tsts

🌏 About Project

How a Config File is Structured
- Basic Usage
- General Settings
- Input & Output
- Training
- Optimizer
- Learning Rate Scheduler
- Trainer
- Model
- Local Scaler
- Loss Function
- Metric
- Dataset
- Pipeline
- Scaler
- Dataloader
- Logger
- Putting It Together

⚡️ Getting Started

Train & Test on a Custom Dataset
- Using tools/train.py & tools/test.py

📗 Catalog

Data Augmentation¶

Data augmentation can be applied during training by adding a PIPELINE section to the configuration. Multiple data augmentations can be applied in combination by adding multiple data augmentations to the list.

Catalog

Gaussian Noise

Gaussian Noise ¶

Add gaussian noise to input values.

PIPELINE:
  TRANSFORMS_TRAIN:
    [{"name": "GaussianNoise", "args": {"mean": 0.0, "std": 0.1}}]