google colab crashed because of using all the memory

[Mohamad-Hallak]

Search before asking

• I have searched the HUB issues and discussions and found no similar questions.

Question

when I use google colab, I am getting the message "Your session crashed. Automatically restarting." how to fix this. thank you in advance.

Additional

No response

[UltralyticsAssistant]

👋 Hello @Mohamad-Hallak, thank you for raising an issue about Ultralytics HUB 🚀! This is an automated response to help you get started—an Ultralytics engineer will review your report and assist you soon.
Please visit our HUB Docs to learn more:

• Quickstart. Start training and deploying YOLO models with HUB in seconds.
• Datasets: Preparing and Uploading. Learn how to prepare and upload your datasets to HUB in YOLO format.
• Projects: Creating and Managing. Group your models into projects for improved organization.
• Models: Training and Exporting. Train YOLOv5 and YOLOv8 models on your custom datasets and export them to various formats for deployment.
• Integrations. Explore different integration options for your trained models, such as TensorFlow, ONNX, OpenVINO, CoreML, and PaddlePaddle.
• Ultralytics HUB App. Learn about the Ultralytics App for iOS and Android, which allows you to run models directly on your mobile device.
  • iOS. Learn about YOLO CoreML models accelerated on Apple's Neural Engine on iPhones and iPads.
  • Android. Explore TFLite acceleration on mobile devices.
• Inference API. Understand how to use the Inference API for running your trained models in the cloud to generate predictions.
  If this is a 🐛 Bug Report, please provide screenshots and steps to reproduce your problem to help us get started working on a fix.
  If this is a ❓ Question (as it appears), please share more details to help us assist you faster:
• Exact steps and code you ran in Google Colab (cell-by-cell if possible) 💻
• Dataset size and source, plus how it’s loaded (local upload, GDrive, HUB, etc.)
• Model and settings used (task, image size, batch size, epochs, augmentations)
• Runtime environment details (Colab runtime type, GPU/CPU, RAM shown)
• Any logs, screenshots, or crash messages from the Colab session
  We try to respond to all issues as promptly as possible. Thank you for your patience! 🙏

[pderrenger]

Hi @Mohamad-Hallak — thanks for reaching out!

On Colab, “Your session crashed. Automatically restarting.” almost always means the runtime ran out of GPU VRAM or system RAM. A few quick steps usually fix it:

Quick checks

• Use the official Ultralytics HUB Colab notebook so defaults are sane and training streams to HUB: open the HUB Colab notebook and follow the Train dialog instructions. See Start via Colab under Models in the docs.
• Enable GPU: Runtime > Change runtime type > GPU. Colab free tier has limits; Colab Pro extends them. See the Colab integration guide for context and tips.
• Update to the latest package in a fresh runtime:
  pip install -U ultralytics

Lower memory usage

• Choose a smaller model: start with YOLO11n to minimize VRAM.
• Reduce image size: try imgsz=512 (or 416/384 if needed).
• Reduce batch size: start small (batch=2 or 4).
• Reduce dataloader workers: workers=0 to lower RAM pressure.
• Avoid caching the dataset in RAM: cache=False.
• Don’t enable multi-scale during early runs (it increases memory).

Minimal example (Colab cell)
from ultralytics import YOLO

model = YOLO("yolo11n.pt")
model.train(
data="path/to/data.yaml",
epochs=50,
imgsz=512,
batch=4,
workers=0,
cache=False,
device=0, # ensure GPU
)

Monitoring and recovery

• Watch Colab’s RAM/Disk indicators; if they spike, lower batch/imgsz further. The Colab guide covers session limits and monitoring.
• If training stops mid-run, you can resume from the last checkpoint. If you launched training from Ultralytics HUB, use the Model page’s Resume feature. See the HUB Models docs for details.
• If you’re exporting/converting models in the same session, restarting the runtime can free memory. This and other tips are in Troubleshooting Common YOLO Issues.

If issues persist, please share:

• The exact notebook or a minimal reproducible snippet
• Your model variant, imgsz, batch, workers, and whether cache was enabled
• Dataset size (images) and task (detect/segment/classify)
• A screenshot of Colab RAM/GPU usage or brief logs before the crash

Alternative

• If Colab limits are blocking you, consider HUB Cloud Training (managed GPUs with live logs and easy resume). See Cloud Training.

Helpful references

• Accelerating YOLO11 Projects with Google Colab
• HUB: Train with Google Colab
• Troubleshooting Common YOLO Issues

Happy to help you tune the settings once we have a few more details 🙂