**The lightweight PyTorch wrapper for high-performance AI research. Scale your models, not the boilerplate.** ---

WebsiteKey FeaturesHow To UseDocsExamplesCommunityGrid AILicence

[![PyPI - Python Version](https://img.shields.io/pypi/pyversions/pytorch-lightning)](https://pypi.org/project/pytorch-lightning/) [![PyPI Status](https://badge.fury.io/py/pytorch-lightning.svg)](https://badge.fury.io/py/pytorch-lightning) [![PyPI Status](https://pepy.tech/badge/pytorch-lightning)](https://pepy.tech/project/pytorch-lightning) [![Conda](https://img.shields.io/conda/v/conda-forge/pytorch-lightning?label=conda&color=success)](https://anaconda.org/conda-forge/pytorch-lightning) [![DockerHub](https://img.shields.io/docker/pulls/pytorchlightning/pytorch_lightning.svg)](https://hub.docker.com/r/pytorchlightning/pytorch_lightning) [![codecov](https://codecov.io/gh/PyTorchLightning/pytorch-lightning/branch/master/graph/badge.svg)](https://codecov.io/gh/PyTorchLightning/pytorch-lightning) [![ReadTheDocs](https://readthedocs.org/projects/pytorch-lightning/badge/?version=stable)](https://pytorch-lightning.readthedocs.io/en/stable/) [![Slack](https://img.shields.io/badge/slack-chat-green.svg?logo=slack)](https://join.slack.com/t/pytorch-lightning/shared_invite/zt-f6bl2l0l-JYMK3tbAgAmGRrlNr00f1A) [![Discourse status](https://img.shields.io/discourse/status?server=https%3A%2F%2Fforums.pytorchlightning.ai)](https://forums.pytorchlightning.ai/) [![license](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://github.com/PytorchLightning/pytorch-lightning/blob/master/LICENSE)
###### *Codecov is > 90%+ but build delays may show less --- ## PyTorch Lightning is just organized PyTorch Lightning disentangles PyTorch code to decouple the science from the engineering. ![PT to PL](docs/source/_static/images/general/pl_quick_start_full_compressed.gif) --- ## Lightning Design Philosophy Lightning structures PyTorch code with these principles:
Lightning forces the following structure to your code which makes it reusable and shareable: - Research code (the LightningModule). - Engineering code (you delete, and is handled by the Trainer). - Non-essential research code (logging, etc... this goes in Callbacks). - Data (use PyTorch Dataloaders or organize them into a LightningDataModule). Once you do this, you can train on multiple-GPUs, TPUs, CPUs and even in 16-bit precision without changing your code! Get started with our [2 step guide](https://pytorch-lightning.readthedocs.io/en/stable/new-project.html) --- ## Continuous Integration Lightning is rigurously tested across multiple GPUs, TPUs CPUs and against major Python and PyTorch versions.
Current build statuses
| System / PyTorch ver. | 1.4 (min. req.)* | 1.5 | 1.6 | 1.7 (latest) | 1.8 (nightly) | | :---: | :---: | :---: | :---: | :---: | :---: | | Conda py3.7 [linux] | [![PyTorch & Conda](https://github.com/PyTorchLightning/pytorch-lightning/workflows/PyTorch%20&%20Conda/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22PyTorch+%26+Conda%22+branch%3Amaster) | [![PyTorch & Conda](https://github.com/PyTorchLightning/pytorch-lightning/workflows/PyTorch%20&%20Conda/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22PyTorch+%26+Conda%22+branch%3Amaster) | [![PyTorch & Conda](https://github.com/PyTorchLightning/pytorch-lightning/workflows/PyTorch%20&%20Conda/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22PyTorch+%26+Conda%22+branch%3Amaster) | [![PyTorch & Conda](https://github.com/PyTorchLightning/pytorch-lightning/workflows/PyTorch%20&%20Conda/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22PyTorch+%26+Conda%22+branch%3Amaster) | [![PyTorch & Conda](https://github.com/PyTorchLightning/pytorch-lightning/workflows/PyTorch%20&%20Conda/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22PyTorch+%26+Conda%22+branch%3Amaster) | | Linux py3.7 [GPUs**] | - | - | [![Build Status](https://dev.azure.com/PytorchLightning/pytorch-lightning/_apis/build/status/PyTorchLightning.pytorch-lightning?branchName=master)](https://dev.azure.com/PytorchLightning/pytorch-lightning/_build/latest?definitionId=2&branchName=master) | - | - | | Linux py3.{6,7} [TPUs***] | - | - | [![TPU tests](https://github.com/PyTorchLightning/pytorch-lightning/workflows/TPU%20tests/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22TPU+tests%22+branch%3Amaster) | [![TPU tests](https://github.com/PyTorchLightning/pytorch-lightning/workflows/TPU%20tests/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22TPU+tests%22+branch%3Amaster) | | Linux py3.{6,7} | [![CI complete testing](https://github.com/PyTorchLightning/pytorch-lightning/workflows/CI%20complete%20testing/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22CI+testing%22) | - | - | [![CI complete testing](https://github.com/PyTorchLightning/pytorch-lightning/workflows/CI%20complete%20testing/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22CI+testing%22) | - | | OSX py3.{6,7,8} | - | [![CI complete testing](https://github.com/PyTorchLightning/pytorch-lightning/workflows/CI%20complete%20testing/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22CI+testing%22) | - | [![CI complete testing](https://github.com/PyTorchLightning/pytorch-lightning/workflows/CI%20complete%20testing/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22CI+testing%22) | - | | Windows py3.{6,7,8} | [![CI complete testing](https://github.com/PyTorchLightning/pytorch-lightning/workflows/CI%20complete%20testing/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22CI+testing%22) | - | - | [![CI complete testing](https://github.com/PyTorchLightning/pytorch-lightning/workflows/CI%20complete%20testing/badge.svg?branch=master&event=push)](https://github.com/PyTorchLightning/pytorch-lightning/actions?query=workflow%3A%22CI+testing%22) | - | - _\** tests run on two NVIDIA K80_ - _\*** tests run on Google GKE TPUv2/3_ - _TPU w/ py3.6/py3.7 means we support Colab and Kaggle env._
Bleeding edge build status (1.2)
![CI base testing](https://github.com/PyTorchLightning/pytorch-lightning/workflows/CI%20base%20testing/badge.svg?branch=release%2F1.2-dev&event=push) ![CI complete testing](https://github.com/PyTorchLightning/pytorch-lightning/workflows/CI%20complete%20testing/badge.svg?branch=release%2F1.2-dev&event=push) ![PyTorch & Conda](https://github.com/PyTorchLightning/pytorch-lightning/workflows/PyTorch%20&%20Conda/badge.svg?branch=release%2F1.2-dev&event=push) ![TPU tests](https://github.com/PyTorchLightning/pytorch-lightning/workflows/TPU%20tests/badge.svg?branch=release%2F1.2-dev&event=push) ![Docs check](https://github.com/PyTorchLightning/pytorch-lightning/workflows/Docs%20check/badge.svg?branch=release%2F1.2-dev&event=push)
--- ## How To Use ### Step 0: Install Simple installation from PyPI ```bash pip install pytorch-lightning ```
Other installation options #### Install with optional dependencies (CPU) ```bash pip install pytorch-lightning['cpu-extra'] ``` #### Install with optional dependencies (GPU, TPU) ```bash pip install pytorch-lightning['extra'] ``` #### Conda ```bash conda install pytorch-lightning -c conda-forge ``` #### Install bleeding-edge - future 1.2 Install future release from the source (no guarantees) ```bash pip install git+https://github.com/PytorchLightning/pytorch-lightning.git@release/1.2-dev --upgrade ``` or nightly from testing PyPI ```bash pip install -iU https://test.pypi.org/simple/ pytorch-lightning ```
### Step 1: Add these imports ```python import os import torch from torch import nn import torch.nn.functional as F from torchvision.datasets import MNIST from torch.utils.data import DataLoader, random_split from torchvision import transforms import pytorch_lightning as pl ``` ### Step 2: Define a LightningModule (nn.Module subclass) A LightningModule defines a full *system* (ie: a GAN, autoencoder, BERT or a simple Image Classifier). ```python class LitAutoEncoder(pl.LightningModule): def __init__(self): super().__init__() self.encoder = nn.Sequential(nn.Linear(28 * 28, 128), nn.ReLU(), nn.Linear(128, 3)) self.decoder = nn.Sequential(nn.Linear(3, 128), nn.ReLU(), nn.Linear(128, 28 * 28)) def forward(self, x): # in lightning, forward defines the prediction/inference actions embedding = self.encoder(x) return embedding def training_step(self, batch, batch_idx): # training_step defined the train loop. It is independent of forward x, y = batch x = x.view(x.size(0), -1) z = self.encoder(x) x_hat = self.decoder(z) loss = F.mse_loss(x_hat, x) self.log('train_loss', loss) return loss def configure_optimizers(self): optimizer = torch.optim.Adam(self.parameters(), lr=1e-3) return optimizer ``` **Note: Training_step defines the training loop. Forward defines how the LightningModule behaves during inference/prediction.** ### Step 3: Train! ```python dataset = MNIST(os.getcwd(), download=True, transform=transforms.ToTensor()) train, val = random_split(dataset, [55000, 5000]) autoencoder = LitAutoEncoder() trainer = pl.Trainer() trainer.fit(autoencoder, DataLoader(train), DataLoader(val)) ``` ## Advanced features Lightning has over [40+ advanced features](https://pytorch-lightning.readthedocs.io/en/stable/trainer.html#trainer-flags) designed for professional AI research at scale. Here are some examples:
Highlighted feature code snippets ```python # 8 GPUs # no code changes needed trainer = Trainer(max_epochs=1, gpus=8) # 256 GPUs trainer = Trainer(max_epochs=1, gpus=8, num_nodes=32) ``` Train on TPUs without code changes ```python # no code changes needed trainer = Trainer(tpu_cores=8) ``` 16-bit precision ```python # no code changes needed trainer = Trainer(precision=16) ``` Experiment managers ```python from pytorch_lightning import loggers # tensorboard trainer = Trainer(logger=TensorBoardLogger('logs/')) # weights and biases trainer = Trainer(logger=loggers.WandbLogger()) # comet trainer = Trainer(logger=loggers.CometLogger()) # mlflow trainer = Trainer(logger=loggers.MLFlowLogger()) # neptune trainer = Trainer(logger=loggers.NeptuneLogger()) # ... and dozens more ``` EarlyStopping ```python es = EarlyStopping(monitor='val_loss') trainer = Trainer(callbacks=[es]) ``` Checkpointing ```python checkpointing = ModelCheckpoint(monitor='val_loss') trainer = Trainer(callbacks=[checkpointing]) ``` Export to torchscript (JIT) (production use) ```python # torchscript autoencoder = LitAutoEncoder() torch.jit.save(autoencoder.to_torchscript(), "model.pt") ``` Export to ONNX (production use) ```python # onnx with tempfile.NamedTemporaryFile(suffix='.onnx', delete=False) as tmpfile: autoencoder = LitAutoEncoder() input_sample = torch.randn((1, 64)) autoencoder.to_onnx(tmpfile.name, input_sample, export_params=True) os.path.isfile(tmpfile.name) ```
### Pro-level control of training loops (advanced users) For complex/professional level work, you have optional full control of the training loop and optimizers. ```python class LitAutoEncoder(pl.LightningModule): def training_step(self, batch, batch_idx, optimizer_idx): # access your optimizers with use_pl_optimizer=False. Default is True (opt_a, opt_b) = self.optimizers(use_pl_optimizer=True) loss_a = ... self.manual_backward(loss_a, opt_a) opt_a.step() opt_a.zero_grad() loss_b = ... self.manual_backward(loss_b, opt_b, retain_graph=True) self.manual_backward(loss_b, opt_b) opt_b.step() opt_b.zero_grad() ``` --- ## Advantages over unstructured PyTorch * Models become hardware agnostic * Code is clear to read because engineering code is abstracted away * Easier to reproduce * Make fewer mistakes because lightning handles the tricky engineering * Keeps all the flexibility (LightningModules are still PyTorch modules), but removes a ton of boilerplate * Lightning has dozens of integrations with popular machine learning tools. * [Tested rigorously with every new PR](https://github.com/PyTorchLightning/pytorch-lightning/tree/master/tests). We test every combination of PyTorch and Python supported versions, every OS, multi GPUs and even TPUs. * Minimal running speed overhead (about 300 ms per epoch compared with pure PyTorch). --- ## Examples ###### Hello world - [MNIST hello world](https://colab.research.google.com/github/PytorchLightning/pytorch-lightning/blob/master/notebooks/01-mnist-hello-world.ipynb) - [MNIST on TPUs](https://colab.research.google.com/github/PytorchLightning/pytorch-lightning/blob/master/notebooks/06-mnist-tpu-training.ipynb) ###### Contrastive Learning - [BYOL](https://pytorch-lightning-bolts.readthedocs.io/en/latest/self_supervised_models.html#byol) - [CPC v2](https://pytorch-lightning-bolts.readthedocs.io/en/latest/self_supervised_models.html#cpc-v2) - [Moco v2](https://pytorch-lightning-bolts.readthedocs.io/en/latest/self_supervised_models.html#moco-v2) - [SIMCLR](https://pytorch-lightning-bolts.readthedocs.io/en/latest/self_supervised_models.html#simclr) ###### NLP - [BERT](https://colab.research.google.com/github/PytorchLightning/pytorch-lightning/blob/master/notebooks/04-transformers-text-classification.ipynb) - [GPT-2](https://pytorch-lightning-bolts.readthedocs.io/en/latest/convolutional.html#gpt-2) ###### Reinforcement Learning - [DQN](https://pytorch-lightning-bolts.readthedocs.io/en/latest/reinforce_learn.html#dqn-models) - [Dueling-DQN](https://pytorch-lightning-bolts.readthedocs.io/en/latest/reinforce_learn.html#dueling-dqn) - [Reinforce](https://pytorch-lightning-bolts.readthedocs.io/en/latest/reinforce_learn.html#reinforce) ###### Vision - [GAN](https://colab.research.google.com/github/PytorchLightning/pytorch-lightning/blob/master/notebooks/03-basic-gan.ipynb) ###### Classic ML - [Logistic Regression](https://pytorch-lightning-bolts.readthedocs.io/en/latest/classic_ml.html#logistic-regression) - [Linear Regression](https://pytorch-lightning-bolts.readthedocs.io/en/latest/classic_ml.html#linear-regression) --- ## Community The lightning community is maintained by - [16 core contributors](https://pytorch-lightning.readthedocs.io/en/latest/governance.html) who are all a mix of professional engineers, Research Scientists, Ph.D. students from top AI labs. - 280+ community contributors. Lightning is also part of the [PyTorch ecosystem](https://pytorch.org/ecosystem/) which requires projects to have solid testing, documentation and support. ### Asking for help If you have any questions please: 1. [Read the docs](https://pytorch-lightning.rtfd.io/en/latest). 2. [Search through the Discussions](https://github.com/PyTorchLightning/pytorch-lightning/discussions). 3. [Look it up in our forum (or add a new question)](https://forums.pytorchlightning.ai) 4. [Join our slack](https://join.slack.com/t/pytorch-lightning/shared_invite/zt-f6bl2l0l-JYMK3tbAgAmGRrlNr00f1A). ### Funding [We're venture funded](https://techcrunch.com/2020/10/08/grid-ai-raises-18-6m-series-a-to-help-ai-researchers-and-engineers-bring-their-models-to-production/) to make sure we can provide around the clock support, hire a full-time staff, attend conferences, and move faster through implementing features you request. --- ## Grid AI Grid AI is our native platform for training models at scale on the cloud! **Sign up for [early access here](https://www.grid.ai/)** To use grid, take your regular command: ``` python my_model.py --learning_rate 1e-6 --layers 2 --gpus 4 ``` And change it to use the grid train command: ``` grid train --grid_gpus 4 my_model.py --learning_rate 'uniform(1e-6, 1e-1, 20)' --layers '[2, 4, 8, 16]' ``` The above command will launch (20 * 4) experiments each running on 4 GPUs (320 GPUs!) - by making ZERO changes to your code. --- ## Licence Please observe the Apache 2.0 license that is listed in this repository. In addition the Lightning framework is Patent Pending. ## BibTeX If you want to cite the framework feel free to use this (but only if you loved it 😊): ```bibtex @article{falcon2019pytorch, title={PyTorch Lightning}, author={Falcon, WA}, journal={GitHub. Note: https://github.com/PyTorchLightning/pytorch-lightning}, volume={3}, year={2019} } ```