# lightning-ai/pytorch-lightning architecture

> Architecture diagram of the GitHub repository lightning-ai/pytorch-lightning, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-09). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/lightning-ai/pytorch-lightning
- Repository: https://github.com/lightning-ai/pytorch-lightning

## Overview

Lightning provides two training abstractions: PyTorch Lightning automates model training and evaluation, while Fabric offers more direct control over distributed execution. The documented workflow defines a LightningModule and data, then passes both to Trainer.fit. Trainer orchestrates training loops; model hooks, optimization, metrics, callbacks, logging, and checkpointing provide distinct capabilities. Fabric exposes a lower-level entry point with configurable strategies, precision, and device support. Several subsystems are represented from the README and tree, but their internal wiring was not sampled, so uncertain connections are omitted.

Main entry points: [`src/lightning/pytorch/trainer/trainer.py`](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/trainer/trainer.py) (Trainer), [`src/lightning/fabric/fabric.py`](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/fabric.py) (Fabric API), [`src/lightning/pytorch/loops/training_epoch_loop.py`](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/training_epoch_loop.py) (Epoch Loop), [`src/lightning/pytorch/core/module.py`](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/core/module.py) (LightningModule)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_pytorch["PyTorch Lightning"]
  node_trainer["Trainer<br/>[trainer.py]"]
end

subgraph group_model["Model and Data"]
  node_module["LightningModule<br/>[module.py]"]
  node_datamodule["DataModule<br/>[datamodule.py]"]
end

subgraph group_execution["Training Execution"]
  node_fitloop["Fit Loop<br/>[fit_loop.py]"]
  node_epochloop["Epoch Loop"]
  node_eval["Evaluation Loop<br/>[evaluation_loop.py]"]
  node_predict["Prediction Loop<br/>[prediction_loop.py]"]
  node_optimization["Optimization<br/>[automatic.py]"]
end

subgraph group_support["Training Support"]
  node_metrics["Metric Results<br/>[result.py]"]
  node_callbacks["Callbacks<br/>[callback.py]"]
  node_averaging["Weight Averaging"]
  node_checkpoint["Checkpointing"]
  node_loggers["Loggers<br/>[logger.py]"]
end

subgraph group_fabric["Lightning Fabric"]
  node_fabric_api["Fabric API<br/>[fabric.py]"]
  node_connector["Device Connector<br/>[connector.py]"]
  node_strategy["Execution Strategy<br/>[strategy.py]"]
  node_precision["Precision Plugins<br/>[precision.py]"]
  node_accelerators["Accelerators<br/>[accelerator.py]"]
end

node_user(("Engineer"))

node_user -->|"defines model"| node_module
node_user -->|"provides data"| node_datamodule
node_user -->|"starts training"| node_trainer
node_trainer -->|"fits model"| node_module
node_trainer -->|"uses data"| node_datamodule
node_trainer -->|"dispatches fit"| node_fitloop
node_trainer -->|"dispatches evaluation"| node_eval
node_fitloop -->|"runs epochs"| node_epochloop
node_epochloop -->|"performs optimization"| node_optimization
node_epochloop -->|"updates results"| node_metrics
node_module -->|"logs metrics"| node_metrics
node_trainer -.->|"dispatches prediction"| node_predict
node_trainer -->|"configures callbacks"| node_callbacks
node_callbacks -.->|"extends callbacks"| node_averaging
node_trainer -.->|"configures checkpointing"| node_checkpoint
node_trainer -->|"configures logging"| node_loggers
node_user -->|"chooses expert control"| node_fabric_api
node_fabric_api -->|"configures devices"| node_connector
node_fabric_api -->|"uses strategy"| node_strategy
node_fabric_api -->|"uses precision"| node_precision
node_connector -.->|"selects accelerator"| node_accelerators

click node_module "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/core/module.py"
click node_datamodule "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/core/datamodule.py"
click node_trainer "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/trainer/trainer.py"
click node_fitloop "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/fit_loop.py"
click node_epochloop "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/training_epoch_loop.py"
click node_eval "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/evaluation_loop.py"
click node_predict "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/prediction_loop.py"
click node_optimization "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/optimization/automatic.py"
click node_metrics "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/trainer/connectors/logger_connector/result.py"
click node_callbacks "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/callbacks/callback.py"
click node_averaging "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/callbacks/weight_averaging.py"
click node_checkpoint "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/callbacks/model_checkpoint.py"
click node_loggers "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loggers/logger.py"
click node_fabric_api "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/fabric.py"
click node_connector "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/connector.py"
click node_strategy "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/strategies/strategy.py"
click node_precision "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/plugins/precision/precision.py"
click node_accelerators "https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/accelerators/accelerator.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_trainer toneBlue
class node_module,node_datamodule toneAmber
class node_fitloop,node_epochloop,node_eval,node_predict,node_optimization toneMint
class node_metrics,node_callbacks,node_averaging,node_checkpoint,node_loggers toneRose
class node_fabric_api,node_connector,node_strategy,node_precision,node_accelerators,node_user toneIndigo
```

## Components

### PyTorch Lightning

- [Trainer](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/trainer/trainer.py): `src/lightning/pytorch/trainer/trainer.py`

### Model and Data

- [LightningModule](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/core/module.py): `src/lightning/pytorch/core/module.py`
- [DataModule](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/core/datamodule.py): `src/lightning/pytorch/core/datamodule.py`

### Training Execution

- [Fit Loop](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/fit_loop.py): `src/lightning/pytorch/loops/fit_loop.py`
- [Epoch Loop](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/training_epoch_loop.py): `src/lightning/pytorch/loops/training_epoch_loop.py`
- [Evaluation Loop](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/evaluation_loop.py): `src/lightning/pytorch/loops/evaluation_loop.py`
- [Prediction Loop](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/prediction_loop.py): `src/lightning/pytorch/loops/prediction_loop.py`
- [Optimization](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loops/optimization/automatic.py): `src/lightning/pytorch/loops/optimization/automatic.py`

### Training Support

- [Metric Results](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/trainer/connectors/logger_connector/result.py): `src/lightning/pytorch/trainer/connectors/logger_connector/result.py`
- [Callbacks](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/callbacks/callback.py): `src/lightning/pytorch/callbacks/callback.py`
- [Weight Averaging](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/callbacks/weight_averaging.py): `src/lightning/pytorch/callbacks/weight_averaging.py`
- [Checkpointing](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/callbacks/model_checkpoint.py): `src/lightning/pytorch/callbacks/model_checkpoint.py`
- [Loggers](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/pytorch/loggers/logger.py): `src/lightning/pytorch/loggers/logger.py`

### Lightning Fabric

- [Fabric API](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/fabric.py): `src/lightning/fabric/fabric.py`
- [Device Connector](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/connector.py): `src/lightning/fabric/connector.py`
- [Execution Strategy](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/strategies/strategy.py): `src/lightning/fabric/strategies/strategy.py`
- [Precision Plugins](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/plugins/precision/precision.py): `src/lightning/fabric/plugins/precision/precision.py`
- [Accelerators](https://github.com/lightning-ai/pytorch-lightning/blob/master/src/lightning/fabric/accelerators/accelerator.py): `src/lightning/fabric/accelerators/accelerator.py`

### Other components

- **Engineer**

## Connections

- Engineer → LightningModule: defines model
- Engineer → DataModule: provides data
- Engineer → Trainer: starts training
- Trainer → LightningModule: fits model
- Trainer → DataModule: uses data (evidence: `src/lightning/pytorch/trainer/trainer.py`)
- Trainer → Fit Loop: dispatches fit (evidence: `src/lightning/pytorch/trainer/trainer.py`)
- Trainer → Evaluation Loop: dispatches evaluation (evidence: `src/lightning/pytorch/trainer/trainer.py`)
- Fit Loop → Epoch Loop: runs epochs (evidence: `src/lightning/pytorch/loops/training_epoch_loop.py`)
- Epoch Loop → Optimization: performs optimization (evidence: `src/lightning/pytorch/loops/training_epoch_loop.py`)
- Epoch Loop → Metric Results: updates results (evidence: `src/lightning/pytorch/loops/training_epoch_loop.py`)
- LightningModule → Metric Results: logs metrics (evidence: `src/lightning/pytorch/core/module.py`)
- Trainer → Prediction Loop: dispatches prediction
- Trainer → Callbacks: configures callbacks (evidence: `src/lightning/pytorch/trainer/trainer.py`)
- Callbacks → Weight Averaging: extends callbacks (evidence: `src/lightning/pytorch/callbacks/weight_averaging.py`)
- Trainer → Checkpointing: configures checkpointing
- Trainer → Loggers: configures logging (evidence: `src/lightning/pytorch/trainer/trainer.py`)
- Engineer → Fabric API: chooses expert control
- Fabric API → Device Connector: configures devices (evidence: `src/lightning/fabric/fabric.py`)
- Fabric API → Execution Strategy: uses strategy (evidence: `src/lightning/fabric/fabric.py`)
- Fabric API → Precision Plugins: uses precision (evidence: `src/lightning/fabric/fabric.py`)
- Device Connector → Accelerators: selects accelerator
