Expand description
Keras-style model building API for QuantRS2-ML
This module provides a Keras-like interface for building quantum machine learning models, with both Sequential and Functional API patterns familiar to Keras users.
Modules§
- utils
- Utility functions for building models
Structs§
- Activation
- Activation layer
- Batch
Normalization - Batch normalization layer
- Bidirectional
- Bidirectional wrapper
- CSVLogger
- CSV logger callback
- Conv2D
- Conv2D layer (Keras-compatible)
- Cosine
Decay - Cosine decay schedule
- Dense
- Dense (fully connected) layer
- Dropout
- Dropout layer for regularization
- Early
Stopping - Early stopping callback
- Embedding
- Embedding layer
- Exponential
Decay - Exponential decay schedule
- Flatten
- Flatten layer to reshape inputs
- GRU
- GRU layer (Keras-compatible)
- Global
Average Pooling2D - GlobalAveragePooling2D layer
- Input
- Model input specification
- LSTM
- LSTM layer (Keras-compatible)
- Layer
Info - Layer information for summary
- MaxPooling2D
- MaxPooling2D layer
- Model
Checkpoint - Model checkpoint callback
- Model
Summary - Model summary information
- Multi
Head Attention - Multi-head attention layer (Keras-compatible)
- Piecewise
Constant Decay - Piecewise constant schedule
- Polynomial
Decay - Polynomial decay schedule
- Quantum
Dense - Quantum Dense layer
- ReduceLR
OnPlateau - Reduce learning rate on plateau callback
- Sequential
- Sequential model
- Training
History - Training history
Enums§
- Activation
Function - Activation function types
- Data
Type - Data types
- Initializer
Type - Weight initializer types
- Loss
Function - Loss functions
- Metric
Type - Metric types
- Optimizer
Type - Optimizer types
- Quantum
Ansatz Type - Quantum ansatz types
- Regularizer
Type - Regularizer types
Traits§
- Callback
- Callback trait for training
- Keras
Layer - Keras-style layer trait
- Learning
Rate Schedule - Learning rate schedule trait