Metadata-Version: 2.1
Name: catalyst
Version: 21.9rc1
Summary: Catalyst. Accelerated deep learning R&D with PyTorch.
Home-page: https://github.com/catalyst-team/catalyst
Author: Sergey Kolesnikov
Author-email: scitator@gmail.com
License: Apache License 2.0
Download-URL: https://github.com/catalyst-team/catalyst
Project-URL: Bug Tracker, https://github.com/catalyst-team/catalyst/issues
Project-URL: Documentation, https://catalyst-team.github.io/catalyst
Project-URL: Source Code, https://github.com/catalyst-team/catalyst
Keywords: Machine Learning,Distributed Computing,Deep Learning,Reinforcement Learning,Computer Vision,Natural Language Processing,Recommendation Systems,Information Retrieval,PyTorch
Platform: UNKNOWN
Classifier: Environment :: Console
Classifier: Natural Language :: English
Classifier: Development Status :: 4 - Beta
Classifier: Operating System :: OS Independent
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Science/Research
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Scientific/Engineering :: Image Recognition
Classifier: Topic :: Scientific/Engineering :: Information Analysis
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: Implementation :: CPython
Requires-Python: >=3.6.0
Description-Content-Type: text/markdown
Provides-Extra: albu
Provides-Extra: cv
Provides-Extra: deepspeed
Provides-Extra: dev
Provides-Extra: fairscale
Provides-Extra: hydra
Provides-Extra: ml
Provides-Extra: mlflow
Provides-Extra: neptune
Provides-Extra: nifti
Provides-Extra: onnx-gpu
Provides-Extra: onnx
Provides-Extra: optuna
Provides-Extra: wandb
Provides-Extra: all
License-File: LICENSE


<div align="center">

[![Catalyst logo](https://raw.githubusercontent.com/catalyst-team/catalyst-pics/master/pics/catalyst_logo.png)](https://github.com/catalyst-team/catalyst)

**Accelerated Deep Learning R&D**

[![CodeFactor](https://www.codefactor.io/repository/github/catalyst-team/catalyst/badge)](https://www.codefactor.io/repository/github/catalyst-team/catalyst)
[![Pipi version](https://img.shields.io/pypi/v/catalyst.svg)](https://pypi.org/project/catalyst/)
[![Docs](https://img.shields.io/badge/dynamic/json.svg?label=docs&url=https%3A%2F%2Fpypi.org%2Fpypi%2Fcatalyst%2Fjson&query=%24.info.version&colorB=brightgreen&prefix=v)](https://catalyst-team.github.io/catalyst/index.html)
[![PyPI Status](https://pepy.tech/badge/catalyst)](https://pepy.tech/project/catalyst)

[![Twitter](https://img.shields.io/badge/news-twitter-499feb)](https://twitter.com/CatalystTeam)
[![Telegram](https://img.shields.io/badge/channel-telegram-blue)](https://t.me/catalyst_team)
[![Slack](https://img.shields.io/badge/Catalyst-slack-success)](https://join.slack.com/t/catalyst-team-devs/shared_invite/zt-d9miirnn-z86oKDzFMKlMG4fgFdZafw)
[![Github contributors](https://img.shields.io/github/contributors/catalyst-team/catalyst.svg?logo=github&logoColor=white)](https://github.com/catalyst-team/catalyst/graphs/contributors)

![codestyle](https://github.com/catalyst-team/catalyst/workflows/codestyle/badge.svg?branch=master&event=push)
![docs](https://github.com/catalyst-team/catalyst/workflows/docs/badge.svg?branch=master&event=push)
![catalyst](https://github.com/catalyst-team/catalyst/workflows/catalyst/badge.svg?branch=master&event=push)
![integrations](https://github.com/catalyst-team/catalyst/workflows/integrations/badge.svg?branch=master&event=push)

[![python](https://img.shields.io/badge/python_3.6-passing-success)](https://github.com/catalyst-team/catalyst/workflows/catalyst/badge.svg?branch=master&event=push)
[![python](https://img.shields.io/badge/python_3.7-passing-success)](https://github.com/catalyst-team/catalyst/workflows/catalyst/badge.svg?branch=master&event=push)
[![python](https://img.shields.io/badge/python_3.8-passing-success)](https://github.com/catalyst-team/catalyst/workflows/catalyst/badge.svg?branch=master&event=push)

[![os](https://img.shields.io/badge/Linux-passing-success)](https://github.com/catalyst-team/catalyst/workflows/catalyst/badge.svg?branch=master&event=push)
[![os](https://img.shields.io/badge/OSX-passing-success)](https://github.com/catalyst-team/catalyst/workflows/catalyst/badge.svg?branch=master&event=push)
[![os](https://img.shields.io/badge/WSL-passing-success)](https://github.com/catalyst-team/catalyst/workflows/catalyst/badge.svg?branch=master&event=push)
</div>

Catalyst is a PyTorch framework for Deep Learning Research and Development.
It focuses on reproducibility, rapid experimentation, and codebase reuse
so you can create something new rather than write yet another train loop.
<br/> Break the cycle – use the Catalyst!

- [Project Manifest](https://github.com/catalyst-team/catalyst/blob/master/MANIFEST.md)
- [Framework architecture](https://miro.com/app/board/o9J_lxBO-2k=/)
- [Catalyst at AI Landscape](https://landscape.lfai.foundation/selected=catalyst)
- Part of the [PyTorch Ecosystem](https://pytorch.org/ecosystem/)

<details>
<summary>Catalyst at PyTorch Ecosystem Day</summary>
<p>

[![Catalyst poster](https://raw.githubusercontent.com/catalyst-team/catalyst-pics/master/pics/Catalyst-PTED21.png)](https://github.com/catalyst-team/catalyst)

</p>
</details>

----

## Getting started

```bash
pip install -U catalyst
```

```python
import os
from torch import nn, optim
from torch.utils.data import DataLoader
from catalyst import dl, utils
from catalyst.data import ToTensor
from catalyst.contrib.datasets import MNIST

model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=0.02)

train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
loaders = {
    "train": DataLoader(train_data, batch_size=32),
    "valid": DataLoader(valid_data, batch_size=32),
}

runner = dl.SupervisedRunner(
    input_key="features", output_key="logits", target_key="targets", loss_key="loss"
)

# model training
runner.train(
    model=model,
    criterion=criterion,
    optimizer=optimizer,
    loaders=loaders,
    num_epochs=1,
    callbacks=[
        dl.AccuracyCallback(input_key="logits", target_key="targets", topk_args=(1, 3, 5)),
        dl.PrecisionRecallF1SupportCallback(
            input_key="logits", target_key="targets", num_classes=10
        ),
    ],
    logdir="./logs",
    valid_loader="valid",
    valid_metric="loss",
    minimize_valid_metric=True,
    verbose=True,
    load_best_on_end=True,
)

# model evaluation
metrics = runner.evaluate_loader(
    loader=loaders["valid"],
    callbacks=[dl.AccuracyCallback(input_key="logits", target_key="targets", topk_args=(1, 3, 5))],
)
assert "accuracy" in metrics.keys()

# model inference
for prediction in runner.predict_loader(loader=loaders["valid"]):
    assert prediction["logits"].detach().cpu().numpy().shape[-1] == 10

features_batch = next(iter(loaders["valid"]))[0]
# model stochastic weight averaging
model.load_state_dict(utils.get_averaged_weights_by_path_mask(logdir="./logs", path_mask="*.pth"))
# model tracing
utils.trace_model(model=runner.model, batch=features_batch)
# model quantization
utils.quantize_model(model=runner.model)
# model pruning
utils.prune_model(model=runner.model, pruning_fn="l1_unstructured", amount=0.8)
# onnx export
utils.onnx_export(model=runner.model, batch=features_batch, file="./logs/mnist.onnx", verbose=True)
```

### Step-by-step Guide
1. Start with [Catalyst — A PyTorch Framework for Accelerated Deep Learning R&D](https://medium.com/pytorch/catalyst-a-pytorch-framework-for-accelerated-deep-learning-r-d-ad9621e4ca88?source=friends_link&sk=885b4409aecab505db0a63b06f19dcef) introduction.
1. Check the [minimal examples](#minimal-examples).
1. Try [notebook tutorials with Google Colab](#notebooks).
1. Read the [blog posts](#notable-blog-posts) with use-cases and guides.
1. Learn machine learning with our ["Deep Learning with Catalyst" course](https://catalyst-team.com/#course).
1. If you would like to contribute to the project, follow our [contribution guidelines](https://github.com/catalyst-team/catalyst/blob/master/CONTRIBUTING.md).
1. If you are motivated by Catalyst vision, you could [support our initiative](https://opencollective.com/catalyst) or [write us](#user-feedback) for collaboration.
1. **And finally, [join our slack](https://join.slack.com/t/catalyst-team-core/shared_invite/zt-d9miirnn-z86oKDzFMKlMG4fgFdZafw) if you want to chat with the team and contributors**.


## Table of Contents
- [Getting started](#getting-started)
  - [Step-by-step Guide](#step-by-step-guide)
- [Table of Contents](#table-of-contents)
- [Overview](#overview)
  - [Installation](#installation)
  - [Minimal Examples](#minimal-examples)
  - [Features](#features)
  - [Tests](#tests)
- [Catalyst](#catalyst)
  - [Documentation](#documentation)
  - [Notebooks](#notebooks)
  - [Notable Blog Posts](#notable-blog-posts)
  - [Talks](#talks)
  - [Projects](#projects)
    - [Examples, Notebooks, and Starter Kits](#examples-notebooks-and-starter-kits)
    - [Competitions](#competitions)
    - [Research Papers](#research-papers)
    - [Toolkits](#toolkits)
- [Community](#community)
  - [Contribution Guide](#contribution-guide)
  - [User Feedback](#user-feedback)
  - [Acknowledgments](#acknowledgments)
    - [Catalyst.Team](#catalystteam)
    - [Catalyst.Contributors](#catalystcontributors)
  - [Trusted by](#trusted-by)
  - [Citation](#citation)


## Overview
Catalyst helps you implement compact
but full-featured Deep Learning pipelines with just a few lines of code.
You get a training loop with metrics, early-stopping, model checkpointing,
and other features without the boilerplate.


### Installation

Generic installation:
```bash
pip install -U catalyst
```

<details>
<summary>Specialized versions, extra requirements might apply</summary>
<p>

```bash
pip install catalyst[ml]         # installs ML-based Catalyst
pip install catalyst[cv]         # installs CV-based Catalyst
# master version installation
pip install git+https://github.com/catalyst-team/catalyst@master --upgrade
# all extensions are listed here:
# https://github.com/catalyst-team/catalyst/blob/master/setup.py#L87#L99
```
</p>
</details>

Catalyst is compatible with: Python 3.6+. PyTorch 1.3+. <br/>
Tested on Ubuntu 16.04/18.04/20.04, macOS 10.15, Windows 10, and Windows Subsystem for Linux.


### Minimal Examples

<details>
<summary>CustomRunner – PyTorch for-loop decomposition</summary>
<p>

```python
import os
from torch import nn, optim
from torch.nn import functional as F
from torch.utils.data import DataLoader
from catalyst import dl, metrics
from catalyst.data import ToTensor
from catalyst.contrib.datasets import MNIST

model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
optimizer = optim.Adam(model.parameters(), lr=0.02)

train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
loaders = {
    "train": DataLoader(train_data, batch_size=32),
    "valid": DataLoader(valid_data, batch_size=32),
}

class CustomRunner(dl.Runner):
    def predict_batch(self, batch):
        # model inference step
        return self.model(batch[0].to(self.device))

    def on_loader_start(self, runner):
        super().on_loader_start(runner)
        self.meters = {
            key: metrics.AdditiveMetric(compute_on_call=False)
            for key in ["loss", "accuracy01", "accuracy03"]
        }

    def handle_batch(self, batch):
        # model train/valid step
        # unpack the batch
        x, y = batch
        # run model forward pass
        logits = self.model(x)
        # compute the loss
        loss = F.cross_entropy(logits, y)
        # compute the metrics
        accuracy01, accuracy03 = metrics.accuracy(logits, y, topk=(1, 3))
        # log metrics
        self.batch_metrics.update(
            {"loss": loss, "accuracy01": accuracy01, "accuracy03": accuracy03}
        )
        for key in ["loss", "accuracy01", "accuracy03"]:
            self.meters[key].update(self.batch_metrics[key].item(), self.batch_size)
        # run model backward pass
        if self.is_train_loader:
            loss.backward()
            self.optimizer.step()
            self.optimizer.zero_grad()

    def on_loader_end(self, runner):
        for key in ["loss", "accuracy01", "accuracy03"]:
            self.loader_metrics[key] = self.meters[key].compute()[0]
        super().on_loader_end(runner)

runner = CustomRunner()
# model training
runner.train(
    model=model,
    optimizer=optimizer,
    loaders=loaders,
    logdir="./logs",
    num_epochs=5,
    verbose=True,
    valid_loader="valid",
    valid_metric="loss",
    minimize_valid_metric=True,
)
# model inference
for logits in runner.predict_loader(loader=loaders["valid"]):
    assert logits.detach().cpu().numpy().shape[-1] == 10
```
</p>
</details>

<details>
<summary>ML - linear regression</summary>
<p>

```python
import torch
from torch.utils.data import DataLoader, TensorDataset
from catalyst import dl

# data
num_samples, num_features = int(1e4), int(1e1)
X, y = torch.rand(num_samples, num_features), torch.rand(num_samples)
dataset = TensorDataset(X, y)
loader = DataLoader(dataset, batch_size=32, num_workers=1)
loaders = {"train": loader, "valid": loader}

# model, criterion, optimizer, scheduler
model = torch.nn.Linear(num_features, 1)
criterion = torch.nn.MSELoss()
optimizer = torch.optim.Adam(model.parameters())
scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, [3, 6])

# model training
runner = dl.SupervisedRunner()
runner.train(
    model=model,
    criterion=criterion,
    optimizer=optimizer,
    scheduler=scheduler,
    loaders=loaders,
    logdir="./logdir",
    valid_loader="valid",
    valid_metric="loss",
    minimize_valid_metric=True,
    num_epochs=8,
    verbose=True,
)
```
</p>
</details>


<details>
<summary>ML - multiclass classification</summary>
<p>

```python
import torch
from torch.utils.data import DataLoader, TensorDataset
from catalyst import dl

# sample data
num_samples, num_features, num_classes = int(1e4), int(1e1), 4
X = torch.rand(num_samples, num_features)
y = (torch.rand(num_samples,) * num_classes).to(torch.int64)

# pytorch loaders
dataset = TensorDataset(X, y)
loader = DataLoader(dataset, batch_size=32, num_workers=1)
loaders = {"train": loader, "valid": loader}

# model, criterion, optimizer, scheduler
model = torch.nn.Linear(num_features, num_classes)
criterion = torch.nn.CrossEntropyLoss()
optimizer = torch.optim.Adam(model.parameters())
scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, [2])

# model training
runner = dl.SupervisedRunner(
    input_key="features", output_key="logits", target_key="targets", loss_key="loss"
)
runner.train(
    model=model,
    criterion=criterion,
    optimizer=optimizer,
    scheduler=scheduler,
    loaders=loaders,
    logdir="./logdir",
    num_epochs=3,
    valid_loader="valid",
    valid_metric="accuracy03",
    minimize_valid_metric=False,
    verbose=True,
    callbacks=[
        dl.AccuracyCallback(input_key="logits", target_key="targets", num_classes=num_classes),
        # uncomment for extra metrics:
        # dl.PrecisionRecallF1SupportCallback(
        #     input_key="logits", target_key="targets", num_classes=num_classes
        # ),
        # dl.AUCCallback(input_key="logits", target_key="targets"),
        # catalyst[ml] required ``pip install catalyst[ml]``
        # dl.ConfusionMatrixCallback(
        #     input_key="logits", target_key="targets", num_classes=num_classes
        # ),
    ],
)
```
</p>
</details>


<details>
<summary>ML - multilabel classification</summary>
<p>

```python
import torch
from torch.utils.data import DataLoader, TensorDataset
from catalyst import dl

# sample data
num_samples, num_features, num_classes = int(1e4), int(1e1), 4
X = torch.rand(num_samples, num_features)
y = (torch.rand(num_samples, num_classes) > 0.5).to(torch.float32)

# pytorch loaders
dataset = TensorDataset(X, y)
loader = DataLoader(dataset, batch_size=32, num_workers=1)
loaders = {"train": loader, "valid": loader}

# model, criterion, optimizer, scheduler
model = torch.nn.Linear(num_features, num_classes)
criterion = torch.nn.BCEWithLogitsLoss()
optimizer = torch.optim.Adam(model.parameters())
scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, [2])

# model training
runner = dl.SupervisedRunner(
    input_key="features", output_key="logits", target_key="targets", loss_key="loss"
)
runner.train(
    model=model,
    criterion=criterion,
    optimizer=optimizer,
    scheduler=scheduler,
    loaders=loaders,
    logdir="./logdir",
    num_epochs=3,
    valid_loader="valid",
    valid_metric="accuracy",
    minimize_valid_metric=False,
    verbose=True,
    callbacks=[
        dl.BatchTransformCallback(
            transform=torch.sigmoid,
            scope="on_batch_end",
            input_key="logits",
            output_key="scores"
        ),
        dl.AUCCallback(input_key="scores", target_key="targets"),
        # uncomment for extra metrics:
        # dl.MultilabelAccuracyCallback(input_key="scores", target_key="targets", threshold=0.5),
        # dl.MultilabelPrecisionRecallF1SupportCallback(
        #     input_key="scores", target_key="targets", threshold=0.5
        # ),
    ]
)
```
</p>
</details>


<details>
<summary>ML - multihead classification</summary>
<p>

```python
import torch
from torch import nn, optim
from torch.utils.data import DataLoader, TensorDataset
from catalyst import dl

# sample data
num_samples, num_features, num_classes1, num_classes2 = int(1e4), int(1e1), 4, 10
X = torch.rand(num_samples, num_features)
y1 = (torch.rand(num_samples,) * num_classes1).to(torch.int64)
y2 = (torch.rand(num_samples,) * num_classes2).to(torch.int64)

# pytorch loaders
dataset = TensorDataset(X, y1, y2)
loader = DataLoader(dataset, batch_size=32, num_workers=1)
loaders = {"train": loader, "valid": loader}

class CustomModule(nn.Module):
    def __init__(self, in_features: int, out_features1: int, out_features2: int):
        super().__init__()
        self.shared = nn.Linear(in_features, 128)
        self.head1 = nn.Linear(128, out_features1)
        self.head2 = nn.Linear(128, out_features2)

    def forward(self, x):
        x = self.shared(x)
        y1 = self.head1(x)
        y2 = self.head2(x)
        return y1, y2

# model, criterion, optimizer, scheduler
model = CustomModule(num_features, num_classes1, num_classes2)
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters())
scheduler = optim.lr_scheduler.MultiStepLR(optimizer, [2])

class CustomRunner(dl.Runner):
    def handle_batch(self, batch):
        x, y1, y2 = batch
        y1_hat, y2_hat = self.model(x)
        self.batch = {
            "features": x,
            "logits1": y1_hat,
            "logits2": y2_hat,
            "targets1": y1,
            "targets2": y2,
        }

# model training
runner = CustomRunner()
runner.train(
    model=model,
    criterion=criterion,
    optimizer=optimizer,
    scheduler=scheduler,
    loaders=loaders,
    num_epochs=3,
    verbose=True,
    callbacks=[
        dl.CriterionCallback(metric_key="loss1", input_key="logits1", target_key="targets1"),
        dl.CriterionCallback(metric_key="loss2", input_key="logits2", target_key="targets2"),
        dl.MetricAggregationCallback(metric_key="loss", metrics=["loss1", "loss2"], mode="mean"),
        dl.OptimizerCallback(metric_key="loss"),
        dl.SchedulerCallback(),
        dl.AccuracyCallback(
            input_key="logits1", target_key="targets1", num_classes=num_classes1, prefix="one_"
        ),
        dl.AccuracyCallback(
            input_key="logits2", target_key="targets2", num_classes=num_classes2, prefix="two_"
        ),
        # catalyst[ml] required ``pip install catalyst[ml]``
        # dl.ConfusionMatrixCallback(
        #     input_key="logits1", target_key="targets1", num_classes=num_classes1, prefix="one_cm"
        # ),
        # dl.ConfusionMatrixCallback(
        #     input_key="logits2", target_key="targets2", num_classes=num_classes2, prefix="two_cm"
        # ),
        dl.CheckpointCallback(
            logdir="./logs/one",
            loader_key="valid", metric_key="one_accuracy", minimize=False, save_n_best=1
        ),
        dl.CheckpointCallback(
            logdir="./logs/two",
            loader_key="valid", metric_key="two_accuracy03", minimize=False, save_n_best=3
        ),
    ],
    loggers={"console": dl.ConsoleLogger(), "tb": dl.TensorboardLogger("./logs/tb")},
)
```
</p>
</details>


<details>
<summary>ML – RecSys</summary>
<p>

```python
import torch
from torch.utils.data import DataLoader, TensorDataset
from catalyst import dl

# sample data
num_users, num_features, num_items = int(1e4), int(1e1), 10
X = torch.rand(num_users, num_features)
y = (torch.rand(num_users, num_items) > 0.5).to(torch.float32)

# pytorch loaders
dataset = TensorDataset(X, y)
loader = DataLoader(dataset, batch_size=32, num_workers=1)
loaders = {"train": loader, "valid": loader}

# model, criterion, optimizer, scheduler
model = torch.nn.Linear(num_features, num_items)
criterion = torch.nn.BCEWithLogitsLoss()
optimizer = torch.optim.Adam(model.parameters())
scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, [2])

# model training
runner = dl.SupervisedRunner(
    input_key="features", output_key="logits", target_key="targets", loss_key="loss"
)
runner.train(
    model=model,
    criterion=criterion,
    optimizer=optimizer,
    scheduler=scheduler,
    loaders=loaders,
    num_epochs=3,
    verbose=True,
    callbacks=[
        dl.BatchTransformCallback(
            transform=torch.sigmoid,
            scope="on_batch_end",
            input_key="logits",
            output_key="scores"
        ),
        dl.CriterionCallback(input_key="logits", target_key="targets", metric_key="loss"),
        # uncomment for extra metrics:
        # dl.AUCCallback(input_key="scores", target_key="targets"),
        # dl.HitrateCallback(input_key="scores", target_key="targets", topk_args=(1, 3, 5)),
        # dl.MRRCallback(input_key="scores", target_key="targets", topk_args=(1, 3, 5)),
        # dl.MAPCallback(input_key="scores", target_key="targets", topk_args=(1, 3, 5)),
        # dl.NDCGCallback(input_key="scores", target_key="targets", topk_args=(1, 3, 5)),
        dl.OptimizerCallback(metric_key="loss"),
        dl.SchedulerCallback(),
        dl.CheckpointCallback(
            logdir="./logs", loader_key="valid", metric_key="loss", minimize=True
        ),
    ]
)
```
</p>
</details>


<details>
<summary>CV - MNIST classification</summary>
<p>

```python
import os
from torch import nn, optim
from torch.utils.data import DataLoader
from catalyst import dl
from catalyst.data import ToTensor
from catalyst.contrib.datasets import MNIST

model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=0.02)

train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
loaders = {
    "train": DataLoader(train_data, batch_size=32),
    "valid": DataLoader(valid_data, batch_size=32),
}

runner = dl.SupervisedRunner()
# model training
runner.train(
    model=model,
    criterion=criterion,
    optimizer=optimizer,
    loaders=loaders,
    num_epochs=1,
    logdir="./logs",
    valid_loader="valid",
    valid_metric="loss",
    minimize_valid_metric=True,
    verbose=True,
# uncomment for extra metrics:
#     callbacks=[
#         dl.AccuracyCallback(input_key="logits", target_key="targets", num_classes=10),
#         dl.PrecisionRecallF1SupportCallback(
#             input_key="logits", target_key="targets", num_classes=10
#         ),
#         dl.AUCCallback(input_key="logits", target_key="targets"),
#         # catalyst[ml] required ``pip install catalyst[ml]``
#         dl.ConfusionMatrixCallback(
#             input_key="logits", target_key="targets", num_classes=num_classes
#         ),
#     ]
)
```
</p>
</details>


<details>
<summary>CV - MNIST segmentation</summary>
<p>

```python
import os
import torch
from torch import nn
from torch.utils.data import DataLoader
from catalyst import dl
from catalyst.data import ToTensor
from catalyst.contrib.datasets import MNIST
from catalyst.contrib.nn import IoULoss


model = nn.Sequential(
    nn.Conv2d(1, 1, 3, 1, 1), nn.ReLU(),
    nn.Conv2d(1, 1, 3, 1, 1), nn.Sigmoid(),
)
criterion = IoULoss()
optimizer = torch.optim.Adam(model.parameters(), lr=0.02)

train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
loaders = {
    "train": DataLoader(train_data, batch_size=32),
    "valid": DataLoader(valid_data, batch_size=32),
}

class CustomRunner(dl.SupervisedRunner):
    def handle_batch(self, batch):
        x = batch[self._input_key]
        x_noise = (x + torch.rand_like(x)).clamp_(0, 1)
        x_ = self.model(x_noise)
        self.batch = {self._input_key: x, self._output_key: x_, self._target_key: x}

runner = CustomRunner(
    input_key="features", output_key="scores", target_key="targets", loss_key="loss"
)
# model training
runner.train(
    model=model,
    criterion=criterion,
    optimizer=optimizer,
    loaders=loaders,
    num_epochs=1,
    callbacks=[
        dl.IOUCallback(input_key="scores", target_key="targets"),
        dl.DiceCallback(input_key="scores", target_key="targets"),
        dl.TrevskyCallback(input_key="scores", target_key="targets", alpha=0.2),
    ],
    logdir="./logdir",
    valid_loader="valid",
    valid_metric="loss",
    minimize_valid_metric=True,
    verbose=True,
)
```
</p>
</details>


<details>
<summary>CV - MNIST model distillation</summary>
<p>

```python
import os
import torch
from torch import nn, optim
from torch.nn import functional as F
from torch.utils.data import DataLoader
from catalyst import dl
from catalyst.data import ToTensor
from catalyst.contrib.datasets import MNIST

# [!] teacher model should be already pretrained
teacher = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
student = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
criterion = {"cls": nn.CrossEntropyLoss(), "kl": nn.KLDivLoss(reduction="batchmean")}
optimizer = optim.Adam(student.parameters(), lr=0.02)

train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
loaders = {
    "train": DataLoader(train_data, batch_size=32),
    "valid": DataLoader(valid_data, batch_size=32),
}

class DistilRunner(dl.Runner):
    def handle_batch(self, batch):
        x, y = batch

        self.model["teacher"].eval()  # let's manually set teacher model to eval mode
        with torch.no_grad():
            t_logits = self.model["teacher"](x)

        s_logits = self.model["student"](x)
        self.batch = {
            "t_logits": t_logits, "s_logits": s_logits, "targets": y,
            "s_logprobs": F.log_softmax(s_logits, dim=-1), "t_probs": F.softmax(t_logits, dim=-1)
        }

runner = DistilRunner()
callbacks = [
    dl.AccuracyCallback(
        input_key="t_logits", target_key="targets", num_classes=2, prefix="teacher_"
    ),
    dl.AccuracyCallback(
        input_key="s_logits", target_key="targets", num_classes=2, prefix="student_"
    ),
    dl.CriterionCallback(
        input_key="s_logits", target_key="targets", metric_key="cls_loss", criterion_key="cls"
    ),
    dl.CriterionCallback(
        input_key="s_logprobs", target_key="t_probs", metric_key="kl_div_loss", criterion_key="kl"
    ),
    dl.MetricAggregationCallback(metric_key="loss", metrics=["kl_div_loss", "cls_loss"], mode="mean"),
    dl.OptimizerCallback(metric_key="loss", model_key="student"),
    dl.CheckpointCallback(
        logdir="./logs", loader_key="valid", metric_key="loss", minimize=True, save_n_best=3
    ),
]
# model training
runner.train(
    model={"teacher": teacher, "student": student},
    criterion=criterion,
    optimizer=optimizer,
    loaders=loaders,
    num_epochs=1,
    logdir="./logs",
    verbose=True,
    callbacks=callbacks,
)
```
</p>
</details>


<details>
<summary>CV - MNIST metric learning</summary>
<p>

```python
import os
from torch.optim import Adam
from torch.utils.data import DataLoader
from catalyst import data, dl
from catalyst.contrib import datasets, models, nn
from catalyst.data.transforms import Compose, Normalize, ToTensor


# 1. train and valid loaders
transforms = Compose([ToTensor(), Normalize((0.1307,), (0.3081,))])

train_dataset = datasets.MnistMLDataset(root=os.getcwd(), download=True, transform=transforms)
sampler = data.BatchBalanceClassSampler(
    labels=train_dataset.get_labels(), num_classes=5, num_samples=10, num_batches=10
)
train_loader = DataLoader(dataset=train_dataset, batch_sampler=sampler)

valid_dataset = datasets.MnistQGDataset(root=os.getcwd(), transform=transforms, gallery_fraq=0.2)
valid_loader = DataLoader(dataset=valid_dataset, batch_size=1024)

# 2. model and optimizer
model = models.MnistSimpleNet(out_features=16)
optimizer = Adam(model.parameters(), lr=0.001)

# 3. criterion with triplets sampling
sampler_inbatch = data.HardTripletsSampler(norm_required=False)
criterion = nn.TripletMarginLossWithSampler(margin=0.5, sampler_inbatch=sampler_inbatch)

# 4. training with catalyst Runner
class CustomRunner(dl.SupervisedRunner):
    def handle_batch(self, batch) -> None:
        if self.is_train_loader:
            images, targets = batch["features"].float(), batch["targets"].long()
            features = self.model(images)
            self.batch = {"embeddings": features, "targets": targets,}
        else:
            images, targets, is_query = \
                batch["features"].float(), batch["targets"].long(), batch["is_query"].bool()
            features = self.model(images)
            self.batch = {"embeddings": features, "targets": targets, "is_query": is_query}

callbacks = [
    dl.ControlFlowCallback(
        dl.CriterionCallback(input_key="embeddings", target_key="targets", metric_key="loss"),
        loaders="train",
    ),
    dl.ControlFlowCallback(
        dl.CMCScoreCallback(
            embeddings_key="embeddings",
            labels_key="targets",
            is_query_key="is_query",
            topk_args=[1],
        ),
        loaders="valid",
    ),
    dl.PeriodicLoaderCallback(
        valid_loader_key="valid", valid_metric_key="cmc01", minimize=False, valid=2
    ),
]

runner = CustomRunner(input_key="features", output_key="embeddings")
runner.train(
    model=model,
    criterion=criterion,
    optimizer=optimizer,
    callbacks=callbacks,
    loaders={"train": train_loader, "valid": valid_loader},
    verbose=False,
    logdir="./logs",
    valid_loader="valid",
    valid_metric="cmc01",
    minimize_valid_metric=False,
    num_epochs=10,
)
```
</p>
</details>


<details>
<summary>CV - MNIST GAN</summary>
<p>

```python
import os
import torch
from torch import nn
from torch.utils.data import DataLoader
from catalyst import dl
from catalyst.contrib.datasets import MNIST
from catalyst.contrib.nn.modules import Flatten, GlobalMaxPool2d, Lambda
from catalyst.data import ToTensor

latent_dim = 128
generator = nn.Sequential(
    # We want to generate 128 coefficients to reshape into a 7x7x128 map
    nn.Linear(128, 128 * 7 * 7),
    nn.LeakyReLU(0.2, inplace=True),
    Lambda(lambda x: x.view(x.size(0), 128, 7, 7)),
    nn.ConvTranspose2d(128, 128, (4, 4), stride=(2, 2), padding=1),
    nn.LeakyReLU(0.2, inplace=True),
    nn.ConvTranspose2d(128, 128, (4, 4), stride=(2, 2), padding=1),
    nn.LeakyReLU(0.2, inplace=True),
    nn.Conv2d(128, 1, (7, 7), padding=3),
    nn.Sigmoid(),
)
discriminator = nn.Sequential(
    nn.Conv2d(1, 64, (3, 3), stride=(2, 2), padding=1),
    nn.LeakyReLU(0.2, inplace=True),
    nn.Conv2d(64, 128, (3, 3), stride=(2, 2), padding=1),
    nn.LeakyReLU(0.2, inplace=True),
    GlobalMaxPool2d(),
    Flatten(),
    nn.Linear(128, 1),
)

model = {"generator": generator, "discriminator": discriminator}
criterion = {"generator": nn.BCEWithLogitsLoss(), "discriminator": nn.BCEWithLogitsLoss()}
optimizer = {
    "generator": torch.optim.Adam(generator.parameters(), lr=0.0003, betas=(0.5, 0.999)),
    "discriminator": torch.optim.Adam(discriminator.parameters(), lr=0.0003, betas=(0.5, 0.999)),
}
train_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
loaders = {"train": DataLoader(train_data, batch_size=32)}

class CustomRunner(dl.Runner):
    def predict_batch(self, batch):
        batch_size = 1
        # Sample random points in the latent space
        random_latent_vectors = torch.randn(batch_size, latent_dim).to(self.device)
        # Decode them to fake images
        generated_images = self.model["generator"](random_latent_vectors).detach()
        return generated_images

    def handle_batch(self, batch):
        real_images, _ = batch
        batch_size = real_images.shape[0]

        # Sample random points in the latent space
        random_latent_vectors = torch.randn(batch_size, latent_dim).to(self.device)

        # Decode them to fake images
        generated_images = self.model["generator"](random_latent_vectors).detach()
        # Combine them with real images
        combined_images = torch.cat([generated_images, real_images])

        # Assemble labels discriminating real from fake images
        labels = \
            torch.cat([torch.ones((batch_size, 1)), torch.zeros((batch_size, 1))]).to(self.device)
        # Add random noise to the labels - important trick!
        labels += 0.05 * torch.rand(labels.shape).to(self.device)

        # Discriminator forward
        combined_predictions = self.model["discriminator"](combined_images)

        # Sample random points in the latent space
        random_latent_vectors = torch.randn(batch_size, latent_dim).to(self.device)
        # Assemble labels that say "all real images"
        misleading_labels = torch.zeros((batch_size, 1)).to(self.device)

        # Generator forward
        generated_images = self.model["generator"](random_latent_vectors)
        generated_predictions = self.model["discriminator"](generated_images)

        self.batch = {
            "combined_predictions": combined_predictions,
            "labels": labels,
            "generated_predictions": generated_predictions,
            "misleading_labels": misleading_labels,
        }


runner = CustomRunner()
runner.train(
    model=model,
    criterion=criterion,
    optimizer=optimizer,
    loaders=loaders,
    callbacks=[
        dl.CriterionCallback(
            input_key="combined_predictions",
            target_key="labels",
            metric_key="loss_discriminator",
            criterion_key="discriminator",
        ),
        dl.CriterionCallback(
            input_key="generated_predictions",
            target_key="misleading_labels",
            metric_key="loss_generator",
            criterion_key="generator",
        ),
        dl.OptimizerCallback(
            model_key="generator",
            optimizer_key="generator",
            metric_key="loss_generator"
        ),
        dl.OptimizerCallback(
            model_key="discriminator",
            optimizer_key="discriminator",
            metric_key="loss_discriminator"
        ),
    ],
    valid_loader="train",
    valid_metric="loss_generator",
    minimize_valid_metric=True,
    num_epochs=20,
    verbose=True,
    logdir="./logs_gan",
)

# visualization (matplotlib required):
# import matplotlib.pyplot as plt
# %matplotlib inline
# plt.imshow(runner.predict_batch(None)[0, 0].cpu().numpy())
```
</p>
</details>


<details>
<summary>CV - MNIST VAE</summary>
<p>

```python
import os
import torch
from torch import nn, optim
from torch.nn import functional as F
from torch.utils.data import DataLoader
from catalyst import dl, metrics
from catalyst.contrib.datasets import MNIST
from catalyst.data import ToTensor

LOG_SCALE_MAX = 2
LOG_SCALE_MIN = -10

def normal_sample(loc, log_scale):
    scale = torch.exp(0.5 * log_scale)
    return loc + scale * torch.randn_like(scale)

class VAE(nn.Module):
    def __init__(self, in_features, hid_features):
        super().__init__()
        self.hid_features = hid_features
        self.encoder = nn.Linear(in_features, hid_features * 2)
        self.decoder = nn.Sequential(nn.Linear(hid_features, in_features), nn.Sigmoid())

    def forward(self, x, deterministic=False):
        z = self.encoder(x)
        bs, z_dim = z.shape

        loc, log_scale = z[:, : z_dim // 2], z[:, z_dim // 2 :]
        log_scale = torch.clamp(log_scale, LOG_SCALE_MIN, LOG_SCALE_MAX)

        z_ = loc if deterministic else normal_sample(loc, log_scale)
        z_ = z_.view(bs, -1)
        x_ = self.decoder(z_)

        return x_, loc, log_scale

class CustomRunner(dl.IRunner):
    def __init__(self, logdir, device):
        super().__init__()
        self._logdir = logdir
        self._device = device

    def get_engine(self):
        return dl.DeviceEngine(self._device)

    def get_loggers(self):
        return {
            "console": dl.ConsoleLogger(),
            "csv": dl.CSVLogger(logdir=self._logdir),
            "tensorboard": dl.TensorboardLogger(logdir=self._logdir),
        }

    @property
    def stages(self):
        return ["train"]

    def get_stage_len(self, stage: str) -> int:
        return 3

    def get_loaders(self, stage: str):
        train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
        valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
        loaders = {
            "train": DataLoader(train_data, batch_size=32),
            "valid": DataLoader(valid_data, batch_size=32),
        }
        return loaders

    def get_model(self, stage: str):
        model = self.model if self.model is not None else VAE(28 * 28, 64)
        return model

    def get_optimizer(self, stage: str, model):
        return optim.Adam(model.parameters(), lr=0.02)

    def get_callbacks(self, stage: str):
        return {
            "optimizer": dl.OptimizerCallback(metric_key="loss"),
            "checkpoint": dl.CheckpointCallback(
                self._logdir, loader_key="valid", metric_key="loss", minimize=True
            ),
        }

    def on_loader_start(self, runner):
        super().on_loader_start(runner)
        self.meters = {
            key: metrics.AdditiveMetric(compute_on_call=False)
            for key in ["loss_ae", "loss_kld", "loss"]
        }

    def handle_batch(self, batch):
        x, _ = batch
        x = x.view(x.size(0), -1)
        x_, loc, log_scale = self.model(x, deterministic=not self.is_train_loader)

        loss_ae = F.mse_loss(x_, x)
        loss_kld = (-0.5 * torch.sum(1 + log_scale - loc.pow(2) - log_scale.exp(), dim=1)).mean()
        loss = loss_ae + loss_kld * 0.01

        self.batch_metrics = {"loss_ae": loss_ae, "loss_kld": loss_kld, "loss": loss}
        for key in ["loss_ae", "loss_kld", "loss"]:
            self.meters[key].update(self.batch_metrics[key].item(), self.batch_size)

    def on_loader_end(self, runner):
        for key in ["loss_ae", "loss_kld", "loss"]:
            self.loader_metrics[key] = self.meters[key].compute()[0]
        super().on_loader_end(runner)

    def predict_batch(self, batch):
        random_latent_vectors = torch.randn(1, self.model.hid_features).to(self.device)
        generated_images = self.model.decoder(random_latent_vectors).detach()
        return generated_images

runner = CustomRunner("./logs", "cpu")
runner.run()
# visualization (matplotlib required):
# import matplotlib.pyplot as plt
# %matplotlib inline
# plt.imshow(runner.predict_batch(None)[0].cpu().numpy().reshape(28, 28))
```
</p>
</details>


<details>
<summary>CV - MNIST multistage finetuning</summary>
<p>

```python
import os
from torch import nn, optim
from torch.utils.data import DataLoader
from catalyst import dl, utils
from catalyst.contrib.datasets import MNIST
from catalyst.data import ToTensor


class CustomRunner(dl.IRunner):
    def __init__(self, logdir, device):
        super().__init__()
        self._logdir = logdir
        self._device = device

    def get_engine(self):
        return dl.DeviceEngine(self._device)

    def get_loggers(self):
        return {
            "console": dl.ConsoleLogger(),
            "csv": dl.CSVLogger(logdir=self._logdir),
            "tensorboard": dl.TensorboardLogger(logdir=self._logdir),
        }

    @property
    def stages(self):
        return ["train_freezed", "train_unfreezed"]

    def get_stage_len(self, stage: str) -> int:
        return 3

    def get_loaders(self, stage: str):
        train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
        valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
        loaders = {
            "train": DataLoader(train_data, batch_size=32),
            "valid": DataLoader(valid_data, batch_size=32),
        }
        return loaders

    def get_model(self, stage: str):
        model = (
            self.model
            if self.model is not None
            else nn.Sequential(nn.Flatten(), nn.Linear(784, 128), nn.ReLU(), nn.Linear(128, 10))
        )
        if stage == "train_freezed":
            # freeze layer
            utils.set_requires_grad(model[1], False)
        else:
            utils.set_requires_grad(model, True)
        return model

    def get_criterion(self, stage: str):
        return nn.CrossEntropyLoss()

    def get_optimizer(self, stage: str, model):
        if stage == "train_freezed":
            return optim.Adam(model.parameters(), lr=1e-3)
        else:
            return optim.SGD(model.parameters(), lr=1e-1)

    def get_scheduler(self, stage: str, optimizer):
        return None

    def get_callbacks(self, stage: str):
        return {
            "criterion": dl.CriterionCallback(
                metric_key="loss", input_key="logits", target_key="targets"
            ),
            "optimizer": dl.OptimizerCallback(metric_key="loss"),
            # "scheduler": dl.SchedulerCallback(loader_key="valid", metric_key="loss"),
            # "accuracy": dl.AccuracyCallback(
            #     input_key="logits", target_key="targets", topk_args=(1, 3, 5)
            # ),
            # "classification": dl.PrecisionRecallF1SupportCallback(
            #     input_key="logits", target_key="targets", num_classes=10
            # ),
            # "confusion_matrix": dl.ConfusionMatrixCallback(
            #     input_key="logits", target_key="targets", num_classes=10
            # ),
            "checkpoint": dl.CheckpointCallback(
                self._logdir, loader_key="valid", metric_key="loss", minimize=True, save_n_best=3
            ),
        }

    def handle_batch(self, batch):
        x, y = batch
        logits = self.model(x)

        self.batch = {
            "features": x,
            "targets": y,
            "logits": logits,
        }

runner = CustomRunner("./logs", "cpu")
runner.run()
```
</p>
</details>


<details>
<summary>CV - MNIST multistage finetuning (distributed)</summary>
<p>

```python
import os
from torch import nn, optim
from torch.utils.data import DataLoader
from catalyst import dl, utils
from catalyst.contrib.datasets import MNIST
from catalyst.data import ToTensor


class CustomRunner(dl.IRunner):
    def __init__(self, logdir):
        super().__init__()
        self._logdir = logdir

    def get_engine(self):
        return dl.DistributedDataParallelEngine()

    def get_loggers(self):
        return {
            "console": dl.ConsoleLogger(),
            "csv": dl.CSVLogger(logdir=self._logdir),
            "tensorboard": dl.TensorboardLogger(logdir=self._logdir),
        }

    @property
    def stages(self):
        return ["train_freezed", "train_unfreezed"]

    def get_stage_len(self, stage: str) -> int:
        return 3

    def get_loaders(self, stage: str):
        train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
        valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
        loaders = {
            "train": DataLoader(train_data, batch_size=32),
            "valid": DataLoader(valid_data, batch_size=32),
        }
        return loaders

    def get_model(self, stage: str):
        model = nn.Sequential(nn.Flatten(), nn.Linear(784, 128), nn.ReLU(), nn.Linear(128, 10))
        if stage == "train_freezed":  # freeze layer
            utils.set_requires_grad(model[1], False)
        else:
            utils.set_requires_grad(model, True)
        return model

    def get_criterion(self, stage: str):
        return nn.CrossEntropyLoss()

    def get_optimizer(self, stage: str, model):
        if stage == "train_freezed":
            return optim.Adam(model.parameters(), lr=1e-3)
        else:
            return optim.SGD(model.parameters(), lr=1e-1)

    def get_callbacks(self, stage: str):
        return {
            "criterion": dl.CriterionCallback(
                metric_key="loss", input_key="logits", target_key="targets"
            ),
            "optimizer": dl.OptimizerCallback(metric_key="loss"),
            "accuracy": dl.AccuracyCallback(
                input_key="logits", target_key="targets", topk_args=(1, 3, 5)
            ),
            "classification": dl.PrecisionRecallF1SupportCallback(
                input_key="logits", target_key="targets", num_classes=10
            ),
            # catalyst[ml] required ``pip install catalyst[ml]``
            # "confusion_matrix": dl.ConfusionMatrixCallback(
            #     input_key="logits", target_key="targets", num_classes=10
            # ),
            "checkpoint": dl.CheckpointCallback(
                self._logdir,
                loader_key="valid",
                metric_key="loss",
                minimize=True,
                save_n_best=3,
                # here is the main trick:
                load_on_stage_start={
                    "model": "best",
                    "global_epoch_step": "last",
                    "global_batch_step": "last",
                    "global_sample_step": "last",
                },
            ),
            "verbose": dl.TqdmCallback(),
        }

    def handle_batch(self, batch):
        x, y = batch
        logits = self.model(x)

        self.batch = {
            "features": x,
            "targets": y,
            "logits": logits,
        }


if __name__ == "__main__":
    runner = CustomRunner("./logs")
    runner.run()
```
</p>
</details>


<details>
<summary>AutoML - hyperparameters optimization with Optuna</summary>
<p>

```python
import os
import optuna
import torch
from torch import nn
from torch.utils.data import DataLoader
from catalyst import dl
from catalyst.data import ToTensor
from catalyst.contrib.datasets import MNIST


def objective(trial):
    lr = trial.suggest_loguniform("lr", 1e-3, 1e-1)
    num_hidden = int(trial.suggest_loguniform("num_hidden", 32, 128))

    train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
    valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
    loaders = {
        "train": DataLoader(train_data, batch_size=32),
        "valid": DataLoader(valid_data, batch_size=32),
    }
    model = nn.Sequential(
        nn.Flatten(), nn.Linear(784, num_hidden), nn.ReLU(), nn.Linear(num_hidden, 10)
    )
    optimizer = torch.optim.Adam(model.parameters(), lr=lr)
    criterion = nn.CrossEntropyLoss()

    runner = dl.SupervisedRunner(input_key="features", output_key="logits", target_key="targets")
    runner.train(
        model=model,
        criterion=criterion,
        optimizer=optimizer,
        loaders=loaders,
        callbacks={
            "accuracy": dl.AccuracyCallback(
                input_key="logits", target_key="targets", num_classes=10
            ),
            # catalyst[optuna] required ``pip install catalyst[optuna]``
            "optuna": dl.OptunaPruningCallback(
                loader_key="valid", metric_key="accuracy01", minimize=False, trial=trial
            ),
        },
        num_epochs=3,
    )
    score = trial.best_score
    return score

study = optuna.create_study(
    direction="maximize",
    pruner=optuna.pruners.MedianPruner(
        n_startup_trials=1, n_warmup_steps=0, interval_steps=1
    ),
)
study.optimize(objective, n_trials=3, timeout=300)
print(study.best_value, study.best_params)
```
</p>
</details>


<details>
<summary>Config API - minimal example</summary>
<p>

```python
import torch
from torch.utils.data import TensorDataset
from catalyst import dl

NUM_SAMPLES, NUM_FEATURES, NUM_CLASSES = int(1e4), int(1e1), 4
LOGDIR = "./logs"

class CustomConfigRunner(dl.SupervisedConfigRunner):
    def get_datasets(self, stage: str):
        # sample data
        X = torch.rand(NUM_SAMPLES, NUM_FEATURES)
        y = (torch.rand(NUM_SAMPLES,) * NUM_CLASSES).to(torch.int64)

        # pytorch dataset
        dataset = TensorDataset(X, y)
        datasets = {"train": dataset, "valid": dataset}
        return datasets


runner = CustomConfigRunner(
    input_key="features",
    output_key="logits",
    target_key="targets",
    loss_key="loss",
    config={
        "args": {
            "logdir": LOGDIR,
            "valid_loader": "valid",
            "valid_metric": "accuracy01",
            "minimize_valid_metric": False,
            "verbose": False,
        },
        "model": {
            "_target_": "Linear",
            "in_features": NUM_FEATURES,
            "out_features": NUM_CLASSES,
        },
        "engine": {"_target_": "DeviceEngine"},
        "loggers": {
            "console": {"_target_": "ConsoleLogger"},
            "csv": {"_target_": "CSVLogger", "logdir": LOGDIR},
            "tensorboard": {"_target_": "TensorboardLogger", "logdir": LOGDIR},
        },
        "stages": {
            "stage1": {
                "num_epochs": 10,
                "criterion": {"_target_": "CrossEntropyLoss"},
                "optimizer": {"_target_": "Adam", "lr": 1e-3},
                "scheduler": {"_target_": "MultiStepLR", "milestones": [2]},
                "loaders": {"batch_size": 32, "num_workers": 1},
                "callbacks": {
                    "accuracy": {
                        "_target_": "AccuracyCallback",
                        "input_key": "logits",
                        "target_key": "targets",
                        "num_classes": NUM_CLASSES,
                    },
                    "classification": {
                        "_target_": "PrecisionRecallF1SupportCallback",
                        "input_key": "logits",
                        "target_key": "targets",
                        "num_classes": NUM_CLASSES,
                    },
                    "criterion": {
                        "_target_": "CriterionCallback",
                        "input_key": "logits",
                        "target_key": "targets",
                        "metric_key": "loss",
                    },
                    "optimizer": {"_target_": "OptimizerCallback", "metric_key": "loss"},
                    "scheduler": {"_target_": "SchedulerCallback"},
                    "checkpointer": {
                        "_target_": "CheckpointCallback",
                        "logdir": LOGDIR,
                        "loader_key": "valid",
                        "metric_key": "accuracy01",
                        "minimize": False,
                        "save_n_best": 3,
                    },
                },
            },
        },
    },
)
runner.run()
```
</p>
</details>

- [More Catalyst examples](./examples/).
- [Config/Hydra API – full example](https://github.com/catalyst-team/catalyst/tree/master/examples/mnist_stages).


### Features
- Universal train/inference loop.
- Configuration files for model and data hyperparameters.
- Reproducibility – all source code and environment variables are saved.
- Callbacks – reusable train/inference pipeline parts with easy customization.
- Training stages support.
- Deep Learning best practices: SWA, AdamW, Ranger optimizer, OneCycle, and more.
- Workflow best practices: fp16 support, distributed training, slurm support.
- Any hardware backend supported: [AMP, Apex, DeepSpeed, FairScale, XLA](./examples/engines)


### Tests
All Catalyst code, features, and pipelines [are fully tested](./catalyst/tests).
We also have our own [catalyst-codestyle](https://github.com/catalyst-team/codestyle) and a corresponding pre-commit hook.

During testing, we train a variety of different models: image classification,
image segmentation, text classification, GANs, and much more.
We then compare their convergence metrics in order to verify
the correctness of the training procedure and its reproducibility.

As a result, Catalyst provides fully tested and reproducible
best practices for your deep learning research and development.


## Catalyst

### Documentation
- [master](https://catalyst-team.github.io/catalyst/)
- [21.07](https://catalyst-team.github.io/catalyst/v21.07/index.html)
- [21.06](https://catalyst-team.github.io/catalyst/v21.06/index.html)
- [21.05](https://catalyst-team.github.io/catalyst/v21.05/index.html) ([Catalyst — A PyTorch Framework for Accelerated Deep Learning R&D](https://medium.com/pytorch/catalyst-a-pytorch-framework-for-accelerated-deep-learning-r-d-ad9621e4ca88?source=friends_link&sk=885b4409aecab505db0a63b06f19dcef))
- [21.04/21.04.1](https://catalyst-team.github.io/catalyst/v21.04/index.html), [21.04.2](https://catalyst-team.github.io/catalyst/v21.04.2/index.html)
- [21.03](https://catalyst-team.github.io/catalyst/v21.03/index.html), [21.03.1/21.03.2](https://catalyst-team.github.io/catalyst/v21.03.1/index.html)
- [20.12](https://catalyst-team.github.io/catalyst/v20.12/index.html)
- [20.11](https://catalyst-team.github.io/catalyst/v20.11/index.html)
- [20.10](https://catalyst-team.github.io/catalyst/v20.10/index.html)
- [20.09](https://catalyst-team.github.io/catalyst/v20.09/index.html)
- [20.08.2](https://catalyst-team.github.io/catalyst/v20.08.2/index.html)
- [20.07](https://catalyst-team.github.io/catalyst/v20.07/index.html) ([dev blog: 20.07 release](https://medium.com/pytorch/catalyst-dev-blog-20-07-release-fb489cd23e14?source=friends_link&sk=7ab92169658fe9a9e1c44068f28cc36c))
- [20.06](https://catalyst-team.github.io/catalyst/v20.06/index.html)
- [20.05](https://catalyst-team.github.io/catalyst/v20.05/index.html), [20.05.1](https://catalyst-team.github.io/catalyst/v20.05.1/index.html)
- [20.04](https://catalyst-team.github.io/catalyst/v20.04/index.html), [20.04.1](https://catalyst-team.github.io/catalyst/v20.04.1/index.html), [20.04.2](https://catalyst-team.github.io/catalyst/v20.04.2/index.html)

### Notebooks
- [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/catalyst-team/catalyst/blob/master/examples/notebooks/customizing_what_happens_in_train.ipynb) Introduction tutorial "[Customizing what happens in `train`](./examples/notebooks/customizing_what_happens_in_train.ipynb)"
- [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/catalyst-team/catalyst/blob/master/examples/notebooks/customization_tutorial.ipynb) Demo with [customization examples](./examples/notebooks/customization_tutorial.ipynb)
- [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/catalyst-team/catalyst/blob/master/examples/notebooks/reinforcement_learning.ipynb) [Reinforcement Learning with Catalyst](./examples/notebooks/reinforcement_learning.ipynb)

### Notable Blog Posts
- [Catalyst 2021–Accelerated PyTorch 2.0](https://medium.com/catalyst-team/catalyst-2021-accelerated-pytorch-2-0-850e9b575cb6?source=friends_link&sk=865d3c472cfb10379864656fedcfe762)
- [BERT Distillation with Catalyst](https://medium.com/pytorch/bert-distillation-with-catalyst-c6f30c985854?source=friends_link&sk=1a28469ac8c0e6e6ad35bd26dfd95dd9)
- [Metric Learning with Catalyst](https://medium.com/pytorch/metric-learning-with-catalyst-8c8337dfab1a?source=friends_link&sk=320b95f9b2a9074aab8d916ed78912d6)
- [Pruning with Catalyst](https://medium.com/pytorch/pruning-with-catalyst-50e98f2cef2d?source=friends_link&sk=688e7a2c2e963c69c7e022e3204de5ef)
- [Distributed training best practices](https://catalyst-team.github.io/catalyst/tutorials/ddp.html)
- [Solving the Cocktail Party Problem using PyTorch](https://medium.com/pytorch/addressing-the-cocktail-party-problem-using-pytorch-305fb74560ea)
- [Beyond fashion: Deep Learning with Catalyst (Config API)](https://evilmartians.com/chronicles/beyond-fashion-deep-learning-with-catalyst)
- [Tutorial from Notebook API to Config API (RU)](https://github.com/Bekovmi/Segmentation_tutorial)

### Talks
- [Catalyst-team YouTube channel](https://www.youtube.com/channel/UC39Z1Cwr9n8DVpuXcsyi9FQ)
- [Catalyst.RL – reproducible RL research framework](https://docs.google.com/presentation/d/1U6VWIwQnQDGtu6a1x61tt3AlxCJ1-A1EYKd8lR9tKos/edit?usp=sharing) at [Stachka](https://nastachku.ru/archive/2019_innopolis/index.php?dispatch=products.view&product_id=3650)
- [Catalyst.DL – reproducible DL research framework (rus)](https://youtu.be/EfG8iwFNdWg) and [slides (eng)](https://docs.google.com/presentation/d/1TL7N_H31zDFShVbKzLfMC3DYw4e1psj6ScDN8spKQlk/edit?usp=sharing) at [RIF](http://rifvrn.ru/program/catalyst-dl-fast-reproducible-dl-4-html)
- [Catalyst.DL – reproducible DL research framework (rus)](https://youtu.be/7xyMP_5eA8c?t=8964) and [slides (eng)](https://docs.google.com/presentation/d/1XGubfTWvpiJrMyKNx2G6GtAq68y2__sDmx30eSdSRZs/edit?usp=sharing) at [AI-Journey](https://ai-journey.ru/conference-moscow/broadcast?page=2&per-page=12)
- [Catalyst.DL – fast & reproducible DL](https://docs.google.com/presentation/d/1fbF4PMl092kIdjJTw3olR3wI2cl_P2ttN3c9-WTh1gA/edit?usp=sharing) at [Datastart](https://datastart.ru/msk-autumn-2019/)
- [Catalyst.RL - NeurIPS 2019: Learn to Move - Walk Around](https://www.youtube.com/watch?v=PprDcJHrFdg&feature=youtu.be&t=4020) and [slides (eng)](https://docs.google.com/presentation/d/1g4g_Rxp9M3xAHwpp_hNzC87L9Gvum3H09g2DIQn1Taw/edit?usp=sharing) at RL reading group Meetup
- [Catalyst – accelerated DL & RL (rus)](https://youtu.be/Rmo2rx5V3v8?t=77) and [slides (eng)](https://docs.google.com/presentation/d/1xMZMjSwJfM5mZMK7pHp6hVI0FxPyZOpRtBZ0J2l1AaY/edit?fbclid=IwAR1q4XJVqYdD-a5oO2n68Y4xHvChIeOSjCSmlUYqrjIzneYpehzF8PiNdMc#slide=id.g75815b5293_0_202) at [Facebook Developer Circle: Moscow | ML & AI Meetup](https://www.facebook.com/groups/475428499888062/)
- [Catalyst.RL - Learn to Move - Walk Around 2nd place solution](https://docs.google.com/presentation/d/14UzYAURBulLjuCbQRnNeROhZ74h51-o460DPTkKMrwo/edit?usp=sharing) at NeurIPS competition track
- [Open Source ML 2019 edition](https://docs.google.com/presentation/d/1A-kwek7USA-j2Nn4n8PmLUQ1PdeUzkkViwXST7RyL-w/edit?usp=sharing) at [Datafest.elka](https://datafest.ru/elka/)


### Projects

#### Examples, Notebooks, and Starter Kits
- [CamVid Segmentation Example](https://github.com/BloodAxe/Catalyst-CamVid-Segmentation-Example) - Example of semantic segmentation for CamVid dataset
- [Notebook API tutorial for segmentation in Understanding Clouds from Satellite Images Competition](https://www.kaggle.com/artgor/segmentation-in-pytorch-using-convenient-tools/)
- [Catalyst.RL - NeurIPS 2019: Learn to Move - Walk Around](https://github.com/Scitator/learning-to-move-starter-kit) – starter kit
- [Catalyst.RL - NeurIPS 2019: Animal-AI Olympics](https://github.com/Scitator/animal-olympics-starter-kit) - starter kit
- [Inria Segmentation Example](https://github.com/BloodAxe/Catalyst-Inria-Segmentation-Example) - An example of training segmentation model for Inria Sattelite Segmentation Challenge
- [iglovikov_segmentation](https://github.com/ternaus/iglovikov_segmentation) - Semantic segmentation pipeline using Catalyst
- [Logging Catalyst Runs to Comet](https://colab.research.google.com/drive/1TaG27HcMh2jyRKBGsqRXLiGUfsHVyCq6?usp=sharing) - An example of how to log metrics, hyperparameters and more from Catalyst runs to [Comet](https://www.comet.ml/site/data-scientists/)

#### Competitions
- [Kaggle Quick, Draw! Doodle Recognition Challenge](https://github.com/ngxbac/Kaggle-QuickDraw) - 11th place
- [Catalyst.RL - NeurIPS 2018: AI for Prosthetics Challenge](https://github.com/Scitator/neurips-18-prosthetics-challenge) – 3rd place
- [Kaggle Google Landmark 2019](https://github.com/ngxbac/Kaggle-Google-Landmark-2019) - 30th place
- [iMet Collection 2019 - FGVC6](https://github.com/ngxbac/Kaggle-iMet) - 24th place
- [ID R&D Anti-spoofing Challenge](https://github.com/bagxi/idrnd-anti-spoofing-challenge-solution) - 14th place
- [NeurIPS 2019: Recursion Cellular Image Classification](https://github.com/ngxbac/Kaggle-Recursion-Cellular) - 4th place
- [MICCAI 2019: Automatic Structure Segmentation for Radiotherapy Planning Challenge 2019](https://github.com/ngxbac/StructSeg2019)
  * 3rd place solution for `Task 3: Organ-at-risk segmentation from chest CT scans`
  * and 4th place solution for `Task 4: Gross Target Volume segmentation of lung cancer`
- [Kaggle Seversteal steel detection](https://github.com/bamps53/kaggle-severstal) - 5th place
- [RSNA Intracranial Hemorrhage Detection](https://github.com/ngxbac/Kaggle-RSNA) - 5th place
- [APTOS 2019 Blindness Detection](https://github.com/BloodAxe/Kaggle-2019-Blindness-Detection) – 7th place
- [Catalyst.RL - NeurIPS 2019: Learn to Move - Walk Around](https://github.com/Scitator/run-skeleton-run-in-3d) – 2nd place
- [xView2 Damage Assessment Challenge](https://github.com/BloodAxe/xView2-Solution) - 3rd place

#### Research Papers
- [Hierarchical Attention for Sentiment Classification with Visualization](https://github.com/neuromation/ml-recipe-hier-attention)
- [Pediatric Bone Age Assessment](https://github.com/neuromation/ml-recipe-bone-age)
- [Implementation of the paper "Tell Me Where to Look: Guided Attention Inference Network"](https://github.com/ngxbac/GAIN)
- [Implementation of the paper "Filter Response Normalization Layer: Eliminating Batch Dependence in the Training of Deep Neural Networks"](https://github.com/yukkyo/PyTorch-FilterResponseNormalizationLayer)
- [Implementation of the paper "Utterance-level Aggregation For Speaker Recognition In The Wild"](https://github.com/ptJexio/Speaker-Recognition)
- [Implementation of the paper "Looking to Listen at the Cocktail Party: A Speaker-Independent Audio-Visual Model for Speech Separation"](https://github.com/vitrioil/Speech-Separation)
- [Implementation of the paper "ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks"](https://github.com/leverxgroup/esrgan)

#### Toolkits
- [Catalyst.RL](https://github.com/Scitator/catalyst-rl-framework) – A Distributed Framework for Reproducible RL Research by [Scitator](https://github.com/Scitator)
- [Catalyst.Classification](https://github.com/catalyst-team/classification) - Comprehensive classification pipeline with Pseudo-Labeling by [Bagxi](https://github.com/bagxi) and [Pdanilov](https://github.com/pdanilov)
- [Catalyst.Segmentation](https://github.com/catalyst-team/segmentation) - Segmentation pipelines - binary, semantic and instance, by [Bagxi](https://github.com/bagxi)
- [Catalyst.Detection](https://github.com/catalyst-team/detection) - Anchor-free detection pipeline by [Avi2011class](https://github.com/Avi2011class) and [TezRomacH](https://github.com/TezRomacH)
- [Catalyst.GAN](https://github.com/catalyst-team/gan) - Reproducible GANs pipelines by [Asmekal](https://github.com/asmekal)
- [Catalyst.Neuro](https://github.com/catalyst-team/neuro) - Brain image analysis project, in collaboration with [TReNDS Center](https://trendscenter.org)
- [MLComp](https://github.com/catalyst-team/mlcomp) – Distributed DAG framework for machine learning with UI by [Lightforever](https://github.com/lightforever)
- [Pytorch toolbelt](https://github.com/BloodAxe/pytorch-toolbelt) - PyTorch extensions for fast R&D prototyping and Kaggle farming by [BloodAxe](https://github.com/BloodAxe)
- [Helper functions](https://github.com/ternaus/iglovikov_helper_functions) - An assorted collection of helper functions by [Ternaus](https://github.com/ternaus)
- [BERT Distillation with Catalyst](https://github.com/elephantmipt/bert-distillation) by [elephantmipt](https://github.com/elephantmipt)

See other projects at [the GitHub dependency graph](https://github.com/catalyst-team/catalyst/network/dependents).

If your project implements a paper,
a notable use-case/tutorial, or a Kaggle competition solution, or
if your code simply presents interesting results and uses Catalyst,
we would be happy to add your project to the list above!
Do not hesitate to send us a PR with a brief description of the project similar to the above.

## Community

### Contribution Guide

We appreciate all contributions.
If you are planning to contribute back bug-fixes, there is no need to run that by us; just send a PR.
If you plan to contribute new features, new utility functions, or extensions,
please open an issue first and discuss it with us.

- Please see the [Contribution Guide](CONTRIBUTING.md) for more information.
- By participating in this project, you agree to abide by its [Code of Conduct](CODE_OF_CONDUCT.md).


### User Feedback

We've created `feedback@catalyst-team.com` as an additional channel for user feedback.

- If you like the project and want to thanks us, this the right place.
- If you would like to start a collaboration between your team and Catalyst team to improve Deep Learning R&D, you are always welcome.
- If you just don't like Github Issues and this prefer email, feel free to email us.
- Finally, if you do not like something, please, share it with us and we can see how to improve it.

We appreciate any type of feedback. Thank you!


### Acknowledgments

Since the beginning of the Сatalyst development, a lot of people have influenced it in a lot of different ways.

#### Catalyst.Team
- [Dmytro Doroshenko](https://www.linkedin.com/in/dmytro-doroshenko-05671112a/) ([ditwoo](https://github.com/Ditwoo))
- [Eugene Kachan](https://www.linkedin.com/in/yauheni-kachan/) ([bagxi](https://github.com/bagxi))
- [Nikita Balagansky](https://www.linkedin.com/in/nikita-balagansky-50414a19a/) ([elephantmipt](https://github.com/elephantmipt))
- [Sergey Kolesnikov](https://www.scitator.com/) ([scitator](https://github.com/Scitator))

#### Catalyst.Contributors
- [Aleksey Grinchuk](https://www.facebook.com/grinchuk.alexey) ([alexgrinch](https://github.com/AlexGrinch))
- [Aleksey Shabanov](https://linkedin.com/in/aleksey-shabanov-96b351189) ([AlekseySh](https://github.com/AlekseySh))
- [Alex Gaziev](https://www.linkedin.com/in/alexgaziev/) ([gazay](https://github.com/gazay))
- [Andrey Zharkov](https://www.linkedin.com/in/andrey-zharkov-8554a1153/) ([asmekal](https://github.com/asmekal))
- [Artem Zolkin](https://www.linkedin.com/in/artem-zolkin-b5155571/) ([arquestro](https://github.com/Arquestro))
- [David Kuryakin](https://www.linkedin.com/in/dkuryakin/) ([dkuryakin](https://github.com/dkuryakin))
- [Evgeny Semyonov](https://www.linkedin.com/in/ewan-semyonov/) ([lightforever](https://github.com/lightforever))
- [Eugene Khvedchenya](https://www.linkedin.com/in/cvtalks/) ([bloodaxe](https://github.com/BloodAxe))
- [Ivan Stepanenko](https://www.facebook.com/istepanenko)
- [Julia Shenshina](https://github.com/julia-shenshina) ([julia-shenshina](https://github.com/julia-shenshina))
- [Nguyen Xuan Bac](https://www.linkedin.com/in/bac-nguyen-xuan-70340b66/) ([ngxbac](https://github.com/ngxbac))
- [Roman Tezikov](http://linkedin.com/in/roman-tezikov/) ([TezRomacH](https://github.com/TezRomacH))
- [Valentin Khrulkov](https://www.linkedin.com/in/vkhrulkov/) ([khrulkovv](https://github.com/KhrulkovV))
- [Vladimir Iglovikov](https://www.linkedin.com/in/iglovikov/) ([ternaus](https://github.com/ternaus))
- [Vsevolod Poletaev](https://linkedin.com/in/vsevolod-poletaev-468071165) ([hexfaker](https://github.com/hexfaker))
- [Yury Kashnitsky](https://www.linkedin.com/in/kashnitskiy/) ([yorko](https://github.com/Yorko))


### Trusted by
- [Awecom](https://www.awecom.com)
- Researchers at the [Center for Translational Research in Neuroimaging and Data Science (TReNDS)](https://trendscenter.org)
- [Deep Learning School](https://en.dlschool.org)
- Researchers at [Emory University](https://www.emory.edu)
- [Evil Martians](https://evilmartians.com)
- Researchers at the [Georgia Institute of Technology](https://www.gatech.edu)
- Researchers at [Georgia State University](https://www.gsu.edu)
- [Helios](http://helios.to)
- [HPCD Lab](https://www.hpcdlab.com)
- [iFarm](https://ifarmproject.com)
- [Kinoplan](http://kinoplan.io/)
- Researchers at the [Moscow Institute of Physics and Technology](https://mipt.ru/english/)
- [Neuromation](https://neuromation.io)
- [Poteha Labs](https://potehalabs.com/en/)
- [Provectus](https://provectus.com)
- Researchers at the [Skolkovo Institute of Science and Technology](https://www.skoltech.ru/en)
- [SoftConstruct](https://www.softconstruct.io/)
- Researchers at [Tinkoff](https://www.tinkoff.ru/eng/)
- Researchers at [Yandex.Research](https://research.yandex.com)


### Citation

Please use this bibtex if you want to cite this repository in your publications:

    @misc{catalyst,
        author = {Kolesnikov, Sergey},
        title = {Accelerated deep learning R&D},
        year = {2018},
        publisher = {GitHub},
        journal = {GitHub repository},
        howpublished = {\url{https://github.com/catalyst-team/catalyst}},
    }


