# ultralytics/yolov5 architecture

> Architecture diagram of the GitHub repository ultralytics/yolov5, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-10). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/ultralytics/yolov5
- Repository: https://github.com/ultralytics/yolov5

## Overview

YOLOv5 is a PyTorch computer-vision toolkit for object detection, instance segmentation, and image classification. Users provide images or video for inference, or labeled datasets for task-specific training and validation; trained models can then be exported for deployment. The graph also shows the shared model, data, and evaluation layers and the documented Hub and REST interfaces. Relationships are limited to README statements and source-index references; some task and deployment internals were not sampled.

Main entry points: [`train.py`](https://github.com/ultralytics/yolov5/blob/master/train.py) (Detection Training), [`models/yolo.py`](https://github.com/ultralytics/yolov5/blob/master/models/yolo.py) (YOLO Model), [`utils/dataloaders.py`](https://github.com/ultralytics/yolov5/blob/master/utils/dataloaders.py) (Image Data), [`segment/train.py`](https://github.com/ultralytics/yolov5/blob/master/segment/train.py) (Segmentation Training)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_tasks["Vision Tasks"]
  node_detect["Detection Inference<br/>[detect.py]"]
  node_classify_predict["Classification Inference<br/>[predict.py]"]
  node_segment_predict["Segmentation Inference<br/>[predict.py]"]
  node_train["Detection Training<br/>[train.py]"]
  node_classify_train["Classification Training<br/>[train.py]"]
  node_segment_train["Segmentation Training<br/>[train.py]"]
  node_validate["Detection Validation<br/>[val.py]"]
  node_classify_validate["Classification Validation<br/>[val.py]"]
  node_segment_validate["Segmentation Validation<br/>[val.py]"]
end

subgraph group_shared["Shared Runtime"]
  node_model["YOLO Model<br/>[yolo.py]"]
  node_common["Model Components<br/>[common.py]"]
  node_experimental["Model Loading<br/>[experimental.py]"]
  node_data["Image Data<br/>[dataloaders.py]"]
  node_segment_data["Mask Data<br/>[dataloaders.py]"]
  node_augment["Image Augmentation<br/>[augmentations.py]"]
  node_loss["Training Losses<br/>[loss.py]"]
  node_metrics["Evaluation Metrics<br/>[metrics.py]"]
  node_segment_metrics["Mask Metrics<br/>[metrics.py]"]
  node_general["Runtime Utilities<br/>[general.py]"]
end

subgraph group_deployment["Deployment"]
  node_export["Model Export<br/>[export.py]"]
  node_hub["PyTorch Hub<br/>[hubconf.py]"]
  node_api["REST API Example<br/>[restapi.py]"]
end

subgraph group_tracking["Training Tracking"]
  node_loggers["Training Loggers<br/>[__init__.py]"]
  node_callbacks["Training Callbacks<br/>[callbacks.py]"]
end

node_user(("User"))

node_user -->|"runs inference"| node_detect
node_user -->|"runs inference"| node_classify_predict
node_user -->|"runs inference"| node_segment_predict
node_user -->|"trains model"| node_train
node_user -->|"exports model"| node_export
node_detect -.->|"loads inputs"| node_data
node_classify_predict -->|"loads inputs"| node_data
node_segment_predict -->|"loads inputs"| node_data
node_train -->|"builds model"| node_model
node_train -->|"loads batches"| node_data
node_train -->|"computes loss"| node_loss
node_train -->|"runs validation"| node_validate
node_train -->|"logs metrics"| node_loggers
node_train -->|"uses callbacks"| node_callbacks
node_classify_train -->|"builds model"| node_model
node_classify_train -->|"runs validation"| node_classify_validate
node_segment_train -->|"builds model"| node_model
node_segment_train -->|"loads masks"| node_segment_data
node_segment_train -->|"uses metrics"| node_segment_metrics
node_segment_train -->|"runs validation"| node_segment_validate
node_validate -->|"loads batches"| node_data
node_validate -->|"computes metrics"| node_metrics
node_classify_validate -->|"loads batches"| node_data
node_segment_validate -->|"loads masks"| node_segment_data
node_segment_validate -->|"computes metrics"| node_segment_metrics
node_data -->|"applies transforms"| node_augment
node_model -->|"uses layers"| node_common
node_model -->|"uses modules"| node_experimental
node_export -->|"exports architecture"| node_model
node_classify_predict -->|"loads model"| node_model
node_segment_predict -->|"loads model"| node_model
node_train -->|"uses utilities"| node_general
node_export -->|"uses utilities"| node_general

click node_detect "https://github.com/ultralytics/yolov5/blob/master/detect.py"
click node_classify_predict "https://github.com/ultralytics/yolov5/blob/master/classify/predict.py"
click node_segment_predict "https://github.com/ultralytics/yolov5/blob/master/segment/predict.py"
click node_train "https://github.com/ultralytics/yolov5/blob/master/train.py"
click node_classify_train "https://github.com/ultralytics/yolov5/blob/master/classify/train.py"
click node_segment_train "https://github.com/ultralytics/yolov5/blob/master/segment/train.py"
click node_validate "https://github.com/ultralytics/yolov5/blob/master/val.py"
click node_classify_validate "https://github.com/ultralytics/yolov5/blob/master/classify/val.py"
click node_segment_validate "https://github.com/ultralytics/yolov5/blob/master/segment/val.py"
click node_model "https://github.com/ultralytics/yolov5/blob/master/models/yolo.py"
click node_common "https://github.com/ultralytics/yolov5/blob/master/models/common.py"
click node_experimental "https://github.com/ultralytics/yolov5/blob/master/models/experimental.py"
click node_data "https://github.com/ultralytics/yolov5/blob/master/utils/dataloaders.py"
click node_segment_data "https://github.com/ultralytics/yolov5/blob/master/utils/segment/dataloaders.py"
click node_augment "https://github.com/ultralytics/yolov5/blob/master/utils/augmentations.py"
click node_loss "https://github.com/ultralytics/yolov5/blob/master/utils/loss.py"
click node_metrics "https://github.com/ultralytics/yolov5/blob/master/utils/metrics.py"
click node_segment_metrics "https://github.com/ultralytics/yolov5/blob/master/utils/segment/metrics.py"
click node_general "https://github.com/ultralytics/yolov5/blob/master/utils/general.py"
click node_export "https://github.com/ultralytics/yolov5/blob/master/export.py"
click node_hub "https://github.com/ultralytics/yolov5/blob/master/hubconf.py"
click node_api "https://github.com/ultralytics/yolov5/blob/master/utils/flask_rest_api/restapi.py"
click node_loggers "https://github.com/ultralytics/yolov5/blob/master/utils/loggers/__init__.py"
click node_callbacks "https://github.com/ultralytics/yolov5/blob/master/utils/callbacks.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_detect,node_classify_predict,node_segment_predict,node_train,node_classify_train,node_segment_train,node_validate,node_classify_validate,node_segment_validate,node_user toneBlue
class node_model,node_common,node_experimental,node_data,node_segment_data,node_augment,node_loss,node_metrics,node_segment_metrics,node_general toneAmber
class node_export,node_hub,node_api toneMint
class node_loggers,node_callbacks toneRose
```

## Components

### Vision Tasks

- [Detection Inference](https://github.com/ultralytics/yolov5/blob/master/detect.py): `detect.py`
- [Classification Inference](https://github.com/ultralytics/yolov5/blob/master/classify/predict.py): `classify/predict.py`
- [Segmentation Inference](https://github.com/ultralytics/yolov5/blob/master/segment/predict.py): `segment/predict.py`
- [Detection Training](https://github.com/ultralytics/yolov5/blob/master/train.py): `train.py`
- [Classification Training](https://github.com/ultralytics/yolov5/blob/master/classify/train.py): `classify/train.py`
- [Segmentation Training](https://github.com/ultralytics/yolov5/blob/master/segment/train.py): `segment/train.py`
- [Detection Validation](https://github.com/ultralytics/yolov5/blob/master/val.py): `val.py`
- [Classification Validation](https://github.com/ultralytics/yolov5/blob/master/classify/val.py): `classify/val.py`
- [Segmentation Validation](https://github.com/ultralytics/yolov5/blob/master/segment/val.py): `segment/val.py`

### Shared Runtime

- [YOLO Model](https://github.com/ultralytics/yolov5/blob/master/models/yolo.py): `models/yolo.py`
- [Model Components](https://github.com/ultralytics/yolov5/blob/master/models/common.py): `models/common.py`
- [Model Loading](https://github.com/ultralytics/yolov5/blob/master/models/experimental.py): `models/experimental.py`
- [Image Data](https://github.com/ultralytics/yolov5/blob/master/utils/dataloaders.py): `utils/dataloaders.py`
- [Mask Data](https://github.com/ultralytics/yolov5/blob/master/utils/segment/dataloaders.py): `utils/segment/dataloaders.py`
- [Image Augmentation](https://github.com/ultralytics/yolov5/blob/master/utils/augmentations.py): `utils/augmentations.py`
- [Training Losses](https://github.com/ultralytics/yolov5/blob/master/utils/loss.py): `utils/loss.py`
- [Evaluation Metrics](https://github.com/ultralytics/yolov5/blob/master/utils/metrics.py): `utils/metrics.py`
- [Mask Metrics](https://github.com/ultralytics/yolov5/blob/master/utils/segment/metrics.py): `utils/segment/metrics.py`
- [Runtime Utilities](https://github.com/ultralytics/yolov5/blob/master/utils/general.py): `utils/general.py`

### Deployment

- [Model Export](https://github.com/ultralytics/yolov5/blob/master/export.py): `export.py`
- [PyTorch Hub](https://github.com/ultralytics/yolov5/blob/master/hubconf.py): `hubconf.py`
- [REST API Example](https://github.com/ultralytics/yolov5/blob/master/utils/flask_rest_api/restapi.py): `utils/flask_rest_api/restapi.py`

### Training Tracking

- [Training Loggers](https://github.com/ultralytics/yolov5/blob/master/utils/loggers/__init__.py): `utils/loggers/__init__.py`
- [Training Callbacks](https://github.com/ultralytics/yolov5/blob/master/utils/callbacks.py): `utils/callbacks.py`

### Other components

- **User**

## Connections

- User → Detection Inference: runs inference (evidence: `README.md`)
- User → Classification Inference: runs inference (evidence: `README.md`)
- User → Segmentation Inference: runs inference (evidence: `README.md`)
- User → Detection Training: trains model (evidence: `README.md`)
- User → Model Export: exports model (evidence: `README.md`)
- Detection Inference → Image Data: loads inputs
- Classification Inference → Image Data: loads inputs (evidence: `classify/predict.py`)
- Segmentation Inference → Image Data: loads inputs (evidence: `segment/predict.py`)
- Detection Training → YOLO Model: builds model (evidence: `train.py`)
- Detection Training → Image Data: loads batches (evidence: `train.py`)
- Detection Training → Training Losses: computes loss (evidence: `train.py`)
- Detection Training → Detection Validation: runs validation (evidence: `train.py`)
- Detection Training → Training Loggers: logs metrics (evidence: `train.py`)
- Detection Training → Training Callbacks: uses callbacks (evidence: `train.py`)
- Classification Training → YOLO Model: builds model (evidence: `classify/train.py`)
- Classification Training → Classification Validation: runs validation (evidence: `classify/train.py`)
- Segmentation Training → YOLO Model: builds model (evidence: `segment/train.py`)
- Segmentation Training → Mask Data: loads masks (evidence: `segment/train.py`)
- Segmentation Training → Mask Metrics: uses metrics (evidence: `segment/train.py`)
- Segmentation Training → Segmentation Validation: runs validation (evidence: `segment/train.py`)
- Detection Validation → Image Data: loads batches (evidence: `val.py`)
- Detection Validation → Evaluation Metrics: computes metrics (evidence: `val.py`)
- Classification Validation → Image Data: loads batches (evidence: `classify/val.py`)
- Segmentation Validation → Mask Data: loads masks (evidence: `segment/val.py`)
- Segmentation Validation → Mask Metrics: computes metrics (evidence: `segment/val.py`)
- Image Data → Image Augmentation: applies transforms (evidence: `utils/dataloaders.py`)
- YOLO Model → Model Components: uses layers (evidence: `models/yolo.py`)
- YOLO Model → Model Loading: uses modules (evidence: `models/yolo.py`)
- Model Export → YOLO Model: exports architecture (evidence: `export.py`)
- Classification Inference → YOLO Model: loads model (evidence: `classify/predict.py`)
- Segmentation Inference → YOLO Model: loads model (evidence: `segment/predict.py`)
- Detection Training → Runtime Utilities: uses utilities (evidence: `train.py`)
- Model Export → Runtime Utilities: uses utilities (evidence: `export.py`)
