# vaseline555/federated-learning-in-pytorch architecture

> Architecture diagram of the GitHub repository vaseline555/federated-learning-in-pytorch, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2025-01-07). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/vaseline555/federated-learning-in-pytorch
- Repository: https://github.com/vaseline555/federated-learning-in-pytorch

## Overview

##### Main Components Identification
The system should be divided into these primary components:

a. Core FL Components:
- Server Module
- Client Module
- Algorithm Module
- Dataset Management
- Model Zoo
- Metrics System

##### Component Relationships

a. Control Flow:
- Main → Server → Clients
- Server ←→ Algorithm
- Clients ←→ Models
- Clients ←→ Datasets

b. Data Flow:
- Datasets → Loaders → Clients
- Models → Clients → Server
- Metrics → Server

##### Specific Implementation Details

a. Algorithm Variants:
- FedAvg
- FedSGD
- FedProx
- FedOpt (FedAdam, FedYogi, FedAdaGrad)

b. Dataset Types:
- LEAF Benchmark
- TorchVision Datasets
- TorchText Datasets
- Custom Datasets

c. Model Categories:
- CNN Models (LeNet, MobileNet, etc.)
- BERT Variants
- Basic Models (LogReg, TwoNN)

##### Notes Section

A notes section explaining:
- Configuration options
- Dataset auto-download capability
- Model compatibility
- Evaluation schemes

Remember to indicate the extensibility points where new algorithms, models, or datasets can be added to the system, as this appears to be a research-oriented implementation.

## Diagram (Mermaid)

```mermaid
graph TB
    %% Styles
    classDef core fill:#3498db,stroke:#2980b9,color:white
    classDef data fill:#2ecc71,stroke:#27ae60,color:white
    classDef model fill:#e67e22,stroke:#d35400,color:white
    classDef metric fill:#9b59b6,stroke:#8e44ad,color:white
    classDef algorithm fill:#f1c40f,stroke:#f39c12,color:black
    classDef entry fill:#34495e,stroke:#2c3e50,color:white

    %% Entry Layer
    Main["Main Application"]:::entry
    Config["Configuration System"]:::entry
    click Main "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/main.py"

    %% Core FL Layer
    subgraph CoreFL["Core FL Components"]
        Server["Server Module"]:::core
        Client["Client Module"]:::core
        AlgoBase["Algorithm Base"]:::algorithm
        click Server "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/server/baseserver.py"
        click Client "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/client/baseclient.py"
        click AlgoBase "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/algorithm/basealgorithm.py"
    end

    %% Algorithm Variants
    subgraph Algorithms["FL Algorithms"]
        FedAvg["FedAvg"]:::algorithm
        FedSGD["FedSGD"]:::algorithm
        FedProx["FedProx"]:::algorithm
        subgraph FedOpt["FedOpt Variants"]
            FedAdam["FedAdam"]:::algorithm
            FedYogi["FedYogi"]:::algorithm
            FedAdaGrad["FedAdaGrad"]:::algorithm
        end
        click FedAvg "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/algorithm/fedavg.py"
        click FedSGD "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/algorithm/fedsgd.py"
        click FedProx "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/algorithm/fedprox.py"
        click FedAdam "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/algorithm/fedadam.py"
        click FedYogi "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/algorithm/fedyogi.py"
        click FedAdaGrad "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/algorithm/fedadagrad.py"
    end

    %% Data & Models Layer
    subgraph DataModels["Data & Models Layer"]
        subgraph Datasets["Dataset Management"]
            DataCore["Dataset Core"]:::data
            LEAF["LEAF Datasets"]:::data
            DataLoaders["Data Loaders"]:::data
            click DataCore "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/datasets/__init__.py"
            click LEAF "https://github.com/vaseline555/Federated-Learning-in-PyTorch/tree/main/src/datasets/leaf"
            click DataLoaders "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/loaders/data.py"
        end

        subgraph Models["Model Zoo"]
            CNNModels["CNN Models"]:::model
            BERTModels["BERT Variants"]:::model
            BasicModels["Basic Models"]:::model
            click CNNModels "https://github.com/vaseline555/Federated-Learning-in-PyTorch/tree/main/src/models"
            click BERTModels "https://github.com/vaseline555/Federated-Learning-in-PyTorch/tree/main/src/models"
            click BasicModels "https://github.com/vaseline555/Federated-Learning-in-PyTorch/tree/main/src/models"
        end
    end

    %% Support Systems
    subgraph Support["Support Systems"]
        Metrics["Metrics System"]:::metric
        Utils["Utility Functions"]:::core
        ModelUtils["Model Utilities"]:::core
        DataSplit["Data Splitting"]:::data
        click Metrics "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/metrics/basemetric.py"
        click Utils "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/utils.py"
        click ModelUtils "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/models/model_utils.py"
        click DataSplit "https://github.com/vaseline555/Federated-Learning-in-PyTorch/blob/main/src/loaders/split.py"
    end

    %% Relationships
    Main --> Server
    Main --> Config
    Server --> Client
    Server <--> AlgoBase
    AlgoBase --> FedAvg & FedSGD & FedProx & FedOpt
    Client <--> Models
    Client <--> Datasets
    DataLoaders --> Client
    Models --> Client
    Metrics --> Server
    Utils --> Server & Client
    ModelUtils --> Models
    DataSplit --> DataLoaders

    %% Legend
    subgraph Legend
        L1["Core Component"]:::core
        L2["Data Component"]:::data
        L3["Model Component"]:::model
        L4["Metric Component"]:::metric
        L5["Algorithm Component"]:::algorithm
        L6["Entry Point"]:::entry
    end
```
