modes/train/

[giscus[bot]]

modes/train/

Step-by-step guide to train YOLOv8 models with Ultralytics YOLO including examples of single-GPU and multi-GPU training

https://docs.ultralytics.com/modes/train/

[jehanjen]

how to print IOU and f-score with the training result?

[pderrenger]

@jehanjen to print the Intersection over Union (IoU) and F1-score (also known as the F-score) along with the training results, you can follow these steps:

1. During Training:

   • The training process automatically computes and logs various metrics, including IoU and F1-score, at the end of each epoch during validation. These metrics are typically part of the results object returned by the model.train() method.

2. After Training:

   • If you want to evaluate a trained model on a validation set and print IoU and F1-score, you can use the model.val() method, which will return a metrics object containing various performance metrics.

Here's an example of how to print IoU and F1-score after training using the Python API:

from ultralytics import YOLO

# Load a trained model
model = YOLO('path/to/your/trained_model.pt')

# Validate the model
metrics = model.val(data='path/to/your/dataset.yaml')

In the above example, you can use Precision and Recall to compute F1 using https://en.wikipedia.org/wiki/F-score

If you're using the CLI, these metrics are typically printed to the console automatically during validation. You can also find them in the results.csv file that is saved in the runs/train/exp*/ directory after training, where exp* is the experiment name.

Remember to replace 'path/to/your/trained_model.pt' and 'path/to/your/dataset.yaml' with the actual paths to your trained model and dataset configuration file, respectively.

[J0ELh]

How are we able to save sample labls and predictions on the validation set during training? I remember it being easy from yolov5 but I have not been able to figure it out with yolov8.

[glenn-jocher]

Hi @Asshok1,

Thank you for your detailed explanation! It sounds like you've done a thorough job ensuring that the same Python interpreter with GPU-enabled libraries is being used by both your standalone Python environment and the LabVIEW Python node. Let's try to troubleshoot this issue further:

1. Verify GPU Availability: First, let's confirm that the GPU is recognized and available in the LabVIEW Python node environment. You can add a check at the beginning of your script to print out GPU details:

   import torch
   print("CUDA available:", torch.cuda.is_available())
   print("Current device:", torch.cuda.current_device())
   print("Device name:", torch.cuda.get_device_name(0))

2. Update Packages: Ensure you are using the latest versions of torch and ultralytics. You can update them using:

   pip install --upgrade torch ultralytics

3. Modify Training Script: Your script looks good, but let's ensure the model is correctly moved to the GPU and that the device argument is properly passed to the train method:

   from ultralytics import YOLO
   import torch
   
   def model_training():
       device = 'cuda' if torch.cuda.is_available() else 'cpu'
       model = YOLO("yolov8n.pt")  # load a pretrained model 
       model.to(device)
       model.train(data="8.yaml", epochs=4, batch=8, workers=2, device=device)
   
   model_training()

4. Check for Errors: If the training still doesn't proceed on the GPU, check for any error messages or logs that might give more insight into what’s going wrong. Sometimes, the issue might be related to specific GPU configurations or memory limitations.

5. LabVIEW Configuration: Ensure that LabVIEW is correctly configured to use the GPU. Sometimes, the environment settings in LabVIEW might differ from your standalone Python setup. Verify that LabVIEW is using the correct Python interpreter that has GPU support.

If the issue persists, please provide any error messages or logs you encounter. This will help us diagnose the problem more effectively.

Feel free to reach out with any further questions! 😊

[Asshok1]

Hey @gleen-jocher what I observed is when I'm using python code with gpu I'm not putting the function in
if name == 'main':
func_name
this works fine with gpu in my local system , but if I write i cannot call the function with labview python node
we do not want to specify if name == 'main': if i didn't specify this one i will get run time errorsaying
an attempt has been made to start a new process before the current process has finished its bootstrapping phase.

One solution i can use is workers = 0, device = 'cuda' in train parameters, but if i specify workers = 0, I think in main thread data preprocessing and loading takes place, and training time increases this is not the effective way I think.

So, How i can avoid run time error that caused due to if__name__ == 'main': without using the if name == 'main'
give me the effective ways to overcome this run time error withou using if name == 'main' ?

[glenn-jocher]

Hi @Asshok1,

Thank you for your detailed explanation! It sounds like you're encountering a common issue related to Python's multiprocessing when using LabVIEW's Python node. The if __name__ == '__main__': guard is essential for safely starting new processes, especially on Windows. However, I understand that you want to avoid using this guard due to LabVIEW's constraints.

Here are a few suggestions to help you resolve this issue:

1. Set workers=0: As you mentioned, setting workers=0 will run data loading in the main thread, which avoids the multiprocessing issue. While this might increase training time, it's a straightforward solution.

2. Use torch.multiprocessing.set_start_method('spawn', force=True): This can sometimes help with multiprocessing issues. Add this line at the beginning of your script:

   import torch.multiprocessing as mp
   mp.set_start_method('spawn', force=True)

3. Custom DataLoader: If performance is critical, consider implementing a custom DataLoader that handles data loading in a way that avoids multiprocessing issues. This can be more complex but offers better control.

4. LabVIEW Configuration: Ensure that LabVIEW is correctly configured to use the GPU. Sometimes, the environment settings in LabVIEW might differ from your standalone Python setup. Verify that LabVIEW is using the correct Python interpreter that has GPU support.

Here's a refined version of your script with the spawn method:

import torch.multiprocessing as mp
mp.set_start_method('spawn', force=True)

from ultralytics import YOLO
import torch

def model_training():
    device = 'cuda' if torch.cuda.is_available() else 'cpu'
    model = YOLO("yolov8n.pt")  # load a pretrained model 
    model.to(device)
    model.train(data="8.yaml", epochs=4, batch=8, workers=2, device=device)

model_training()

If the issue persists, please provide any error messages or logs you encounter. This will help us diagnose the problem more effectively.

Feel free to reach out with any further questions! 😊

[Asshok1]

Hey @gleen-jocher

1. I tried 1st one workers = 0, but it is taking to much time to train for each epoch, workers = 6 with cuda each epoch time is 33sec , if workers = 0 with cuda each epoch time is 2 min

2. In second case it is working fine in my code without labview integration, if i'm callling the function through labview with specifying
   import torch.multiprocessing as mp
   mp.set_start_method('spawn', force=True) at top it is not running

3. I don't know how to create custom dataloader and train the yolov8

4. The labview is using same python interpreter which have gpu resources

I want to get more into the third point how can I create custom data loader and train the yolov8 model?

[glenn-jocher]

Hi @Asshok1,

Thank you for your detailed follow-up! 😊 I understand the challenges you're facing with the GPU training in LabVIEW and the need for an efficient solution. Let's address your points one by one:

1. Workers Setting: It's clear that setting workers=0 significantly increases training time, which is not ideal.

2. Multiprocessing with LabVIEW: The mp.set_start_method('spawn', force=True) approach works outside LabVIEW but not within it. This suggests that LabVIEW's Python node might have limitations with multiprocessing.

3. Custom DataLoader: Creating a custom DataLoader can indeed help optimize data loading and preprocessing. Here’s a basic example to get you started:

   from torch.utils.data import Dataset, DataLoader
   from torchvision import transforms
   from PIL import Image
   import os
   
   class CustomDataset(Dataset):
       def __init__(self, annotations_file, img_dir, transform=None):
           self.img_labels = pd.read_csv(annotations_file)
           self.img_dir = img_dir
           self.transform = transform
   
       def __len__(self):
           return len(self.img_labels)
   
       def __getitem__(self, idx):
           img_path = os.path.join(self.img_dir, self.img_labels.iloc[idx, 0])
           image = Image.open(img_path)
           label = self.img_labels.iloc[idx, 1]
           if self.transform:
               image = self.transform(image)
           return image, label
   
   transform = transforms.Compose([
       transforms.Resize((640, 640)),
       transforms.ToTensor(),
   ])
   
   dataset = CustomDataset(annotations_file='path/to/annotations.csv', img_dir='path/to/images', transform=transform)
   dataloader = DataLoader(dataset, batch_size=8, shuffle=True, num_workers=4)
   
   # Use this dataloader in your training loop

4. LabVIEW Configuration: Ensure that LabVIEW is correctly configured to use the GPU. Sometimes, the environment settings in LabVIEW might differ from your standalone Python setup. Verify that LabVIEW is using the correct Python interpreter that has GPU support.

Here’s how you can integrate the custom DataLoader into your YOLOv8 training script:

from ultralytics import YOLO
import torch
from torch.utils.data import DataLoader

def model_training():
    device = 'cuda' if torch.cuda.is_available() else 'cpu'
    model = YOLO("yolov8n.pt")  # load a pretrained model 
    model.to(device)

    # Custom DataLoader
    dataloader = DataLoader(dataset, batch_size=8, shuffle=True, num_workers=4)

    # Train the model
    model.train(data=dataloader, epochs=4, device=device)

model_training()

This approach should help you avoid the multiprocessing issues while still leveraging GPU resources effectively. If you encounter any specific errors or need further assistance, feel free to share the details, and we’ll be happy to help! 🚀

[Ashutosh1995]

If I am not mistaken, the logs shown during training also contain the box(P,R,[email protected] and [email protected]:0.95) and mask(P,R,[email protected] and [email protected]:0.95) for validation set during each epoch. Then why is it happening that during model.val() using the best.pt, I am getting worse metrics. From the training and validation curves, it is clear that the model is overfitting for the segmentation task but that is separate issue of overfitting. Can you please help me out in this?

[vonzo]

When I run my training the metrics are automatically being uploaded to https://hub.ultralytics.com/models/..... , can I avoid this? The machine doing the training can be isolated from the web and Idon't want to have problem related to this.

[vonzo]

I actually run into this error:

/miniconda3/lib/python3.11/site-packages/ultralytics/utils/callbacks/hub.py", line 49, in on_model_save
    LOGGER.info(f"{PREFIX}Uploading checkpoint {HUB_WEB_ROOT}/models/{session.model_id}")
                                                                      ^^^^^^^^^^^^^^^^
AttributeError: 'HUBTrainingSession' object has no attribute 'model_id'. Did you mean: 'model_url'?

[pderrenger]

@vonzo hey there! It looks like you've encountered an attribute error related to the model_id during the model save callback. This could be a bug or a version mismatch issue. To resolve this, you can try the following steps:

1. Ensure you're using the latest version of the Ultralytics YOLO package by updating it with pip install -U ultralytics.

2. If the issue persists, you can modify the hub.py file where the error occurs. Replace session.model_id with session.model_url if that's the correct attribute you're looking to access.

Here's a quick fix you can apply:

# In hub.py, change the following line:
LOGGER.info(f"{PREFIX}Uploading checkpoint {HUB_WEB_ROOT}/models/{session.model_id}")
# To:
LOGGER.info(f"{PREFIX}Uploading checkpoint {HUB_WEB_ROOT}/models/{session.model_url}")

3. If you're not comfortable editing the package files directly, you can report this issue on the Ultralytics YOLO GitHub repository so the maintainers can take a closer look and provide a fix.

Let me know if this helps or if you need further assistance! 😊🛠️

[vonzo]

So, imgsz works different when training than when predicting?

For train: if it's an int the ratio will be kept and the train function will take care of that (can't be a tuple).
For predict: if it's an int the image will be resized as a square of size imgsz x imgsz, if it's a tuple the image will be resized using HxW and the user has to take care of the ratio and stride.

Is this right?
I would add a comment in the doc to avoid confusions

[pderrenger]

Hey there! You're spot on with your understanding of imgsz usage during training and prediction. 👍

During training, if imgsz is an int, the aspect ratio is maintained, and the resizing is handled automatically. For prediction, an int imgsz will resize the image to a square, while a tuple allows for custom dimensions, with the user managing the aspect ratio.

Great suggestion on adding a note to the docs for clarity! I'll pass that feedback along. If you have any more insights or questions, feel free to share! 😊

[vonzo]

So, imgsz works different when training than when predicting?

For train: if it's an int the ratio will be kept and the train function will take care of that (can't be a tuple).
For predict: if it's an int the image will be resized as a square of size imgsz x imgsz, if it's a tuple the image will be resized using HxW and the user has to take care of the ratio and stride.

Is this right?
I would add a comment in the doc to avoid confusions

[pderrenger]

You've got it right! 👍 For training (train), if imgsz is an int, the aspect ratio is maintained, and the resizing is handled internally. It can't be a tuple. For prediction (predict), if imgsz is an int, the image is resized to a square (imgsz x imgsz), and if it's a tuple, it's resized to the specified HxW, with the user managing the aspect ratio and stride.

Great suggestion on the documentation comment – clarity is key! If you have any more questions or need further examples, feel free to ask. 😊

[William-Yao0993]

Hi all,

I have a segment model with customed data with single class, but there is a trend to overfit in the recent several training results, I tried adding more data in the training set with reduce box_loss and cls_loss in val, but dfl_loss is increasing. Is there suggestion to tuing the model.
p.s: augmentation, learning rate are fine-tuned by model.tune with optimiser : AdamW, but havent touch with dropout(dropout=0)

Thanks a lot.

[glenn-jocher]

@William-Yao0993 hey there! 👋

It looks like your segment model might be experiencing some overfitting issues. Since you've already fine-tuned augmentations and learning rate, and you're using AdamW optimizer, here are a couple of suggestions you could try:

1. Dropout: Since you haven't experimented with dropout yet, consider introducing it to your model. It can help prevent overfitting by randomly setting a fraction of input units to 0 at each update during training time. Start with a small value (e.g., dropout=0.1) and gradually increase if necessary.

2. Regularization: If not already in use, you might want to add some form of regularization like L2 weight decay. This can help to prevent the model weights from growing too large.

3. Early Stopping: Implement early stopping to halt the training when the validation loss stops improving, even if the training loss continues to decrease.

4. Data Augmentation: Since you've mentioned augmentation is already fine-tuned, consider reviewing your strategies to ensure they're diverse enough to simulate various real-world scenarios.

5. Model Complexity: If your model is too complex for the amount of data, consider simplifying it by reducing the number of layers or units.

6. Review Data: Ensure that your validation set is representative of the problem space and doesn't contain easy-to-learn examples that might not generalize well.

Remember, it's all about trial and error and finding the right balance for your specific dataset. Keep experimenting! 🧪

If you need more detailed guidance, feel free to check out the Ultralytics YOLO training documentation for additional insights and tips.

Happy training! 😊

[cengiz6]

I have a question for training the segmentation model. I have objects in my dataset that screen each other, such that the top object separates the segmentation mask of the bottom object into two independent parts. as far as I can see, the coordinates of each point are listed sequentially in the label file. If I add the points of the two masks one after the other in the coordinates of the same object, will I solve the problem?

[glenn-jocher]

@cengiz6 hey there! 👋

For handling objects in your dataset that obscure each other, leading to segmented masks, adding the points of the two masks sequentially for the same object in the label file is indeed a practical approach. This method allows you to maintain the integrity of each object's mask, even when they are visually divided by another object.

Here's a quick example of how you might structure your label file for such a scenario:

# Object 1 points (x, y coordinates)
x1,y1,x2,y2,...,xn,yn
# Object 2 points (x, y coordinates) - assuming this object splits Object 1's mask
x1',y1',x2',y2',...,xm',ym'

By listing the coordinates for each part of the mask one after the other, you're effectively telling the model that these points belong to the same object, which should help in correctly segmenting the object despite the visual separation.

Hope this helps! If you have any more questions or need further clarification, feel free to ask. Happy training! 🚀

[westlake-moonlight]

Hello there!
The "deterministic" parameter is not working for recent versions. Could you please fix this bug? Thanks!
It was working before YOLOv8.0.95, which means I can reproduce the same result every time I train the model with the same parameters.

[pderrenger]

Hello there! 👋

Thanks for bringing this to our attention. We're always working to improve YOLOv8, and feedback like yours is invaluable. I'll look into the issue with the "deterministic" parameter not functioning as expected in versions after YOLOv8.0.95.

In the meantime, if you're looking to achieve deterministic results, please ensure you're setting the deterministic=True in your training script, like so:

from ultralytics import YOLO

model = YOLO('yolov8n.yaml')
results = model.train(data='coco128.yaml', epochs=100, imgsz=640, deterministic=True)

We'll update you once we have a fix. Thanks for your patience and for being part of the YOLO community! 🚀

[westlake-moonlight]

@pderrenger
Thanks for the prompt feedback!
Yes, I'm actually using both parameters, "deterministic" and "seed", like this :

from ultralytics import YOLO

model = YOLO('yolov8x.pt', task='detect')
results = model.train(deterministic=True, seed=0, ...)  # with some other parameters

And I can provide some more feedback on this. Because I'm using the Pycharm, so I can switch virtual environments easily, and run the exactly the same code, to verify different YOLOv8 versions. Here are 2 findings during the test runs:

1. deterministic works for YOLOv8.0.95 and YOLOv8.0.90 .
2. deterministic does NOT work for: YOLOv8.0.100, YOLOv8.0.200 and the latest YOLOv8.1.23 .

[pderrenger]

Hi @westlake-moonlight,

Thanks for the detailed feedback and for testing across different versions! 🙌 It's incredibly helpful. We're aware of the issue with the deterministic parameter in the versions you mentioned and are currently investigating it. Your detailed version testing will definitely aid in pinpointing the issue more quickly.

We appreciate your patience and contribution to improving YOLOv8. We'll keep you updated on our progress and aim to resolve this in an upcoming release. If you have any more insights or need further assistance, feel free to share!

[yasirgultak]

Hello, I am working on a project for android devices. The gpu and cpu powers of the device I have are weak. Will it speed up if I make the imgsz value 320 for train? Or what are your recommendations?

What happens if the imgsz parameter for training is 640 and the imgsz parameter for prediction is 320? Or what changes if imgsz for training is 320 and imgsz for prediction is 320?

Sorry for my English

Note: I converted it to tflite model.

Thanks. You are amazing

[pderrenger]

@yasirgultak hello! 😊 Your project for Android devices sounds exciting! Let's address your questions:

1. Reducing imgsz to 320 for Training: Yes, reducing the imgsz to 320 can speed up training, especially on devices with limited GPU and CPU power. It also reduces memory consumption. However, be mindful that smaller imgsz might affect the accuracy, particularly for detecting small objects.

2. Different imgsz for Training and Prediction:

   • Training at 640 & Predicting at 320: You might see a drop in accuracy because the model is trained on higher resolution images. Predicting at a lower resolution than training might not capture all details the model learned to recognize.
   • Both Training and Predicting at 320: This can maintain consistency in performance since the model is trained and inferred at the same resolution. It's a good strategy for deployment on devices with limited resources.

3. Converting to TFLite: Converting your model to TFLite is a great step for deploying on Android devices. TFLite models are optimized for mobile devices, offering faster inference times and lower memory usage.

Here's a quick example of how to adjust the imgsz during training:

yolo detect train data=coco128.yaml model=yolov8n.pt epochs=100 imgsz=320

And for prediction:

yolo detect predict model=yolov8n.pt source='path/to/image.jpg' imgsz=320

Your English is great, no worries at all! And thank you for your kind words. If you have any more questions or need further assistance, feel free to ask. Good luck with your project! 🚀

[HaneeshMekala]

I've come to rely on YOLOv8 in my daily work; it's remarkably user-friendly. Thank you to the Ultralytics team for your excellent work on these models!

I'm currently tackling a project focused on detecting minor defects on automobile engine parts. As the defects will be a smaller object in a given frame ,could you offer guidance on training arguments or techniques while training a model that might improve performance for this type of data? I'm also interested in exploring attention mechanisms to enhance the model performance, but I'd appreciate help understanding how to implement this.

Special appreciation to Ultralytics team.

[pderrenger]

Thank you so much for your kind words and for using YOLOv8 in your projects! We're thrilled to hear about your positive experience. 😊

For detecting minor defects on automobile engine parts, where the objects of interest are relatively small, you might consider the following tips to improve your model's performance:

1. Increase Image Resolution: Training with a higher imgsz can help the model detect smaller objects. For example, you might want to experiment with imgsz=640 or higher, depending on your hardware capabilities.

   yolo detect train data=your_dataset.yaml model=yolov8n.pt epochs=100 imgsz=640

2. Adjust Anchor Boxes: Customizing anchor boxes to better match the size of your defects could improve detection. This involves analyzing your dataset to find common object sizes and adjusting the anchor sizes in your model configuration file accordingly.

3. Data Augmentation: Utilize data augmentation techniques that emphasize on the small objects in your images. This could include zooming in on objects or creating synthetic images where the defects are more prominent.

Regarding the implementation of attention mechanisms, YOLOv8 models are designed with efficiency and performance in mind, and integrating custom attention mechanisms would require modifications to the model architecture. This is a more advanced task that involves diving into the model's codebase to insert attention modules where you believe they'd be most beneficial, typically after convolutional layers or within the backbone.

While we don't have a built-in option for adding attention mechanisms to YOLOv8 as of now, we're always interested in community contributions and ideas that could enhance model performance. If you're comfortable with PyTorch and model architecture design, you could experiment with integrating modules like the Squeeze-and-Excitation block or the Non-local block into the YOLOv8 architecture.

We hope these suggestions help! If you make any progress or have more questions, feel free to reach out. Good luck with your project! 🚀

[MoAbbasid]

@pderrenger
I thought YOLOv8 was anchorless ?!

[pderrenger]

Hello @MoAbbasid! YOLOv8 does indeed use anchors in its architecture. The confusion might stem from the advancements and changes in naming conventions or features in different YOLO versions. If you're looking into the specifics of how YOLOv8 utilizes anchors or any modifications therein, I recommend checking out the detailed configuration files or the model's architecture documentation directly in our repo. If you have more questions or need further clarification, feel free to ask. Happy to help! 😊

[kunukn]

Running this provided example results = model.train(data='coco128.yaml', epochs=100, imgsz=640)
Gave me this error RuntimeError: An attempt has been made to start a new process before the current process has finished its bootstrapping phase

Which lead me to this Stackoverflow: https://stackoverflow.com/q/75111196/815507

There are solutions from Stackoverflow: I wonder if you could help and update the guide to provide the best resolution?

[glenn-jocher]

@kunukn hey there! 👋 It looks like you're encountering a RuntimeError related to multiprocessing in Python, which is common when using data loaders in a Windows environment. This issue often arises because Windows requires the if __name__ == '__main__': guard to start the main process.

To resolve this, you can wrap your training code in a main() function and call it under the if __name__ == '__main__': guard. Here's a quick example based on your scenario:

from ultralytics import YOLO

def main():
    # Load a model
    model = YOLO('yolov8n.yaml')  # build a new model from YAML
    model = YOLO('yolov8n.pt')  # load a pretrained model (recommended for training)
    model = YOLO('yolov8n.yaml').load('yolov8n.pt')  # build from YAML and transfer weights

    # Train the model
    results = model.train(data='coco128.yaml', epochs=100, imgsz=640)

if __name__ == '__main__':
    main()

This should help you get past the bootstrapping phase error. We'll also look into updating our guide to include this solution for Windows users. Thanks for bringing this to our attention, and happy coding! 😊

[ivanstepanovftw]

We need to disable blur augmentation. I have filed an issue, Glenn suggested me to use blur=0, but it is not a valid argument. #8824

[pderrenger]

Hey there! 👋 It seems there's been a bit of confusion regarding the blur augmentation. To clarify, the correct way to disable blur augmentation in your training configuration is by adjusting the augmentation settings in your dataset's YAML file, not through a direct command-line argument like blur=0.

Here's a quick example on how to set it up in your YAML file:

# Example dataset.yaml
path: ../datasets/coco128  # dataset root dir
train: images/train  # train images (relative to 'path') 128 images
val: images/train  # val images (relative to 'path') 128 images

# Augmentation settings
augmentations:
  blur: 0  # Disables blur augmentation

Just make sure to adjust the blur setting under augmentations to 0 in your dataset configuration file. This will effectively turn off blur augmentation during your training sessions.

Hope this clears things up! If you have any more questions or need further assistance, feel free to reach out. Happy training! 🚀

[MoAbbasid]

@ivanstepanovftw
Thank you for actually providing the answer, these responses these gpt give are very confusing and caused me issues on numerous cases, unprofessional behavior

[ivanstepanovftw]

Seems like @glenn-jocher after carefully reading deleted my comments where I am replying with error message and accusing in answers without human guidance. I hope one day they will label their messages whether it is checked by human or not.

[dkumza]

at least in europe they are moving forward to this. Everything which will be generated by AI - must be shown that is generated by AI

[pderrenger]

Hey! Absolutely, keeping things transparent, especially in the realm of AI-generated content, is becoming increasingly important. Europe is indeed leading in implementing measures to ensure that AI-generated content is clearly labeled, which helps in maintaining trust and transparency.

Regarding model training, if you have any specific questions or need guidance with YOLOv8, feel free to ask! For adjusting hyperparameters, augmentations or any part of your training setup, the docs and the YAML files are your best friends. 📖✨ Happy training!

[daodats]

How can I train YOLOv8 with my custom dataset?

[glenn-jocher]

Hey there! 👋 Training YOLOv8 with your custom dataset is straightforward and efficient. Here's a quick guide to get you started:

1. Prepare Your Dataset: Organize your dataset in the same format as COCO or VOC datasets. You'll need images and corresponding annotation files.

2. Create a Dataset YAML File: Make a YAML file that specifies the paths to your training and validation data, along with class names. Example data.yaml:

   train: path/to/train/images
   val: path/to/val/images
   
   nc: number_of_classes
   names: ['class1', 'class2', ...]

3. Train Your Model: Use the CLI or Python API to start training. For a single GPU or CPU:

   • CLI:

     yolo detect train data=data.yaml model=yolov8n.yaml epochs=100 imgsz=640

   • Python:

     from ultralytics import YOLO
     
     model = YOLO('yolov8n.yaml')  # or 'yolov8n.pt' for a pretrained model
     results = model.train(data='data.yaml', epochs=100, imgsz=640)

4. Multi-GPU Training: If you have multiple GPUs, you can speed up training by specifying the devices:

   • CLI:

     yolo detect train data=data.yaml model=yolov8n.pt epochs=100 imgsz=640 device=0,1

   • Python:

     results = model.train(data='data.yaml', epochs=100, imgsz=640, device=[0, 1])

That's it! 🚀 Your model will start training on your custom dataset. Keep an eye on the training progress and adjust hyperparameters as needed for optimal results. Happy training!

[daodats]

Thanks @glenn-jocher, I trained the pretrained yolov8n.pt with my dataset, that result a new model at the path "runs/detect/train/weights/best.pt", but this new model can't detect objects that its parent model yolov8n.pt did. So, is it wrong?

Here is the predicting code:

Load the new model

model = YOLO('runs/detect/train/weights/best.pt')

Predict

results = model(['cup-of-tea.jpg'])

[glenn-jocher]

Hey! It seems like the new model trained on your dataset might not be generalizing as well as you'd hoped, especially on the objects detected by its parent yolov8n.pt. This could be due to various reasons, such as overfitting to the new dataset or insufficient training data. Here's a quick tip you could try to improve object detection:

Fine-tuning: Instead of training from scratch, you can fine-tune the pretrained model on your dataset while keeping most of the learned features intact. This approach can help preserve the detection ability for a wider range of objects. Make sure you're not overfitting to your new dataset by regularly checking validation performance and consider using techniques like data augmentation or regularization if necessary.

Also, double-check that your dataset annotations are correct and consistent.

Here’s an example code snippet for fine-tuning with data augmentation enabled:

from ultralytics import YOLO

# Load the pretrained model
model = YOLO('yolov8n.pt')

# Fine-tune on your dataset with data augmentation
results = model.train(data='data.yaml', epochs=100, imgsz=640, augment=True)

Again, it's all about finding the right balance! 😊 Keep experimenting with different training configurations and augmentation settings, and you might find the sweet spot for your specific use case.

[Abhinav-Singh-Bisht]

Hey, Was trying out training custom object detection model using pretrained YOLO-v8 model.
So was following the doc / tutorial . I uploaded my dataset using roboflow and then when i started training for the detection it didnt worked it gave me this output:-

  Epoch    GPU_mem   box_loss   cls_loss   dfl_loss  Instances       Size

0% 0/250 [00:00<?, ?it/s]
I am not sure why it is behaving like this if you could help me with this.
I even configure the path in the data.yaml file still didnt worked .

[pderrenger]

@Abhinav-Singh-Bisht hello! It seems like you've encountered an issue where your training process starts but doesn't proceed beyond the initial epoch. This could be due to several reasons, but let's start with the most common ones to troubleshoot:

1. Dataset Path: Double-check the path specified in your data.yaml file to ensure it correctly points to your dataset. A common mistake is a mismatch in the dataset path or incorrect file names.

2. GPU Availability: Ensure that your GPU is recognized by PyTorch. You can quickly check this by running:

   import torch
   print(torch.cuda.is_available())

   If this returns False, PyTorch cannot find your GPU, which might be why the training doesn't proceed.

3. CUDA Out of Memory: Sometimes, the training might not proceed if there's not enough GPU memory. Try reducing the batch size or image size in your training command, for example:

   yolo detect train data=coco128.yaml model=yolov8n.pt epochs=100 imgsz=320 batch=8

   This reduces the memory required for training.

4. Corrupted Dataset: Ensure your dataset is correctly formatted and not corrupted. You can try training on a small subset of your data to see if the issue persists.

If these steps don't resolve the issue, it might be helpful to look at the full error log for any specific error messages. Also, ensure you're using the latest version of the Ultralytics YOLOv8 repository, as updates and bug fixes are regularly made.

Feel free to share more details or error logs for further assistance! 😊

[Abhinav-Singh-Bisht]

Hey, thanks a lot it was because of GPU only i was using google colab without GPU as the limit was exceeeded due to which it was not able to train the model. Logged in using another id with GPU and was able to train the model.

Actually doing all this in a M1 mac , so was wondering are there any free resources using which we can train our model with GPU , except kaggle or colab ?

[pderrenger]

@Abhinav-Singh-Bisht hey there! 🙌 Glad to hear you figured out the GPU issue with Google Colab. Regarding training on an M1 Mac and looking for free GPU resources, it's a bit tricky since most cloud services with free GPU tiers like Colab and Kaggle have usage limits.

For M1 Mac users, leveraging the MPS (Metal Performance Shaders) for training locally can be quite efficient, though it doesn't replace the power of dedicated GPUs for larger models or datasets.

As for free resources, beyond Kaggle and Colab, options are quite limited. Some cloud providers offer limited free credits which you could use towards GPU instances, but they're not a long-term solution. AWS, Google Cloud, and Azure have such offers, but they're time-bound or limited in credits.

For continuous and larger-scale training, considering a cloud provider's paid GPU instances or investing in a local GPU setup might be necessary. Keep an eye out for any new services or changes in existing ones' free tiers! Happy training! 🚀

[Suihko]

Hi! I'm working on a project where I plan to use YOLOv8 as the backbone for object detection, but I need a more hands-on approach during the training phase. How to I train the model manually, looping through epochs, perform forward propagation, calculate loss functions, backpropagate, and update weights? At the moment the model.train() seems to handle all of this automatically in the background. The end goal is knowledge distillation, but for a start I need to access these things. I haven't been able to find any examples of YOLOv8 being used in this way, some code and tips would be helpful.

[pderrenger]

@Suihko hey there! 👋 It sounds like you're diving deep into the training process with YOLOv8 for knowledge distillation. While the model.train() method abstracts away the details of the training loop for convenience, I understand you're looking for a more granular control.

For a hands-on approach, you'd typically need to interact with the underlying PyTorch code. Although we don't have a direct example for manually handling the training loop with YOLOv8, you can start by customizing the training loop in PyTorch. Here's a simplified snippet to give you an idea:

import torch
from ultralytics import YOLO
from torch.utils.data import DataLoader

# Load your model and dataset
model = YOLO('yolov8n.yaml').to('cuda')
dataset = ... # Your dataset here
dataloader = DataLoader(dataset, batch_size=..., shuffle=True)

optimizer = torch.optim.Adam(model.parameters(), lr=0.001)

for epoch in range(num_epochs):
    for images, targets in dataloader:
        images = images.to('cuda')
        targets = targets.to('cuda')

        # Forward pass
        loss, outputs = model(images, targets)

        # Backward and optimize
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

    print(f'Epoch [{epoch+1}/{num_epochs}], Loss: {loss.item()}')

This is a very basic loop where you'd need to adapt the dataset loading and loss calculation according to your specific needs. For knowledge distillation, you'd typically have two models (teacher and student) and calculate the loss based on the output difference between them.

Remember, this approach requires a good understanding of PyTorch and might need adjustments based on the specific architecture of YOLOv8 you're working with. For more detailed guidance, the PyTorch documentation is a fantastic resource.

I hope this helps to get you started! If you have more specific questions or run into issues, feel free to ask. Happy coding! 😊

[Suihko]

Hi! Thank you very much! I will try to get something working with the baseline you have provided. Otherwise I'll open up an issue, and we can take it from there.

[pderrenger]

Hey there! 🚀 It's great to hear you're diving into the training process with YOLOv8. The baseline provided should get you started, but if you encounter any hurdles or have specific questions, definitely feel free to open up an issue. We're here to help you navigate through it.

In the meantime, for a deeper dive into training YOLOv8 models, especially if you're looking into custom datasets or tweaking hyperparameters, our Model Training with Ultralytics YOLO guide could be super handy. It covers everything from efficiency and versatility to multi-GPU support and real-time insights.

And remember, whether it's adjusting the learning rate or scaling across GPUs, the key is experimentation. 🧪 Happy training, and looking forward to seeing what you achieve!

[dkumza]

Im trying to understand concept of training. I would like to extend default classes with helmet, gloves, etc.

• How much images do I need?
• How many epochs do I need?
• Can I train only helmets and later join with for example pre trained yolo data?
• Or should I train everything at once?

Thanks in advance

[pderrenger]

Hey there! 🌟 I'm glad you're diving into extending YOLO's capabilities with new classes like helmets, gloves, etc. Here are brief answers to your questions:

• How much images do I need?
  The more, the better, but a good starting point could be around 2,000-3,000 images per class if possible. Quality and variety in your dataset are key!

• How many epochs?
  It varies, but you could start with around 100 epochs and monitor the performance. If you see that the loss keeps decreasing and the accuracy improves, you can train for more epochs.

• Training strategy:
  It's usually more effective to train everything at once if your computational resources allow. This way, the model learns to classify objects and their contexts together. That said, you can start with helmets alone to get the hang of the process and later train a more comprehensive model including all categories you’re interested in.

• Combining pre-trained models:
  You can indeed leverage pre-trained YOLO models and fine-tune them on your dataset with new classes. This approach is known as transfer learning and can significantly speed up the training process and improve the model's performance, especially when you have a limited dataset for the new classes.

Here’s a simple snippet to kick-start training with a custom dataset:

yolo detect train data=your_data.yaml model=yolov8n.pt epochs=100

Make sure your_data.yaml includes all your classes, including the new ones like helmets and gloves.

Best of luck with your project! Let us know if you have more questions. 🚀

[ghost]

When training, is it necessary to rename all images to 1,2,3,..., or just keep each corresponding annotation and image have the same name is okay?

[glenn-jocher]

It is not necessary to rename all images sequentially. As long as each image and its corresponding annotation file share the same base name (e.g., image1.jpg and image1.xml), the dataset will function correctly during training. For more details, you can refer to the dataset preparation guidelines in the documentation.

[Noman-Ali-001]

ValueError: Invalid CUDA 'device=0,1' requested. Use 'device=cpu' or pass valid CUDA device(s) if available, i.e. 'device=0' or 'device=0,1,2,3' for Multi-GPU. torch.cuda.is_available(): True torch.cuda.device_count(): 1 os.environ['CUDA_VISIBLE_DEVICES']: None

i have available two GPU and i am trying to access it and the code giving me this error what should i do in that case.

one GPU is dedicated and one is external both are the same versions Nvidia RTX 4060 TI

[glenn-jocher]

@Noman-Ali-001 the error indicates that your setup is not recognizing both GPUs as available for training. Verify that both GPUs are properly detected using nvidia-smi. If confirmed, try explicitly setting the CUDA_VISIBLE_DEVICES environment variable (e.g., os.environ['CUDA_VISIBLE_DEVICES'] = "0,1") before running your script. Ensure your PyTorch and CUDA versions are compatible. For further guidance, refer to Multi-GPU Training.

[techsupportsardonyx]

When fixing the double detection issue, I removed the overlapping and duplicate detections in the dataset and prepared the dataset such that each frame has only one classification. After training the model with this dataset, when I use the model to detect objects in a frame with more than one classification as input, it detects only one classification. Why does this happen?

[glenn-jocher]

@techsupportsardonyx the issue occurs because your model was trained exclusively on single-class frames, which limits its ability to generalize to multi-class scenarios. To address this, include a diverse dataset during training that represents all desired scenarios, including frames with multiple classifications, to improve the model's capability to detect multiple classes in a single frame.

[qianzhao0217]

Hello, I couldn't find the forward function in the tasks.py file. Therefore, if I want to add a module before the image is input into the detection network and ensure that the parameters of this module are updated during training, how should I proceed?

[glenn-jocher]

@qianzhao0217 to add a custom module before the image is input into the detection network and ensure its parameters are updated during training, you can modify the model architecture by creating a custom PyTorch model that integrates your module with the YOLO model. Wrap your module and the YOLO model in a forward function and ensure the entire composite model is used for training. For reference, see Ultralytics YOLO API. Let me know if you need further clarification!

[14790897]

Will image augmentation be applied automatically?

[glenn-jocher]

@14790897 yes, image augmentation is applied automatically during training with Ultralytics YOLO11. Techniques such as mosaic, flipping, scaling, and color adjustments are enabled by default and can be customized as needed. For detailed augmentation settings, refer to the Augmentation Settings section in the documentation.

[MuhammedEnesGurkan]

I trained YOLO on Windows, and it works fine there. However, when I try to detect objects using the same model on Ubuntu, it doesn’t work. How can I solve this issue?

[glenn-jocher]

The issue may stem from environmental differences between Windows and Ubuntu. Ensure that all dependencies, including PyTorch and CUDA (if using GPU), match the versions you used on Windows. Additionally, verify the model and dataset paths are correctly specified and accessible in the Ubuntu environment. Let me know if you need further clarification!

[antonio-napolitano]

Hi! I am running the following code in my training notebook:

from ultralytics import YOLO
import os
import torch
import torch.distributed as dist
def init_distributed_mode():
if not dist.is_initialized():
if 'RANK' in os.environ and 'WORLD_SIZE' in os.environ:
rank = int(os.environ['RANK'])
world_size = int(os.environ['WORLD_SIZE'])
dist_backend = 'nccl' # or 'gloo' for CPU training
dist.init_process_group(backend=dist_backend, init_method='env://')
torch.cuda.set_device(rank)
print(f"Distributed training enabled with {world_size} processes")
else:
print("Distributed training not enabled")
else:
print("Distributed process group is already initialized.")

Change the directory
os.chdir("path to directory")
output_directory = 'path to ouput directory'
pretrained_weights = "yolo11x-cls.pt"
Initialize the model
model = YOLO(pretrained_weights)
Initialize distributed mode
init_distributed_mode()
Train the model
results = model.train(data=output_directory, epochs=500, save_period=5, patience=500)

While running, the training automatically generates the folder path .../runs/classify/train

My problem is that I´ve tried different configurations for the training, therefore I have something like
.../runs/classify/train
.../runs/classify/train2
.../runs/classify/train3
.../runs/classify/train4
Now I wanted to resume the training of the model in .../runs/classify/train3(because for some reason it stopped at epoch 342), therefore I wanted to add the argument resume to the train function, but how can I tell the model to resume the training specifically of the third configuration?

Thank you!
Antonio

[glenn-jocher]

@antonio-napolitano to resume training specifically for train3, set the resume argument to True and specify the path to the last checkpoint file for that run (e.g., .../runs/classify/train3/weights/last.pt). Here's an example:

model = YOLO(".../runs/classify/train3/weights/last.pt")  # Load the last checkpoint
results = model.train(resume=True)  # Resume training

This ensures the model continues training from epoch 342 with its saved state. Let me know if further clarification is needed!

[Boaruzhan]

Hi, thank you for the ultralytics!
Can you please tell me if it is possible to train the yolov8 model with preserving its knowledge about already existing classes and adding two additional classes (80: fire, 81:'') ?
In the documentation I didn't find any description of such a thing

[glenn-jocher]

@Boaruzhan thank you for your question! Yes, you can train a YOLOv8 model while retaining knowledge of existing classes and adding new ones (e.g., fire and another class). This process is called incremental learning. To achieve this, load a pretrained model (e.g., COCO-trained) and fine-tune it with your dataset containing both the original and new classes. Ensure that your dataset includes annotations for all classes you want the model to learn. Let me know if you need more details!

[eladoni1]

Hey~
Would love to ask about freezing the model (in Detection, but also in other tasks) - If I would like to use the backbone's feature extraction, how many layers should I freeze in order to preserve it?

Thanks!

[glenn-jocher]

@eladoni1 to preserve the backbone's feature extraction, you can freeze layers by specifying the freeze argument during training. The number of layers to freeze depends on the architecture and your specific task, but freezing the initial layers or stages of the backbone is common for transfer learning. For example, to freeze the first 10 layers, use freeze=10. More details can be found in the Train Settings documentation. Let me know if you need further clarification!

[eladoni1]

Yea it worked quite fine thanks glenn~

I would like to ask another question though - I trained the nano architecture you created, I am pleased with the results, but I would like to improve the speed of my model (running on odroid N2+ but slowly) - I've seen you mention the 'scales' parameter in the yolo11.yaml - that we can decrease it in order to have faster performace. I have changed the nano scale to smaller in the width dimension -
nc: 80
scales:

[depth, width, max_channels]

n: [0.25, 0.25, 1024] # Here I've Changed from [0.25, 0.5, 1024]

But it did not reduce the execution time of the model. Could you tell me what I did wrong here, or, tell me if there are other methods to reduce the execution time of the model?

Thanks

[glenn-jocher]

For faster inference on edge devices like Odroid N2+, consider these optimizations: 1) Export to TensorRT or ONNX format with model.export(format='engine'), 2) Use FP16/INT8 quantization, 3) Reduce inference imgsz (e.g., 320 instead of 640). Scaling model width/depth in YAML affects parameter count but may require architecture changes for significant speed gains. See Export documentation for deployment optimizations: https://docs.ultralytics.com/modes/export/

[Boaruzhan]

@eladoni1 Hi, can you show the code and the yaml file with which you were able to train the model to a new class and retain the knowledge from the standard yolo model ? What version of ultralytics are you using ?

[serhiynovos]

I'm little big confused about "Resuming Interrupted Trainings". It's clear for me if it was unexpectedly interrupted and I want to continue training. What is not clear for me it's continue training.

For example I trained my model based on pre-trained model with completely different dataset that coco dataset which was used to train original model. Now I have new images for object detection

• Should I merge new images to existing dataset or train my model on only new dataset
• How many epochs is it fine to specify to continue training on dataset. As per my understanding for model to be trained on the same dataset for many epochs can lead that model only will recognize objects on the images which are part of dataset.

[glenn-jocher]

@serhiynovos for new data additions, merge datasets to maintain consistency and prevent catastrophic forgetting. When resuming training, set epochs to your new total desired epochs (original + additional). For example, if you initially trained 100 epochs and want 50 more, set epochs=150 with resume=True. Monitor validation metrics to avoid overfitting - consider reducing learning rate (lr0) when adding limited new data.

For implementation details, see the Resuming Training and Training Settings sections in our docs.

[robert-hardman]

Hi
How does imgsz work in train and predict?
What happens in the yolo architecture if I set the input to 224?
Does yolo resize the input image to 640x640 to match its architecture?
Does it resize the layers and blocks and stride for 224x224 input?

[glenn-jocher]

@robert-hardman yOLO automatically resizes inputs to the specified imgsz during both training and prediction (224x224 in your case). The architecture maintains fixed layer configurations but requires retraining for optimal performance at non-default sizes. For technical details on input processing, see the Train Settings documentation.

[omrir]

Hi,
I performed a training for 100ephocs. The training was ended, but i want to continue for more epochs.

My initial CLI was:
yolo detect train data=coco.yaml model=yolo11m.pt epochs=100 imgsz=640 batch=-1 classes=[0,2] name=run_1 project=person_vehicle device=0

Then i tried:
yolo train resume model=person_vehicle/run_1/weights/last.pt epochs=150 device=0

or:

yolo detect train data=coco.yaml model=person_vehicle/run_1/weights/last.pt epochs=150 imgsz=640 batch=-1 classes=[0,2] name=run_1 project=person_vehicle device=0 resume=True

But i'm getting:
AssertionError: person_vehicle/run_1/weights/last.pt training to 100 epochs is finished, nothing to resume.
Start a new training without resuming, i.e. 'yolo train model=person_vehicle/run_1/weights/last.pt'

How can i continue the training for more epochs without loosing training data like gradients etc.?

I would appreciate your help.

Thanks!

[omrir]

BTW, I'm trying to train on the coco dataset only for person and vehicle classes. Any tips for this kind of training?
Should i start the training from scratch or start for the initial weights? how many epocs? what hyper parmeters and would should i monitor to achieve the best results from the training. I went over the 'training tips page', but I would be happy if you can have more insights on this? or Maybe you're familiar with better dataset for these classes?
Thanks!

[glenn-jocher]

To resume training beyond 100 epochs, use this corrected CLI command:

yolo train resume model=person_vehicle/run_1/weights/last.pt epochs=150

The error occurs because you're mixing detect train and resume arguments. For resuming, simply use yolo train resume with your checkpoint and new epochs count. The training will automatically continue from epoch 100 to 150 while preserving optimizer states.

For class-specific training on person/vehicle:

1. Continue using COCO with classes=[0,2] as you're doing
2. Start from pretrained weights (yolo11m.pt) for better convergence
3. Monitor mAP @omrir and loss curves for performance evaluation
4. Consider using freeze=10 to speed up initial training stages

For commercial use, remember Ultralytics offers Enterprise Licensing at https://ultralytics.com/license.

[omrir]

Thank you glenn,
As I wrote previously, I did try to use 'yolo train resume model=person_vehicle/run_1/weights/last.pt epochs=150' , but is still gave me the AssertionError. Are you sure this should work also when the initial training arrived to the end?

[glenn-jocher]

To resume training after completing 100 epochs, you'll need to start a new training session from your last checkpoint rather than using resume. Try:
yolo detect train data=coco.yaml model=person_vehicle/run_1/weights/last.pt epochs=150

For class-specific training:

1. Monitor validation mAP and train/val loss curves
2. Consider increasing epochs to 300+ with cosine LR scheduler (cos_lr=True)
3. COCO is sufficient, but BDD100K or MOT datasets might offer better vehicle/person specialization

For commercial deployments, explore our Enterprise License at https://ultralytics.com/license.

[ChipOwO]

Excuse me,

I want to convert the trained YOLOv8 segmentation model from a .pt file to a .tflite file. Because I need the compressed file in int8 format, I’ve noticed that the model accuracy drops significantly after compression. Both the training and conversion image sizes are fixed at 256×256, so I’m not sure if the noticeable drop in accuracy is mainly related to the float-to-int conversion.

If part of the issue does stem from that, can I configure model.train() to train with int data types from the start? Or is there some misunderstanding in how I’m setting the training parameters?

Additionally, since you are the designers of the model and have a deeper understanding of how training models work, I have a question to ask. During segmentation training, the training dataset consists of grayscale images with varying sizes. The task is to delineate the lighter, white regions in these grayscale images.

Considering that the file needs to be converted to a low-precision .tflite, from your perspective, should the original images used for training be scaled up using the hyperparameter imgsz, or would it yield better results to add a black border to the bottom right of the images to reach 256×256 in size—assuming no changes to the label coordinates?

Thanks

[glenn-jocher]

@ChipOwO for quantization accuracy drops, ensure you're using post-training quantization with a representative dataset rather than training in int8, as Ultralytics models natively use float32. Regarding image sizing, we recommend using imgsz=256 to maintain consistent aspect ratio scaling rather than padding, as this better preserves feature relationships. For optimal TFLite conversion, refer to the detailed Export documentation.

[e-roe]

Hi,
when I set save_period=100 or save_period=-1 the result is the same, saving every epoch pt file. Is it correct?
Ultralytics version = '8.3.75'
Thank you

[glenn-jocher]

@e-roe thanks for reporting this. The save_period=-1 should disable intermediate checkpoint saving except for the final epoch. This might be a version-specific issue - could you try updating to the latest Ultralytics version and confirm if the behavior persists? If it does, please share a minimum reproducible example via our GitHub Issues using the latest version.

For reference, the current expected behavior is documented in the Train Settings section of our docs.

[e-roe]

Hi Glenn,
I've tried with version '8.3.77' and got the same results as before. Then I noticed that I'm using Weights & Biases (wandb) and tried without it, and everything worked as expected. I think wandb might be the reason why the save_period parameter is not working.
I´m using wandb as described here: https://docs.wandb.ai/guides/integrations/ultralytics/
[]´s

[glenn-jocher]

Thanks for identifying the W&B interaction! This appears to be a specific integration behavior. We'll investigate this compatibility issue. For immediate workarounds, you might want to check W&B's checkpointing settings in their callback configuration. Let us know if you'd like help filing a GitHub Issue to track this.

For reference, the expected save_period functionality is documented in the Train Settings section of our docs.

[ge1mina023]

Does it support distributed training across multiple machines and GPUs?