Skip to main content

ToolBox to make using Pytorch much easier.

Project description

Pytorch-Toolbox

Stable Version: v0.1.4(recommend to use.)

Automatic upload to PyPI has been finished.

This is toolbox project for Pytorch. Aiming to make you write Pytorch code more easier, readable and concise.

You could also regard this as a auxiliary tool for Pytorch. It will contain what you use most frequently tools.

Installing

A easy way to install this is by using pip:

pip install torchtoolbox

If you want to install the nightly version:

pip install -U git+https://github.com/PistonY/torch-toolbox.git@master

Todo

  • <input type="checkbox" checked="" disabled="" /> cv2 transforms.
  • <input type="checkbox" disabled="" /> Integrate with albumentations
  • <input type="checkbox" disabled="" /> Prepare tensorboard support with metric collection.

Usage

Toolbox have two mainly parts:

  1. Additional tools to make you use Pytorch easier.
  2. Some fashion work which don't exist in Pytorch core.

Tools

0. Now CV2 transforms have been released.

Support as list:(need test)

__all__ = ["Compose", "ToTensor", "ToCVImage", "Normalize", "Resize", "Scale", "CenterCrop", "Pad",
           "Lambda", "RandomApply", "RandomChoice", "RandomOrder", "RandomCrop", "RandomHorizontalFlip",
           "RandomVerticalFlip", "RandomResizedCrop", "RandomSizedCrop", "FiveCrop", "TenCrop", "LinearTransformation",
           "ColorJitter", "RandomRotation", "RandomAffine", "Grayscale", "RandomGrayscale",
           "RandomPerspective", "RandomErasing", "RandomGaussianNoise", "RandomPoissonNoise", "RandomSPNoise",
           "RandomTransforms", "Cutout"]

Part of this code refers to opencv_transforms_torchvision

  • albumentations is also recommend to use which is cv2 backend transform tools.

1. Show your model parameters and FLOPs.

import torch
from torchtoolbox.tools import summary
from torchvision.models.mobilenet import mobilenet_v2
model = mobilenet_v2()
summary(model, torch.rand((1, 3, 224, 224)))

Here are some short outputs.

        Layer (type)               Output Shape          Params    FLOPs(M+A) #
================================================================================
            Conv2d-1          [1, 64, 112, 112]            9408       235225088
       BatchNorm2d-2          [1, 64, 112, 112]             256         1605632
              ReLU-3          [1, 64, 112, 112]               0               0
         MaxPool2d-4            [1, 64, 56, 56]               0               0
          ...                      ...                      ...              ...
          Linear-158                  [1, 1000]         1281000         2560000
     MobileNetV2-159                  [1, 1000]               0               0
================================================================================
        Total parameters: 3,538,984  3.5M
    Trainable parameters: 3,504,872
Non-trainable parameters: 34,112
Total flops(M)  : 305,252,872  305.3M
Total flops(M+A): 610,505,744  610.5M
--------------------------------------------------------------------------------
Parameters size (MB): 13.50

2. Metric collection

When we train a model we usually need to calculate some metrics like accuracy(top1-acc), loss etc. Now toolbox support as below:

  1. Accuracy: top-1 acc.
  2. TopKAccuracy: topK-acc.
  3. NumericalCost: This is a number metric collection which support mean, max, min calculate type.
  4. FeatureVerification.
    • This is widely used in margin based algorithm.
from torchtoolbox import metric

# define first
top1_acc = metric.Accuracy(name='Top1 Accuracy')
top5_acc = metric.TopKAccuracy(top=5, name='Top5 Accuracy')
loss_record = metric.NumericalCost(name='Loss')

# reset before using
top1_acc.reset()
top5_acc.reset()
loss_record.reset()

...
model.eval()
for data, labels in val_data:
    data = data.to(device, non_blocking=True)
    labels = labels.to(device, non_blocking=True)

    outputs = model(data)
    losses = Loss(outputs, labels)
    # update/record
    top1_acc.update(outputs, labels)
    top5_acc.update(outputs, labels)
    loss_record.update(losses)
    test_msg = 'Test Epoch {}: {}:{:.5}, {}:{:.5}, {}:{:.5}\n'.format(
    epoch, top1_acc.name, top1_acc.get(), top5_acc.name, top5_acc.get(),
    loss_record.name, loss_record.get())

print(test_msg)

Then you may get outputs like this

Test Epoch 101: Top1 Accuracy:0.7332, Top5 Accuracy:0.91514, Loss:1.0605

3. Model Initializer

Now ToolBox support XavierInitializer and KaimingInitializer.

from torchtoolbox.nn.init import KaimingInitializer

model = XXX
initializer = KaimingInitializer()
model.apply(initializer)

4. AdaptiveSequential

Make Pytorch nn.Sequential could handle multi input/output layer.

from torch import nn
from torchtoolbox.nn import AdaptiveSequential
import torch


class n_to_n(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(3, 3, 1, 1, bias=False)
        self.conv2 = nn.Conv2d(3, 3, 1, 1, bias=False)

    def forward(self, x1, x2):
        y1 = self.conv1(x1)
        y2 = self.conv2(x2)
        return y1, y2


class n_to_one(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(3, 3, 1, 1, bias=False)
        self.conv2 = nn.Conv2d(3, 3, 1, 1, bias=False)

    def forward(self, x1, x2):
        y1 = self.conv1(x1)
        y2 = self.conv2(x2)
        return y1 + y2


class one_to_n(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(3, 3, 1, 1, bias=False)
        self.conv2 = nn.Conv2d(3, 3, 1, 1, bias=False)

    def forward(self, x):
        y1 = self.conv1(x)
        y2 = self.conv2(x)
        return y1, y2

seq = AdaptiveSequential(one_to_n(), n_to_n(), n_to_one()).cuda()
td = torch.rand(1, 3, 32, 32).cuda()

out = seq(td)
print(out.size())

# output
# torch.Size([1, 3, 32, 32])

5. Make and Use LMDB dataset.

If you meet IO speed limit, you may think about LMDB format dataset. LMDB is a tiny database with some excellent properties.

Easy to generate a LMDB format dataset.

from torchtoolbox.tools.convert_lmdb import generate_lmdb_dataset, raw_reader
from torchvision.datasets import ImageFolder

dt = ImageFolder(..., loader=raw_reader)
save_dir = XXX 
dataset_name = YYY
generate_lmdb_dataset(dt, save_dir=save_dir, name=dataset_name)

Then if you use ImageFolder like dataset you can easily use ImageLMDB to load you dataset.

from torchtoolbox.data import ImageLMDB

dt = ImageLMDB(db_path=save_dir, db_name=dataset_name, ...)

6. Non-Lable dataset

This dataset only return images.

More details please refers to codes

7. Activation Layer

A Activation Layer is provided to select a activation by string.

from torchtoolbox.nn import Activation

relu = Activation('relu', auto_optimize=True)

8. FeatureVerification Metric

A FeatureVerification Metric used to test feature based accuracy. More details refers to code.

Fashion work

1. LabelSmoothingLoss

from torchtoolbox.nn import LabelSmoothingLoss
# The num classes of your task should be defined.
classes = 10
# Loss
Loss = LabelSmoothingLoss(classes, smoothing=0.1)

...
for i, (data, labels) in enumerate(train_data):
    data = data.to(device, non_blocking=True)
    labels = labels.to(device, non_blocking=True)

    optimizer.zero_grad()
    outputs = model(data)
    # just use as usual.
    loss = Loss(outputs, labels)
    loss.backward()
    optimizer.step()

2. CosineWarmupLr

Cosine lr scheduler with warm-up epochs.It's helpful to improve acc for classification models.

from torchtoolbox.optimizer import CosineWarmupLr

optimizer = optim.SGD(...)
# define scheduler
# `batches_pre_epoch` means how many batches(times update/step the model) within one epoch.
# `warmup_epochs` means increase lr how many epochs to `base_lr`.
# you can find more details in file.
lr_scheduler = CosineWarmupLr(optimizer, batches_pre_epoch, epochs,
                              base_lr=lr, warmup_epochs=warmup_epochs)
...
for i, (data, labels) in enumerate(train_data):
    ...
    optimizer.step()
    # remember to step/update status here.
    lr_scheduler.step()
    ...

3. SwitchNorm2d/3d

from torchtoolbox.nn import SwitchNorm2d, SwitchNorm3d

Just use it like Batchnorm2d/3d. More details please refer to origin paper Differentiable Learning-to-Normalize via Switchable Normalization OpenSourse

4. Swish activation

from torchtoolbox.nn import Swish

Just use it like Relu. More details please refer to origin paper SEARCHING FOR ACTIVATION FUNCTIONS

5. Lookahead optimizer

A wrapper optimizer seems better than Adam. Lookahead Optimizer: k steps forward, 1 step back

from torchtoolbox.optimizer import Lookahead
from torch import optim

optimizer = optim.Adam(...)
optimizer = Lookahead(optimizer)

5. Mixup training

Mixup method to train a classification model. mixup: BEYOND EMPIRICAL RISK MINIMIZATION

from torchtoolbox.tools import mixup_data, mixup_criterion

# set beta distributed parm, 0.2 is recommend.
alpha = 0.2
for i, (data, labels) in enumerate(train_data):
    data = data.to(device, non_blocking=True)
    labels = labels.to(device, non_blocking=True)

    data, labels_a, labels_b, lam = mixup_data(data, labels, alpha)
    optimizer.zero_grad()
    outputs = model(data)
    loss = mixup_criterion(Loss, outputs, labels_a, labels_b, lam)

    loss.backward()
    optimizer.update()

6. Cutout

A image transform method. Improved Regularization of Convolutional Neural Networks with Cutout

from torchvision import transforms
from torchtoolbox.transform import Cutout

_train_transform = transforms.Compose([
    transforms.RandomResizedCrop(224),
    Cutout(),
    transforms.RandomHorizontalFlip(),
    transforms.ColorJitter(0.4, 0.4, 0.4),
    transforms.ToTensor(),
    normalize,
])

7. No decay bias

If you train a model with big batch size, eg. 64k, you may need this, Highly Scalable Deep Learning Training System with Mixed-Precision: Training ImageNet in Four Minutes

from torchtoolbox.tools import split_weights
from torch import optim

model = XXX
parameters = split_weights(model)
optimizer = optim.SGD(parameters, ...)

8. Margin based classification loss

Now support:

  1. ArcLoss
  2. CosLoss
  3. L2Softmax
from torchtoolbox.nn.loss import ArcLoss, CosLoss, L2Softmax

You could use this like nn.CrossEntropyLoss

9. DCNv2

10. FocalLoss

11. SigmoidCrossEntropy

12. CircleLoss

13. EvoNrom

14. Activation Layer

  1. Swish
  2. HardSwish
  3. HardSigmoid Usage:
from torchtoolbox.nn import Swish, HardSwish, HardSigmoid

swish = Swish()
hswish = HardSwish()
hsigmoid = HardSigmoid()

15. Zero LastGamma Init

from torchtoolbox.nn.init import ZeroLastGamma

model == XXX
init = ZeroLastGamma(block_name='Bottleneck', bn_name='bn3')
model.apply(init)

15. SGC_GC

Gradient Centralization: A New Optimization Technique for Deep Neural Networks

from torchtoolbox.optimizer import SGD_GC

optimizer = SGD_GC(model.parameters())
optimizer.zero_grad()
loss.backward()
optimizer.step()

Contribution

Welcome pull requests and issues!!!

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for torchtoolbox, version 0.1.5
Filename, size File type Python version Upload date Hashes
Filename, size torchtoolbox-0.1.5-py3-none-any.whl (58.6 kB) File type Wheel Python version py3 Upload date Hashes View
Filename, size torchtoolbox-0.1.5.tar.gz (52.0 kB) File type Source Python version None Upload date Hashes View

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring DigiCert DigiCert EV certificate Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page