graph LR
Model_Architectures_Baselines_["Model Architectures (Baselines)"]
baselines_ModelName_["baselines.<ModelName>"]
baselines_ModelName_arch["baselines.<ModelName>.arch"]
baselines_ModelName_arch_ModelName__Arch["baselines.<ModelName>.arch.<ModelName>_Arch"]
Internal_Sub_components["Internal Sub-components"]
Model_Architectures_Baselines_ -- "contains" --> baselines_ModelName_
baselines_ModelName_ -- "contains" --> baselines_ModelName_arch
baselines_ModelName_arch -- "houses" --> baselines_ModelName_arch_ModelName__Arch
baselines_ModelName_arch_ModelName__Arch -- "composes and processes data using" --> Internal_Sub_components
click Model_Architectures_Baselines_ href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/BasicTS/Model_Architectures_Baselines_.md" "Details"
The baselines subsystem is structured as a collection of independent time series forecasting model implementations. Each model is encapsulated within its own top-level package, providing a modular and extensible design. The core of each model's implementation resides in a dedicated arch module, which defines the PyTorch computational graph and layers. The main model class, typically named _Arch, serves as the runnable instance, composed of various internal sub-components that provide reusable building blocks for complex architectures. This clear separation of concerns facilitates the development, testing, and integration of diverse forecasting models.
Model Architectures (Baselines) [Expand]
The top-level conceptual component representing the entire repository of time series forecasting model implementations. It serves as the primary entry point for accessing and managing different baseline models within the toolkit.
Related Classes/Methods:
A specific top-level package dedicated to a single time series forecasting model implementation. Each such package encapsulates all components, configurations, and specific logic related to that particular model.
Related Classes/Methods:
A dedicated module within each specific model package responsible for housing the core PyTorch model implementation. This module defines the computational graph and layers of the model.
Related Classes/Methods:
The main PyTorch nn.Module class that implements the core logic and structure of a specific time series forecasting model. This is the actual runnable model instance.
Related Classes/Methods:
Various internal sub-modules, layers, or helper classes that constitute the specific model's architecture. These are typically defined within the arch module or as separate helper classes, providing reusable building blocks for complex models.
Related Classes/Methods: