From a372dfcf5fde56c5efbe9681515490b184975375 Mon Sep 17 00:00:00 2001 From: RenChu Wang Date: Tue, 30 Nov 2021 02:47:44 +0800 Subject: [PATCH] =?UTF-8?q?=E2=9C=8F=EF=B8=8F=20Have=20written=20a=20littl?= =?UTF-8?q?e=20bit=20more=20on=20README.?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Writing why I've made this project instead of using pytorch-lightning. --- README.md | 32 +++++++++++++++++++------------- 1 file changed, 19 insertions(+), 13 deletions(-) diff --git a/README.md b/README.md index 1d135ee..8e671ac 100644 --- a/README.md +++ b/README.md @@ -24,19 +24,19 @@ ## 🤔 Why Koila? Ever encountered `RuntimeError: CUDA error: out of memory`? -We all love PyTorch because of its speed, efficiency, and transparency, but that means it doesn't do extra things. Things like preventing a very common error that has been bothering every user since [2017](https://github.com/pytorch/pytorch/issues/958#issuecomment-285090162). +We all love `PyTorch` because of its speed, efficiency, and transparency, but that means it doesn't do extra things. Things like preventing a very common error that has been bothering many users since [2017](https://github.com/pytorch/pytorch/issues/958#issuecomment-285090162). -This library aims to prevent that by being a light-weight wrapper over native PyTorch. When a tensor is wrapped, the library **automatically computes the amount of remaining GPU memory and uses the right batch size**, saving everyone from having to manually finetune the batch size whenever a model is used. +This library aims to prevent that by being a light-weight wrapper over native `PyTorch`. When a tensor is wrapped, the library **automatically computes the amount of remaining GPU memory and uses the right batch size**, saving everyone from having to manually finetune the batch size whenever a model is used. Also, the library automatically uses the right batch size to GPU. Did you know that using bigger batches doesn't always speed up processing? It's handled automatically in this library too. -Because `Koila` code is PyTorch code, as it runs PyTorch under the hood, you can use both together without worrying compatibility. +Because `Koila` code is `PyTorch` code, as it runs `PyTorch` under the hood, you can use both together without worrying compatibility. Oh, and all that in 1 line of code! 😊 ## ⬇️ Installation -Koila is available on [PyPI](https://pypi.org/project/koila/). To install, run the following command. +`Koila` is available on [PyPI](https://pypi.org/project/koila/). To install, run the following command. ```bash pip install koila @@ -44,7 +44,7 @@ pip install koila ## 🏃 Getting started -The usage is dead simple. For example, you have the following PyTorch code (copied from PyTorch's tutorial) +The usage is dead simple. For example, you have the following `PyTorch` code (copied from `PyTorch`'s [tutorial](https://pytorch.org/tutorials/beginner/basics/buildmodel_tutorial.html)) Define the input, label, and model: @@ -71,8 +71,6 @@ class NeuralNetwork(Module): x = self.flatten(x) logits = self.linear_relu_stack(x) return logits - -nn = NeuralNetwork().to(DEVICE) ``` Define the loss function, calculate output and losses. @@ -108,24 +106,32 @@ See `examples/getting-started.py` for the full example. `CUDA error: out of memory` generally happens in forward pass, because temporary variables will need to be saved in memory. -Koila is a thin wrapper around PyTorch. It is inspired by TensorFlow's static/lazy evaluation. By building the graph first, and run the model only when necessarily, the model has access to all the information necessarily to determine how much resources is really need to compute the model. +`Koila` is a thin wrapper around `PyTorch`. It is inspired by TensorFlow's static/lazy evaluation. By building the graph first, and run the model only when necessarily, the model has access to all the information necessarily to determine how much resources is really need to compute the model. In terms of memory usage, only **shapes of temporary variables are required to calculate the memory usage of those variables used in the model**. For example, `+` takes in two tensors with equal sizes, and outputs a tensor with a size equal to the input size, and `log` takes in one tensor, and outputs another tensor with the same shape. Broadcasting makes it a little more complicated than that, but the general ideas are the same. By tracking all these shapes, one could easily tell how much memory is used in a forward pass. And select the optimal batch size accordingly. ## 🐌 It sounds slow. Is it? -**NO**. Indeed, calculating shapes and computing the size and memory usage sound like a lot of work. However, keep in mind that even a gigantic model like GPT-3, which has 96 layers, has only a few hundred nodes in its computing graph. Because Koila's algorithms run in linear time, any modern computer will be able to handle a graph like this instantly. +**NO**. Indeed, calculating shapes and computing the size and memory usage sound like a lot of work. However, keep in mind that even a gigantic model like GPT-3, which has 96 layers, has only a few hundred nodes in its computing graph. Because `Koila`'s algorithms run in linear time, any modern computer will be able to handle a graph like this instantly. -Most of the computing is spent on computing individual tensors, and transferring tensors across devices. And bear in mind that those checks happen in vanilla PyTorch anyways. So no, not slow at all. +Most of the computing is spent on computing individual tensors, and transferring tensors across devices. And bear in mind that those checks happen in vanilla `PyTorch` anyways. So no, not slow at all. -## 🔊 How to pronounce Koila? +## 🔊 How to pronounce koila? -This project was originally named _koala_, the laziest species in the world, and this project is about lazy evaluation of tensors. However, as that name is taken on [PyPI](https://pypi.org/project/koala/), I had no choice but to use another name. Koila is a word made up by me, pronounced similarly to _voila_ (It's a French word), so sounds like koala. +This project was originally named _koala_, the laziest species in the world, and this project is about lazy evaluation of tensors. However, as that name is taken on [PyPI](https://pypi.org/project/koala/), I had no choice but to use another name. `Koila` is a word made up by me, pronounced similarly to _voila_ (It's a French word), so sounds like koala. ## ⭐ Give me a star! If you like what you see, please consider giving this a star (★)! +## 🏗️ Why did I build this? + +Batch size search is not new. In fact, the mighty popular [PyTorch Lightning](https://github.com/PyTorchLightning/pytorch-lightning) has it. So why did I go through the trouble and build this project? + +PyTorch Lightning's batch size search is deeply integrated in its own ecosystem. You have to use its `DataLoader`, subclass from their models, and train your models accordingly. While it works well with supervised learning tasks, it's really painful to use in a reinforcement learning task, where interacting with the environment is a must. + +In comparison, because `Koila` is a super lightweight PyTorch wrapper, it works when PyTorch works, thus providing maximum flexibility and minimal changes to existing code. + ## 📝 Todos - 🧩 Provide an extensible API to write custom functions for the users. @@ -134,7 +140,7 @@ If you like what you see, please consider giving this a star (★)! ## 🚧 Warning -The code works on many cases, but it's still a work in progress. This is not (yet) a fully pytorch compatible library due to limited time. +The code works on many cases, but it's still a work in progress. This is not (yet) a fully `PyTorch` compatible library due to limited time. ## 🥰 Contributing