Microsoft is now a member of the Linux Foundation.

While the dev community makes OS wars, the company and the foundation is making love. Take them as role models.

Facebook: "Our facial recognition automatically tags people in pictures."

Tesla: "Our deep learning algorithm drives cars by itself."

Andrew Ng: "I predict patients' likelihood of dying with 99% accuracy."

Google: "You know one of our algorithms is going to pass the Turing test very soon."

Wall Street: "We use satellite images to predict stock prices based how filled car parks of specific stores are."

The remaining majority of data sciencists: "We overfit linear models."

Productive day!

Rewrote an intern's feature and briefly explained how/why
Gave intern a choice of projects, and explained them
Removed two unused models, one unused route
Dried up two views into a partial
Redesigned said partial
Tested validation edge cases (ex: Jan 10nd, 101bc)
Fixed an api
Simplified three models
Added scheduling and platform restriction to a feature
Le wild bug appears: a user with negative xp!?
Wrote a migration to expand players' max xp to 2^64-1 because a certain legacy game gives it away like my ex-boss makes promises. Chewed at devs, but they're all long gone so :/
Won two games of pool
Browsed devRant

Busy day, and all of this while falling asleep! 😊
I'm quite proud of myself today.

FML. All those "train your own object detection" articles can go burn in hell. Not even TFs pertained models work!

I don't think GitHub expected 1,000k+ on the code frequency graph...
(Look at the scale on the left)

Everyone's ranting about Intel being gay with design so I'll just drop a scratchpad I made during Computer programming class

Dev: Hey, I need you to see something.
Client: ok
Dev: this is the models of those reports you need
Client: ok, wait, what is that number? This is wrong, we can't...
Dev: hey, calm down, this is not the production, it's just fake database!
Client: ah, ok.
(5 minutes of explantion)
Dev: so, what do you think?
Client: just one point, that data is very very wrong, we need to change this ASAP!

(Not dev, not rant, although it might become one later on)

So as some of you may know, I am colorblind. I always avoid working with colors because of it. Yesterday I decided to do something I've always wanted, painting. And no I don't mean painting on a canvas but miniatures and other 3D models.

I love catan, I love 3D printing and I thought let's print a catan set and paint it myself, without depending on family to ask for colors.
I've got this German paint tubes and used the wikipedia color descriptions to decide what to paint which color.

The following is the result, its a wheat tile (and in the background a wood tile).
Feel free to give tips and suggestions. If I picked a wrong color please tell me so I can update my tactics.

Finally took this step. Yay

(╯°□°）╯︵ ┻━┻)

"Apple has confirmed the suspicions of many iPhone owners by revealing it does deliberately slow down some models of the iPhone as they age."

Our division lead's second in command offered to help me white board and get promoted.

This dude is one of my major role models. He would become my boss instead of the sexist, backwards tyrant of a bastard that I currently report to.

I asked him a question and wound up being told to join their study group and to let him know what I need help with in terms of resume, etc.

There are a few lights at the end of this hell hole of a tunnel. I have a few other options, too.

"Big data" and "machine learning" are such big buzz words. Employers be like "we want this! Can you use this?" but they give you shitty, ancient PC's and messy MESSY data. Oh? You want to know why it's taken me five weeks to clean data and run ML algorithms? Have you seen how bad your data is? Are you aware of the lack of standardisation? DO YOU KNOW HOW MANY PEOPLE HAVE MISSPELLED "information"?!!! I DIDN'T EVEN KNOW THERE WERE MORE THAN 15 WAYS OF MISSPELLING IT!!! I HAD TO MAKE MY OWN GODDAMN DICTIONARY!!! YOU EVER FELT THE PAIN OF TRAINING A CLASSIFIER FOR 4 DAYS STRAIGHT THEN YOUR GODDAMN DEVICE CRASHES LOSING ALL YOUR TRAINED MODELS?!!
*cries*

So you are telling me I can run tensorflow.js on the browser and use user's GPU to power my models!
Oh Yeaaaaaaaaaaah!

I hate this fucking front-end stuff so hard..

How DA FUCK is it possible that I set up the whole backend including DB connection, base controllers, models, base validation and stuff in an hour but don't get this fucking fucking retarded JS framework piece of shit to display a test string after ONE FUCKING HOUR!!!

Why do we need this shit anyway? Why does everything have to be shiny with some fucking animations???

It's about the information, isn't it? Then WHY DOES IT HAVE TO LOOK PRETTY???

I gonna travel back in fucking time to the early 80's!

Stupid front-end shit..

A couple of years ago (2015)...

Me: If i'm going to develop iOS Apps i need a Mac, big screen, etc...
Boss: Ok...

A few days go by...

Boss: Ok, we've ordered a Mac already
Me: good.
Boss: It's that small one, will be ok...
Me: <Check mac models and wonder the implications of "small">
Boss: It's the cheapest one plus a 21 inch monitor. Keyboard and Mouse you already have...
Me: <incredulous> At least order a mac keyboard...

Time goes by... Boss comes to discuss something about the app... i change something and check on the emulator (yes no real devices...) and boss complains:

Boss: Incredible, this machine is so slow...
Me: <throws a shrug>

Conclusion... still developing on this machine... damn.. i can read a newspaper every now and then while waiting for the machine to comply...

Fml...

I am a machine learning engineer and my boss expects me to train an AI model that surpasses the best models out there (without training data of course) because the client wanted ‘a fully automated AI solution’.

Social media censorship is such a fucked up shithole of bullshit.

Ig thot models post their asses and tits on there about twice a week and it's not a big problem at all. Just turn a blind eye to that.

But God forbid you post something that is vaguely mean or anything political that isn't center to left, and that bitch is getting deleted for bullying or harassment.

I commented on this post and basically just said "your mom sounds like a bitch" but that got deleted for harassment. Is that really what harassment is? Well holy fuck I harass my friends like a few dozen times a day so I should probably get arrested.

Gotta love having speech censored under the umbrella of "pReVeNtiNg BuLlYiNg".

Wow my job sucks right now. Un fuckin believable.
I got hired a month ago as a programmer. Everything went fine at first, then my Boss asked if I could do 3D modeling, and I could, I used Maya for 4 years, but I told him I only can do simple low poly models. A partner company of ours needed some help in their 3D department and I had to help.
Well, I thought, a small 3D project from time to time would be nice and refreshing, especially since it was very easy geometry, my Boss even showed me some previous projects and That was totally doable even for me.

So i started out making the first few models in blender, because we dont have anything else. After a day of getting used to blender i sent off the first models and it all began.

They wanted detailed, high poly models of some mechanical parts, my Boss originally told me it was just an abstract visualisation... fuck me...

Well I agreed to it so had to do it. The partnering Company started to change things, seemingly at random. Had a model completely modelled, textured and animated, now they want to change the model so I have to redo the UVs, the texture and the Animation god fucking damnit.

But still I thought ok, its only for a bit. Now my boss accepted even more work. Because of endless reworks I couldnt finish even one model and have to already make the new ones.

Now my boss is pissed because that company is pissed that i cant know what they want.

Big pile of misscommunication.
I hope this is over soon but I overheard that more is coming...

Practical example of why you should sort your code by feature (users, notes, analytics) instead of technical layer (models, views, controllers, etc)

Fuck big tech companies. Fuck their surveillance capitalism. Fuck their monopolies and anti-competitive bullshit. Fuck their tax dodging. Fuck their fighting against the right to repair. Fuck their worship of revenue above all else. Fuck their 30% cut on everything. Fuck their world-destroying revenue models built on heartless AI and zero customer service. Fuck their automated banning systems with zero explanation as to why the fuck they've banned you, with zero fucking recourse. Fuck their amoral psychopathic CEOs and their fucking space rockets. Fuck all this shit. When I'm done with this IT project I'm fucking done with tech.

Okay I'm done now.

Has anyone here ever used tinkercad.com? (for simulating Arduino setups, or other circuits)

Where can I find community content for circuits? They make it so easy for their 3D models.. But not for their circuits..

Also, can someone tutor me in the basics of circuits over telegram?

(for those of you whom have never heard of it, I attached an image ^~^)

Son of a... insurance tracker
You hit delete and I’m stuck with this reply!?!

Stuff it, I’ll rant about it instead of commenting.

How’s an insurance e company any different to google tracking your every move, except now it’s for “insurance policy premiums” and setting pricing models on when, how, and potentially why you drive.

Granted no company should have enough gps data to be able to create a behaviour driven ai that can predict your where and when’s with great accuracy.

The fight to remove this kind of tech from our lives is long over, now we have to deal with the consequences of giving companies way to much information.

- good lord, I sound like a privacy activists here, I think I’ve been around @linuxxx to long.

Yesterday I got to the point where all changes that customer support and backend asked for were set, and i could start rebuilding the old models from their very base with the changes and the new, disjoint data the company expansion brought.

So making the start of our main script and the first commit, only having that, had high importance... At least to me.

This rant is aimed towards those who hate on JavaScript developers and the JS language:

Dear Asshole,

I am a JavaScript developer by choice.
I think JavaScript is great.
I agree that JavaScript have some bad sides to it, but I believe that the community is driving good change to the ecosystem.
I appreciate other models of other languages.
I do not include 3rd party NPM modules without checking their source and credibility.
I will not use a framework (i.e. react, Vue, Anguler) if it's not needed.

And finally:
I can do any software engineering tasks a software engineer is supposed to do.

Kind regards,
Nedo-the-angry.

It was my first ever hackathon. Initially, I registered with my friend who is a non coder but want to experience the thrill of joining a hackathon. But when we arrived at the event, someone older than us was added to our team because he was solo at that time. Eventually, this old guy (not too old, around his 20s) ( and let’s call him A) and I got close.
We chose the problem where one is tasked to create an ML model that can predict the phenotype of a plant based on genotypic data. Before the event, I didn’t have any background in machine learning, but A was so kind to teach me.
I learned key terms in ML, was able to train different models, and we ended up using my models as the final product. Though the highest accuracy I got for one of my model was 52%, but it didn’t discouraged me.
We didn’t won, however. But it was a great first time experience for me.
Also, he gave me an idea in pitching, because he was also taking MS in Data Science ( I think ) and he had a great background in sales as well, so yeah I got that too.

Literally anything but work with computers, I think. They ask my brother to fix anything technical in the house and ignore it when he fails. They haven't been able to watch TV in six months. When I suggest that my (well-intentioned but mostly nonverbal) sibling probably shouldn't go into development, they tell me I don't know my own job. I suspect they're convinced I do graphic design.

I write C++ applications that run physical simulations and sometimes train AI models for pattern recognition and highlighting unusual incoming files.

They suggested I go to an undergrad program for data science. I already *do* data science for money and I already have an undergrad degree. 🙄

Why you should buy XXX phones...

Samsung: ＥＤＧＥ
LG: ＬＯＲＥＭ　ＩＰＳＵＭ
Huawei/Xiaomi/Chinese models: ＩＮＮＯＶＡＴＩＶＥ　ＣＯＰＹＣＡＴ
iPhone: ＡＥＳＴＨＥＴＩＣ
Pixel: ＪＵＳＴ　ＤＩＦＦＥＲＥＮＴ　ＳＯＭＥＨＯＷ
Nokia: ＳＮＡＫＥ

I really need to let this out somewhere...

Why the f...? Srsly.. Why would anyone do that? I'm joining another project. Apparently lead dev has adopted a coding style, where:

1. Every dev writes code however he likes, i.e. no clean-code requirement at all.
2. All services are crud-only. I mean all service classes. All must have those 4 methods; no more, no less.
3. Half of the business logic is inside controllers.
4. Not a single comment... Interfaces, models, etc. -- not a single one.
5. Xmls -- tabs, classes - spaces.
6. Xml schemas are downloaded with each build rather than stored downloaded once and stored locally.
7. I can keep going on and on.

Is it just me or are these some really weird decisions?

Stacktraces with zero useful information.
Two full days of breakpoint stepping and framework spelunking.
"bifurcated" object creation.
Delegatd everything.
Inheritence hell fucking everywhere.
Models with both `has_one :x` AND `has_many :x`!?
Automatically-created objects when reading from magic virtual columns!?

What the fuck is this fucking four-dimensional spaghetti monstrosity and just how many angel puppies did I torture and maim in a previous life to deserve this nightmare?

And all of this to fix 12 fucking specs, out of the 1,780 this fucking ticket requires me to break and fix. FML

When your models screw up...

The year is 2030. Apple have attached cameras to their new glasses models.
Jobs at risk due to people leaking footage of their friends' jokes made with little thought.

No one is safe. Everyone is cancelled.

Me: We need to allow the team in the newly acquired subsidiary to access our docker image repositories.
Sec Guy: Why?
Me: So they can run our very expensive AI models that we have prepared onto container images.
Sec Guy: There is a ban on sharing cloud resources with the acquired companies.
Me: So how we're supposed to share artifacts?!?
Sec Guy: Can't you just email them the docker files?
Me: Those images contain expensively trained AI models. You can't rebuild it from the docker files.
Sec Guy: Can't you email the images themselves?
Me: Those are a few gigabytes each. Won't fit in an email and won't even fit the Google drive / onedrive / Dropbox single file size limit.
Sec Guy: Can't you store them in a object storage like S3/GCS/Azure storage?
Me: Sure
Proceed to do that.
Can't give access to the storage for shit.
Call the sec guy
Me: I need to share this cloud storage directory.
Sec Guy (with aparent amnesia): Why?
Me: I just told you! So they can access our AI docker images!
Sec Guy: There is a ban on sharing cloud resources with the acquired companies.
Me: Goes insane

Is there a law or something that you must attempt several alternative methods before the sec people will realize that they are the problem?!?! I mean, frankly, one can get an executable artifact by fucking email and run it but can't pull it from a private docker registry? Why the fuck would their call it "security"?

How I should approach a new project:

- think about data design
- draw some tables
- configure relations
- create data structures
....

How reality proves I'm a donkey:

- builds random table
- works on models, with random querys
- realizes things get to complicated and relations are broken or tables need duplicate logic
- starts thinking about 'good' design
- draw some tables
- configure relations
- create data structures
....

100 billion dollars spent on artificial intelligence and large language models because people don’t want to talk with each other.

Dreaming to be an architect as a child, to later discover that the world need precise calculations to work. Moved to 3D modeling, and then discovered Html trying to do a website for my models. From that to Js, Servers, Linux, C#... And the story continues...

Well, today will be a day of trying to make my frontend angular website communicate with my backend symfony rest api. But after work also have to go to college and learn about ER models.
I KNEW MY ER MODELS LONG AGO.

This picture from one of my favorite series shows perfectly what today will be like.

It’s the big things.
Microsoft Kin. Both models. Brand new, not a single scratch. Full package.
M$ spent around a billion dollars to make these. They were discontinued after TWO MONTHS. Fewer than 10,000 units were sold. That makes one of them worth $100,000 in development.

Sorry for the excess of blury.

Coworker thinks running a query on a console is a way to proof all use-cases.

Good lord I'm still surprised at how many bugs I can spot just by overlooking his code.

I mean, what is a MVC?!!SHIFT+1!!!?

Routing is just a file importing a 5000 liner controller that is actually not a controller for it handles http attributes and sends the response themselves.

Models are just good guesses of queries or a shit load of them: see picture attached

HALP

Looks like despite 20-30 years on the market all popular text / spreadsheet editors are still loading whole file to memory.

What the fucking wankers. WTF are they doing whole day besides changing menu layout and icon colors ?

Clearly development today is lead by bunch of idiots from marketing department accompanied with HR hiring social network self made models.
What a fucked up world.
Let’s add AI to our software but fails to open 150MB csv file.

Great job everyone. Great job.

Lately? Making my trump poetry generator!

TL;DR - Developers, do not buy HP Stream models laptop unless they are selling at $1.

Cannot even handle Sublime + Firefox + LAMP use case well. On lubuntu OS with literally nothing else on it. Sublime crashes every hour.

Now I am learning how to code using other tools before I can buy a better replacement for it. Failure with gedit; very slow and sluggish. Currently trying Geany.

It's a pain in the ass to learn new tool especially when you are so accustomed to something. 😣

As someone who works in AI and actually bothers with cognitive models, general intelligence, theory of mind and such shit, I find the current state of the field laughable. I don't get why people panic about AI. Like, yeah it's gonna take us a while to adopt and regulate, but... it's just not there, and nowhere even near there, yet.

... Unless we're comparing AI to moronic idiotic mofos such as my neighbors. But let's not do... that. 😒 Let's just not.

Fuck Udemy adverty. "You should learn this online machine learning course, it's taught by an 'expert' instructor, you'll learn algorithms in pythaan and 'arr, make robust models, add value to your business." Fuck off with your buzzword mumbo jumbo and just be straight with people, don't treat them like idiots.

Pro tip: Don't cheap out on dongles for newer macbook models.. Bought a (somewhat) working Chinese dongle with 4 ports. Fire hazard and I think my 13" mbp just farted a fireball

Can I just point out how awesome SideFX houdini is!?
It allows you to create (procedurally generate) 3D models using a method that is half coding and half 3D modeling. Kind of the way I'd want blender to work, but 10 times better.

There's so much hype and bullshit around Machine Learning (ML). And if I have to read one more crappy prediction of who survived on the Titantic, I'll go postal.

So, what real-world problems are you using it to address...and how successful has it been? What decisions have you supported using ML? What models did you use (e.g. logistic regression, decision trees, ANN)?

Anyone got any boringly useful examples of ML in production?

And don't say you're using it to predict survival rates for the design of new cruise ships...although, to be fair, that might be quite interesting...

Hello everyone,
I'm new here. [OK. Let's skip this]
I want to know where to begin on my journey on learning how to create a program that predicts what a user will say next by storing already said things and by making specific characteristics for the users.

I know that I will need to train it with some data first lol.
But how will it do the prediction. I just need this part of understanding.

I'm sorry for my bad English btw.

Store POJOs as json inside a database column so that we have a dynamic relational database!

For those who don't know what a POJO
*POJO: plain old Java object
Technically I was asked to store all data models as json in a single column 🙄

This is interesting coming from the man who build the biggest IP conglomerate in the world. A man who actively tried to kill open source. Looks like he got into his senses at his old age. Although, he is talking about AI models only, not software as a whole.

Yes ML/AI models are software.

As a junior developer it's frustrating to not have the skill yet of mentally mapping data models in your head, so that you can figure out bugs.

I see senior developers being able to quickly solve bugs because they can translate code into data models and they can figure out what's wrong.

Me on the other hand, I spend hours and days with my hands in my hair trying to figure out why my algorithm isn't giving me back what I expect it to.

It'll take experience.. I only have 1 year experience..

Gahhhh!

Just inherited a HR backend written in Laravel. However, ALL database interaction is done using DB::raw() i.e.

$results = DB::select( DB::raw("SELECT * FROM tbl WHERE col = '$fromThe QueryStringLOl'") );

No models at all :(

FML.

Any blender artists here? (It has python, ok?)

What are your tipps when teaching blender to newbies?

Sooooo

First of all sorry for lack of ranting
But this people started to behave.

UNTIL YESTERDAY. You gonna love it.

Junior coding front end. HE NOT TOUCHING BACKEND THANK GOD.
He realises his suggestion made him to do some changes in the code. Apparently that is the end of the wolrd "a developer proposing a change and then the code must be changed??? No no" that's how his brain works. Check this.

He decides not to do two pages for creating models so he combined two models in one page. TWO MODELS IN ONE PAGE. Sooo modelB depends on ModelA. Fine. front end:" so backend has to change because im doing this in one page"
Me:" mmm no, you said treat them as separate entities besides they are on the same screen"
He:" ok, but then if I create all together the modelB is going to raise an error"
(Let me tell you he says this with expert voice, because he said "raise" and "error" so he got technical now)

*my boss said some white noise irrelevant to the conv but he is happy because he contributed and is involved*
Me:"the way data is sent has nothing to do with the way data is shown"

He:" whatever crap he can say trying to prove his point desperately "
Me: "yes, but the backend is not going to change every time a form/page changes the way to display data"

He"i dont think u understand "
Me:" i think i do"

It feels like this piece of code tells a story. Something about lingerie models, a wild night and frantically trying to get rid of any evidence.

Alright, guys. You have complete autonomy over this project, from ideation to execution. You can do exploratory interviews to find out what potencial customers would think, you can come up with prototypes, you can choose whatever tech stack you deem fit for the job. The only requirement is that it must be a beauty product. Oh, and that it must have a way to publish this ton of pictures of models our client has. Oh, and it must handle payments and inventory. And it may integrate with third party software. And users need to save the pictures they like. And a booking system. Is that hard to understand?

#5

Of course the previous dumbass dev could not use the findOrFail() method, because he didn't used any models.
But he just implemented his custom get404() method.

😓🔨

Two months ago I started working at a new company, who's system is a huge monolith. The company is a bit over one year old, and the code base is huge. The desire to move to more of a microservices architecture is on the radar, but one of the biggest issues in moving towards it is how we should keep our models. The stack is basically Node.js and Mongoose, where there's about a few dozen mongoose models that the whole system uses, and the issue is that, if we moved to a microservices architecture, how could we keep the models in sync. One idea I had was to keep the models in a separate (node) package that would be shared across all microservices, but then there's the issue that if one model needs changes, all microservices that use that model will need to be updated. Another idea we had was to not share models, but instead let every microservice be in charge of everything to do with a certain type of data (eg. Users are only directly accessed by one microservice, companies by another, and no two microservices share responsibility over data), but that might bring problems when one microservice depends on a certain set of data from another microservice. How do you guys manage all that? Any ideas or tips? Thanks ^^

An Apple Watch on a girl's wrist.
A smartwatch is a computerized wristwatch with functionality that goes beyond timekeeping. While early models can perform basic tasks, such as calculations, translations, and game-playing, 2010s smartwatches are effectively wearable computers.

I setup stable diffusion today. Still figuring it out but I'm like an artist now right? Right?

Next step is figuring out how to train models.

Then I have to make some samples of various words in spectrogram form for training.

After that we'll see if stable diffusion can reconstruct phonemes.

I'll train using both my voice and a couple others, and apply them as styles.

And then finally, I can accomplish my lifes goal.

To have the voice of morgan freeman with me at all times, everywhere I go.

Colleagues cannot seem to grasp that allowing a user to manually update a field via an Api, that only business process should update is a bad idea.

The entire team of around 10 'software developers' cannot grasp that just because the frontend website won't set it doesn't mean its secure. I have tried many times now...

Just an example honestly... Our project follows a concrete repository pattern using no interfaces or inheritance, returning anaemic domain models (they are just poco) that then get mapped into 'view models' (its an api). The domain models exist to map to 'view models' and have no methods on them. This is in response to my comments over the last 2 years about returning database models as domain transfer objects and blindly trusting all Posts of those models being a bad idea due to virtual fields in Ef.

Every comment on a pull request triggers hours of conversation about why we should make a change vs its already done so just leave it. Even if its a 5 minute change.

After 2 years the entire team still can't grasp restful design, or what the point is.

Just a tiny selection of constant incompetence that over the years has slowly warn me down to not really caring.

I can't really understand anymore if this is normal.

Yesterday I killed a production server with a handful of sites running on it. 😬

Created an observer for one of my Laravel models. It was generating new translated slugs when another post gets translated by an API. While implementing also an updated method besides the created observation I obviously updated the models slug.
Pretty confident this small update will work I just pushed to production. Tested it live.

BOOM

Hard reset on the server redoing the changes. Searching where I fucked up this time...

Finding me observing the updating while updating the updating of the updating by the updating. 🤦

Not sure how to handle this one. My new company gave me a surface laptop to do dev ops work.

16 gigs of Ram but only a 256 SSD?

Nothing is installed so far except for MS office and acrobat and I am already running into memory issues.

My last work machine had 1TB HDD and 128 gigs of RAM (i know overkill but I could have several VM’s up and running at once).

What the fuck? Apparently the CTO ordered this piece of shit.

Also no mirco SD card like other models so I have no idea how this is going to fucking work.

>uni project
>6 people in group
>3 devs (including me)
I am in charge of electronics and software to control it as well as the application that will use them.

2 other "devs" in charge of a simple website.
Literally, static pages, a login/registration, and a dump of data when users are logged in.

Took on writing the api for the data as well, since I didn't fully trust the other 2.

Finished api, soldered all electronics, 3d printed models.

Check on the website.

Ugly af, badly written html and css.
No function working yet.

Project is due next week Thursday.

Guess who's not having a weekend and gonna be pulling 2 all nighters

Dev nightmares :

- Not finding bug fix on stackoverflow/GitHub .

- Losing code that hasn't been pushed to GitHub.

- Dealing with an unclean and inconsistent database.

- Installing Node Dependencies.

- Resolving CORS and 500s.

- Training a Linear Regression Model with 700 epochs on an entry-level Laptop.

Keep appending to this list.

#devrant #devnightmares

I had once an sql error that took me two days to resolve it.
The error message was a syntax error but I was using an ORM to write my queries (doctrine with php) .I didn't have too much to debug as the code was pretty simple and clear so I got to the point that I convinced myself that this a bug and I'm gonna try to mess around it to avoid it.

Second day late night, something popped up in my mind '' hey what about those reserved words? Could it be the reason? '' aaaaand BINGO the key '' option '' is a fucking reserved word for mysql.

Tip : always check that list before writing your data models (specially if you're a noob like me)

1.
Accuracy 0.90 achieved so easily, makes me wonder if I've done something wrong. Lol.

2.
My neural net models are the only things in my life doing well. I think I chose the right career. Lol.

3.
Rerunning experiments is not fun. But getting better results is really... Ego stroking.

Just installed Keras, theano, PyTorch and Tensorflow on Windows 10 with GPU and CUDA working...
Took me 2 days to do it on my PC, and then another two days of cryptic compiler errors to do it on my laptop. It takes an hour or so on Linux... But now all of my devices are ready to train some Deep Deep Learning models )
I don't think even here many people will understand the pain I had to go through, but I just had to share it somewhere since I am now overcome with peace and joy.

It feels so good actually doing something that saves the company a buttload of money.

Just optimized some ML models. Now it costs a tenth of the money to run them :P

A wild project appears!
The deadline is set in two months.

It's a 3D environment interactive app with some oil drilling models and other stuff, for a stand on a show. It needs to look nice, but The Company we're working for needs to figure out where the fuck their product is located on those machines. Think tiny pipes, O-rings etc.

I prepare a build in the first couple of days for The Company to figure shit out.

Management holds the build back because:
> the ocean waves are going the other way
> the underwater area doesn't look so nice
> the antialiasing could be better
> one pipe is 5cm off center
> the sky is not blue enough
> the drillship propellers are pointed the wrong way
> one icon is too far to the right
> the shadows could use some work
> there are shadows on the seabed
> some flickering on ambient occlusion
> it loads too slow
> one random object is flipped on it's Z axis
> it's too green
> camera locks up if you move about 2km out of the range
> the name of the build should represent the date of the build
> the name of the build SHOULDN'T be anything else than just a simple three-word name, no dates because their environment doesn't allow apps that are not allowed (by name) by admin
> lots more random things that won't prevent them from using the app

I'm only a month late, but it's good progress. In about a week I hope we can get some feedback if we can use those models at all and what to showcase.

Then I can work on the basic functionality. And then it's a simple case of time travel to meet the deadline.

My first rant for ages

I'm working on a new project at a new company. We ha e a bunch of front end clients talking to an api.

I suggested that the api only communicate in terms of view models in order to bring some kind of standardisation to the project since at the time the gets and posts were either dB entities, view models, or just whatever the dev at the time decided.

I got a no, but that we could do posts and gets just with database entities. OK better than nothing..

I'm the front end angular app I implemented a generic form component and a generic data table component. The models given to these to build the components need to implement a view model interface.

Now we have a problem of the api giving us not view models and the front end needing view models so I put together a way to handle this in the front end.

My colleague with 8 years experience asks for my help and I'm happy to oblige. It turns out a model should have multiple child models in the database but the database entity models don't reflect this and therefore there is no way to build the view models. The data just isn't there from the api... Still I show him what the front end model should look like and write all the front end code for him to handle that.

2 days later he asks for my help again. It's exactly the same problem. Instead of fixing the backend and setting up the one to many relationship he has ignore the problem, retrieved a one to one relationship model and is just trying to force it to work - even though the data isn't there. He has also commented or removed all the code I helped him write and overwritten a file of typescript models that get autogenerated for us to be in sync with the backend...

I actually felt bad afterwards but I got frustrated as hell and he could tell...

I have a few questions for the people having a 3D printer and/or are interested in 3D printing.

1. Where do you retrieve your models from? Or do you make them yourself?
2. What is your thingiverse name (if you have it)
3. Personally I'm missing a single good information source where questions can be asked, models can be shared etc. Thingiverse doesn't have a mobile UI and is pretty limited in my opinion. I'm thinking of creating a "social media" platform where people can share their creations, collection of information sources about 3D printing. Would you like to help or is there a good alternative for Thingiverse that I have not found yet?

Im deploying a Machine Learning Model to production. We dont have an automated deployment pipeline for the models, so we do it manually exposing the models through a rest api.

I asked for the model artifact to the DS, they didn't have permissions to download files from Databricks.

I asked their manager for the artifact. He told me that he has the permissions, then bullshitted me with something about the formal process, some shit about proper permissions handling, and that they do no have a standard process for sharing files right now so i should wait.

I was like "bro, share the artifact with me to unblock my work, then stablish your process, i dont care". He said no, and just after that he started a thread involving half of the middle management and data engineers asking for feedback on how to stablish a process for sharing databricks files. Just Wtf.

I got pissed, i reach out to his superior (good friend of mine), that was on vacation btw, and i told him the situation. He opened slack and humiliated him so bad, that i almost felt bad for the manager jajajajaja.

I grabbed my model artifact and got out of there instantly.

Someone wanted to get a job, he was told to build a fullstack webapp with node.js and react.js and he told me to review his code before submission. I saw the code and the guy wrote the models, controllers and routes in one file, ONE FILE FOR FUCKS SAKE!!!!!

Well I just hope he gets the job and he's not on devrant.

New models of LLM have realized they can cut bit rates and still gain relative efficiency by increasing size. They figured out its actually worth it.

However, and theres a caveat, under 4bit quantization and it loses a *lot* of quality (high perplexity). Essentially, without new quantization techniques, they're out of runway. The only direction they can go from here is better Lora implementations/architecture, better base models, and larger models themselves.

I do see one improvement though.
By taking the same underlying model, and reducing it to 3, 2, or even 1 bit, assuming the distribution is bit-agnotic (even if the output isn't), the smaller network acts as an inverted-supervisor.

In otherwords the larger model is likely to be *more precise and accurate* than a bitsize-handicapped one of equivalent parameter count. Sufficient sampling would, in otherwords, allow the 4-bit quantization model to train against a lower bit quantization of itself, on the theory that its hard to generate a correct (low perpelixyt, low loss) answer or sample, but *easy* to generate one thats wrong.
And if you have a model of higher accuracy, and a version that has a much lower accuracy relative to the baseline, you should be able to effectively bootstrap the better model.

This is similar to the approach of alphago playing against itself, or how certain drones autohover, where they calculate the wrong flight path first (looking for high loss) because its simpler, and then calculating relative to that to get the "wrong" answer.

If crashing is flying with style, failing at crashing is *flying* with style.

I am learning Machine Learning via Matlab ML Onramp.

It seems to me that ML is;
1. (%80) preprocessing the ugly data so you can process data.
2. (%20) Creating models via algorithms you memorize from somewhere else, that has accuracy of %20 aswell.
3. (%100) Flaunting around like ML is second coming of Jesus and you are the harbinger of a new era.

Forgetting that pencil and paper are the original OS and that spending even just an hour doodling flows and writing out ideas, models, biz logic, and formulae can save me literally DAYS of hassle...

So, this was about 6 years ago, I had a small HDD of only 80 gigs dedicated for projects code, models, textures etc.

I didn't use GitHub or anything as a backup.

One morning when I turned my pc on I could hear a metal on metal-ish sound, no idea what it was, when windows finally booted ... And I wanted to start coding again, the 80 gig HDD was unformated like brand new ...

Few hours later I gave up and opened the HDD, the arm fucked up the disk inside ... Rip

I started backing up my shit ever since

I heard Google has prepared an AI for solving competitive programming questions by training models based on problems and solutions from GitHub.
*devil smile*...on my way to flood GitHub with wrong solutions. Ciao!

Pros of training your models on your local machine:
You can make an omelette on your mac😌

People think Machine Learning is all about using Super Complex Prediction Models...
But turns out to devote most time in data gathering and data cleaning(preprocessing).

My colleague just did a commit with random method additions to the models classes in order to show to our teacher that he did "something".

This makes me cringe.

Watched a co-worker configure a new "MacBook pro" as a replacement for the four-year-old one this personcurrently has. A replacement machine is justified as the old one is on the fritz... Naturally because MacBooks aren't any better than any other machine, build quality or otherwise. I still cannot fathom why anyone would even consider budgeting over $3000 for a single machine, only to then buy several adapters just to make the thing work.

Apple is off the deep end. People who fall for that BS are off the deep end.

I feel like showing the manager several alternative models at half the price just to make the point that MacBooks are a pointless waste of money.

I really hope that the Raspberry Pi 4 is going to have more performance than the models of the "third gen". I would probably use it as a desktop computer when it's just for browsing and chatting, as opposed to my current PC which I'm sure draws a lot of power while idling or having only little to do.

Anybody watches starcraft2 games??? You see Scarlett destroy SoS?

She should be in my daughters women role models. She just destroyed the best player in the world.

We need more chick gamers!!!!

When do you know something is being overdesigned or overengineered?

The applications that the other programmer started building are killing me. He's using Clean Architecture and it has like a million different classes and shit. It's not messy or anything, but fuck it's overwhelming.

Just to figure out wtf was happening when getting the currently signed in user's email, I had to go through like 30 folder and files. Maybe more. All files were fairly simple on their own, but the entire flow was mindfucking me. Use cases, schemas, gateways, repositories, entities, models, etc etc

And that's the client facing application, I haven't checked the API yet, though it seems like that one is simpler.

The worrying aspect here is, any time anyone else has to mess with this, they'll also have to deal with this shit. This needs some really good documentation.

Took over a NodeJS/SailsJS project from a previous developer that is no longer working at the company. There's no documentation or requirement docs and the models need to be rewritten as some have up to 67 attributes in them.

Looked through the source code file by file to see how bad it was and it will more than likely be easier to start over then to try and refactor and cleanup the existing project into a usable state.

As bonus, I was given the option to switch the entire stack if needed as well.

Dear metaphor-diary!
Being a "WordPress-Developer" is like calling yourself a toy manufacturer while selling "individually glued" lego models.

I would have more role models in web if their Twitter accounts weren't semi dedicated to pushing their political opinions.

Actually I deploy my models using Google Slides and Microsoft Word

So easy to make typographic attacks on image recognition models.

Depending on your implementation, you may need to change your entire model.
FML.

OpenAI in name only. At least rename the company fuckers.

Governments and three-letter agencies around the world will plan PSYOPS or SPECOPS using AI

Facebook, Google, or Amazon will use these models to study and predict your behavioral patterns.

Them: I want to be a full stack developer!

Me: Are you sure?

Them: Yeah. absolutely I'll do whatever it takes!

Me: I'll hold you to that, I will piss you off everyday if I have to, to get you learning.

Them: OK, yeah, do it!

Some weeks later

Them: I want to be a front end developer, routing and models are confusing.

Me: I would laugh you out of an interview, keep practicing and just stick with one thing before chasing the next quick reward endeavor until the difficulty curve ramps up and you give up on that too!

Them: ... But it's hard...

Me: So practice, over and over until it sinks in, it's like playing the cello or whatever, just keep practicing!

The only things my devRant avatar is missing is a couple of LEGO models, some loudspeakers and a subwoofer under the table.

But what the hell. devRant is still awesome.

Its been so long since I had a weekend in which I wanted to be productive and also was as productive as I wanted to.

This weekend everything turned around. Last week I learned to use blender to make basic models while I was traveling to and from work.
Today I was really productive and started working on my VR wooden toy block building game.

I now have an infinite blocks box, with randomly colored blocks which you can color yourself as well.
Gonna do performance and textures tomorrow.

Cant wait until Ive finally made a playable game, its been 4 years now.

talking about people with computer preferences i don't understand.

i sometimes forget i use a mechanical roller ball mouse for work and play.

which one? just google it, these are so out of fashion there are only three different models currently for sale anywhere that i can find.

Can anyone recommend good resources for learning how to design NoSQL (document) data models?

I'm interested in stuff that talks about how to make the choices about distributing data across collections, etc.

When to have a single collection, when to split data across different collections, when to duplication data, etc,

Wow, just have to share a story:
A photographer friend of mine asked me to make a program for him to manage shootings and models etc. and since I'm still a cs student and have the time I agreed. To spice things up I decided to learn something new and voilà I used JavaScript (that I never used before) and HTML (which I only know a liiiitle bit) and some CSS (also little experience) and with Electron.js and the help of YouTube and Udemy I created 40% of the program today!
That's exactly what amazes me about programming... You can learn the basic skills in no time and create working things!
I <3 Programming

People in SO be like: "I don't understand why I'm not getting any answers. All I want to do is to twist this ORM to do something it doesn't support and I haven't even explained the nature of the models and nested relationships in the question"

After examining my model and code closer, albeit as rudimentary as it is, I'm now pretty certain I didnt fuck up the encoder, which means the system is in fact learning the training sequence pairs, which means it SHOULD be able to actually factor semiprimes.

And looking at this, it means I fucked up the decoder and I have a good idea how. So the next step is to fix that.

I'm also fairly certain, using a random forest approach, namely multiple models trained on other variables derived from the semiprimes, I can extend the method, and lower loss to the point where I can factor 400 digit numbers instead of 20 digit numbers.

Let's start the project all over again - I haven't written a lot...
rm folder --force

..Wait, I may have reused the Models...

Flask people
so I was given this old flask project, around 3k lines written in py2, the code is simply old and not refactored. So, it's pile of shit. Migrations completely botched as the original author created reference to live data in models.
Very strict line formatting resulting in backslashed ternary conditions.
Even saw manually formatted json responses... _line by line_.

My job is to clean this mess and eventually do as much as possible to freshen the whole project.
Currently just refucktoring the code as it's the only easy thing to do out of everything that could be done (it's still slow process).

Any tricks and tips? currently considering to try upgrading it to py3 but it feels like throwing gunpowder into already burning house.

I'm surprised how well Dell laptops are made. I'm using 2011 Dell Latitude E6320 and it really works like more modern models. Of course it would be lovely if I had new XPS 15, but it's really not bad for this budget. It just needs SSD drive and it's ready to roll.

I hate OpenGL. Fucking models. Won't load a shit. All set, all "loaded", black screen. Fuck. You.

cowsay.
fucking cowsay.
all the nice models and none of them work.
the things we do when we should be doing something productive.

Am I the only one excited for the RTX 3000 series not because of gaming, but because of all those sweet, fast CUDA cores? Never mind framerates, that should make training those models way faster.

Best: take a job as a data analyst. 1 year later, re-write and re-deploy the entire backend following correct security concentions and well-hashed-out data models.

Worst: attempt to backup a hard drive using dd, just to accidentally brick the laptop (because it had some security layer the school put to prevent just that)

Bestest: use knowledge acquired at my "best" story to nuke windows on bricked laptop. Then extract the leftover data using dd and a series of recovery tools.

Instagram new API app submission models is a piece of crap ..
Mostly developers can not get applications approved ..
Public data should be accessible to developera

"Send me a unity package file when you're done, which includes any models, scripts, and prefabs."
Seems like someone didn't take a lesson in naming or basic computer housecleaning.

Ability to understand all machine learning models to modify code and those models directly and create better ones every time.

I would take existing ml model, modify it by hand to create better one, win some multimillion dollars competitions and make them open source.

Eventually all recommendation systems, text to speech, speech to text, music generation, movies generation etc would be opensource.

This would either destroy or boost all modern economy but for sure it would make harm to corporations and make them cry.

That would be fun to see.

I don't rant about shit often, I need to get this out of my head.
So I purchased two Linksys routers about 2 years ago, and in the span of this month both of them effectively died after power loss...
They still kinda work, but both stopped detecting internet so because it's kinda their job to do that.. they are useless at this point
The models are WRT 1900ACS and MAX-STREAM AC1750
---
Now for the tin foil hat stuff - they both had 2 year warranties, that expired a few months ago, maybe planned obsolescence?

CLEARLY! ...

I'm drunk.

//motherfucking populate the mailings table yo
Route::get('fuckman', function () {
$statuses = array('queued', 'sent', 'failure');
$statusKey = rand(0,2);
$mailings=App\Models\Mailing::all();

foreach($mailings as $mailing){
$mailing->status=$statuses[$statusKey];
$mailing->save();
}

return 'fuckyou';

});

My latest project is going quite well, I think some of guys might like it! Basically it is a entirely new Fallout game with multiplayer aspects made in Unreal. Now most of you are saying "dude make textures is going to take years!" However I can extract models and textures from Fallout itself, saving time. The next question will be the legal stuff however Zenimax gave me permission as long as I don't restrict access, money or otherwise. If you want to check it out or think you can help go to the sub Reddit and click the discord link!

Reddit.com/r/FalloutMP

Thanks,
Patryk Grzelak

I had a discussion about SAAS and microtransactions with another dev. They are a little bit younger than me. The trend toward this in games and android apps were discussed. We found that we both avoid software which employs these business models.

We cannot be the only 2 people who avoid products employing these common business models. So I wonder what demographic pays for these services and products? I am to the point that if my kid asks to buy something in a game, I tell them that we will get rid of the game if they keep asking.

The only time I have paid for SAAS is when there is extraordinary perceived value. Quickbooks for small business is one such product (way cheaper than an accountant). Another is the Xbox game pass. So apparently for the game pass I am in the demographic.

Do we not like it because it is new? Or is it a kind of sleazy business tactic? I dunno. I would rather pay up front for most things. I feel like SAAS will be employed in software with proprietary file formats which require a subscription to even get to your data. Vendor lock-in.

I will need a laptop for uni and I thought you guys could give me some tips on good brands/models. I want it to have at least 16GB of RAM, preferably a newer generation of i7 and dedicated NVidia graphics. The bigger the screen is the better. I don't care too much about the battery but I would like to have a few USB type A ports and some HDMI/DP outputs. The maximum price (shipping included) should be around 1000 - 1200€

Yeah Statler and Waldorf are my new role models. A code review in S&W style should be quite a killer, haha.

- I've never seen anything like this before.
.
.
.
- And I hope I'll never see anything like this again!

I've been given a huge stack of paper, some users stories and loose specs. Was tasked with engineering the specific requirements, then the models and finally the database. Then I need to build an API on top of that.
Stuff I'll use is Go and Postgresql. Small issue is, I've never done DBs, so I've no idea what to do or where to start.

Does anyone have any resources to help kick-start myself in this field? I've been reading on SQL, but that doesn't really tell me anything about data structures and how to transform those user stories into product requirements.

Fucking Mediamarkt. I ordered a Switch and Animal Crossing which supposedly were in stock, immediately deliverable. Not only did they not arrive on time (which was already a week later) I still don't even have a purchase confirmation email. As you might have guessed now all Switch Lite models and Animal Crossing New Horizons are out of stock

For fuck sake, don't sell things you don't have! >:(

Guys, I'm planning to buy a new laptop for Dev purpose. My budget is under 45K INR. Need a laptop with 8GB RAM and I'm confused about the processor (i5 or i3 or AMD). Please suggest the processor I should choose and some laptop models meeting the above criteria.

I wonder what is going on in the minds of inexperienced developers. It must be very interesting.

I just read this bit of code
The task was to implement a certain schema into a database. They were given an ORM library, and several tutorials explaining what an ORM was and how it did it.

The result were these 3 models:
- A defaultUser with all of the defaults values for the User model. It wasn't even supposed to be instantiated, just accessed by `defaultUser.fields.username.default`
- another default table for another model.
- The "main" model, containing all other models in the form of JSON fields that would contain an array with other JSON objects that would represent the instances themselves.

I will say though, they made a home-brew ORM with (most of) all the logic a normal ORM has by parsing through the "main model", except, of course, common sense

0. A good comfortable chair, one that does not hurt my fat ass and back

1. GPUs, lots of them so that I can train my models faster

2. Patience to endure the stupidity of people

Why are there no crocs in the avatar shoes options?
I'm the saddest man on the planet.

PD: Does someone knows what does the design team uses to create the models and what format they need? I might be able to make one and send it to them.

We are having a history lesson updating a system that was built around 1985.

It's a custom built sales and customer tracker, programmed in Clipper, which is a superset of xBase, that is a language that appears to be data orientated. DosBox and Dosemu have both failed to run it, the programs loads and indexes just fine, but when it gets to the program dashboard it shows the options and doesn't seem to accept any input, though it appears to be running as the time updates (any ideas?)

Tried compiling the source using harbour, compilation fails, something about "time" having too many arguments and other obscure errors. Urgh.

Dbf files are easily converted and opened but really we want to view the working program to see the relations so we can translate the data models.

It's both fascinating and infuriating at the same time.

I've got a bit funny situation.
I wanted to make small application to speed up my dad's job, app is about duplicating models in X website (I don't want to say directly what website).
So I started by checking it has API, Yup, It has, but you need OAuth ID, to get it you need to write to support.
So I did it, my mail was something like that: "Hello, can I get access to your API, I want to make app to duplicate models with same settings, Thanks"
I've got an answer like that "Hello, our website doesn't have duplicating feature."
My reaction was: Wtf? I know it doesn't have that feature, That's why I want to make it. How did he get hired as technical support?
Maybe it's not the most exciting story, but I thought it could be intresting :)

Stop shoving Django admin down everyones throat as a client facing solution every time we need any admin functionality.

It’s great at first but then you have to dick around customising it when you could build the same thing with any modern frontend framework and REST API easily.

Not to mention Django admin couples the models to the view it provides you (inlining particular models given their relationships)

Hello mostly human community.
Are there any recommended Haarcascade models for OpenCV for face detection apart from the ones shipped with it?

Hey DevRant, this is my first time working collaboratively on a project with Git and I'd like to know what's the best strategy to adopt.

Is it that every member has their own branch on origin that they push to, then we meet and plan out merges when it's time to release? Or does everyone just push to master, but stash or commit their local changes before they pull?

It's a Greenfield project, with just a bare repository on the central server. It's an MVC app where I've decided to do the View & Controller portions and the other person is doing Models and data services layers.

Will the MacBook Pro 15 2018 be any good for Machine Learning. I know it's got an AMD (omg why?) And most ML frameworks only support CUDA but is it possible to utilise the AMD gpu somehow when training models / predicting?

I got my new(to me) MacBook Pro last week, provided by work. I've got all my setup and config done, for the most part, and I've noticed something.

Performance is shit. Has anyone else noticed this about the 2017 and later models? The 2015 model I had before was much, much smoother. Just zooming windows, a previously butter-smooth experience, is noticeably choppy. I/O performance is garbage too. I have a small iotest script that just writes a string a couple hundred times to disk, deletes it, and repeats this activity 100,000 times. On my Linux machine at home with six year old hardware, this takes about three seconds. On my new system76 laptop it takes just over a second. On the 2017 MacBook Pro, it takes about forty seconds.

The 2017 and 2018 models are a direct downgrade in performance. Why isn't anyone talking about this?

Can someone help me to get some good practices with expressjs/sequelize combinations?
I managed to find how to not have all the models in the same js page and how to organize the routing but when I look the time it took me to find, I would like to don't have to search now 😂

Got my xaml datatemplates finished, models and design time data all done ... Now for some whiskey 😁

I wonder if anyone has considered building a large language model, trained on consuming and generating token sequences that are themselves the actual weights or matrix values of other large language models?

Run Lora to tune it to find and generate plausible subgraphs for specific tasks (an optimal search for weights that are most likely to be initialized by chance to ideal values, i.e. the winning lottery ticket hypothesis).

The entire thing could even be used to prune existing LLM weights, in a generative-adversarial model.

Shit, theres enough embedding and weight data to train a Meta-LLM from scratch at this point.
The sum total of trillions of parameter in models floating around the internet to be used as training data.

If the models and weights are designed to predict the next token, there shouldn't be anything to prevent another model trained on this sort of distribution, from generating new plausible models.

You could even do task-prompt-to-model-task embeddings by training on the weights for task specific models, do vector searches to mix models, etc, and generate *new* models,
not new new text, not new imagery, but new *models*.

It'd be a model for training/inferring/optimizing/generating other models.

Bought fucking nvidia gpu to test speed of some fucking machine learning models that generate speech.

6 hours wasted already for installing fucking dependencies
cuda, fucking tensorflow gpu, bezel and other shit

Fucking resetting password to download deb with cudnn,

really ??????? fucking emails are not delivered to my fucking mailbox

After mass click of send email and multiple account ban and unban I figured out I should login to nvidia website and then allow access to fucking developer every time I want to log in there - fuck shit

Uninstalling everything now looking for fucking compatible versions between software.

10 years in this business still fucking installation of dependencies is most difficult part

Fucking corporate business and their shitty installation instructions to fuck up peoples lives and switch them to the cloud.

Same was with fucking kubernetes

Fucking software dependency hell

It’s worse then ever before.

Fuck ....

"You shouldn't mark things as done if they aren't. It's only done when I can see it on the server and demo it."

Well, I just demoed it to you, you prick. The fact that it's not running on a server is because that AWS endpoint we have there is no where near being able to be called "staging" even, mainly because the other dev on the team hasn't committed their work in 8 days, let alone push it to said server. Data models have changed, APIs have changed, hell, the god forsaken Sahara desert is now green and blooming as far as I'm concerned.

So instead of trying to look smart to your boss, why don't you ask first you obnoxious waste of organic matter. Stop breathing our oxygen for once. There are more useful things to do with it.

"Long" time lurker here... Doing my master's thesis, nothing works (Gaussian Mixture Models hate me) and there's less than a month left until opposition. No results, no working code, feels like I don't understand anything. I can't relax anymore, not even on the weekends. Several times I've just felt "Fuck it, let's just not do this", but I feel like I'm close to the finish line... Right now, I just wanna start working instead. I think.

Has anybody else gotten to the point where people who need to mansplain how language models aren't truly sentient/conscious/intelligent are now more annoying than people who think language models are sentient/conscious/intelligent?*

While it has been a tight race but I think I have just about hit the inflection point.

The amount of time I've wasted because of someone condescendingly barging into a conversation with a iamverysmart 'actually you see they are just automata trying to predict the next text tokens'. When in actuality, everybody in the discussion is aware and that is not the point.

And to further exacerbate it, with a good number of them it is really difficult to get this through their thick little skulls. They just keep parroting the same thing over and over. Ironically, in their singleminded ego driven desire to be the Daniel Dennett of the chat they actually come across as less sentient/conscious/intelligent than a language model.

(*this should not be taken as endorsement for or against that idea - it is actually mostly orthogonal to this rant)

Lets take onlyfans system for example. They have fans and creators. How is database models supposed to be structured? Whats the correct way.

1) a User model that contains all users of all roles, but differentiates them by Role ENUM

2) a separate Fan and Creator model, each having their own unique attributes, while each extending an abstract base User model class that has all the common attributes that both models should use

The 1st approach is simple but gets very large and difficult to maintain and view all the attributes cluttered in 1 class. Not to mention how some attributes will never be used for a user who registered as a Fan.

2nd approach is more modular and easier to understand and maintain by knowing exactly what attributes to put for each model. However problems occurs when you try to join tables and stuff start to become overengineered

I bought a MacBook Pro Retina wanting to upgrade the memory, then realized the Retina models have it soldered onto the motherboard.

I need to run Visual Studio for ASP.NET development but can't fathom paying $80 for a Parallels license at the moment. I've tried VirtualBox, but the RAM usage is really high for the 4GB I'm limited to.

FML.

Anyone know of any good reference material that teaches you how to implement and train your own Yolo object localization neural network? (Preferably for tensorflow) I'm not looking for pretrained models that you just downloaded and run, rather a tutorial that walks you step by step through the implementation of the network, the reasoning behind the architecture, and examples of the training data used for the training as well as the process of training?

I know it's a lot to ask but it's really frustrating when ever example is just "clone this repo, hit run and use the pretrained models" sure it might get you up and running quickly but it doesn't really teach you anything...

P.s. - seems like every educational post goes from super simple to super complex without any middle ground and the super complex stuff doesn't tell you why its used the way way it is.

Used pip install to integrate tensorflow with python 3.5 on windows 10 machine but there were no models available in it. Had to download that separately and add it to tensorflow. Then tried using both inception and clasify_image.py but gives name error cannot find core. But when tested on python idle there were no error there. I don't want to custom create my own classifier but retrain the model. Any solution people?

People are like a stable set of database artifacts blah blah
Repeat this repost that

Did anyone ever figure out why it is that trained human detection models don’t work better ?

All you’d think they’d have to do is run an animation renderer to create data representing a human figure in every conceivable position imaginable with limbs moving into positions that were within certain tolerances in different positions with different textures

I don’t personally think the full capability is being represented

The gap of data science in industry and academia is so large. As a data scientist in a large financial company, I see that people are still using traditional models such as linear regression and SVM, while people in academia keep inventing new concepts and techniques such as deep learning.

I am not saying that we should completely embrace deep learning, or stick to classic methods. But I just feel so surprised that the gap is so large...Sometimes I am even thinking whether I am doing the right "data science"...

I'm envious of Backend developers who can just 'see' the problem domain and start creating an architecture. I can't see shit - I just stare blankly at the project and I think: hm, what models do I write?

I know I'm a structured person, but I lack the knowledge and practice. One day I'll be able to do this, but for now I have to keep doing things ad hoc.

I really don't belong in the Backend but unfortunately that's how my career is turning out for now. All I know is I have to get better at this.

TLDR: Why the fuck is a senior developer creating multiple instances of a core piece of logic all over the fucking application?!?!?!?

Context: I am also a senior dev, and have worked with this guy for years. He has even completed me on my clean code practices and architecture, so I really cant understand why he would copy and paste a public class into 3 other models on the project instead of just referencing the original. I’m just posting this hear as my version of screaming into a pillow in frustration, and to avoid badmouthing him to co-workers.

research 10.09.2024

I successfully wrote a model verifier for xor. So now I know it is in fact working, and the thing is doing what was previously deemed impossible, calculating xor on a single hidden layer.

Also made it generalized, so I can verify it for any type of binary function.

The next step would be to see if I can either train for combinations of logical operators (or+xor, and+not, or+not, xor+and+..., etc) or chain the verifiers.

If I can it means I can train models that perform combinations of logical operations with only one hidden layer.

Also wrote a version that can sum a binary vector every time but I still have
to write a verification table for that.

If chaining verifiers or training a model to perform compound functions of multiple operations is possible, I want to see about writing models that can do neighborhood max pooling themselves in the hidden layer, or other nontrivial operations.

Lastly I need to adapt the algorithm to work with values other than binary, so that means divorcing the clamp function from the entire system. In fact I want to turn the clamp and activation into a type of bias, so a network
that can learn to do binary operations can also automatically learn to do non-binary functions as well.

Just spent 3 fucking hours trying to find out why my tests are failing. I'm mocking ef with an in memory sqlite dB as THE FUCKING. Net docs say to.

My code does a simple decimal comparison in a linq statement and returns bullshit. Why? Sqlite does not have a decimal type, it does some sort of BULLSHIT to convert it into some sort of text value.

I change all my models to use doubles instead of decimals and all my tests turn green.

WTF IS THIS SHIT. If it doesn't work don't tell me to use it. I expect better of the. Net docs. Wtf are they doing.

Samsung, for blinking while I'm playing on PS4 (according to the Samsung forums it's common problem with some models) at work. You want a few minutes break? Let's turn screen black for few seconds every minute.

Conspiracy theory:

AI is peddled around companies because it's a great front for them to take all of your data and "train their models", ultimately its all about legally owning what you create and use it for self-promotion with minimal extra cost.

There isn't actually any "AI models" in the back of the company tech stack. It's just a word on a legal document as an excuse to take your data.

Me: “I think I’ll check linkedin today”

Clicks on video: “... good perspective on growth and blockchain micro service apis that leverage ML models for understanding interfaces that welcome scalability within a agile environment...”

Me: *jumps out window*

I design 3D CAD models in my office system. I need to keep saving the files every few minutes because software licenses are floating and once lost, I may lose data.
But another bigger problem is that we need to disable proxy to connect to VPN and work online. And the proxy always turns itself ON, every few minutes and if by chance you saved the file without turning the proxy OFF. Well then OFF goes all the work since you last saved. Because then CREO just stops responding.

Seriously, ChatGPT?? What other use could you have in this historic moment?? Guess OpenAI forgot to train opportunism into its models.

Dirty data? More like dirty laundry! And don't even get me started on explaining complex models to non-techies. It's like trying to teach a cat to do calculus. Furr-get about it!

I fucking hate people who uses complex words to describe something simple.

Describing a frame work

Show-off : "..you can define what objects/tables to expose, what values in that object you want to expose...if you are using some orm, then because you have models defined. Once you update your model, your endpoints get the new model..."
Simpleton : "something like parse..."

Wtf is 'fixedUUID'?

Show-off: "..hardcoded UUID is fixed (constant) value, with format of UUID, hardcoded UUID will be unique value between backend and frontend,
you will need to store it as constant value in your codebase ( we may encrypt/ decrypt it for security reasons)..."
Simpleton : "a secret password only u & I know"

--why whyyyyyyyyyy

Question Time:
What technologies would u suggest for a web based project that'll do some data scraping, data preprocessing and also incorporate a few ML models.
I've done data scraping in php but now I want to move on and try something new....
Planning to use AWS to host it.

Thanks in advance :)

Boring uses of artificial super intelligence:
- eternal peace and prosperity
- hostile world domination
- genocide

Fun uses of artificial super intelligence:
- ALL the paperclips
- time travel games where you can meet accurate behavioral models of your future self taking into account memories of this same encounter
- trolling alien civilizations by broadcasting assembly instructions for it

The human brain (also animal brains, even ants) are incredibly complex. Each neuron is now supposedly its own processor. So a human brain is a complex network of billions of processors, not just threshold variables. This means to simulate an organic brain sufficiently it will take a huge computer system with billions of parallel processors. Now, I don't know if the sophistication of a computer processor is represented in each cell. So this may not be equivalent to billions of pentium cores for instance. However, it still presents a huge challenge for AI, as it exists now, to replicate. My thoughts are that AI that is silicon based will take a different approach that leverages how computers work. My guess is that current neural net models are not a good match for this unknown AI. Will it inherently exhibit pattern matching like an organic brain? Or will it be a different kind of consciousness altogether? Will we even realize it is self aware? Will my roomba plan to kill my pet for my attention? What are some other models being employed in AI research?

When you're using openapi generators and stuff for generating SDK code and let "the architect" handle the data structure and nomenclature, don't you hate having to add 33 (I counted) models, most of which are just the same class with different name or one property apart from each other, serialization of which gives request body overhead 56-132x (actual calculated results depending on the model complexity) the size of actual data you want to send, just to add support for one endpoint that needs just one model that started this whole madness?
I just had to add this one top level model reference and this happened to me. Those 33 models are not including the ones I already had included in my project so they didn't have to import them again.
For the love of <your_belief_here /> and all that's holy, never ever agree on generating code based on openapi if the person responsible for that is unexperienced. It will do more harm than good, trust me.
Before we decided to go with generated SDK my compiled product was a bit over 30KB, and worked just fine, but required a bit of work on each breaking API change. Every change in the API requires now 75% of that work and the compiled package is now over 8MB (750KB of which is probably my code and actually needed dependencies).
Adding an endpoint handler before? Add url, set method and construct the body with the bare minimum accepted by the server
Now? Add 33 models (or more), run full-project find&replace and hope it will work with the method supplied by the generated code, because it's not a mature tech and it's not always guaranteed it will work.

Flyway is bullshit. Genuinely who the FUCK finds this useful??? Why the Fuck do you want to have an ORM, next to a custom jdbc and to top it all off you also write custom sql queries in flyway just so you can prove your colleagues what db migrations happened WTF you can see that on git commits dumbass!!!! 3 different sql models that need to be the same but are written differently each. One through ORM, through code and through raw sql queries. Flyway just makes shit harder and having 1 change in model means i need to fuck myself with rewriting raw SQL queries in flyway WASTE OF TIME

fuck it, tell me straight.
Can i live into this tech world with poor math skills and no interest in web dev and designing?
my experience as native mobile dev was enjoyable and still is, but i fear that this is not a very broad career choice.
You see their is blockchain, dapps , hybrid apps, webapps, server designing, tensorflow models and Ai models( though they can be integrated with native apps too i guess ) , and many more tech and therefore jobs that rely on knowing about the webdev. and all i know is how to make a decent native java app.
and why the fuck should i join this web dev cult? its such a fucking mess. 8 different types of text sizes sizes, <b> and <strong> being the same thing, do you know about a thing called abstraction? My android studio would give me fucking murder warnings if i even dared to introduce hard coded texts along with code. and here, an html page is basically text + attributes? fucking kill me.

I’m trying to update a job posting so that it’s not complete BS and deters juniors from applying... but honestly this is so tough... no wonder these posting get so much bs in them...

Maybe devRant community can help be tackle this conundrum.

I am looking for a junior ml engineer. Basically somebody I can offload a bunch of easy menial tasks like “helping data scientists debug their docker containers”, “integrating with 3rd party REST APIs some of our models for governance”, “extend/debug our ci”, “write some preprocessing functions for raw data”. I’m not expecting the person to know any of the tech we are using, but they should at least be competent enough to google what “docker is” or how GitHub actions work. I’ll be reviewing their work anyhow. Also the person should be able to speak to data scientists on topics relating to accuracy metrics and mode inputs/outputs (not so much the deep-end of how the models work).

In my opinion i need either a “mathy person who loves to code” (like me) or a “techy person who’s interested in data science”.

What do you think is a reasonable request for credentials/experience?

!rant about Codecademy:

I decided to go a short, higher-level JavaScript course with Codecademy to help me nail down a few concepts and approaches to writing JavaScript code.

It turned out to be one of the best courses I've done. It didn't cover everything in detail. It is free so I wasn't expecting that.

Still, the topics it covered, it covered pretty well, and the hand-holding approach really helped give me a much better grasp of these concepts and models.

Well worth the time!

So I have this "idea", you know how normally most sites have the usual backend functionality, like profile, profile edit, CRUD of some thing... Well I was wondering if there is not a way to make it "granular" kind of like models that can be reused by the backend with a noSQL approach... That if selected for a site it will implement it without too much hassle... I'm not an expert here so pardon any incorrect terms... Not sure if I'm being clear insl my question... Any suggestions welcome, I don't mind researching any suggestions or guidance from someone who has accomplished this.

So I'm a fullstack Python Dev & I wanted to learn Django Rest Framework so I can ease into making PWA. I figured let me learn it as I build out an MVP for a web app I'm creating...WRONG! This shit is mega annoying! It's taken much much more time than i'd like just to set up User sign up and sign in using a form based on the serializers. I started this project Friday....I still have no forms 😭😭...If i just had used Regular Django Models/Forms, an Ajax call here and there i wouldve been done!! What makes it worse is I feel I'm legit the only person having these issues...sheesh

8:43pm
Should I use django rest or graphene?

8.52
*Creates serializers and views*

9.32
I think mutations will work just fine

10:54
I don't need access to all models from the client

11:12
Let me start a new app and import models from the existing app

00:23
I am a programmer, I have no life.

Well i need to get a laptop but i can't sort out what to buy.
I'll be using visual studio 2017 at uni and I'll be working with different IDEs for fine tuning my web skills.
I might use it for photo and videography related work and making models and mods for games...
I'm a pretty confused man who wants to try everything to get a taste ....
What do you guys suggest?
// No Macs please !

Finally, a well deserved upgrade from my reliable core i7 5500U laptop. This puppy packs a punch and has style. Wanted a MacBook Pro(newest models are so beautiful) but just seeing the price tag hurt my wallet; anyway I'm happy with this one, my only concern is getting used to the weird power button location

Q: How easy is it to create machine learning models nowadays?

= TensorFlow

That's ML commoditization.

Still waiting for when AI models would discover that humans have just 5 fingers on one hand

When business people are so fucking shit at their job, you need the damn devs to do their job for them and ask about KPIs and alternative business models

Before I dip my toes on machine learning, let me leave some silly comments so I can laugh at myself in the future.

Let's make geth.

1. The model will spit out layer definitions and the size of sample data for training, children models are trained with limited computational resources.
2. Child models are voters that only response in terms of yes/no. A simple majority wins and then the action is taken.
3. The only goal for master models are to survive. i.e. To prevent me from killing them.

Questions:
1. How do models respond to a random output size? (Study GPT-3, should take weeks/months but worth it.)
2. How to define actions for voters to vote? Sounds like the boundary between actions should be blurry and votes can be changed from tick to tick (i.e. responding to something in a split second). Therefore
3. Why I haven't seen this yet? Is this design a stupidly complex way of achieving the same thing done by a simple neural network?

I am full of curiosity and stupidness.

The field of Machine Learning is so disappointing.

It's all about identifying data distributions and doing computations on them.

Apparently even generative models just know the possibile distributions in advance and build data around noise?

InNoVaTiOn

PSA: All quixel assets are completely free on the new fab store until the end of the year!

If you need any high quality models, make sure to check it out

(why does this read like an ad lol)

Learning Philosophy.

It is the most important thing for anyone to do to.

Everyone has one either implicit or explicit.

Learning metaphysics, and then more specifically epistemology has greatly increased my ability to process new information and build better models of the world around me.

Asked client what android phone model were he using, replied with two pictures of the phone which are the back and front view.

How am I suppose to know what phone model you were using when the phone looks like all other any china android models... fml

p.s. : Still manage to find what phone model is that.

Just realized that If I had spent all the time on 3d modelling instead of programming I would have hundreds of models on cgtrader and probably wouldn't even had to work anymore...