Model Training
Ultralytics Platform provides comprehensive tools for training YOLO models, from organizing experiments to running cloud training jobs with real-time metrics streaming.
Overview
The Training section helps you:
- Organize models into projects for easier management
- Train on cloud GPUs with a single click
- Monitor real-time metrics during training
- Compare model performance across experiments
Workflow
graph LR
A[📁 Project] --> B[⚙️ Configure]
B --> C[🚀 Train]
C --> D[📈 Monitor]
D --> E[📦 Export]
style A fill:#4CAF50,color:#fff
style B fill:#2196F3,color:#fff
style C fill:#FF9800,color:#fff
style D fill:#9C27B0,color:#fff
style E fill:#00BCD4,color:#fff| Stage | Description |
|---|---|
| Project | Create a workspace to organize related models |
| Configure | Select dataset, base model, and training parameters |
| Train | Run on cloud GPUs or your local hardware |
| Monitor | View real-time loss curves and metrics |
| Export | Convert to 17 deployment formats |
Training Options
Ultralytics Platform supports multiple training approaches:
| Method | Description | Best For |
|---|---|---|
| Cloud Training | Train on Platform cloud GPUs | No local GPU, scalability |
| Remote Training | Train locally, stream metrics to Platform | Existing hardware, privacy |
| Colab Training | Use Google Colab with Platform integration | Free GPU access |
GPU Options
Available GPUs for cloud training:
| GPU | VRAM | Performance | Cost |
|---|---|---|---|
| RTX 3090 | 24GB | Good | $0.44/hr |
| RTX 4090 | 24GB | Excellent | $0.74/hr |
| L40S | 48GB | Very Good | $1.14/hr |
| A100 40GB | 40GB | Excellent | $1.29/hr |
| A100 80GB | 80GB | Excellent | $1.99/hr |
| H100 80GB | 80GB | Best | $3.99/hr |
Free Training
New accounts receive credits for training. Check Billing for details.
Real-Time Metrics
During training, view live metrics:
- Loss Curves: Box, class, and DFL loss
- Performance: mAP50, mAP50-95, precision, recall
- System Stats: GPU utilization, memory usage
- Checkpoints: Automatic saving of best weights
Quick Links
- Projects: Organize your models and experiments
- Models: Manage trained checkpoints
- Cloud Training: Train on cloud GPUs
FAQ
How long does training take?
Training time depends on:
- Dataset size (number of images)
- Model size (n, s, m, l, x)
- Number of epochs
- GPU type selected
A typical training run with 1000 images, YOLO11n, 100 epochs on RTX 4090 takes about 30-60 minutes.
Can I train multiple models simultaneously?
Cloud training currently supports one concurrent training job per account. For parallel training, use remote training from multiple machines.
What happens if training fails?
If training fails:
- Checkpoints are saved at each epoch
- You can resume from the last checkpoint
- Credits are only charged for completed compute time
How do I choose the right GPU?
| Scenario | Recommended GPU |
|---|---|
| Small datasets (<5000 images) | RTX 4090 |
| Medium datasets (5000-50000 images) | A100 40GB |
| Large datasets or batch sizes | A100 80GB or H100 |
| Budget-conscious | RTX 3090 |
📅 Created 0 days ago ✏️ Updated 0 days ago
