search

ultralytics / yolov5

YOLOv5 🚀 in PyTorch > ONNX > CoreML > TFLite
https://docs.ultralytics.com
GNU Affero General Public License v3.0
51.13k stars 16.43k forks source link

Memory Leak in yolov5 #3855

Closed m-vaibhav closed 3 years ago

m-vaibhav commented 3 years ago

❔Question

yolov5: Memory is increasing till GPU got out of memory (during inference).

Additional context

Tensorrt: 7.2.2 PyTorch: 1.6 OS: ubuntu 20.04.2 LTS

Logs:

github-actions[bot] commented 3 years ago

👋 Hello @m-vaibhav, 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://www.ultralytics.com or email Glenn Jocher at glenn.jocher@ultralytics.com.

Requirements

Python 3.8 or later with all requirements.txt dependencies installed, including torch>=1.7. To install run:

$ pip install -r requirements.txt

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):

Status

CI CPU testing

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), testing (test.py), inference (detect.py) and export (export.py) on MacOS, Windows, and Ubuntu every 24 hours and on every commit.

glenn-jocher commented 3 years ago

@m-vaibhav 👋 hi, thanks for letting us know about this problem with YOLOv5 🚀. We've created a few short guidelines below to help users provide what we need in order to get started investigating a possible problem.

How to create a Minimal, Reproducible Example

When asking a question, people will be better able to provide help if you provide code that they can easily understand and use to reproduce the problem. This is referred to by community members as creating a minimum reproducible example. Your code that reproduces the problem should be:

In addition to the above requirements, for Ultralytics to provide assistance your code should be:

If you believe your problem meets all of the above criteria, please close this issue and raise a new one using the 🐛 Bug Report template and providing a minimum reproducible example to help us better understand and diagnose your problem.

Thank you! 😃

m-vaibhav commented 3 years ago

The model is yolov5m finetuned on custom data with 3 object classes. Due to official reasons, I won't be able to share the entire code or the model but I hope the following details would help:

The model is ported to ONNX to be used in TensosrRT using the official code provided in this repository.

Relevant code snippets:

Model Initialization:

{
    auto builder = TRTUniquePTR<nvinfer1::IBuilder>(nvinfer1::createInferBuilder(common::gLogger.getTRTLogger()));
    if (!builder) {
        return false;
    }
    const auto explicitBatch = 1U << static_cast<uint32_t>(NetworkDefinitionCreationFlag::kEXPLICIT_BATCH);
    auto network = TRTUniquePTR<nvinfer1::INetworkDefinition>(builder->createNetworkV2(explicitBatch));
    if (!network) {
        return false;
    }
    auto parser = TRTUniquePTR<nvonnxparser::IParser>(
            nvonnxparser::createParser(*network, common::gLogger.getTRTLogger()));
    if (!parser) {
        return false;
    }
    auto config = TRTUniquePTR<nvinfer1::IBuilderConfig>(builder->createBuilderConfig());
    if (!config) {
        return false;
    }

    auto constructed = constructNetwork(builder, network, config, parser);
    if (!constructed) {
        return false;
    }
    engine = std::shared_ptr<nvinfer1::ICudaEngine>(builder->buildEngineWithConfig(*network, *config),
                                                    common::InferDeleter());
    if (!engine) {
        return false;
    }
    context = std::shared_ptr<nvinfer1::IExecutionContext>(engine->createExecutionContext(), common::InferDeleter());

    // ... other code
}

Preprocess Images:

{
    // ... other code
    batchTensor = torch::zeros({batchSize, 3,  640,  640}, options);

    Loop over batch size {
    // <Pre-process image in OpenCV>
    batchTensor[batchIndex] = torch::from_blob(FinalOutput, {640, 640, 3}, torch::kFloat32).contiguous().permute({2, 0, 1});
}

    if (!batchTensor.is_contiguous())
    {
        batchTensor = batchTensor.contiguous();
    }
    buffers.copyInputToDeviceAsync(inputNames[0], batchTensor.data_ptr(), cuStream);
}

Forward Pass:

{
    // ... other code

    context->enqueue(batchSize, buffers.getDeviceBindings().data(), cuStream, nullptr);

    // ... other code
}
glenn-jocher commented 3 years ago

@m-vaibhav as noted above:

Also your pipeline does not follow recommended practices. See TensorRT Deployment tutorial:

YOLOv5 Tutorials

github-actions[bot] commented 3 years ago

👋 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:

Access additional Ultralytics ⚡ resources:

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 ⭐!