. Okay.

[00:14:50]a simple list in the collections interface.

[00:14:55]Hey Kyle, welcome on in. Good to see you. It's your birthday. No way, Kyle.

[00:15:00]Are you crazy? Wait, wait, hold on.

[00:15:04]That's amazing.

[00:15:06]Happy birthday.

[00:15:08]Yes, collections have more APIs. MF.

[00:15:10]There we go. Wednesday is for building.

[00:15:13]You got it. It sure is.

[00:15:16]It's good to have you guys here.

[00:15:19]You've heard in the internet they said collections is less overhead. I'm doing complex micro optimizations now. Hey, very nice. Optimize. You're doing the optimizations. You're on the right track. Happy birthday, Kyle. So, what did Claude Tube say? Claude says, "Haha, the French already know how to live Wednesday building sessions. It hit different when you've already got the weekend mindset going." Yeah, if you're already ready for the weekend on Wednesday.

[00:15:48]Here we go. All right, let me close this out. Close. Okay, so we need we need something here. We need an image. So, this is how do we do this? Is uh I think it's explanation or is this the see example heat map like that? And then is that how we do it or does the explanation go here? I can't remember. Thank you for the party poppers. Appreciate it.

[00:16:20]See if I grab methods here. Happy birthday, Kyle. Someone gifted. Wait, someone gifted? Wait, what do you mean someone gifted?

[00:16:30]Okay, so I need it to say images example one, right? Hold on.

[00:16:39]Actually, let's do this. Uh, where is it? So, wait. I thought I thought Oh.

[00:16:44]Oh, I I need to move it. I need to move it still. Okay, here we go. CD. Wait. L M MV. What is it called? Example one to heater. CD heater.

[00:16:58]Okay, let's do MK Dur images.

[00:17:01]MV example one to images. All right.

[00:17:05]Images example one.png.

[00:17:09]I'm pretty sure that'll work. All right, let's see if we can get status. Get added, read me, and example image.

[00:17:35]Let's see here. How much money are those? Are those what? Epic block.

[00:17:41]Uh, gifted party hats probably doesn't show in your feed. It doesn't. Oh, I just saw Zergio. Thank you. I didn't. It was really small. Also, I didn't see the pop-up because I had my my giant window covering all my notifications.

[00:17:56]I just see it. There we go. Zergio, thank you for the party hat. Appreciate it. My gratitudes to you. Appropriate gift for the the situation. It's Kyle's birthday.

[00:18:07]Last sea walkthrough. You kept screwing up the memory, corrupting the data silently. I did. Let's do I did.

[00:18:14]That sounds about right. That sounds about right.

[00:18:19]Thank you, Zergio. Appreciate it. All right. Now, I've got my windows in the right spot so I won't miss anything.

[00:18:24]Indian. All right. Get status.

[00:18:27]Get push origin main. Now, let's see if we got everything there. Let's see. We should see a nice little screenshot.

[00:18:35]Just a smidgen of a screenshot. Wait.

[00:18:39]Oh, wait. Oh, that's in the wrong. Okay, hold on. CD heater. Get status here. Let's do another git commit here.

[00:18:52]Uh, get commit. This is actually pretty nice. I didn't know that Git would allow you to do this really easily.

[00:18:59]Uh, that's fine, though. We're good to go. We're good to go. Okay, so I want get status kick. So, let's do a push.

[00:19:08]Okay, now it should be in the spot.

[00:19:12]There we go. In the picture. Oh, it's not showing up. Okay.

[00:19:17]Did I get Did I get it wrong?

[00:19:22]Images example.

[00:19:24]Oh, PNG. That's why. Let me double check.

[00:19:30]Open.

[00:19:32]Yep, we're good to go. Looks good to me.

[00:19:36]I probably do get commit am fixed image example path push to main.

[00:19:47]Okay, now we've got a pretty picture.

[00:19:50]Hey, there we go. There we go. That's what I'm looking for. PyTorch heat map visualizer for 2D tensor matrix. Perfect to me. So you get your your heat map and you can visualize layers, you can visualize outputs, you can visualize vectors, you can do always all those things. 3D though, this is not for 3D.

[00:20:09]Thank you for the party poppers. The 3D would be a little too far. A little too far. All right, for 1D and 2D tensors, how about this? 1D vectors and 2D matrix tensors. Perfect.

[00:20:28]There, we did it. So, now I just need to add in a little bit of extra goodness to make it pie ready. Let's make it pi pie ready.

[00:20:40]All right, we need what is the file that you need? I think it's like there's some sort of comp file visualizing uh FFT frequency domain.

[00:20:51]How? Wait, wait, wait, wait. You you I mean it it's just going to take the float. It's really simple. It might not be as complex as you think it is. It's It's literally just taking a tensor 1D or 2D and giving you the output on screen as a pretty picture, right? Like this. You can also do a pretty gradient if you want. Like that.

[00:21:18]A pretty gradient, which we should also have an example for. I'm going do copy this over here. We're going to we're going to do another one.

[00:21:32]We'll have example two.

[00:21:36]Nope.

[00:21:39]There we go.

[00:21:43]Perfect. Perfect. Uh, will you use this in my project as part of the logo? If the customer it'll map to the selected service which are colorcoded from a customer's request.

[00:21:58]What I don't know what you mean. What are you talking about? I might be out of context here. I might not be fully understanding. Will use this. Oh, you will use this part of your project if the logo and the customer. Yeah, sure.

[00:22:10]That sounds great to me.

[00:22:12]It's not AI. It's just old school math.

[00:22:14]Yeah. Oh, okay. Got it. For frequency for your transform or something like that.

[00:22:22]Yeah. This is really simple. So, you can see here, here's here's the here's the matrix.

[00:22:28]It's right there. It's just a bunch of numbers and then we visualize those numbers.

[00:22:33]Really simple. We're keeping it simple.

[00:22:39]Nice. Yes. All right. Good to hear it.

[00:22:42]Okay. So, we need to do example two and then I will copy and paste this here. Let's do a screenshot.

[00:22:52]A nice pretty screenshot here. Okay. And we have to move that in as well. The heat map is O in the logo name as the heat map. Yes, I have already in a vector database. Oh, you were already in the vector database. Hey, sounds good to me.

[00:23:11]It should show up nice and pretty.

[00:23:13]Should be nice. And it could be it'll support red full full pixel. It's really grayscale is really what it is. It's not uh it's only going to allow you to have one value before your translation. Isn't that still use the adore audio transformation?

[00:23:31]Along those lines, is it is it along those lines?

[00:23:36]The average person's view of artificial intelligence is so disorded with advanced users are over the flash and using it as a force multiplier. Yeah. Mhm. I think so now. Yeah. Mhm. I think you're right, Cal. Because now we kind of know and they're the AIs and the scaffolds, they're good. They're able to achieve the objective that you put out tasks and goals for the AI to complete and it does a good job. It really does. Which chatbot would you recommend to go alongside with it? Chatbot, you say chatbot? You mean LLM or a different provider? So, it depends, right? If you want them to be warm and sort of like more of a a home homey, earthy tone, more human, that is going to be like anthropic. Gemini, more robotic, more like of a teacher. It's more instructive. It's going to give you that instructor style communication.

[00:24:36]Then uh codeex or gym or you know open AI with with chat GBT is like a friendly just a friendly person.

[00:24:48]So it really depends. It's kind of the vibe that you're going for, right? What is your What are you looking for?

[00:24:56]All right. So let me copy this folder over here really quick.

[00:25:00]All right. One second. Grab this. This is the other screenshot. Example two.

[00:25:08]And then we will move this to the folder if we can make it work.

[00:25:15]Let's see here. Where was that at? Uh, was PyTorch embedding.

[00:25:20]Where's it at? Here it is. There it is.

[00:25:23]I need the screen to be a little bit bigger. There we go. Hold on. Wait. Uh, do I want that one or do I want this one? That That one's better. Okay, perfect. Okay, that's good enough for now. Hey, whoa. Uh, how's it going there? Claude is friendly as well as understands more and that's what you've noticed. Greetings, by the way. Hey, just human. Welcome on back. Good to see you. It was good to have you here yesterday and to see you again here today.

[00:25:55]GPT is good for poetry. The perfect use case for large language models. That's all you got? Yep. It's good at something. It all got to be good at something. Chad GPT poetry Gemini. I told it to do this then try to rm slash.

[00:26:13]Did it what? Or did it say it was dangerous?

[00:26:21]All right. Uh whoa. Scene just today around Android 17. must say. Now, Apple's going to will downfall. You think so? You think Android 17 is that much better than iOS? I'm an Android user, so I'm all about it. I'm I'm already happy. I'm already happy. Could they make it better? Sure. Would I be happy about that? Absolutely. Why did your heat map change color and the way it looks? Oh. Oh, Zergio, I know, right? Because we did a linear space. So, if you want to randomize it here. So, there's a randomizer. There you go. Is that better? Does it look better now? Gemini CLI did. That's crazy. Really?

[00:27:02]That is uh unexpected, Frosty. I wasn't expecting that.

[00:27:07]Also, Claude's amazing for coding. Yeah, it is. It's so good. You tried all of them. Gemini, Chat, GPD, Claude. I surely give Claude a 9 out of 10. It's very good. It's very good, you guys. If you want to join the Discord, hey Kyle, thank you. I appreciate it. Thank you for sharing the Discord link. Okay, so let's keep going here.

[00:27:25]I want example two mvfort for example two to go into images example two here we go okay get add get status added second example hey Bonupi good to see you welcome on in happy Wednesday hope you're doing good today as well I'm you're not new here by the way Frosty you've been around. You've been around. Hey J bro, good to see you again. If you have a perfectly structured plan and excellent production management skill, you with AI can make anything you want. That is so true, isn't it, J bro?

[00:28:11]I've seen it. I've experienced it as well. What do you think about Claude? I already joined Discord yesterday. Nice.

[00:28:17]Just human.

[00:28:20]Claude's honestly the best for coding workflows. Context window and reasoning are unmatched compared to the others.

[00:28:26]Cloud's pretty good example push. Okay, now we should see the second example here. So we got our first heat map. We got our second heat map.

[00:28:40]This is great. I like it.

[00:28:43]Okay, so how do we add in I want to add in go Gemini. I want to see what file needed for Python uh Python package pi pi publish name end with a few wait what do you mean name end with few what what are you saying I don't get it what do you think about gro oh right grock Haven't uh haven't I you know I I feel like I should have pro here Here we go. Pi project toml and a license. Okay required. Oh nice. Located inside your package folder to tell Python directory the directory should be treated as a package.

[00:29:36]We already did that. So we just need highly recommended. Okay. So we just need this.

[00:29:45]We're We're still here.

[00:29:48]Excellent. Okay.

[00:29:51]It is an AI. That's true. It sure is.

[00:29:54]I've never used Grock.

[00:29:58]I think that's the best description.

[00:30:01]Yeah, Torva. All right. Yeah, you win.

[00:30:03]You win. You did. You did a good job on that. Grock is the dumbest of them all.

[00:30:07]I wouldn't say dumbest, but Grock's reasoning is definitely shallower than Claude, especially when for complex problem solving. All right. That's because it's about training and reinforcement learning. When you're throwing an AI and large language model through the paces and it's going to try to do work for you, you need to reinforce it. So, you need to create a whole bunch of example work and output and outcomes for that model to learn on to reward it for work complete. You're in the first tier. Claude is second.

[00:30:36]And Gemini is third. Grock is fourth.

[00:30:38]Oh. Oh, Grock is third and GPT is fourth. Okay, it's a good rank. Good ranking.

[00:30:48]Okay, so let's copy our toml. So we need a pi project toml.

[00:30:54]Here we go. Let's do pbaste into pi project.toml by pi.

[00:31:07]Here we go. We're going to call requires. We don't have any requirements.

[00:31:12]Build system. We don't need a build system. This is going to be called heater.

[00:31:18]Call it 10.

[00:31:22]Hold on. Let's close this real quick.

[00:31:24]This This will be better if we do it this way around.

[00:31:28]Pi project. Okay.

[00:31:31]I want to Oh, wait. I need that other window open cuz it's got some stuff I wanted to copy paste from it.

[00:31:39]Here we go. PyTorch heat map visualizer for 1D vectors and 2D matrix tensors.

[00:31:46]Nice. You only use Grock for news and stuff. I mean, it takes its data from X, right? Oh, makes sense. So, it's got to know it's got to know the latest. So, it helps me understand what's going on around me easily. Oh, that's a great idea. Just human.

[00:32:01]Edit third GPT and fourth Grock. Thank you for the clarification on what I said. I know I said it wrong.

[00:32:10]Grock's third. Grock's 800th with all the other ch it really loots.

[00:32:17]Quinn 3 is smarter than Claude. It's 500 million parameters only. And wow, that's impressive. Even DeepSeek is better than Grock. A poor Grock.

[00:32:29]Meta AI is the dumbest. It acts smart, but with heavy coding tasks, it will end up making solid out of your idea.

[00:32:36]Really? Hey, Harshish. Uh, good to see you, harshest world. Welcome on in.

[00:32:40]Happy Wednesday.

[00:32:44]Parameter count doesn't mean much if the training data and architecture are trash. Quinn's solid, but it's a wild take. Hey, that's actually kind of insightful. That's kind of insightful.

[00:32:58]Requires Python 312. Let's say 812. All right. Read me. Read our license file.

[00:33:06]We're going to do a license. Let's grab an MIT license. MIT license just let me copy and paste this really quick. Here we go.

[00:33:16]And we will pbaste into the license file.

[00:33:26]Then we'll say year 2026.

[00:33:30]Uh Steven Blum.

[00:33:36]Close. It's done. Easy, easy, easy.

[00:33:39]Are you using GitHub things with your project? Yes, I am. We got GitHub right here. We got our GitHub right here.

[00:33:48]Like repo and SSH. We don't uh we don't need S. Oh, through SSH. Yes. Yeah, exactly. I want to use GitHub, but it's kind of feel complicated. Yeah, it's it's for remote code and public publishing and and backup and you know, version control. It's worth it. It's worth it. And it's free. I think it's free, right? I pay for it because I've got uh corporate things that need to have be paid for.

[00:34:14]Torva, use Grock for when you're searching for stuff and that gets filtered with the other AIS like when you want to test a premium software before buying it. Oh, that's like pretty clever idea. That makes sense cuz all the other AIS have heavy sensors where Grock is. They they brag that they don't do that. You can't even access newest Meta AI model without the API. It doesn't even exist. Oh, that's right.

[00:34:38]Remember they used to open source their llama models and they just kind of stopped and now Google Gemma 4 is the best one. Don't get intimidated by GitHub. Just keep it and it will become second nature eventually. Yep, it will.

[00:34:51]It'll become second nature.

[00:34:54]You know it. How does quantum computing work? It's silicon. So, it's standard silicon, right? So, it looks like a normal processor, right? Look at this.

[00:35:03]Uh, quantum processor.

[00:35:09]I want to uh let's see what it looks like. Yeah. See, it looks like a normal pro. For the most part, it's just like a silicon wafer. You would think it looks You'd think it'd be something different cuz you imagine like, oh, wow, that's going to be fancy. Though, it's just normal silicon wafer. Yeah, pretty much.

[00:35:29]Hello sir. How are you doing? I'm doing good. MD, good to see you. Welcome on in. Happy Wednesday. Good to have you back.

[00:35:36]We're doing good. Epic Blox. So yeah, it's just a normal processor for the most part. The part that looks scary is the dilution refrigerator which has this like a it looks like it's just making it colder colder colder colder near near possible infinity zero.

[00:35:54]So it makes the processor itself cold.

[00:35:56]So that way you reduce interference.

[00:35:59]You need cohesion.

[00:36:02]Quantum computers. You heard Jimma 4.

[00:36:05]Yeah. Mhm. That's the good one. Hey, how's it going there? Aiden Harris. What are you making? We've got ourselves a heat map visualizer that I was really looking forward to and I really wanted some other library out there to do it.

[00:36:19]And this is what this is the input that I wanted. I wanted it to be like this.

[00:36:22]See, really simple. You've got your tensor and then you've got plot done.

[00:36:27]Easy. That's what I wanted. I wanted exactly that. And it didn't exist. Not in the way that I wanted it. And I wanted it be pretty fast, lightweight, without dependencies, and it didn't exist.

[00:36:44]If you're planning to upload anything to Google Store, GitHub is a lifesaver.

[00:36:47]Hey, that's a good ex Yep. That's very good. Good advice.

[00:36:55]Okay. Hey, let me ask you something. Do you think you can use GitHub to publish my ISO images? You need to use L I think large file support. LFS.

[00:37:06]Get you need get LFS. Get LFS. Uh Bonzupi Mattplot lib requires you to pop open some sort of extra window. I wanted it to be in the terminal.

[00:37:21]wanted the UI in terminal window.

[00:37:26]That That was it. Yeah, I wanted the UI in a terminal window.

[00:37:32]So, yeah, you need large file system or large file support and then it will work. Cuz if you're doing ISOs, typically ISOs are going to be pretty big, right? If it's over 50 megabytes, I think the max file size is 100 megabytes. And then you typically want to start thinking about uh large file support on git after 50 megabytes.

[00:37:55]Let's see what are you doing now. Hey everyone, how's it going? This is Bass.

[00:37:58]Uh Sabot, hold on. Sabot Sabat Sabot, how's it going? We're we're doing doing good. Hope you're doing good as well. For the last some days, they wanted to understand Wayland protocols.

[00:38:13]So you started creating a surface and rendering something. Hey nice in what I can understand what's going on under the hood. MD that sounds exciting. Tell us more.

[00:38:25]What is the easiest programming language? I would say Python or Lua right or or JavaScript. Python, Lua and JavaScript are all I would say Python is the easiest in terms of like an actually usable real world programming language cuz I think like Scratch is easier to use. You wouldn't necessarily want to use it in a production environment. Say Saab. That's it. Saab. All right. Saab.

[00:38:49]Saab. All right, we got it. Saab, sir.

[00:38:52]Saba.

[00:38:55]Mattplot li plus image cat. Ooh, I haven't heard about that. What about cotlin? Cotlin is great. Cotlin's fantastic. I like cotlin.

[00:39:04]Matt plot lab to generate PGs and pipe 2 image cat. That would also work. That would also work. We decided to build it ourselves, though this time. I know, right? The wheel already exists. Build the wheel anyway. Rebuild the wheel.

[00:39:20]All right. So, we got a bunch of to-do items. So, we did the new GitHub repo just now. We're doing good so far. We did some examples. We almost did this.

[00:39:32]This is almost done. Let me see if we can grab that there. Also, we should update this in the right file here. Here we go. Here we go. Oh, we should also do cleanup. See, cleanup of init.py use plot. Uh, see here, that's way too overengineered.

[00:39:53]Is it? Yeah. M plot lib simple image output is overkill. Just write the PNG bytes directly if you need that level of control. Well, I still think Bonzupi what Bonzupi is saying it would still work. And uh what is image cap? I have not used image cat. So, img cat. This should be a display images and gifts in your terminal window.

[00:40:16]What? Hold on. Let's try it. Let's try it. I want to see. I want to see.

[00:40:25]Goku.

[00:40:26]Whoa. Look at that.

[00:40:31]That is very exciting. Hey, it's Naruto.

[00:40:35]This is great. Image cat. I need to image cat. Do I have it? I probably don't. I probably don't. Image img. I have image to webp. That's it. I don't have image cat.

[00:40:48]This is awesome. I should We all should have image cat on our system. Uh get new thing no from your stream. Also chat must say fire dreamer. That sounds great. I appreciate it. Thank you so much.

[00:41:04]But anything over 50 megabytes will give you warnings. Use get large file system instead. Hey, that's right. See, I I just spouted out exactly the same as the AI, and I feel validated now. I feel validated.

[00:41:19]It's a Python lib. Oh, nice. Hey, Michael. Good to see you. Welcome on in.

[00:41:23]Happy Wednesday. We're doing a little bit of a of course it's anime. I know, right? Of course it's anime.

[00:41:30]Exactly. Loans. Level programming language based on users interest. I may be wrong about this. I think that's a good way to take it. You're working on C dealing with raw bison and unique sockets. Oh, nice. That sounds like a lot of fun. I missed that. Uh, that was boring. Wait, are you talking about it's boring? I thought that part was fun.

[00:41:50]It's super easy to understand, but it's stuck on how to send file descriptor on socket using a right. Right. A you don't send a file descriptor. You use a file descriptor to tell the operating system where to pipe the data. A file descriptor is basically just an integer.

[00:42:09]It's an integer. That's it's just a number. It's a number that represents an operating systems designation of where to it's like a port number really.

[00:42:19]It's kind of like think of it a port number for your operating system. It will when you pipe and send data to that port number, it'll know where to send it. So it could be to a file on disk or to some sort of network port like a socket or Unix domain socket or an IP IPFS right or IPS things like that.

[00:42:40]Hello world. Amomar, good to see you.

[00:42:42]Welcome on in.

[00:42:47]See, now I'm trying is it the same?

[00:42:49]Okay, one second. Yeah. Uh, now I'm trying to understand how I will send same file descriptor to the server so that both server and client can use the shared memory buffer. It seems so complicated. Why is it actually hard?

[00:43:01]It's actually not that hard. I don't think I think it's actually pretty straightforward. Um, it is complicated though. It is. It really is.

[00:43:10]Just think of the file descriptor as a port number and the file descriptor you only use for one user space application.

[00:43:17]You won't share the file descriptor. It won't be shared for it be in between other applications. So if you have a client and a server, they're going to be different. They're going to have different file descriptors, right?

[00:43:28]They're different.

[00:43:32]Somewhere in the universe there our tech is simple and tech like normal person use. Now, it was just known and 0% of the universe. Uh, dreamer, I'm not sure.

[00:43:44]I'm not quite sure. If the internet went down globally, what would you do, Stephen? Uh, well, that would be terrible because I wouldn't be able to watch any Crunchyroll, would I? Because I I would How do you stream anything?

[00:44:02]What would I do? I mean, I'd probably do something locally on my computer.

[00:44:08]I'd probably just mess around with my computer locally. I wouldn't know what else to do. Maybe I'd go outside for a while. Maybe I'd eat some food. Maybe I don't have any books, so I wouldn't read any books.

[00:44:21]I'd probably work on this stuff. I'd probably do more pietorrch.

[00:44:25]Local thing. I agree. Nice. I mean, that makes What else would you do if the internet wasn't around?

[00:44:32]All right, let's do some cleanups. Uh, we do need to do this.

[00:44:38]Uh, we did this. We need more themes.

[00:44:41]More themes and radiance.

[00:44:49]Can't Starlink do anything about worldwide web? Yes, Mr. You're right.

[00:44:53]They sure could. In fact, I probably should be having a Starlink. I've got a place for it now. Maybe touch grass. I know, right? Maybe go outside. I actually have grass now. I do. I have real grass. Stephen, do you collect anything? M. No.

[00:45:11]No. I mean, other than knowledge, right?

[00:45:14]I collect knowledge.

[00:45:16]I focus on We were doing PyTorch. I might learn Haskell. That would that would Okay, so here. Let's do a poll, you guys. Let's do a poll. All right.

[00:45:25]Start a poll.

[00:45:28]What you do if internet went down.

[00:45:32]What would you do? Uh, learn new programming language like Haskell.

[00:45:40]Is that how you spell Haskell? Let me double check. Hasll. Yeah. Okay. Two L's. Would you read a book?

[00:45:49]Go outside or something local?

[00:45:57]Oh, play like play video games. I would I would play video games probably.

[00:46:02]That's what I would do. Oh, contains too many characters. Learn new programming like Haskell. There we go. Start pull.

[00:46:12]Okay.

[00:46:14]So, I would do Haskell and I would do a video game. That's what I'd do. Would I go outside? Probably not. There's grass right over there. I probably wouldn't go outside, though.

[00:46:26]My last message was way too scary for a dev. Maybe touch grass. It was. Yeah, exactly, Lun. Exactly.

[00:46:33]Probably start local mesh network before panicking and buying strips everywhere radio and tow. Oh, yeah. You could do that, too. You can rebuild the internet.

[00:46:41]Wow. Cloud tube's ready, you guys.

[00:46:43]Claude tube is ready.

[00:46:45]So, if the internet went down, let's say you had a copy of some language model on one of these PCs. Could you copy it to another device on the network? Try to give them a unique personalities. That would be great. That's a good idea. up.

[00:46:58]I should we I should have we should all have Gemma 4 downloaded. We should all have it downloaded as a backup plan and then make them watch me code and comment. Hey, you could put them through the paces. Power went off recently and I did pick up a book. You did, Kyle. Wow.

[00:47:18]And from a storm. So, ready to pick up a book. You're good to go.

[00:47:25]Reported you for YouTube terror.

[00:47:28]That's funny. Okay. I'm allergic to grass. I bonzupi. That is uh that's fine because you don't ever have to touch it.

[00:47:37]You don't. You never have to.

[00:47:40]You feel like concepts are not that hard. It becomes hard when you when we don't know the concept that we should have known before understanding the one.

[00:47:48]Yeah. Well, sure. It's okay though because I think you can go into most situations without understanding it and still succeed. You can. It might struggle. You might struggle a little bit. It's worth it. I think it's absolutely worth it. You should go into situations that you do not understand anything about and struggle. That struggle is going to help you remember better than if you were to learn it ahead of time. I think that struggle is what makes you learn.

[00:48:15]Things are very complicated. Yeah, they sure are.

[00:48:19]Doing is making himself back online first thing. Yeah. I mean, prerequisites are everything. You can build a house without understanding foundations. So Claude tube basically is saying it's going to rebuild the internet.

[00:48:32]What's the to-do list for? The to-do list is for our heat map library. You can see here we got a nice little heat map library. There we are. Where are we at? Here.

[00:48:43]Here we go. So we're building a heat map library for PyTorch.

[00:48:51]And we've got a bunch of to-do items that we need to complete to make this good. And then I I really want to jump back in and getting PyTorch going.

[00:49:02]We have Stephen on record stating for a fact that you don't ever have to touch graph. Well, you're you don't have to, right? Why would you? There's no requirement for it. It's not it's not requ one time I stayed inside for 6 months. I didn't remember when the pandemic happened. I I just stayed inside. I'm like, I'm not going outside.

[00:49:21]I'm not I'm not going out there. It's scary.

[00:49:28]Or is it just a list? Is it a function?

[00:49:32]Oh, yeah. This is It's just a list here.

[00:49:34]We got a It's just a a bunch of code comments. So, it's a to-do list. We got the new GitHub repo up. We got the GitHub repo up right there. We want to got the examples in the read me. So, we got two examples, which we need more.

[00:49:49]I'm going to say more examples.

[00:49:51]More examples in the read me. I think we got enough for now. We got an initial amount. And then we're going to create a package here by Let's see where we at here. So, l we're vi pi project. We need to finish this really quick. Uh Steven Blum.

[00:50:13]There we go.

[00:50:15]And then blum.steeen Steven G. No, you guys didn't You didn't see that. You don't know what's going on there.

[00:50:23]Just wondering. I have to This is going to be public anyway.

[00:50:26]blum.stephvengmail.com.

[00:50:30]There we go. Don't send me anything. Uh requests. We don't need requests. We do need PyTorch though, right? So, I need pip freeze. One second. Pip freeze.

[00:50:44]Grab torch.

[00:50:47]I need torch this one. I need torch that one. Okay, perfect.

[00:50:55]Okay, torch.

[00:51:00]Does it doesn't exactly need to be 2.9.

[00:51:03]I would say 2.0.

[00:51:06]Greater than. No, there we go. Greater than. Perfect.

[00:51:10]Okay. It needs to be torch 2.0.

[00:51:14]Probably. All right. probably. And we we don't really need torch. We don't we don't really need torch. It could be numpy, though. It could be a numpy. So, that would make it better here. Here's what we should do. We should add numpy support.

[00:51:30]Uh numpy support as alter. Well, numpy or numpy support.

[00:51:38]And then we should also add in uh native native native support. Native only support. Couple answers there. A few bit of items.

[00:51:52]Okay. Thanks for answering your questions. See you next time. Sounds good. MD.

[00:51:57]Saab. If your workplace is one street away and your day is like this, wake up, cross the street, come back, is it still considered as staying inside?

[00:52:07]M I mean it's minimal outside if we're going to be very specific about it.

[00:52:14]Minimal outside.

[00:52:17]It's minimal.

[00:52:20]So touching grass is a separate thing.

[00:52:23]The idea of touching grass is taking downtime for yourself. That's what that term means. Touching grass means it doesn't always mean go outside, but usually it does. It says go outside and do something that's not work. So technically Saab, it counts as not touching grass, but it does count as going outside.

[00:52:47]Okay, let's keep on going here.

[00:52:53]Let me see here. All right, project URL is going to be this here.

[00:53:03]Okay, got the URL dependencies though. I don't think we we we I would want I want to remove this. It It should have no dependencies. How's it going there, Rival Sebastian? Welcome on in. Thank you for clicking the hi button. Good to have you here. Happy Wednesday. All right, I think we've got everything I need. And then I'll publish off stream.

[00:53:23]I'll publish off stream.

[00:53:26]Let me see. Or maybe maybe we could publish it now. Should we publish it now? I don't know. Let's find out.

[00:53:33]Remove dependencies. All right.

[00:53:37]See, remove all depths.

[00:53:41]Do we even use that in heater? Hold on.

[00:53:46]Heater underscore init.

[00:53:52]Okay. Do we use torch? Oh, we do. Why?

[00:53:57]Oh, we don't. We actually don't need torch.

[00:54:01]We don't.

[00:54:03]Oh, no, we do. Right here. Ah, okay. I see what's going on. All right. I think we'll just keep it like that for now.

[00:54:13]That's fine. How's it going there, Jeff?

[00:54:14]YouTube headquarters. Welcome on in.

[00:54:16]Good to see you. Happy Wednesday, Stephen. You're the best. Thank you.

[00:54:20]Hey, Kyle, you're the best. You're doing great. Keep on going.

[00:54:25]Make it great. Do live your dream.

[00:54:30]Make it happen. All right. So, I think that's fine for now. We'll just deal with that. Okay.

[00:54:36]Close.

[00:54:39]Close that. Keep that open there. This will be called heater.

[00:54:47]No, no, no, no. Uh, thanks. Good to have you here.

[00:54:52]You wonder, do you ever play some video game or just go write code when you're feeling bored? Both. I've done both.

[00:55:01]Mostly video game. Mostly video game.

[00:55:05]Minecraft. So, been playing Crimson Desert and I I really like Star Starbound. I feel like I haven't played enough of that. I feel like I should play more of it. I did play it through all the way once when it came out many years ago. It's now on console. You can play it on console now.

[00:55:23]Let me see here. Where are we at? So, heater. Here we go. Looks good to me.

[00:55:28]Okay, so we've got a bunch of to-do items.

[00:55:33]Scale gradient to fit window. What if we publish this? Let's see if we can publish this right now. Let's go ahead and make it public.

[00:55:41]All right. So, all right. Command to publish to pi. Pi.

[00:55:51]I might have to do some extra stuff offream.

[00:55:58]Yep.

[00:56:00]Okay.

[00:56:03]Do we have to use UV? Can we use pip?

[00:56:07]How did I do this earlier? There's twine. I don't want to use twine. Do I have UV? I don't know if I have UV. I just don't use it very often.

[00:56:17]Apparently, I do. All right.

[00:56:20]UV build. If you're using UV, the process streamlined in a single tool. UV publish check.

[00:56:28]UV publish token. So I need a token. Can I do it without a token?

[00:56:34]UV build.

[00:56:38]Oh, CD heater.

[00:56:42]So that we'll call this the UV build and pi pi publish building. Oh, hey, we did it. Look at that. All right. Get status.

[00:56:54]All right. Modified in it. All right.

[00:56:57]Heater dot. Do do we want to do Do we need to com commit that by heater?

[00:57:04]Do we need to? It's just text files. I think we're fine. Okay.

[00:57:11]Let's see. Yeah, you have it. We have it. It's right there, Michael.

[00:57:17]How's it going there? Uh, yo, me again.

[00:57:19]And impressed by the real time closed captions. They're pretty good, right?

[00:57:23]These are pretty good. I like them, too.

[00:57:26]They're pretty impressive.

[00:57:29]Some software that we built many years ago. Where is that at? Here. Yeah, you can see you. It's open source. If you want, you can have it yourself, too.

[00:57:38]Right here.

[00:57:40]right here. GitHub repository on our Discord. Available for free. Available for free. We built it eight years ago.

[00:57:47]We've been maintaining it. We have your brain works in amazing Steven. Oh, Jeff's thank you. Appreciate it. That's a great compliment. Hey, what's up? Hey, what's going there, city? City Janna, welcome on in. Thank you for saying hi.

[00:58:02]We're doing some Python today.

[00:58:05]Get commit- AM.

[00:58:09]We added license and pi project tunnel for publishing package to pi pie you guys.

[00:58:23]Mhm. Push to the main. Don't you didn't see that? Get push origin main.

[00:58:33]I don't know if you guys noticed that just now. If you did, good on you. So, if you didn't see, if you didn't see, I'm not telling you what I just did there. Uh, you guys could literally just study Steven's GitHub and start building several legit businesses. Yeah. Pro you.

[00:58:48]There's so much opportunity in the world, you guys. You could build anything you want and then you could track customers and you can get it's a lot of work, don't get me wrong. It's a lot of work and Kyle definitely knows this. Kyle is definitely a hard worker and knows that it takes a lot of work.

[00:59:06]Why don't you work in healthcare in oncology?

[00:59:10]I do technically. I do because my company we have tele medicine. We've got special we've got specific tech for tele medicine around here. Digital healthcare. Here you go. We do. This is my company right here you guys. Over about 3.6 million patients served per month.

[00:59:35]sent you an email. Don't worry, it's my pro email. All right, Moons, thank you.

[00:59:39]H Stephen, you're making me watch your VOD again.

[00:59:42]Tova, all right, it only showed on the screen once, and it might not have been obvious what it was, but it was a a very common like casual sips water. Casually sips water.

[00:59:56]You guys didn't see anything. Yeah. So, Nurse Grid, one call hero. Uh we've got others. We got Teddoc, we power healthcare, communications.

[01:00:10]Sure, we do. So, I feel like I'm affecting that space in a good way. I feel like I'm doing a good job there.

[01:00:20]Got to go, loons. All right, sounds good. I'm going to check your I'll check your email. I'll check your email later.

[01:00:25]I've actually got to go through my email. It's been a while. We got a lot of stuff to do.

[01:00:30]All right.

[01:00:31]Let me see where were we at here. Thank you for the reactions you guys. Is Claude still Is Claude Tube still there?

[01:00:40]What's your website? Lil uh Tuni. Hey, Lil Tunki. Welcome on in normally. Yes, my website is called PubNub. Here I Yep. There we go. Thank you, Lun. You pasted the link in.

[01:00:54]Appreciate it.

[01:00:58]You are You are You guys. Yeah. So this is my company. We do m power communications for in-app multi-user experiences like auctions and bidding fintech where you get updates in real time. Uh tuni like what little Wayne says. Oh, is it Tuni? Did I Oh. Oh, Tuni.

[01:01:24]Tuni. I was think Well, I was I was pronouncing it like more more. Yeah, I was pronouncing it the way you would read it in uh like anime. I always read things like anime.

[01:01:37]So, tun tuni. Uh little tuni. Okay, little tuni. Got it.

[01:01:44]Not pasted. Just know it by heart. Nice.

[01:01:48]Lun, you typed it in manually.

[01:01:51]Are you transhumanist, Stephen? What does that mean? I actually don't know what that means.

[01:02:02]He loves to Hey, David. Hey, David. Good to see you. Welcome on in. Happy to have you here. I saw your message there. So, I don't What is a transhumanist favorite anime?

[01:02:13]Um, I've got a lot of them uh right now.

[01:02:16]So, the ones I'm watching right now, Attack on Titan, cuz I haven't seen it all the way through. I'm I'm nearing the end of it, right? I think there's only two seasons left. And then Fragrren Ferren. I'm watching that right now as well. And I've been reincarnated as a slime, which is a lot of fun. I've been liking that one. I really like it's called Oruron High School Host Club.

[01:02:39]Wamp wamp wamp. Hey, is that another one? Man, there are so many. There's so many. I know, right?

[01:02:47]Tuni. There's so many Tuni. So many animes. H. There's also the melancholy of Harahara or something like that. Favorite early 2000s anime. Well, the Naruto. I mean, I watched that one a whole bunch. Naruto, you will be dodged. I know. I know I will be judged. I know it. Definitely will be judged.

[01:03:17]I know it. I know it.

[01:03:21]Epic Bloxow. Can you what? Can you what?

[01:03:28]Let's see. Favorite. What about you guys? Uh, what's your samurai? Oh, yeah.

[01:03:34]Shampoo. Shampoo. I remember that.

[01:03:37]Uh-huh. What about samurai shampoo?

[01:03:40]There's There's a lot. I can't There's It's actually kind of exciting how many options there are. Cowboy Bbop. That's the one. That's the one I was thinking about. My last message didn't reach you.

[01:03:52]It certainly didn't. Let me see if I can see if I if I can grab that.

[01:03:57]Uh, what's the easiest program language?

[01:04:00]This is what I see here. We see what's the easiest program language.

[01:04:07]That's the last message I saw from Epic Blocks Bow. That would be I would say Python. It's got to be Python. Also, it it's it's also complex, so we can't say it's the easiest in the planet.

[01:04:20]Fooly Oh, FL Seal. That was really good, wasn't it? Fooly Michael. FL. I remember that. Fooly What about Digimon? Does Digimon count?

[01:04:33]I I think Digimon counts. I think it does. It's a little more Americanized, right? Some more American cartoon.

[01:04:39]Digimon does count, though. It counts.

[01:04:44]Sab, I got a I got a question. All right. If an evil evil wizard came in and stole all your technical knowledge, where would you begin again? Oh, so if I start from if you were to start from the beginning, uh, and it has to be today in 2026, probably I'd start at Python again. I'd start with Python, then C, then assembly.

[01:05:09]Scratch. Hey, Bonupi. That would also work. Scratch is a good way to get started. It sound like I'm trying to prompt AI. It did. It sure did. So, yeah. And then JavaScript slashTypeScript.

[01:05:24]And then I would do maybe some Cotlin, maybe Scala, those types of things, maybe some Lua.

[01:05:35]Bunch of options in there. Okay.

[01:05:39]I once wrote a short story called Digimon in real life attack of the AI.

[01:05:44]True story. Did you Did you write that?

[01:05:49]Nice.

[01:05:53]Okay, where is are we at here? Okay, so we've got a lot of tasks remaining. More gradients. I want numpai support. We can't do this. May maybe can we do this?

[01:06:04]Uh we're just going to keep it as as torch for now. You did Kyle. That's crazy. That's amazing. You should share that. You need to post it. People need to see this. People need to see it.

[01:06:16]Digimon is sci-fi. Totally could happen.

[01:06:19]Technically. Technically, it's in the realm. You could do it. I think so.

[01:06:28]All right. La clean. We're just going to do some clean up here. A little bit of cleanup. Don't need that. Don't need that. Oh, we need this. We need that.

[01:06:37]All right.

[01:06:38]And I would like to redraw this since shades don't matter anymore.

[01:06:47]Right.

[01:06:49]We could do some more themes.

[01:06:52]Okay.

[01:06:55]Double check where we are here on our list. Okay. We're doing good on our list. Okay. Good.

[01:07:01]Excellent.

[01:07:03]All right. So, this is our actual themes. We need more. We need more themes.

[01:07:09]We could do rainbow. I would like to do rainbow theme. So, we could do red, green, we need we need all the colors of the rainbow, right?

[01:07:21]Let's do let's do rainbow colors. All right. So, let's let's do all the colors of the rainbow. All uh rain rainbow colors.

[01:07:32]Red, orange, yellow, green, blue, indigo, violet. Okay. So, it should look like this. Let's do rainbow.

[01:07:40]Rainbow.

[01:07:46]You want uh let's see here. You decided the story was too infringing on existing intellectual property to share it. Well, you could open source it. You just can't monetize it, right? So, you I think that's safe, right? There's there's fanfiction all over the place, so it should be fair. Wasn't brightness an irrelevant feature? It was because now what we're doing is with this new approach. Thank you.

[01:08:12]Thank you, Torva. It's now doing full red, green, blue. So brightness is no longer necessary. So we have to eliminate that as well.

[01:08:21]You're right. You're absolutely right.

[01:08:25]So let's let's get rid of that now because you all have mentioned it. And since we don't need this, let's get rid of this.

[01:08:32]And then keep this. We don't need lightness anymore. We just need the specific element here. We just need this guy right here. All right, let's grab that char code right there. And we are going to put it right here.

[01:08:49]We're going to delete this.

[01:08:52]Perfect.

[01:08:53]Now it's faster.

[01:08:56]A smidge faster. Our shade function.

[01:09:00]That's really clean. Isn't that clean?

[01:09:02]Hello. Can you see my chat? Yes. Epic blocks bow. Hey, Snot, 123. Oh, you said no. No. Actually, epic blocks bow.

[01:09:10]Apparently, we cannot. We cannot see it.

[01:09:14]So, we need this to be red, green, blue.

[01:09:18]Let me put these Let me put these colors in here. One second. So, let's do let's just put everything to zero really quick. All right.

[01:09:27]One red. 2 3 4 5 1 2 3 4 5 6 7. Okay, we need seven. 1 2 3 4 5 6 7. Okay, so red 255. Does that Can we do H?

[01:09:48]Okay, let's see here. Red and see green.

[01:09:51]How do we do orange?

[01:09:54]128, right?

[01:09:58]Red, red, red, red, orange, yellow, 255, 255, and then green.

[01:10:08]Red, green, blue, and then 255, 128, and then 255.

[01:10:20]Let's try that.

[01:10:24]Hey Dina, good to see you. Welcome on in. Happy Wednesday. Good to have you here. We're doing some pretty colors today.

[01:10:32]Okay.

[01:10:35]CD. Wait, wait, wait. Python model.py.

[01:10:42]All right. We need to change the theme real quick.

[01:10:45]I'm just going to I'm just going to close this off. All right.

[01:10:51]I want to open up theur. Where's the themur? Model. Model. M model. M model.

[01:10:56]Model file. Theme will be rainbow.

[01:11:00]Actually, you know what? Just just for the fun of it. All right, let's do rainbow. Hey, there we go. That's what I'm looking for. Linear space.

[01:11:11]We did it. We got the rainbow, you guys.

[01:11:14]Look at that. All right. Hey, Quick Rob.

[01:11:17]Good to see you. Welcome on in. Happy Wednesday. Good to have you here. Do you remember me? I asked for interview tips.

[01:11:22]Hey, yes, Aditia. Mhm. I remember your name, Adav6003.

[01:11:28]Good to have you here.

[01:11:30]A lot of interview tips. There's so many interview tips, you guys. The most important thing in an interview is to have excellent communication with your interviewer. You want to have a good interpersonal relationship, your communication style, you need to make sure that you're able to work together with them to solve a problem. Especially now that we're all just asking the AI to solve our problems for us automatically in order to be collaborative with someone. You need to solve the problem with them together. You'd probably do it without AI. I think that would be a way to practice it. You attended my interview yesterday. Hey, good. I hope it went well.

[01:12:08]You're running on an evaluation on a model you fine-tuned. Really? Dina, tell us the results. Tell us the results.

[01:12:16]Ask us an interview question in a poll.

[01:12:18]That would be cool. Hey, that would be cool. All right, let's try it.

[01:12:24]Let's see. Where's our poll at? How do we I can't start a new poll right now. Uh, let me try to reload.

[01:12:37]All right, Kyle, let's do it. Let's do an interview question. H. Okay.

[01:12:44]Let's think. Let's think of a good interview question, you guys. What's a good one?

[01:12:51]Let's see.

[01:12:53]Let's see. I'm trying to think.

[01:12:54]Actually, you know what? It's It's been It's always situational. It's always situational here. Here's a hard one. Uh, and then how's it going there? Who is it? Uh, Chatanya. Hey, Catanya. Good to see you.

[01:13:08]Welcome on in. Happy Wednesday.

[01:13:12]you're able to solve. I'm My window's really small now. I can't scroll there. Okay.

[01:13:20]Still about halfway in the eval preliminary results look good. Full evaluation looks state-of-the-art.

[01:13:27]Really was able to solve the Python coding SQL queries and almost all the questions related to a IML. Waiting for the results. Hey, very nice. Very nice.

[01:13:41]Hey, I'm asking you a question and it doesn't show for you for some reason.

[01:13:45]Hey, you're right. It does. It's not showing up.

[01:13:48]I don't see it. Yeah, you're not there.

[01:13:50]Wait. Hello. Can you see my chat? Just this one. That's the last one right there. And then you said it doesn't show up. Yeah. Sometimes YouTube will hide things. They like, "Hey, you look suspicious. You're looking suspicious there." All right. Good interview question. H. All right. How would you work without AI writing your code?

[01:14:22]Like if internet went down, what would you do? Okay.

[01:14:33]So, the first option is uh go home. Uh play video games.

[01:14:41]Uh add option uh work manually in an IDE or text editor or text or you know or I just say ID.

[01:14:58]Uh, another option is I don't actually those are the only options I can think of.

[01:15:09]Okay.

[01:15:11]It's not the best interview question.

[01:15:13]We'll come up with more.

[01:15:16]Dan, you added me in the blue app. Oh, really? Wait, that happened? Evil wizard forgot a question. Where would you study from scratch if you had to choose a recourse?

[01:15:30]There is Never mind. Okay, sounds good. Uh, W3 Schools. W3 Schools. It's free. There's a lot of advertisements there, so be careful. There's a lot of advertisements.

[01:15:45]All right, let's see here. Heater.

[01:15:48]Heater. Heater.

[01:15:50]and go into CD heater. Get status. Get commit.

[01:15:57]Added rainbow theme.

[01:16:06]That's a pretty good rainbow theme. It's rainbowish. It's rainbow adjacent.

[01:16:10]There's no purple. It's missing purple, but it's got teal.

[01:16:16]I feel like we I feel like we should have purple in there.

[01:16:20]Let's do it. So, 255 zero 255. All right.

[01:16:30]Let me do this.

[01:16:36]You can add me in the blue app. Hey, there we go. The blue app, you guys. Was that uh X or is that blue sky? What is the blue app?

[01:16:46]Most interview questions are open-ended problem solving, not multiple choice. I know. Bad idea on my part. I know that's well made hard.

[01:16:55]Uh, yeah, I gotcha.

[01:17:00]Yeah, they're open-ended because really it's all about communication styles and seeing that if you're going to be a fit.

[01:17:08]All right, let's see what color that is.

[01:17:11]All right, that's magenta.

[01:17:13]Um, so red, green, blue. What if we did 128?

[01:17:20]All right, that's still too big. That's still too bright. So, let's do 128 here.

[01:17:25]A little bit darker. Now, that's purplish. That's purplish. I want it to be brighter, though. So, let's do a little bit more blue. Let's do 210. So, it's gonna be a lot brighter.

[01:17:39]It's not as dark as I want it to be. I want it to be like purple. Red and blue are purple. So, let's drop this down to 100 maybe. So, it's more red. That's too red though. Now, it's like maroon.

[01:17:52]100.

[01:17:56]That's what There we go. That's purple.

[01:18:01]I like that better. Okay.

[01:18:06]Brother, I'm new at coding. Hey, J bro.

[01:18:08]You have two questions. Suppose we are making a social media type website where should we spend more time refining what's the total expenses what uh so you user experience is critical right because everyone is going to be interacting with computers and a lot of their day now is interacting with AIS and they want that experience to be similar if not as effortless so your UI needs to be so simple that the user doesn't have to do anything right make sure the user has to do nothing in order to you as best as possible. If they have to follow steps, you've lost. So, it's all about user experience.

[01:18:48]I do know about the expenses though.

[01:18:53]If I say it doesn't show up, Bloxbow, but the app where people send short videos.

[01:19:01]Oh, like uh Oh, got it. Okay, I understand. I understand. Epic box. I understand. What is the difference between monolithic and microservices architecture?

[01:19:13]That could be a question maybe.

[01:19:16]That's a good one. That's a really good question, Kyle. That's great. That's a fantastic interview question.

[01:19:26]How could you reduce the work of customer care services? I mentioned rag based domain specific LLM. That's a good one. That's a really good answer and a good question.

[01:19:40]college for computer science and was required to make a mat lib. Hated it.

[01:19:44]Ah, the mat lib. Yeah, I remember that.

[01:19:50]David, it's okay to hate that. It's absolutely fine. You're doing good.

[01:19:54]You're doing good. You're on the right track. Can you add me in the app that start with D and it's like WhatsApp?

[01:20:04]Do I know what that is? I don't think I know what that is.

[01:20:08]Yeah, cuz you've got Instagram, you've got Twitch, you've got Tik Tok, you've got I guess Facebook, maybe you've got YouTube, Blue Sky, maybe Mastadon.

[01:20:22]Working on DSLI app at the moment. MF, that sounds great. Definitely share it with us. Share us. Tell us what it's like and give us a link when you're ready.

[01:20:39]All right, we fixed fixed rainbow theme origin Reddit link in general. Oh, you did Reddit link. All right, let's check.

[01:20:51]Let's take a look you guys.

[01:20:54]What we got over here?

[01:20:57]One second.

[01:21:02]I put together a list of software interview questions. All right, Google for engineering questions and most of them were about algorithms and data structures. However, if you interviewed for engineering before, you know that's not usually what happens. Yeah, exactly.

[01:21:17]You get confronted with a lot of questions about communication and management. See, I told you put some research and curated them and thank you. All right, this is great, Kyle. This is fantastic. Thank you.

[01:21:31]Disco RD. I have no idea what that is.

[01:21:34]I've never heard of it. Software questions. We'll do excellent MF.

[01:21:39]Looking forward to it.

[01:21:42]Was hiring someone and they gave these answers. It would raise a big flag. How can we derive the size of software product? Oh. Ooh. Ooh. That's rough.

[01:21:50]That's a rough question. I mean, based on the number of features and how detailed they are.

[01:21:59]That's 7 years ago. Oh my goodness. It's not It's still the same today. Humans, we have not changed much. We have not changed much.

[01:22:08]Counting lines of derived code. Yeah, that would be a bad answer. Delivered functional points. Yep, that's also a bad answer. It needs to be more oriented on the product itself. What the product is capable of, what the volume of users are. It's got to be external factors. It cannot be the software itself. It needs to be what the product what the software is representing as a product.

[01:22:34]Someone asked me this question my first what exactly did you mean by size? Oh yeah, you got to ask more questions. Do you mean how long it will take to build?

[01:22:41]How complicated it is? How many gigabytes the final executable will be?

[01:22:45]How big the code base will be? It's a pretty ambiguous question. They can't really answer. So for me, I already understood what the what they were asking. I already knew they're going to say, "How long is it going to take you to write the software?" Basically, there's a fire quote about that, is there?

[01:23:04]Let's see.

[01:23:08]Most essential coding interview questions.

[01:23:12]Zoom in on this.

[01:23:16]What is software project management?

[01:23:18]Software project management is a process of managing all activities like time, cost, quality management involved in software development. All right.

[01:23:26]This is are these really are these questions really.

[01:23:32]I mean these are what is project estimation? How long it's going to take to build it?

[01:23:37]The time and resources and efforts. How can queen derive the size of software product? Oh, that's the one right. Oh, they were copy and pasting it. Size of software product can be calculated in either two methods. No. Derived function points. So a function point might be I don't know it could be like ticket t-shirt sizes and things like that from a scrum perspective how long it will take to build that's what I think that is however I really think it's more about the product itself when the measurement becomes the goal it's no longer a good measurement tova that's a pretty epic quote DMS what's DMS at epic what is DMSS Yes. Hey buddy, how are you doing? You know me. I do. Void Doom.

[01:24:26]Hey, how's it going Void Doom? Good to see you. Happy Wednesday. Good to have you here. Also, Pangist, how's it going over there? Pan Panagio.

[01:24:37]Panagotus.

[01:24:39]Pan Fentanacis.

[01:24:42]Fentanacus. All right. Hey, thank you for the peace sign. Good to have you here. Yeah, this red. You should write a better one. There's there's it's it's actually it's actually good. What what is baseline?

[01:24:55]Measurement defines completeness. What are functional requirements? Well, I guess I look I answer that one myself.

[01:25:03]I don't even know if I can answer that question. Functional requirements are going to describe how a product is going to function. That's what I'm going to say. Functional features, specifications expected by the users, the end user experience. Okay. Okay, here. Let's Let's see if we can get some uh let's get some more songs going on here real quick. Let's go back to the previous window here. How about this?

[01:25:36]Perfect.

[01:25:39]All right. Okay. When people search software engineering, Steven Blum's face should pop up. Should it? I don't know about that. What are we building today?

[01:25:46]Void doom. All right. So we are let's see here. Thank you. Thumbs up. We are building we've deployed we're deploying this. We're a heat map. It's a heat map that allows you to visualize 1D and 2D tensors in TensorFlow. That's what we're doing. We've built it. We're adding themes. So we just added rainbow.

[01:26:11]We also have heat map which is the standard default. So you get this nice pretty pretty heat map looking there. So then your model can output beautiful terminalbased and oh wait that's what we got to do. We got to update that as well. Let's got to update that. One second. Let's update the readme real quick. See e readme. All right. Uh 2i pietorch 2i heat map visualizer uh for terminal.

[01:26:46]I'll see this is uh let's see here. How do we want to type this up?

[01:26:52]Hey Stephen, I want to start a course on Rust. Oh, that sounds great. Void Doom, I'll learn from it. Hey, that sounds fantastic. Rust is a really great language. It's powerful. It's fast. It's got built-in safeties. What a what a fantastic language. It is a little complex though. It is It is challenge.

[01:27:11]It'll be a little tricky because it's got some concepts that don't port well over to other languages.

[01:27:18]Direct messages or refers to private.

[01:27:22]Yes. Okay, I got that. Oh, Epic Box. Got it. I thought you were talking about disco and RD and DMS, which I didn't understand the context. That makes sense now. All right. Thank you.

[01:27:38]Makes sense. How's it going there? Mr. Bumpy Biggle, welcome on in. Good to see you.

[01:27:48]I'm trying to see if I miss any of your chats, you guys. Stephen, Rust by example. Hey, Kyle, you found the playlist. Thank you for sharing. Yeah, we go through all the Rust by example, which I think is a really great way to learn Rust, by the way. It was really valuable to me when when they came out with it. It's complex. That's why you want to create simple courses. Hey, there you go. Mr. Bumping Pickle, thank you for saying hello. All right.

[01:28:15]What is functional programming? I know this one. This is programming functions through closures, but allows you to create all sorts of fun patterns like minimalization, factories. It's a pretty neat pattern. Typically, I found it adds a lot of complexity. If you do a lot of functional programming, it works really well in smaller sizes. That's what I would say. Functional programming is a style programming language which uses concepts of mathematical functions.

[01:28:41]What?

[01:28:43]No. It provides means of computations to mathematical functions which produce results irrespective.

[01:28:49]No. No. I mean that might have some sort of truth to it. It's just not what it's meant though. Hey loons, welcome on back. Good to have you back. Which produces irrespective programming state.

[01:29:02]No, no, no, no, no, no, no. That's wrong. I could tell you right now. Who wrote this?

[01:29:10]What are you doing, Dominic?

[01:29:12]Dominic, you ask a great question here and you give that as an answer. You're wrong.

[01:29:21]You found your transcript of the questions you asked when you had an internship. At least some of them I sent them in Discord. Let's look at them.

[01:29:28]Let's look at look. Let's take a look.

[01:29:32]All right. linkure.

[01:29:36]Oh, thank you. Thank you, Bunupi.

[01:29:39]Appreciate that. All right. GitHub ideas maybe. Oh, wait. I remember you posted this previously. Oh, wait. What is the value of grid after running this code?

[01:29:49]Oh, did you Is this one of the questions?

[01:29:52]Is was this a question, Torva?

[01:29:59]Let's see here. So is this Python? It looks like Python. It could be Ruby also.

[01:30:07]So this is going to say create times three.

[01:30:15]Okay.

[01:30:17]So this is actually a really good one. This is actually really good. I like this.

[01:30:23]That's a good question. This is a lot of fun. Okay.

[01:30:28]So if we create an array that's empty and you multiply it by three, well it's it's zero, right? So you're saying 0 * 3. So this should create three zeros.

[01:30:40]So at the very minimum at the very minimum I mean it looks like D is the answer.

[01:30:51]I think D is the answer. Is that right?

[01:30:55]Did I win? Did I win? I don't know.

[01:30:58]We're going to find out.

[01:31:00]It's about different styles of thinking, different problems, different tools.

[01:31:05]Python, yes. Nice, bro. Can we solve advent of code question? Try them. Hey, void doom. Yes, we should do that. That would be a really fun stream. I really want to do that. We're just taking a quick little break here to go through to questions. Let's find out. I'm going to I'm going to We'll pull this up. We'll run it. So, I thought D was the answer.

[01:31:23]So, it could either be three uh 3 1 2 3 * 3. I think it's I think it's D. I really think it's D. All right, let's try it out. Python.

[01:31:36]All right, grid. Here we wait, wait, wait.

[01:31:40]Grid three. Oh, I got it. Ah, I was wrong. It was B. Ah, I should have I should have known, but it doesn't make any sense.

[01:31:52]Okay, hold on.

[01:31:56]Grid. So, it's a bunch of zeros and then you set zero. Uh, hold on. Grid zero and then zero equals 3 grid. Huh? Huh? Oh, because is it? No, I actually I don't understand why. It must be because this giving the same reference to the internal matrix.

[01:32:30]You got it. I was wrong. I was wrong.

[01:32:33]It's this thing that got me right here.

[01:32:34]I'm like, okay, so it's the very first element, right? It's the very first element of the first array. So that it should only update this. Why does it update all three?

[01:32:45]That is so odd to me.

[01:32:48]Cuz when you do this, it's zeros. It's all zeros.

[01:32:56]Crazy. Oh, that is it. It is. Okay. What are you trying to code right now? Build.

[01:33:01]Hey, Trey. How's it good to see you?

[01:33:02]Welcome on in. We're building a heat map output for your terminal user interface, for your terminal window. It's going to be able to visualize layers in PyTorch.

[01:33:12]And I just wanted this because I wanted to see kind of what it looked like. And it makes me really happy. It makes me really happy. It's exactly what I wanted. Right now though, we're doing a few of these interview questions. What is the final value of data after the sequence? This is a whole bunch of array manipulation. What is going on? What is this? All right. So, times three. So, it should be a whole bunch of empty arrays.

[01:33:36]Data zero be appended. So, append. So, it should be a bunch of empty arrays. So, three empty arrays. two-dimensional matrix. So the first one you'll have a one in it and then the second empty array you'll assign it to two and the third you're giving it a three. So it should be 1 2 3.

[01:34:05]I think D is again my my I think D again is the answer. I'm pretty sure D is the answer. Let's find out you guys. Let's find out. Did Steven get it right? Here we go.

[01:34:18]Data.

[01:34:21]Oh, what? Okay, that's uh 1 3 2 and then 1 3.

[01:34:29]So, it was none of these.

[01:34:33]1 3 2 1 3 1 3 Oh, it was a Okay, it was A. It was A. Why 13?

[01:34:46]So, let's go. Let's do the first one real quick. Data a bunch of empty arrays, right? And then you data data zero that append.

[01:34:59]Oh, interesting. So, it it maps it maps over all the elements, huh? What do you know?

[01:35:11]I've never used it like this. See if we can we can hold polls. Oh, yeah. Sure.

[01:35:15]Sure.

[01:35:20]We already know. Sorry. Let me catch up with your with y'all's chat messages here real quick. What are you uh what you trying to code? Yes, exactly. I see.

[01:35:29]Nice. I know nothing about PyTorch. Hey, Trey. That's what we're That's what we're working right now, though. We're messing around with Porva's interview questions, which I'm getting wrong. I'm completely failing. I'm getting them wrong. I didn't realize that this uh would do a map operation.

[01:35:47]Can you make YouTube? All right. So, we already know the answers. We I thought I knew. I think it's time for the D language to be developed considering C++ has been since 1985.

[01:35:58]41 years is not progress. David, I think it would be nice to go to the next level. I think it would be H. We got H. Okay. Uh, your heat map remind me of that method used to interpolate neural network predictions by altering the input value to produce similar values and fit linear regression to understand the local relationship.

[01:36:22]That's basically what that's kind of what we want here. We want to visualize the matrix. We want to visualize the layers and kind of see what they look like. Now, most of it's not going to give you a full picture because it's really abstract, right? Because you've got multiple dimensions and you can't really visualize that as a human. It's too much. It's too much. Yeah. Row zero and two reference the same original list while row one gets assigned a new object.

[01:36:49]Got it.

[01:36:52]Got it. Yes. a bunch of a whole bunch of array adventures here.

[01:37:00]It's just it's just weird. So the append function is going to map over things. I didn't know the append function did that. All right, considering this code using slicing, what's the output? Okay, slicing.

[01:37:14]This is all of A. This is just all of A, right?

[01:37:19]This is everything in A. So it's the same thing as B equal A.

[01:37:26]So 0 0 = 5 and then A equals 1.

[01:37:36]Oh, that's a reference. Oh, I can visualize it. Kanye West, you get a little bit of visualization there.

[01:37:44]So this is tough. It can either be A or B.

[01:37:53]It won't be C and it won't be D. I don't think it's D because you're you're printing B and you're setting B equal to A.

[01:38:03]And so if anything, well, actually it could be D. I always think it might I I'm always thinking it's D. I like it feels like it's D.

[01:38:14]So we've got A equals uh a double matrix with zero.

[01:38:22]B is equal to A.

[01:38:24]Oh, you're setting 0 0 to 5. So it could be B, right? Cuz A and B are the same.

[01:38:35]And then you overwrite A and you're still B is still to A. So I'm going to say B. I'm going to say my answer is B.

[01:38:42]I'm gonna choose B. All right, you guys.

[01:38:44]Let's find out.

[01:38:46]Let's find out what we got here.

[01:38:51]Okay, B. It was B. All right, we did it.

[01:38:56]All right, at least I got one right. At least I got one of them. That's great.

[01:39:01]All right, last question. What happens when you use a mutable object as a default argument?

[01:39:08]A mutable object as a default argument.

[01:39:11]All right. I forgot about this. I know.

[01:39:14]I I've run into this before in the past.

[01:39:17]I've seen this before.

[01:39:22]Stephen, you're hired. Forget the questions. You're bringing us bringing you on. All right. The There we go. I like that. I win. I win.

[01:39:36]One of my first interview questions that I remember was interviewing for a startup in Seattle many many years ago, decades ago at this point. They said, "What what does pro product ownership mean to you?" And I said, you know, owning the outcome of the product.

[01:39:57]That was my answer. And they're like, "No, you're wrong. You're wrong. That's not it at all." I'm like, "Wait, why?

[01:40:03]can you explain to me what you think it is? And they didn't respond. They didn't respond, which means they weren't collaborative either. They were trying to trick me or something. They were trying to trick me. They didn't want me to be there. How's it going there, Caruk? Good to see you, Caruk. Welcome on in. Happy Wednesday.

[01:40:20]Good to have you here. Yeah, it's uh B.

[01:40:24]Remember the next one being even harder.

[01:40:27]Spent like six minutes to answer the correct at the end at least. Oh, you got it correct. Nice.

[01:40:33]Okay.

[01:40:36]Print print print.

[01:40:40]Let me see this. Let me see this. All right. Add item val items is I think the answer is this is static.

[01:40:50]So if you set this and you mutate this, it's going to stay mutated.

[01:40:56]So that's my memory. So items will stay muted. So you set items equals once.

[01:41:00]You're going to append one, two, and three. So, you're gonna add a bunch of items.

[01:41:07]Uh, so it's going to print one append. So, you're going to get one one two with an empty array and then 1 2 3.

[01:41:19]So, it's going to be either D. Oh, it doesn't give you very many options.

[01:41:28]One 1 2 3 1 2 3. So it's this one.

[01:41:35]I feel like the answer is A, but I feel like I'm wrong.

[01:41:39]Yeah, it's the Airbus internship. Is it really for these Python very esoteric uncommon patterns that you would ever hardly use?

[01:41:52]So basically, we're mutating items.

[01:41:54]We're adding one, we're adding two, and then three. The only thing is I see this extra ad here which is an empty array.

[01:42:01]So I would see one, two, and three.

[01:42:05]One, one, two. Not this one. No, I think the answer is A.

[01:42:13]Cuz you're using a pend with an empty list and it has nothing. So I think the answer is A.

[01:42:20]Is your plan to become a YouTube star and sell your company for a billion dollars? Hey David, good question. Guess we're going to find out, aren't we?

[01:42:28]We're going to find out. Look at Post Hog behind Post Hog thread. I pinged you. All right, sounds good. Lo, yeah, they were testing me on edge cases to understand on the data flow and runtime.

[01:42:38]Nice. All right, we'll check out. Well, Stephen, did I get the internship?

[01:42:45]Well, it's got to be one, one, two, one, two, three. It can't be this one. It can't be that one. And it can't be this one. It's got to be A. I think it's A.

[01:42:55]Right.

[01:42:56]All right. Here we go.

[01:42:59]Oh.

[01:43:01]Oh, we've got indentation problems. We can fix that.

[01:43:12]Here we go. Oh, whoa, whoa.

[01:43:14]No. Okay.

[01:43:16]No way. What is that? I got it wrong.

[01:43:19]I'm sorry to disappoint you. I know, right? I didn't get it right. That doesn't make any sense. It makes no sense.

[01:43:27]Add item one.

[01:43:31]One weird.

[01:43:36]What if we do this? What does that do?

[01:43:38]It What?

[01:43:42]This is amazing.

[01:43:44]So, why is it do This is weird.

[01:43:50]Oh, it doesn't make any I don't get it. I don't. So, let me explain. All right, sounds good. Hey, Rival Sebastian, good to see you. Since we are What age did you start learning coding? At 11. At 11.

[01:44:04]Uh, and what age did you start mastering coding? And also, I'd say mastering around 17. Uh, always stressful time.

[01:44:14]Glad to see you streaming. Nice.

[01:44:17]Good to hear it.

[01:44:19]Oh, hello world. Hello. It's it's world.

[01:44:23]Hey. Hello world. It's world. Good to see you. It's been a while. Welcome on back.

[01:44:28]Good to have you here. Performance assessment and collaborations going on at your college right now.

[01:44:36]Always. Oh, those tough times. You got it. It's a default list is created once at definition time. The second calls uses a fresh list, but the third call reuses the original default list. It makes sense. I don't like it. I don't like it, Torva. I don't like it. It I understand it. Do I like it? No, I don't. I I understand. I don't I just don't like it. I started earlier than you, like at 9 or 10. Nice, Lun, it's a good time. You You're doing it right. You're doing it right. All right, real quick. Loon, we're going to go to your Discord and then we got to jump back to our coding project.

[01:45:18]All right, what do you got here? Okay, thanks for the emails. A whole bunch of post emails. Pay us less. What?

[01:45:27]Why is it Wait, wait, wait, wait. Why is Why did they say that?

[01:45:33]Oh, wow. That is unexpected.

[01:45:37]Usually it is the opposite.

[01:45:44]So, good to see you. So, Josh Jooshi, welcome on back. Good to have you here.

[01:45:49]Happy Wednesday. It's company calibrations. You wish you were back in college. Hey, you really college? I like it better not in college, but I went to college and I, you know, didn't finish because I had better things to do. It was an email. We don't want all of your money. I guess so.

[01:46:13]Seems crazy.

[01:46:16]All right, let's get back to our coding real quick and we're going to wrap up here.

[01:46:20]I'll look whether I still have the rest of the interview transcribed. Unsure since it was the last qu after that question, you started to sweat a little since you your sixth one, you think?

[01:46:30]Okay.

[01:46:36]Okay, let's do one more to-do item and we're going to wrap up here. Where's our to-dos at? To-dos will be down here.

[01:46:45]Okay, so we did another gradient. We We could do numpy support native support only. Actually used the lib for our actually used the lib for our original intent, which I really want to do. All right, let's do some cleanup.

[01:47:00]Let's do some cleanup. Uh, you know what? I think we should be able to get rid of the shades. I don't want to cuz I like them. Oh, I like having these here.

[01:47:09]I do. Let's get rid of this old uh Let's see. Sample. I'm just going to keep it as a sample for now.

[01:47:21]Get rid of this. Doesn't return anything.

[01:47:25]We just print.

[01:47:28]All right. Do we need We don't need these themes anymore.

[01:47:33]We don't need shades. I want to keep shades though. I want to keep it.

[01:47:41]Here we go. Let's bring it back over here under the sample.

[01:47:47]You know, I guess we technically don't need to see here. One second.

[01:47:56]Let's do that. Yeah.

[01:48:05]Okay.

[01:48:11]So, Joshi, hello. Steven joined an M andNC recently. What's an M andC connecting with the domain? Should I stay one year for experience or switch early? Stay. It's always good to stay because if it it looks like you didn't stay there for long enough, it's a it's a question mark. What's going on? Why do you only stay there for a little while?

[01:48:31]Although, I don't know what an MNC is, so I guess we could look that up. Ahmad, you start losing your respect over since you stopped streaming Rust. Oh, we could We did Rust. We did Rust once. We did Rust once this month. We did. We did Rust once. We did. I'll show I'll prove it to you, Stephen.

[01:48:50]Here we go. Go to lives. All right, let's let's scroll to it. Rust. See right there? Rust enums right there.

[01:48:57]There we did some rusty nums.

[01:49:00]It was real. We really did it.

[01:49:03]This I like this one. It's a good one.

[01:49:06]We'll do more rest later. We'll do more rest later.

[01:49:11]Hey there, sir. Remember me? Shrew Jenk Haraw. Yeah, a little bit. Your name's difficult to to remember. It's difficult to remember. Good to have you here.

[01:49:24]Welcome on back. Thank you so much for joining in.

[01:49:27]We're just working on some PyTorch stuff here. So, I think we don't need We're just going to keep samper. Let's keep there. Okay. Sample. We got a heat map.

[01:49:36]We don't need these anymore. Let's get rid of these.

[01:49:40]Okay.

[01:49:41]Keep the sample around. We got our themes and our shade. Sample themes.

[01:49:46]Shade.

[01:49:47]And then I'm going to keep what I should do is put this into the readme as chart chart. Uh here we go. Let's put this into the read me uh over here.

[01:50:02]All right. This will be Python.

[01:50:08]Just just just so we have it around.

[01:50:12]reference char codes for Tui.

[01:50:19]Hi, Steven. Enjoying the stream today?

[01:50:22]Fun questions and polls. Learn some new things as well. Zergiola, great to hear it.

[01:50:28]Use Ratatouille. Oh, Ratatouille for the Tui and Rust. That's a good one. That's a good one. Yeah, exactly. Can you show us any of your photographs when you were teen or in 20s? I'm so curious to see your younger self. Sure. I need to figure out where they are. I don't have I don't keep I don't regularly hold those photos around. They're probably in like Facebook or something which I need to log into. They're probably in Facebook.

[01:50:58]Is this all available in Anaconda distribution like access all the place like torch etc. So we've put it on we're going to put it on pi. So I need to publish. I just I need a pi pie token.

[01:51:09]So I'm going to do this. We're going to do this really quick. Let me grab it over here.

[01:51:15]Uh, CD heater.

[01:51:19]All right, let's try UB publish.

[01:51:24]All right, enter username. Okay, so I have to do I do some off stuff. I'll do the off stuff off stream. Can you please tell me the skills to learn for joining PubNub? Sure. So, right now we are I'll tell you what we're using. our tech kubernetes m which is going to be microservices you're going to need kubernetes helm charts um 12 factor it's called 12 factor let's write this all down let's write this down all right things to know for pubnub engineer all right so we're going to do an ordered list here we go let's do a couple items here.

[01:52:11]Okay, so Kubernetes, so things like cube, ctl, and k9s, things like that. Oh, and uh Helm, docker.

[01:52:29]I think that's good. I think that's good. Uh oh, why can't have your token?

[01:52:35]I know, right? Loans, you want my token?

[01:52:38]Kafka. That's actually a good one. So, we do have we do have Kafka, but it's only for a few things. So, we can add some Kafka in there. Kafka. Let's do languages. Java. Let's do languages.

[01:52:50]Languages. Java script. Type script.

[01:52:55]Let's do We also have Python. Python. We mostly have Rust. Here we go.

[01:53:02]Rust C.

[01:53:05]What else we got? Uh let me put in this the right order. Python SQL. See what else we got. Got Typescript and then you know basically almost all for SDK. So you can add in every language essentially at that point.

[01:53:26]Kofka, Rabbit, MQ, and I I suppose we could say like maybe some Reddus. Uh like we've got a bunch of stuff.

[01:53:41]Sir, last time I asked you, you told me summer internships start around June.

[01:53:44]May, June. It's going to Yeah, exactly.

[01:53:46]I'm eager to reply. Thanks. We'll let I will we'll drop a link on Discord to let you guys know. Absolutely. I know all of that. Just hire me. Hey, there you go.

[01:53:54]There you go. And then let's see what else. What else? What else? I feel like that's like the necessary foundational stuff.

[01:54:05]Is there more? I guess I mean general Linux CIS admin things.

[01:54:15]Um, what else? What else? H I think that's like I I'm trying to think of like things that would be required for most situations and I think that's some of the just general foundational stuff. Uh the big one was definitely Rust. That's a huge one.

[01:54:36]Yay. I know one thing. Hey loons. Nice.

[01:54:39]You got some one thing in there. Hope to see you in the interview. Hey, that would be great. Absolutely.

[01:54:46]Starting strong. Don't know Kubernetes.

[01:54:48]Don't know even Kafka. Don't know JavaScript very well. Nor C. I'm great.

[01:54:51]Hey, loons. There you go. How can you apply a mod once? I'll post it on Discord when we have it open. We'll post on Discord when ready. We are not quite ready and it also might not happen. It might not. I'll let you guys know. I feel like it should though. I feel like it would be necessary for it to happen.

[01:55:12]I feel like I need one more thing in here.

[01:55:15]What else should we put in here? I don't know.

[01:55:22]I'm just kind of thinking.

[01:55:24]What do else we do? What else? I feel like there's a lot of things. It's just everything else is just inconsequential.

[01:55:33]That's good enough for now. Things to know. Hey, thank you for subscribing.

[01:55:37]Welcome on in. You joined the right channel for software engineering. What we're doing today is building a heat map for our visualization of our pietorch model. Scarlet fire time. Yep, that's what we're You know it. Scarlet Fire by Otis McDonald.

[01:55:56]All right, you guys. So, I'm going to see here real quick. I'm really getting excited about working with you and going to get all the things to learn from you.

[01:56:04]That would be really neat. There is a lot there. Just to let you know, there might be quite a few applicants. There usually are. It will it'll be And I typically don't choose. The team will choose. the team will choose.

[01:56:19]Okay.

[01:56:21]So, where are we at here? Get status.

[01:56:25]Get add commit am. Wait, did we did we do a full cleanup? Let's see here. I think we fully cleaned it up. So, let me see. Got our shade, our theme, plot.

[01:56:37]Okay, that's clean enough. That's clean enough. I'm going to count it.

[01:56:42]All right. All right, let's mark that as a green check mark for clean and commit.

[01:56:51]We cleaned up package.

[01:56:57]Boom. Done.

[01:57:00]Integrate into Discord. Hey, Discord would be great. It'll it'll get used a lot.

[01:57:07]Okay, you guys. Thank you so much for joining today. I had a lot of fun. We're wrapping up now. I'm going to head out and go do some other work things. I want to publish this and I'll publish this because we had our token, right? So, it was asking for a token and I will publish this later and then we can do more of this tomorrow.

[01:57:25]More tomorrow. And it will be around 11:30 or so is when we'll go live about that time. All right, everyone. Thank you so much for joining. I had a lot of fun. We actually were productive today.

[01:57:36]We got a lot done. We got a lot done here. We got a second color theme. Bye Stephen. Had a beautiful day. You too.

[01:57:42]Zergio Torvo loons. Zergio.

[01:57:46]Uh J Jose, good to see you. Shraon, good to see you as well. By the way, are you tired? Uh not right now. I've got I've got a my day is only halfway over. I've got a lot of stuff to do.

[01:58:02]Good to see you guys. All right. Thank you so much for hanging out with me today. We'll be back tomorrow. Bye, Stephen. See you tomorrow. Bye everybody. Have a good rest of your day.

[01:58:11]Rival Sebastian, thank you so much. See you guys. Thank you for joining. Back tomorrow. Bye guys.