Update Intro YouTube Training with PyTorch tutorial — v2 transforms, super(), formatting

[svekars]

Description

The Training with PyTorch tutorial uses deprecated torchvision transforms, Python 2-style super(), and old string formatting throughout.

Changes needed

Deprecated APIs

• transforms.ToTensor() → v2.ToImage() + v2.ToDtype(torch.float32, scale=True): ToTensor() is deprecated since torchvision 0.18 and lives in torchvision.transforms._deprecated.py. Update Compose and Normalize to v2 namespace as well.

Outdated patterns

• Python 2-style super(GarmentClassifier, self).__init__() → super().__init__()
• .format() → f-strings (9 instances):
  • 'Training set has {} instances'.format(...)
  • 'Validation set has {} instances'.format(...)
  • 'Total loss for this batch: {}'.format(...)
  • ' batch {} loss: {}'.format(...)
  • 'runs/fashion_trainer_{}'.format(timestamp)
  • 'EPOCH {}:'.format(...)
  • 'LOSS train {} valid {}'.format(...)
  • 'model_{}_{}'.format(...)

Files

• beginner_source/introyt/trainingyt.py

cc @subramen

[saurabhkthakur]

/assigntome