HongChuang/ultralytics-ascend
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 3

Update YOLO11 Actions and Docs (#16596)

Browse source 
Signed-off-by: UltralyticsAssistant <web@ultralytics.com>
This commit is contained in:
Ultralytics Assistant 2024-10-01 16:58:12 +02:00 • committed by GitHub
parent 51e93d6111
commit 97f38409fb
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
124 changed files with 1948 additions and 1948 deletions
Download patch file Download diff file Expand all files Collapse all files

78
docs/en/integrations/tensorboard.md
View file

@ -1,14 +1,14 @@
---
comments: true
description: Learn how to integrate YOLOv8 with TensorBoard for real-time visual insights into your model's training metrics, performance graphs, and debugging workflows.
keywords: YOLOv8, TensorBoard, model training, visualization, machine learning, deep learning, Ultralytics, training metrics, performance analysis
description: Learn how to integrate YOLO11 with TensorBoard for real-time visual insights into your model's training metrics, performance graphs, and debugging workflows.
keywords: YOLO11, TensorBoard, model training, visualization, machine learning, deep learning, Ultralytics, training metrics, performance analysis
---
# Gain Visual Insights with YOLOv8's Integration with TensorBoard
# Gain Visual Insights with YOLO11's Integration with TensorBoard
Understanding and fine-tuning [computer vision](https://www.ultralytics.com/glossary/computer-vision-cv) models like [Ultralytics' YOLOv8](https://www.ultralytics.com/) becomes more straightforward when you take a closer look at their training processes. Model training visualization helps with getting insights into the model's learning patterns, performance metrics, and overall behavior. YOLOv8's integration with TensorBoard makes this process of visualization and analysis easier and enables more efficient and informed adjustments to the model.
Understanding and fine-tuning [computer vision](https://www.ultralytics.com/glossary/computer-vision-cv) models like [Ultralytics' YOLO11](https://www.ultralytics.com/) becomes more straightforward when you take a closer look at their training processes. Model training visualization helps with getting insights into the model's learning patterns, performance metrics, and overall behavior. YOLO11's integration with TensorBoard makes this process of visualization and analysis easier and enables more efficient and informed adjustments to the model.
This guide covers how to use TensorBoard with YOLOv8. You'll learn about various visualizations, from tracking metrics to analyzing model graphs. These tools will help you understand your YOLOv8 model's performance better.
This guide covers how to use TensorBoard with YOLO11. You'll learn about various visualizations, from tracking metrics to analyzing model graphs. These tools will help you understand your YOLO11 model's performance better.
## TensorBoard
@ -18,9 +18,9 @@ This guide covers how to use TensorBoard with YOLOv8. You'll learn about various
[TensorBoard](https://www.tensorflow.org/tensorboard), [TensorFlow](https://www.ultralytics.com/glossary/tensorflow)'s visualization toolkit, is essential for [machine learning](https://www.ultralytics.com/glossary/machine-learning-ml) experimentation. TensorBoard features a range of visualization tools, crucial for monitoring machine learning models. These tools include tracking key metrics like loss and accuracy, visualizing model graphs, and viewing histograms of weights and biases over time. It also provides capabilities for projecting [embeddings](https://www.ultralytics.com/glossary/embeddings) to lower-dimensional spaces and displaying multimedia data.
## YOLOv8 Training with TensorBoard
## YOLO11 Training with TensorBoard
Using TensorBoard while training YOLOv8 models is straightforward and offers significant benefits.
Using TensorBoard while training YOLO11 models is straightforward and offers significant benefits.
## Installation
@ -31,13 +31,13 @@ To install the required package, run:
=== "CLI"
```bash
# Install the required package for YOLOv8 and Tensorboard
# Install the required package for YOLO11 and Tensorboard
pip install ultralytics
```
TensorBoard is conveniently pre-installed with YOLOv8, eliminating the need for additional setup for visualization purposes.
TensorBoard is conveniently pre-installed with YOLO11, eliminating the need for additional setup for visualization purposes.
For detailed instructions and best practices related to the installation process, be sure to check our [YOLOv8 Installation guide](../quickstart.md). While installing the required packages for YOLOv8, if you encounter any difficulties, consult our [Common Issues guide](../guides/yolo-common-issues.md) for solutions and tips.
For detailed instructions and best practices related to the installation process, be sure to check our [YOLO11 Installation guide](../quickstart.md). While installing the required packages for YOLO11, if you encounter any difficulties, consult our [Common Issues guide](../guides/yolo-common-issues.md) for solutions and tips.
## Configuring TensorBoard for Google Colab
@ -54,7 +54,7 @@ When using Google Colab, it's important to set up TensorBoard before starting yo
## Usage
Before diving into the usage instructions, be sure to check out the range of [YOLOv8 models offered by Ultralytics](../models/index.md). This will help you choose the most appropriate model for your project requirements.
Before diving into the usage instructions, be sure to check out the range of [YOLO11 models offered by Ultralytics](../models/index.md). This will help you choose the most appropriate model for your project requirements.
!!! example "Usage"
@ -64,7 +64,7 @@ Before diving into the usage instructions, be sure to check out the range of [YO
from ultralytics import YOLO
# Load a pre-trained model
model = YOLO("yolov8n.pt")
model = YOLO("yolo11n.pt")
# Train the model
results = model.train(data="coco8.yaml", epochs=100, imgsz=640)
@ -76,17 +76,17 @@ Upon running the usage code snippet above, you can expect the following output:
TensorBoard: Start with 'tensorboard --logdir path_to_your_tensorboard_logs', view at http://localhost:6006/
```
This output indicates that TensorBoard is now actively monitoring your YOLOv8 training session. You can access the TensorBoard dashboard by visiting the provided URL (http://localhost:6006/) to view real-time training metrics and model performance. For users working in Google Colab, the TensorBoard will be displayed in the same cell where you executed the TensorBoard configuration commands.
This output indicates that TensorBoard is now actively monitoring your YOLO11 training session. You can access the TensorBoard dashboard by visiting the provided URL (http://localhost:6006/) to view real-time training metrics and model performance. For users working in Google Colab, the TensorBoard will be displayed in the same cell where you executed the TensorBoard configuration commands.
For more information related to the model training process, be sure to check our [YOLOv8 Model Training guide](../modes/train.md). If you are interested in learning more about logging, checkpoints, plotting, and file management, read our [usage guide on configuration](../usage/cfg.md).
For more information related to the model training process, be sure to check our [YOLO11 Model Training guide](../modes/train.md). If you are interested in learning more about logging, checkpoints, plotting, and file management, read our [usage guide on configuration](../usage/cfg.md).
## Understanding Your TensorBoard for YOLOv8 Training
## Understanding Your TensorBoard for YOLO11 Training
Now, let's focus on understanding the various features and components of TensorBoard in the context of YOLOv8 training. The three key sections of the TensorBoard are Time Series, Scalars, and Graphs.
Now, let's focus on understanding the various features and components of TensorBoard in the context of YOLO11 training. The three key sections of the TensorBoard are Time Series, Scalars, and Graphs.
### Time Series
The Time Series feature in the TensorBoard offers a dynamic and detailed perspective of various training metrics over time for YOLOv8 models. It focuses on the progression and trends of metrics across training epochs. Here's an example of what you can expect to see.
The Time Series feature in the TensorBoard offers a dynamic and detailed perspective of various training metrics over time for YOLO11 models. It focuses on the progression and trends of metrics across training epochs. Here's an example of what you can expect to see.
![image](https://github.com/ultralytics/docs/releases/download/0/time-series-tensorboard-yolov8.avif)
@ -100,13 +100,13 @@ The Time Series feature in the TensorBoard offers a dynamic and detailed perspec
- **In-Depth Analysis**: Time Series provides an in-depth analysis of each metric. For instance, different learning rate segments are shown, offering insights into how adjustments in learning rate impact the model's learning curve.
#### Importance of Time Series in YOLOv8 Training
#### Importance of Time Series in YOLO11 Training
The Time Series section is essential for a thorough analysis of the YOLOv8 model's training progress. It lets you track the metrics in real time to promptly identify and solve issues. It also offers a detailed view of each metrics progression, which is crucial for fine-tuning the model and enhancing its performance.
The Time Series section is essential for a thorough analysis of the YOLO11 model's training progress. It lets you track the metrics in real time to promptly identify and solve issues. It also offers a detailed view of each metrics progression, which is crucial for fine-tuning the model and enhancing its performance.
### Scalars
Scalars in the TensorBoard are crucial for plotting and analyzing simple metrics like loss and accuracy during the training of YOLOv8 models. They offer a clear and concise view of how these metrics evolve with each training [epoch](https://www.ultralytics.com/glossary/epoch), providing insights into the model's learning effectiveness and stability. Here's an example of what you can expect to see.
Scalars in the TensorBoard are crucial for plotting and analyzing simple metrics like loss and accuracy during the training of YOLO11 models. They offer a clear and concise view of how these metrics evolve with each training [epoch](https://www.ultralytics.com/glossary/epoch), providing insights into the model's learning effectiveness and stability. Here's an example of what you can expect to see.
![image](https://github.com/ultralytics/docs/releases/download/0/scalars-metrics-tensorboard.avif)
@ -130,7 +130,7 @@ Scalars in the TensorBoard are crucial for plotting and analyzing simple metrics
#### Importance of Monitoring Scalars
Observing scalar metrics is crucial for fine-tuning the YOLOv8 model. Variations in these metrics, such as spikes or irregular patterns in loss graphs, can highlight potential issues such as [overfitting](https://www.ultralytics.com/glossary/overfitting), [underfitting](https://www.ultralytics.com/glossary/underfitting), or inappropriate learning rate settings. By closely monitoring these scalars, you can make informed decisions to optimize the training process, ensuring that the model learns effectively and achieves the desired performance.
Observing scalar metrics is crucial for fine-tuning the YOLO11 model. Variations in these metrics, such as spikes or irregular patterns in loss graphs, can highlight potential issues such as [overfitting](https://www.ultralytics.com/glossary/overfitting), [underfitting](https://www.ultralytics.com/glossary/underfitting), or inappropriate learning rate settings. By closely monitoring these scalars, you can make informed decisions to optimize the training process, ensuring that the model learns effectively and achieves the desired performance.
### Difference Between Scalars and Time Series
@ -138,15 +138,15 @@ While both Scalars and Time Series in TensorBoard are used for tracking metrics,
### Graphs
The Graphs section of the TensorBoard visualizes the computational graph of the YOLOv8 model, showing how operations and data flow within the model. It's a powerful tool for understanding the model's structure, ensuring that all layers are connected correctly, and for identifying any potential bottlenecks in data flow. Here's an example of what you can expect to see.
The Graphs section of the TensorBoard visualizes the computational graph of the YOLO11 model, showing how operations and data flow within the model. It's a powerful tool for understanding the model's structure, ensuring that all layers are connected correctly, and for identifying any potential bottlenecks in data flow. Here's an example of what you can expect to see.
![image](https://github.com/ultralytics/docs/releases/download/0/tensorboard-yolov8-computational-graph.avif)
Graphs are particularly useful for debugging the model, especially in complex architectures typical in [deep learning](https://www.ultralytics.com/glossary/deep-learning-dl) models like YOLOv8. They help in verifying layer connections and the overall design of the model.
Graphs are particularly useful for debugging the model, especially in complex architectures typical in [deep learning](https://www.ultralytics.com/glossary/deep-learning-dl) models like YOLO11. They help in verifying layer connections and the overall design of the model.
## Summary
This guide aims to help you use TensorBoard with YOLOv8 for visualization and analysis of machine learning model training. It focuses on explaining how key TensorBoard features can provide insights into training metrics and model performance during YOLOv8 training sessions.
This guide aims to help you use TensorBoard with YOLO11 for visualization and analysis of machine learning model training. It focuses on explaining how key TensorBoard features can provide insights into training metrics and model performance during YOLO11 training sessions.
For a more detailed exploration of these features and effective utilization strategies, you can refer to TensorFlow's official [TensorBoard documentation](https://www.tensorflow.org/tensorboard/get_started) and their [GitHub repository](https://github.com/tensorflow/tensorboard).
@ -154,29 +154,29 @@ Want to learn more about the various integrations of Ultralytics? Check out the
## FAQ
### What benefits does using TensorBoard with YOLOv8 offer?
### What benefits does using TensorBoard with YOLO11 offer?
Using TensorBoard with YOLOv8 provides several visualization tools essential for efficient model training:
Using TensorBoard with YOLO11 provides several visualization tools essential for efficient model training:
- **Real-Time Metrics Tracking:** Track key metrics such as loss, accuracy, precision, and recall live.
- **Model Graph Visualization:** Understand and debug the model architecture by visualizing computational graphs.
- **Embedding Visualization:** Project embeddings to lower-dimensional spaces for better insight.
These tools enable you to make informed adjustments to enhance your YOLOv8 model's performance. For more details on TensorBoard features, check out the TensorFlow [TensorBoard guide](https://www.tensorflow.org/tensorboard/get_started).
These tools enable you to make informed adjustments to enhance your YOLO11 model's performance. For more details on TensorBoard features, check out the TensorFlow [TensorBoard guide](https://www.tensorflow.org/tensorboard/get_started).
### How can I monitor training metrics using TensorBoard when training a YOLOv8 model?
### How can I monitor training metrics using TensorBoard when training a YOLO11 model?
To monitor training metrics while training a YOLOv8 model with TensorBoard, follow these steps:
To monitor training metrics while training a YOLO11 model with TensorBoard, follow these steps:
1. **Install TensorBoard and YOLOv8:** Run `pip install ultralytics` which includes TensorBoard.
2. **Configure TensorBoard Logging:** During the training process, YOLOv8 logs metrics to a specified log directory.
1. **Install TensorBoard and YOLO11:** Run `pip install ultralytics` which includes TensorBoard.
2. **Configure TensorBoard Logging:** During the training process, YOLO11 logs metrics to a specified log directory.
3. **Start TensorBoard:** Launch TensorBoard using the command `tensorboard --logdir path/to/your/tensorboard/logs`.
The TensorBoard dashboard, accessible via [http://localhost:6006/](http://localhost:6006/), provides real-time insights into various training metrics. For a deeper dive into training configurations, visit our [YOLOv8 Configuration guide](../usage/cfg.md).
The TensorBoard dashboard, accessible via [http://localhost:6006/](http://localhost:6006/), provides real-time insights into various training metrics. For a deeper dive into training configurations, visit our [YOLO11 Configuration guide](../usage/cfg.md).
### What kind of metrics can I visualize with TensorBoard when training YOLOv8 models?
### What kind of metrics can I visualize with TensorBoard when training YOLO11 models?
When training YOLOv8 models, TensorBoard allows you to visualize an array of important metrics including:
When training YOLO11 models, TensorBoard allows you to visualize an array of important metrics including:
- **Loss (Training and Validation):** Indicates how well the model is performing during training and validation.
- **Accuracy/Precision/[Recall](https://www.ultralytics.com/glossary/recall):** Key performance metrics to evaluate detection accuracy.
@ -185,9 +185,9 @@ When training YOLOv8 models, TensorBoard allows you to visualize an array of imp
These visualizations are essential for tracking model performance and making necessary optimizations. For more information on these metrics, refer to our [Performance Metrics guide](../guides/yolo-performance-metrics.md).
### Can I use TensorBoard in a Google Colab environment for training YOLOv8?
### Can I use TensorBoard in a Google Colab environment for training YOLO11?
Yes, you can use TensorBoard in a Google Colab environment to train YOLOv8 models. Here's a quick setup:
Yes, you can use TensorBoard in a Google Colab environment to train YOLO11 models. Here's a quick setup:
!!! example "Configure TensorBoard for Google Colab"
@ -198,16 +198,16 @@ Yes, you can use TensorBoard in a Google Colab environment to train YOLOv8 model
%tensorboard --logdir path/to/runs
```
Then, run the YOLOv8 training script:
Then, run the YOLO11 training script:
```python
from ultralytics import YOLO
# Load a pre-trained model
model = YOLO("yolov8n.pt")
model = YOLO("yolo11n.pt")
# Train the model
results = model.train(data="coco8.yaml", epochs=100, imgsz=640)
```
TensorBoard will visualize the training progress within Colab, providing real-time insights into metrics like loss and accuracy. For additional details on configuring YOLOv8 training, see our detailed [YOLOv8 Installation guide](../quickstart.md).
TensorBoard will visualize the training progress within Colab, providing real-time insights into metrics like loss and accuracy. For additional details on configuring YOLO11 training, see our detailed [YOLO11 Installation guide](../quickstart.md).