Train Your Model¶

Step-by-step guide to training your agent.

Before Training¶

Ensure you have: - ✅ Configuration file (config.yaml) - ✅ Training data (data/train.csv) - ✅ Test data (data/test.csv)

Start Training¶

iovalence train --config config.yaml

Or with options:

iovalence train \
  --config config.yaml \
  --epochs 30 \
  --batch-size 32 \
  --gpu  # Use GPU if available

Monitor Training¶

Training output:

Starting training...
Epoch 1/20: loss=0.542, acc=0.78, val_loss=0.498, val_acc=0.82
Epoch 2/20: loss=0.412, acc=0.85, val_loss=0.401, val_acc=0.87
Epoch 3/20: loss=0.321, acc=0.89, val_loss=0.315, val_acc=0.91
...
Training completed!
Best model saved: models/best_model.pkl
Training time: 2 minutes 34 seconds

Training Metrics¶

loss: How wrong predictions are (lower is better)
acc: Training accuracy (higher is better)
val_loss: Validation loss (should decrease)
val_acc: Validation accuracy (should increase)

Troubleshooting Training¶

Low Accuracy¶

# Try:
training:
  epochs: 50          # More training
  learning_rate: 0.01 # Adjust rate
  batch_size: 16      # Smaller batches

Overfitting¶

# Try:
training:
  dropout: 0.5        # More regularization
  epochs: 10          # Fewer epochs
  early_stopping: true
  patience: 3

Training Too Slow¶

# Try:
training:
  batch_size: 128     # Larger batches
  epochs: 10          # Fewer epochs

Save & Load Models¶

Save Trained Model¶

iovalence save --model models/best_model.pkl --tag v1.0

Load Model¶

iovalence load --model models/best_model.pkl

Next Steps¶

Training Deep Dive →