What papers or textbooks do i need to read to have all the basics / background knowledge to use pytorch and understand what I am doing based on solely the documentation pytorch provides?
The easiest way to get the basics is to search for articles, online courses, and youtube videos about the specific modules you’re interested in. Papers are written for people who are already deep in the field. You’ll get there, but they’re not the most efficient way to get up to speed. I have no experience with textbooks.
It helps to think of PyTorch as just a fancy math library. It has some well-documented frameworky structure (nn.Module) and a few differentiation engines, but all the deep learning-specific classes/functions (Conv2d, BatchNorm1d, ReLU, etc.) are just optimized math under the hood.
You can see the math by looking for projects that reimplement everything in numpy, e.g. picoGPT or ConvNet in NumPy.
If you can’t get your head around the tensor operations, I suggest searching for “explainers”. Basically for every impactful module there will be a bunch of " Explained" articles or videos out there. There are also ones for entire models, e.g. The Illustrated Transformer. Once you start googling specific modules’ explainers, you’ll find people who have made mountains of them.
If you’re not getting an explanation of something, just google and find another one. People have done an incredible job of making this information freely accessible in many different formats. I basically learned my way from webdev to an AI career with a couple years of casually watching YouTube videos.
TropicalDingdong@lemmy.world 6 months ago
I mean what are you trying to do?
It’s about the same as tensorfkiw in that it’s a largely abstracted language.
Its lego blocks with ML components. I think the dataset builder/ pipelines are better than tf. I think tf is a bit nicer if you need to do some quick and dirty assements of model performance.
I would say just hop into hugging face and pull down some models/ repos and start playing around.
More broadly it’s more important to understand bug picture concepts in ai ml rather than some language in particular, especially around frameworks and data.
AnarchistsForDemocracy@lemmy.world 6 months ago
I am trying to understand how an example i just coded using pytorch actually works, how the convuluted network works, meaning what are the arguments of conv2 or what it’s called and what is relu. I am digging through the documentation but I am missing a lot of basics as it seems.
My best bet was to read papers, but since this is already a couple years into the whole deep learning thing it is quite a challenge to idenity the foundational papers among the many that just repeat them.
TropicalDingdong@lemmy.world 6 months ago
I think if you haven’t found them yet, some three blue one brown videos might be helpful.
Like it’s important to get an understanding for how the big picture stuff works conceptually, but realistically, you will probably just be making minor modifications to existing frameworks. The framworks, have really ended up being almost more important in these most recent vintages of models, where the previous generations of models were very much architecture solutions.
So in that regard, it’s more important to focus on understanding the frame works around self learning, attention, generative and discriminative approaches etc…
After that, maybe you could answer a question for me.
What is it you want to do? Do you want to build models? Do you want to develop frameworks? Do you want to work on algorithms?
Because each of these really requires it’s own skillset, and while they have some overlap, most people don’t do everything.
howrar@lemmy.ca 6 months ago
I know you said you couldn’t find what you were looking for in the docs, but just in case you were looking in the wrong place:
Besides the convolution operator, I believe all the math should have been covered in high school (summation, max, and basic arithmetics). And convolution is also just defined in terms of these same operations, so you should be able to understand the definition (See the discrete definition in the wiki page under the “cross corrosion of deterministic signals” section).
The math does look daunting if it’s your first time encountering them (I’ve been there), and sometimes all you really need to confirmation that you already have all the requisite knowledge.
jacksilver@lemmy.world 6 months ago
I wouldn’t focus on foundational papers, the current phase of deep learning is far enough along that there are better tutorials/resources that better distill how these models work.
I would actually recommend you look into books on deep learning or something like a udemy course (Harvard or Stanford may also have free courses online, but I’ve never been a fan of their pacing) . I can send you some recommendations if you want, but that’s probably the best/fastest way.