Table of Contents

Interface IFullModel<T, TInput, TOutput>

Namespace
AiDotNet.Interfaces
Assembly
AiDotNet.dll

Represents a complete machine learning model that combines prediction capabilities with serialization and checkpointing support.

public interface IFullModel<T, TInput, TOutput> : IModel<TInput, TOutput, ModelMetadata<T>>, IModelSerializer, ICheckpointableModel, IParameterizable<T, TInput, TOutput>, IFeatureAware, IFeatureImportance<T>, ICloneable<IFullModel<T, TInput, TOutput>>, IGradientComputable<T, TInput, TOutput>, IJitCompilable<T>

Type Parameters

T

The numeric type used for calculations (e.g., double, float).

TInput
TOutput
Inherited Members
Extension Methods

Remarks

For Beginners: This interface combines all the important capabilities that a complete AI model needs.

Think of IFullModel as a "complete package" for a machine learning model. It combines:

  1. The ability to make predictions (from IModel)

    • This is like a calculator that can process your data and give you answers
    • For example, predicting house prices based on features like size and location

  2. The ability to save and load the model (from IModelSerializer and ICheckpointableModel)

    • IModelSerializer: File-based saving/loading for long-term storage
    • ICheckpointableModel: Stream-based checkpointing for training resumption and distillation
    • This is like being able to save your work in a document and open it later
    • It allows you to train a model once and then use it many times without retraining
    • It also lets you share your trained model with others

  3. The ability to compute gradients explicitly (from IGradientComputable)

    • This enables distributed training across multiple GPUs/machines
    • Allows manual control over the training process for advanced users
    • Supports meta-learning algorithms that need fine-grained gradient control

By implementing this interface, a model class provides everything needed for practical use: you can train it, use it for predictions, save it to disk, load it back when needed, checkpoint it during training, and even distribute training across multiple machines for faster learning on large datasets.

This is particularly useful for production environments where models need to be:

  • Trained once (which might take a long time)
  • Saved to disk for deployment
  • Checkpointed during training to prevent data loss
  • Loaded quickly when needed to make predictions
  • Possibly trained in a distributed manner across multiple GPUs
  • Updated with new data periodically
  • Used in knowledge distillation as teacher or student models

Properties

DefaultLossFunction

Gets the default loss function used by this model for gradient computation.

ILossFunction<T> DefaultLossFunction { get; }

Property Value

ILossFunction<T>

Remarks

This loss function is used when calling ComputeGradients(TInput, TOutput, ILossFunction<T>?) without explicitly providing a loss function. It represents the model's primary training objective.

For Beginners: The loss function tells the model "what counts as a mistake". For example: - For regression (predicting numbers): Mean Squared Error measures how far predictions are from actual values - For classification (predicting categories): Cross Entropy measures how confident the model is in the right category

This property provides a sensible default so you don't have to specify the loss function every time, but you can still override it if needed for special cases.

Distributed Training: In distributed training, all workers use the same loss function to ensure consistent gradient computation. The default loss function is automatically used when workers compute local gradients.

Exceptions

InvalidOperationException

Thrown if accessed before the model has been configured with a loss function.

Edit this page