Tensorrt Export of yolov5x - insufficient memory error

[Guruprasadhegde]

Search before asking

• I have searched the YOLOv5 issues and found no similar bug report.

YOLOv5 Component

No response

Bug

I can run inference with batch 16 but when I try to export the same model to tensorrt engine with batch size 16 it gives me insufficient memory error.

Environment

No response

Minimal Reproducible Example

python export.py --weights custom.pt --include engine --device 0 --img 3288 --data custom.yaml --batch 16 --conf-thres=0.01

Additional

No response

Are you willing to submit a PR?

• Yes I'd like to help by submitting a PR!

[github-actions]

👋 Hello @Guruprasadhegde, thank you for your interest in YOLOv5 🚀! Please visit our ⭐️ Tutorials to get started, where you can find quickstart guides for simple tasks like Custom Data Training all the way to advanced concepts like Hyperparameter Evolution.

If this is a 🐛 Bug Report, please provide screenshots and minimum viable code to reproduce your issue, otherwise we can not help you.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset images, training logs, screenshots, and a public link to online W&B logging if available.

For business inquiries or professional support requests please visit https://ultralytics.com or email support@ultralytics.com.

Requirements

Python>=3.7.0 with all requirements.txt installed including PyTorch>=1.7. To get started:

git clone https://github.com/ultralytics/yolov5  # clone
cd yolov5
pip install -r requirements.txt  # install

Environments

YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):

• Google Colab and Kaggle notebooks with free GPU:
• Google Cloud Deep Learning VM. See GCP Quickstart Guide
• Amazon Deep Learning AMI. See AWS Quickstart Guide
• Docker Image. See Docker Quickstart Guide

Status

If this badge is green, all YOLOv5 GitHub Actions Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training (train.py), validation (val.py), inference (detect.py) and export (export.py) on macOS, Windows, and Ubuntu every 24 hours and on every commit.

[glenn-jocher]

@Guruprasadhegde 👋 Hello! Thanks for asking about CUDA memory issues. YOLOv5 🚀 can be trained on CPU, single-GPU, or multi-GPU. When training on GPU it is important to keep your batch-size small enough that you do not use all of your GPU memory, otherwise you will see a CUDA Out Of Memory (OOM) Error and your training will crash. You can observe your CUDA memory utilization using either the nvidia-smi command or by viewing your console output:

CUDA Out of Memory Solutions

If you encounter a CUDA OOM error, the steps you can take to reduce your memory usage are:

• Reduce --batch-size
• Reduce --img-size
• Reduce model size, i.e. from YOLOv5x -> YOLOv5l -> YOLOv5m -> YOLOv5s > YOLOv5n
• Train with multi-GPU at the same --batch-size
• Upgrade your hardware to a larger GPU
• Train on free GPU backends with up to 16GB of CUDA memory:

AutoBatch

You can use YOLOv5 AutoBatch (NEW) to find the best batch size for your training by passing --batch-size -1. AutoBatch will solve for a 90% CUDA memory-utilization batch-size given your training settings. AutoBatch is experimental, and only works for Single-GPU training. It may not work on all systems, and is not recommended for production use.

Good luck 🍀 and let us know if you have any other questions!

[Guruprasadhegde]

@glenn-jocher Thank you for your response!
I am facing insufficient cuda error not during Training but rather when I am using the export.py script to generate Tensorrt Engine file (of my trained network). While val.py with batch 16 runs as expected (with the same Model) indicating that I have enough memory in the system. But I now wanted to checkout inference using Tensorrt engine file. So I am trying to export the engine file with batch 16 it is failing to even to generate the tensorrt engine file.

[glenn-jocher]

@Guruprasadhegde same rules apply to reduce CUDA usage: smaller model, smaller image size, --half, etc. Not very complicated.

[Guruprasadhegde]

@glenn-jocher You are absolutely correct that if I reduce the batch size while exporting Tensorrt it does work. But here is what confuses me (maybe I am thinking about it the wrong way). I can run val.py with batch 16 that would mean that my GPU does fit the model and batch size.
val.py running as expected

cuda usage while it was running

Now if I try to export the same model with the same batch size it errors out. I am not getting why there is CUDA error while exporting a trained model with exact same batch size that fits in my GPU.

Of course the export works when I chose batch size 4, but doing this means I am not fully utilizing my GPU during inference

cuda usage while it was running

[glenn-jocher]

@Guruprasadhegde yeah that is kind of confusing. It appears that TensorRT is using more memory during export to try to run many optimization strategies on the model to find out what works the fastest, so exporting uses more memory than normal inference.

[github-actions]

👋 Hello, this issue has been automatically marked as stale because it has not had recent activity. Please note it will be closed if no further activity occurs.

Access additional YOLOv5 🚀 resources:

• Wiki – https://github.com/ultralytics/yolov5/wiki
• Tutorials – https://docs.ultralytics.com/yolov5
• Docs – https://docs.ultralytics.com

Access additional Ultralytics ⚡ resources:

• Ultralytics HUB – https://ultralytics.com/hub
• Vision API – https://ultralytics.com/yolov5
• About Us – https://ultralytics.com/about
• Join Our Team – https://ultralytics.com/work
• Contact Us – https://ultralytics.com/contact

Feel free to inform us of any other issues you discover or feature requests that come to mind in the future. Pull Requests (PRs) are also always welcomed!

Thank you for your contributions to YOLOv5 🚀 and Vision AI ⭐!