[pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

Author: pre-commit-ci[bot]

.pre-commit-config.yaml

@@ -20,4 +20,4 @@ repos:
     hooks:
       - id: black
         exclude: ^tests/
-        args: [ --safe, --quiet ]
+        args: [ --safe, --quiet ]

docs/source/conf.py

@@ -34,7 +34,7 @@
 author = "Atharva Phatak"
 
 # The full version, including alpha/beta/rc tags
-with open("../../version.txt", "r") as f:
+with open("../../version.txt") as f:
     release = str(f.readline().strip())
 
 

examples/Advanced-Tutorials/KD/vanilla-kd.py

@@ -1,6 +1,7 @@
 """Vanilla Knowledge distillation using Torchflare.
 This example only shows how to modify the training script for KD.
 """
+
 from typing import Dict
 
 import torch
@@ -11,7 +12,7 @@
 
 class KDExperiment(Experiment):
     def __init__(self, temperature, alpha, **kwargs):
-        super(KDExperiment, self).__init__(**kwargs)
+        super().__init__(**kwargs)
         self.temperature = temperature
         self.alpha = alpha
 

examples/Advanced-Tutorials/autoencoders/mnist-vae.py

@@ -1,4 +1,5 @@
 """Generating MNIST digits using Variational Autoencoders."""
+
 import torch
 import torch.nn.functional as F
 from torch import nn
@@ -15,13 +16,13 @@ def __init__(self, d):
         super().__init__()
         self.d = d
         self.encoder = nn.Sequential(
-            nn.Linear(784, self.d ** 2), nn.ReLU(), nn.Linear(self.d ** 2, self.d * 2)
+            nn.Linear(784, self.d**2), nn.ReLU(), nn.Linear(self.d**2, self.d * 2)
         )
 
         self.decoder = nn.Sequential(
-            nn.Linear(self.d, self.d ** 2),
+            nn.Linear(self.d, self.d**2),
             nn.ReLU(),
-            nn.Linear(self.d ** 2, 784),
+            nn.Linear(self.d**2, 784),
             nn.Sigmoid(),
         )
 

examples/Advanced-Tutorials/gans/dcgan.py

@@ -1,4 +1,5 @@
 """Generating MNIST Digits using DCGAN."""
+
 import os
 
 import torch
@@ -20,7 +21,7 @@ def __init__(self, latent_dim, batchnorm=True):
             latent_dim (int): latent dimension ("noise vector")
             batchnorm (bool): Whether or not to use batch normalization
         """
-        super(Generator, self).__init__()
+        super().__init__()
         self.latent_dim = latent_dim
         self.batchnorm = batchnorm
         self._init_modules()
@@ -77,7 +78,7 @@ def __init__(self, output_dim):
         Images must be single-channel and 28x28 pixels.
         Output activation is Sigmoid.
         """
-        super(Discriminator, self).__init__()
+        super().__init__()
         self.output_dim = output_dim
         self._init_modules()  # I know this is overly-organized. Fight me.
 
@@ -127,7 +128,7 @@ def forward(self, input_tensor):
 class DCGANExperiment(Experiment):
     def __init__(self, latent_dim, batch_size, **kwargs):
 
-        super(DCGANExperiment, self).__init__(**kwargs)
+        super().__init__(**kwargs)
 
         self.noise_fn = lambda x: torch.randn((x, latent_dim), device=self.device)
         self.target_ones = torch.ones((batch_size, 1), device=self.device)

examples/Advanced-Tutorials/self-supervision/ssl_byol.py

@@ -65,7 +65,7 @@ def default_augmentation(image_size: Tuple[int, int] = (224, 224)) -> nn.Module:
 # Defining the models.
 class MLPHead(nn.Module):
     def __init__(self, in_channels: int, projection_size: int = 256, hidden_size: int = 4096):
-        super(MLPHead, self).__init__()
+        super().__init__()
 
         self.net = nn.Sequential(
             nn.Linear(in_channels, hidden_size),
@@ -81,7 +81,7 @@ def forward(self, x):
 # Defining resnet encoders.
 class ResnetEncoder(nn.Module):
     def __init__(self, pretrained, mlp_params):
-        super(ResnetEncoder, self).__init__()
+        super().__init__()
         resnet = torchvision.models.resnet18(pretrained=pretrained)
         self.encoder = torch.nn.Sequential(*list(resnet.children())[:-1])
         self.projector = MLPHead(in_channels=resnet.fc.in_features, **mlp_params)
@@ -95,7 +95,7 @@ def forward(self, x):
 # Defining custom training method required as required by Bootstrap your own latent.(SSL)
 class BYOLExperiment(Experiment):
     def __init__(self, momentum, augmentation_fn, image_size, **kwargs):
-        super(BYOLExperiment, self).__init__(**kwargs)
+        super().__init__(**kwargs)
         self.momentum = momentum
         self.augmentation_fn = augmentation_fn(image_size)
 

examples/Basic-Tutorials/fit_methods.py

@@ -14,7 +14,7 @@
 
 class Net(torch.nn.Module):
     def __init__(self, out_features):
-        super(Net, self).__init__()
+        super().__init__()
 
         self.conv1 = nn.Sequential(
             nn.Conv2d(in_channels=1, out_channels=64, kernel_size=3, padding=1),

examples/Basic-Tutorials/text_classification.py

@@ -19,7 +19,7 @@
 class Model(torch.nn.Module):
     def __init__(self, dropout, out_features):
 
-        super(Model, self).__init__()
+        super().__init__()
         self.bert = transformers.BertModel.from_pretrained("prajjwal1/bert-tiny", return_dict=False)
         self.bert_drop = nn.Dropout(dropout)
         self.out = nn.Linear(128, out_features)

setup.py

@@ -1,4 +1,5 @@
 """Setup.py for torchflare."""
+
 # flake8: noqa
 import os
 
@@ -15,15 +16,15 @@
 
 
 readme_file_path = os.path.join(current_file_path, "README.md")
-with open(readme_file_path, "r", encoding="utf-8") as f:
+with open(readme_file_path, encoding="utf-8") as f:
     readme = f.read()
 
 
 version_file_path = os.path.join(current_file_path, "version.txt")
-with open(version_file_path, "r", encoding="utf-8") as f:
+with open(version_file_path, encoding="utf-8") as f:
     version = f.read().strip()
 
-with open(os.path.join(current_file_path, "requirements.txt"), "r") as f:
+with open(os.path.join(current_file_path, "requirements.txt")) as f:
     requirements = f.read().splitlines()
 
 

tests/experiment/test_experiment.py

@@ -13,7 +13,7 @@
 
 class Model(torch.nn.Module):
     def __init__(self, num_features, num_classes):
-        super(Model, self).__init__()
+        super().__init__()
         self.model = torch.nn.Linear(num_features, num_classes)
 
     def forward(self, x):