MMYOLO

https://mmyolo.readthedocs.io/en/latest/
Admin
MMYOLO is an open-source YOLO algorithms toolkit based on PyTorch and MMDetection that provides...
Tokens:323,786
Snippets:1,726
Trust Score:9.6
Update:6 months ago
Show doc for...
Tokens:
Raw
### WandB Visualization Setup

Source: https://mmyolo.readthedocs.io/en/latest/get_started/15_minutes_instance_segmentation.html

Installation, authentication, and configuration steps for integrating WandB visualization.

```bash
pip install wandb
# After running wandb login, enter the API Keys obtained above, and the login is successful.
wandb login
```

```python
visualizer = dict(vis_backends = [dict(type='LocalVisBackend'), dict(type='WandbVisBackend')])
```

```bash
python tools/train.py configs/yolov5/ins_seg/yolov5_ins_s-v61_syncbn_fast_8xb16-300e_balloon_instance.py
```

--------------------------------

### YOLOv5CSPDarknet Model Initialization and Usage Example

Source: https://mmyolo.readthedocs.io/en/latest/_modules/mmyolo/models/backbones/csp_darknet.html

Demonstrates how to initialize and use the YOLOv5CSPDarknet model. This example shows creating a model instance and passing dummy input to get level outputs.

```python
from mmyolo.models import YOLOv5CSPDarknet
import torch
model = YOLOv5CSPDarknet()
model.eval()
inputs = torch.rand(1, 3, 416, 416)
level_outputs = model(inputs)
for level_out in level_outputs:
    print(tuple(level_out.shape))
```

--------------------------------

### Install Wandb for Training Visualization

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/labeling_to_deployment_tutorials.html

Install the wandb library using pip. After installation, log in to your wandb account using your API key to enable visualization.

```bash
pip install wandb
# After running wandb login, enter the API Keys obtained above, and the login is successful.
wandb login

```

--------------------------------

### YOLOXCSPDarknet Example - MMYOLO

Source: https://mmyolo.readthedocs.io/en/latest/_modules/mmyolo/models/backbones/csp_darknet.html

Demonstrates how to instantiate and use the YOLOXCSPDarknet backbone. This example shows creating a model instance, setting it to evaluation mode, and passing random input to get level outputs.

```python
from mmyolo.models import YOLOXCSPDarknet
import torch
model = YOLOXCSPDarknet()
model.eval()
inputs = torch.rand(1, 3, 416, 416)
level_outputs = model(inputs)
for level_out in level_outputs:
    print(tuple(level_out.shape))
```

--------------------------------

### Install Dependencies from Project Directory

Source: https://mmyolo.readthedocs.io/en/latest/get_started/installation.html

Simplified installation command for users already inside the mmyolo project directory.

```bash
cd mmyolo
pip install -U openmim
mim install -r requirements/mminstall.txt
```

--------------------------------

### Install EasyDeploy Dependencies

Source: https://mmyolo.readthedocs.io/en/latest/_downloads/bedb093aa96b0943563f778576c11c5d/15_minutes_object_detection.ipynb

Install the required libraries for model conversion and deployment.

```bash
%pip install onnx
%pip install onnx-simplifier # Install if you want to use simplify
%pip install tensorrt        # If you have GPU environment and need to output TensorRT model you need to continue execution
```

--------------------------------

### Install EasyDeploy Dependencies

Source: https://mmyolo.readthedocs.io/en/latest/get_started/15_minutes_object_detection.html

Install ONNX and optionally ONNX-simplifier and TensorRT based on your environment and needs.

```bash
pip install onnx
pip install onnx-simplifier # Install if you want to use simplify
pip install tensorrt        # If you have GPU environment and need to output TensorRT model you need to continue execution
```

--------------------------------

### Install WandB for Visualization

Source: https://mmyolo.readthedocs.io/en/latest/get_started/15_minutes_object_detection.html

Install the WandB package and log in using your API key to enable real-time visualization during training.

```bash
pip install wandb
# After running wandb login, enter the API Keys obtained above, and the login is successful.
wandb login
```

--------------------------------

### Run Labelme with Example Paths

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/labeling_to_deployment_tutorials.html

Execute the labelme tool with example paths for images and output labels, including flags for autosave and nodata. This command assumes you are in the mmyolo directory.

```bash
cd /path/to/mmyolo
labelme ./data/cat/images --output ./data/cat/labels --autosave --nodata

```

--------------------------------

### Install Tensorboard

Source: https://mmyolo.readthedocs.io/en/latest/get_started/15_minutes_instance_segmentation.html

Command to install the required Tensorboard package.

```bash
pip install tensorboard
```

--------------------------------

### Render documentation

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/contributing.html

Install documentation dependencies and build the HTML documentation locally.

```bash
pip install -r requirements/docs.txt
cd docs/zh_cn/
# or docs/en
make html

```

--------------------------------

### Start Model Training

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/application_examples/ionogram_detection.html

Execute the training script using the specified configuration file.

```bash
python tools/train.py projects/misc/ionogram_detection/yolov5/yolov5_s-v61_fast_1xb96-100e_ionogram.py
```

--------------------------------

### Specific Example of AMP Training Command

Source: https://mmyolo.readthedocs.io/en/latest/common_usage/amp_training.html

A concrete example of how to run the training script with AMP enabled, specifying a configuration file.

```bash
python tools/train.py configs/yolov5/yolov5_s-v61_syncbn_8xb16-300e_coco.py --amp

```

--------------------------------

### Install TensorBoard for Training Visualization

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/labeling_to_deployment_tutorials.html

Install the tensorboard library using pip to enable TensorBoard visualization capabilities.

```bash
pip install tensorboard

```

--------------------------------

### Launch Tensorboard

Source: https://mmyolo.readthedocs.io/en/latest/get_started/15_minutes_instance_segmentation.html

Command to start the Tensorboard server for viewing training metrics.

```bash
tensorboard --logdir=work_dirs/yolov5_ins_s-v61_syncbn_fast_8xb16-300e_balloon_instance
```

--------------------------------

### Install MMYOLO and Dependencies

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/labeling_to_deployment_tutorials.html

Sets up the environment by installing MMYOLO and pycuda within a Docker container.

```bash
export MMYOLO_PATH=/root/workspace/mmyolo # path in the image, which doesn't need to modify
cd ${MMYOLO_PATH}
export MMYOLO_VERSION=$(python -c "import mmyolo.version as v; print(v.__version__)")  # Check the version number of MMYOLO used for training
echo "Using MMYOLO ${MMYOLO_VERSION}"
mim install --no-cache-dir mmyolo==${MMYOLO_VERSION}
pip install --no-cache-dir pycuda==2022.2
```

--------------------------------

### Launch TensorBoard

Source: https://mmyolo.readthedocs.io/en/latest/_downloads/bedb093aa96b0943563f778576c11c5d/15_minutes_object_detection.ipynb

Starts a TensorBoard server to monitor training logs from the specified directory.

```bash
!tensorboard --logdir=work_dirs/yolov5_s-v61_fast_1xb12-40e_cat
```

--------------------------------

### Verify MMYOLO Installation

Source: https://mmyolo.readthedocs.io/en/latest/tutorials/custom_installation.html

Import the mmyolo library and print its version to verify the installation. This is typically done after installing MMYOLO.

```python
import mmyolo
print(mmyolo.__version__)
```

--------------------------------

### Clone and setup repository

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/contributing.html

Clone the forked repository and add the official upstream remote.

```bash
git clone git@github.com:{username}/mmyolo.git
```

```bash
cd mmyolo
git remote add upstream git@github.com:open-mmlab/mmyolo
```

```bash
origin	git@github.com:{username}/mmyolo.git (fetch)
origin	git@github.com:{username}/mmyolo.git (push)
upstream	git@github.com:open-mmlab/mmyolo (fetch)
upstream	git@github.com:open-mmlab/mmyolo (push)
```

--------------------------------

### Install PyTorch

Source: https://mmyolo.readthedocs.io/en/latest/get_started/dependencies.html

Installs PyTorch and torchvision for either GPU or CPU platforms.

```bash
conda install pytorch torchvision -c pytorch
```

```bash
conda install pytorch torchvision cpuonly -c pytorch
```

--------------------------------

### Install MMYOLO from Source

Source: https://mmyolo.readthedocs.io/en/latest/get_started/installation.html

Commands to clone the repository and install MMYOLO in editable mode for development.

```bash
git clone https://github.com/open-mmlab/mmyolo.git
cd mmyolo
# Install albumentations
pip install -r requirements/albu.txt
# Install MMYOLO
mim install -v -e .
```

--------------------------------

### Install Dependencies via MIM

Source: https://mmyolo.readthedocs.io/en/latest/get_started/installation.html

Standard installation commands for MMEngine, MMCV, and MMDetection using the OpenMMLab MIM tool.

```bash
pip install -U openmim
mim install "mmengine>=0.6.0"
mim install "mmcv>=2.0.0rc4,<2.1.0"
mim install "mmdet>=3.0.0,<4.0.0"
```

--------------------------------

### Configure YOLOv5 Plugins

Source: https://mmyolo.readthedocs.io/en/latest/_modules/mmyolo/models/backbones/base_backbone.html

Example usage of the plugins configuration format and the make_stage_plugins method.

```python
>>> plugins=[
...     dict(cfg=dict(type='xxx', arg1='xxx'),
...          stages=(False, True, True, True)),
...     dict(cfg=dict(type='yyy'),
...          stages=(True, True, True, True)),
... ]
>>> model = YOLOv5CSPDarknet()
>>> stage_plugins = model.make_stage_plugins(plugins, 0, setting)
>>> assert len(stage_plugins) == 1
```

--------------------------------

### Execute Training Command

Source: https://mmyolo.readthedocs.io/en/latest/_downloads/bedb093aa96b0943563f778576c11c5d/15_minutes_object_detection.ipynb

Command to start the training process using a specific configuration file.

```bash
!python tools/train.py configs/yolov5/yolov5_s-v61_fast_1xb12-40e_cat.py
```

--------------------------------

### Install sahi

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/visualization.html

Install the sahi library with version 0.11.4 or higher. This is a prerequisite for performing inference on large images.

```bash
pip install -U sahi>=0.11.4

```

--------------------------------

### Visualize and Save Dataset Images

Source: https://mmyolo.readthedocs.io/en/latest/useful_tools/browse_dataset.html

Examples of using browse_dataset.py with various configuration and output options.

```bash
python tools/analysis_tools/browse_dataset.py 'configs/yolov5/yolov5_s-v61_syncbn_8xb16-300e_coco.py' \
                                              --out-dir 'work_dirs/browse_dataset'
```

```bash
python tools/analysis_tools/browse_dataset.py 'configs/yolov5/yolov5_s-v61_syncbn_8xb16-300e_coco.py' \
                                              --out-dir 'work_dirs/browse_dataset' \
                                              --show-interval 10
```

```bash
python tools/analysis_tools/browse_dataset.py 'configs/yolov5/yolov5_s-v61_syncbn_8xb16-300e_coco.py' \
                                              --show-interval 10
```

```bash
python tools/analysis_tools/browse_dataset.py 'configs/yolov5/yolov5_s-v61_syncbn_8xb16-300e_coco.py' \
                                              --out-dir 'work_dirs/browse_dataset' \
                                              --not-show
```

--------------------------------

### Configure pre-commit hooks

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/contributing.html

Install and initialize pre-commit hooks to ensure code style compliance.

```bash
pip install -U pre-commit
pre-commit install

```

--------------------------------

### Configure Optimized Anchors

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/labeling_to_deployment_tutorials.html

Example of the anchor configuration format to be updated in the config file.

```python
anchors = [
    [(68, 69), (154, 91), (143, 162)],  # P3/8
    [(242, 160), (189, 287), (391, 207)],  # P4/16
    [(353, 337), (539, 341), (443, 432)]  # P5/32
]
```

--------------------------------

### Use ConvNeXt-Tiny Backbone from MMClassification in YOLOv5

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/replace_backbone.html

Example configuration to use ConvNeXt-Tiny from MMClassification as a backbone for YOLOv5. Requires installing MMClassification.

```python
_base_ = './yolov5_s-v61_syncbn_8xb16-300e_coco.py'

# please run the command, mim install "mmcls>=1.0.0rc2", to install mmcls

```

--------------------------------

### Install MMYOLO from Source on Google Colab

Source: https://mmyolo.readthedocs.io/en/latest/tutorials/custom_installation.html

Install MMYOLO from its GitHub repository on Google Colab. This involves cloning the repository and installing it in editable mode.

```bash
!git clone https://github.com/open-mmlab/mmyolo.git
%cd mmyolo
!pip install -e .
```

--------------------------------

### Install Conda Environment and PyTorch

Source: https://mmyolo.readthedocs.io/en/latest/get_started/15_minutes_object_detection.html

Create a dedicated conda environment and install PyTorch with torchvision. Use a new environment to avoid conflicts with OpenMMLab 1.0 installations.

```bash
conda create -n mmyolo python=3.8 -y
conda activate mmyolo
# If you have GPU
conda install pytorch torchvision -c pytorch
# If you only have CPU
# conda install pytorch torchvision cpuonly -c pytorch
```

--------------------------------

### Run Deployment Demo

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/labeling_to_deployment_tutorials.html

Execute the `deploy_demo.py` script to perform inference on a sample image using the deployed model. Specify the image, configuration files, output directory, and device.

```bash
cd ${MMYOLO_PATH}/demo
python deploy_demo.py \
    ${MMYOLO_PATH}/data/cat/images/mmexport1633684900217.jpg \
    ${MMYOLO_PATH}/configs/custom_dataset/yolov6_s_syncbn_fast_1xb32-100e_cat.py \
    /root/workspace/mmdeploy/work_dir/yolov6_s_syncbn_fast_1xb32-100e_cat_deploy_dynamic_fp16/end2end.engine \
    --deploy-cfg ${MMYOLO_PATH}/configs/deploy/detection_tensorrt-fp16_dynamic-192x192-960x960.py \
    --out-dir ${MMYOLO_PATH}/work_dirs/deploy_predict_out \
    --device cuda:0 \
    --score-thr 0.5

```

--------------------------------

### Verify PyTorch installation

Source: https://mmyolo.readthedocs.io/en/latest/get_started/dependencies.html

Checks the installed PyTorch version and CUDA availability.

```bash
python -c "import torch; print(torch.__version__); print(torch.cuda.is_available())"
```

--------------------------------

### Download Config and Checkpoint

Source: https://mmyolo.readthedocs.io/en/latest/get_started/installation.html

Download the necessary configuration and model weights for the inference demo.

```bash
mim download mmyolo --config yolov5_s-v61_syncbn_fast_8xb16-300e_coco --dest .
```

--------------------------------

### Install Grad-CAM Package

Source: https://mmyolo.readthedocs.io/en/latest/_downloads/bedb093aa96b0943563f778576c11c5d/15_minutes_object_detection.ipynb

Installs the necessary grad-cam dependency for box-level visualization.

```bash
%pip install "grad-cam"
```

--------------------------------

### Install PyTorch for CPU only

Source: https://mmyolo.readthedocs.io/en/latest/get_started/15_minutes_instance_segmentation.html

Install PyTorch and Torchvision for CPU-only environments using conda.

```bash
# conda install pytorch torchvision cpuonly -c pytorch
```

--------------------------------

### Initialize and Run YOLOXCSPDarknet

Source: https://mmyolo.readthedocs.io/en/latest/api.html

Demonstrates how to instantiate the YOLOXCSPDarknet model, set it to evaluation mode, and perform a forward pass with dummy input data.

```python
>>> from mmyolo.models import YOLOXCSPDarknet
>>> import torch
>>> model = YOLOXCSPDarknet()
>>> model.eval()
>>> inputs = torch.rand(1, 3, 416, 416)
>>> level_outputs = model(inputs)
>>> for level_out in level_outputs:
...     print(tuple(level_out.shape))
...
(1, 256, 52, 52)
(1, 512, 26, 26)
(1, 1024, 13, 13)
```

--------------------------------

### Install MMYOLO via MIM

Source: https://mmyolo.readthedocs.io/en/latest/get_started/installation.html

Command to install MMYOLO as a third-party package using MIM.

```bash
mim install "mmyolo"
```

--------------------------------

### Install MMEngine with Pip

Source: https://mmyolo.readthedocs.io/en/latest/tutorials/custom_installation.html

Install MMEngine using pip. Ensure you have a version greater than or equal to 0.6.0.

```bash
pip install "mmengine>=0.6.0"
```

--------------------------------

### Verify PyTorch Installation

Source: https://mmyolo.readthedocs.io/en/latest/_downloads/bedb093aa96b0943563f778576c11c5d/15_minutes_object_detection.ipynb

Simple Python script to check the installed PyTorch version and GPU availability.

```python
# Check PyTorch version
import torch
print(torch.__version__)
print(torch.cuda.is_available())
```

--------------------------------

### Configure TensorBoard Visualization

Source: https://mmyolo.readthedocs.io/en/latest/_downloads/bedb093aa96b0943563f778576c11c5d/15_minutes_object_detection.ipynb

Installs TensorBoard, updates the configuration file to enable TensorBoard logging, and provides the command to launch the dashboard.

```bash
%pip install tensorboard
```

```python
visualizer = dict(vis_backends=[dict(type='LocalVisBackend'),dict(type='TensorboardVisBackend')])
```

--------------------------------

### Configure WandB Visualization

Source: https://mmyolo.readthedocs.io/en/latest/_downloads/bedb093aa96b0943563f778576c11c5d/15_minutes_object_detection.ipynb

Installs the WandB client, authenticates the user, and updates the configuration file to enable WandB logging.

```bash
%pip install wandb
!wandb login
```

```python
visualizer = dict(vis_backends = [dict(type='LocalVisBackend'), dict(type='WandbVisBackend')])
```

```bash
!python tools/train.py configs/yolov5/yolov5_s-v61_fast_1xb12-40e_cat.py
```

--------------------------------

### Install PyTorch with GPU support

Source: https://mmyolo.readthedocs.io/en/latest/get_started/15_minutes_instance_segmentation.html

Install PyTorch and Torchvision using conda. This command assumes you have a GPU available.

```bash
conda install pytorch torchvision -c pytorch
```

--------------------------------

### YOLOXCSPDarknet Initialization and Usage

Source: https://mmyolo.readthedocs.io/en/latest/_modules/mmyolo/models/backbones/csp_darknet.html

Demonstrates how to initialize and use the YOLOXCSPDarknet model with a sample input.

```APIDOC
## YOLOXCSPDarknet

### Description
CSP-Darknet backbone used in YOLOX.

### Args
- **arch** (str): Architecture of CSP-Darknet, from {P5, P6}. Defaults to P5.
- **plugins** (list[dict]): List of plugins for stages, each dict contains:
    - cfg (dict, required): Cfg dict to build plugin.
    - stages (tuple[bool], optional): Stages to apply plugin, length should be same as 'num_stages'.
- **deepen_factor** (float): Depth multiplier, multiply number of blocks in CSP layer by this amount. Defaults to 1.0.
- **widen_factor** (float): Width multiplier, multiply number of channels in each layer by this amount. Defaults to 1.0.
- **input_channels** (int): Number of input image channels. Defaults to 3.
- **out_indices** (Tuple[int]): Output from which stages. Defaults to (2, 3, 4).
- **frozen_stages** (int): Stages to be frozen (stop grad and set eval mode). -1 means not freezing any parameters. Defaults to -1.
- **use_depthwise** (bool): Whether to use depthwise separable convolution. Defaults to False.
- **spp_kernal_sizes** (tuple[int]): Sequential of kernel sizes of SPP layers. Defaults to (5, 9, 13).
- **norm_cfg** (dict): Dictionary to construct and config norm layer. Defaults to dict(type='BN', momentum=0.03, eps=0.001).
- **act_cfg** (dict): Config dict for activation layer. Defaults to dict(type='SiLU', inplace=True).
- **norm_eval** (bool): Whether to set norm layers to eval mode, namely, freeze running stats (mean and var). Note: Effect on Batch Norm and its variants only.
- **init_cfg** (Union[dict,list[dict]], optional): Initialization config dict. Defaults to None.

### Example
```python
from mmyolo.models import YOLOXCSPDarknet
import torch

model = YOLOXCSPDarknet()
model.eval()
inputs = torch.rand(1, 3, 416, 416)
level_outputs = model(inputs)
for level_out in level_outputs:
    print(tuple(level_out.shape))
```

### Output Example
```
(1, 256, 52, 52)
(1, 512, 26, 26)
(1, 1024, 13, 13)
```
```

--------------------------------

### Run Inference Demo from Source

Source: https://mmyolo.readthedocs.io/en/latest/get_started/installation.html

Execute the image inference script using the downloaded configuration and checkpoint files.

```bash
python demo/image_demo.py demo/demo.jpg \
                          yolov5_s-v61_syncbn_fast_8xb16-300e_coco.py \
                          yolov5_s-v61_syncbn_fast_8xb16-300e_coco_20220918_084700-86e02187.pth
```

--------------------------------

### Clone MMYOLO repository and install dependencies

Source: https://mmyolo.readthedocs.io/en/latest/get_started/15_minutes_instance_segmentation.html

Clone the MMYOLO GitHub repository and install necessary dependencies including openmim, albumentations, and MMYOLO itself in editable mode.

```bash
git clone https://github.com/open-mmlab/mmyolo.git
cd mmyolo
pip install -U openmim
mim install -r requirements/mminstall.txt
# Install albumentations
mim install -r requirements/albu.txt
# Install MMYOLO
mim install -v -e .
```

--------------------------------

### Install MMYOLO and Dependencies

Source: https://mmyolo.readthedocs.io/en/latest/_downloads/bedb093aa96b0943563f778576c11c5d/15_minutes_object_detection.ipynb

Commands to clone the repository and install required dependencies using OpenMIM. Editable mode is used to allow local code modifications.

```bash
!git clone https://github.com/open-mmlab/mmyolo.git
%cd mmyolo
%pip install -U openmim
!mim install -r requirements/mminstall.txt
# Install albumentations
!mim install -r requirements/albu.txt
# Install MMYOLO
!mim install -v -e .
# "-v" means verbose, or more output
# "-e" means installing a project in editable mode,
# thus any local modifications made to the code will take effect without reinstallation.
```

--------------------------------

### Initialize and use YOLOv5CSPDarknet

Source: https://mmyolo.readthedocs.io/en/latest/api.html

Demonstrates how to instantiate the YOLOv5CSPDarknet backbone and perform a forward pass with dummy input data.

```python
>>> from mmyolo.models import YOLOv5CSPDarknet
>>> import torch
>>> model = YOLOv5CSPDarknet()
>>> model.eval()
>>> inputs = torch.rand(1, 3, 416, 416)
>>> level_outputs = model(inputs)
>>> for level_out in level_outputs:
...     print(tuple(level_out.shape))
...
(1, 256, 52, 52)
(1, 512, 26, 26)
(1, 1024, 13, 13)
```

--------------------------------

### Launch TensorBoard to View Training Data

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/labeling_to_deployment_tutorials.html

Run this command in your terminal to launch TensorBoard, which will serve visualizations from the specified log directory. This allows viewing loss, learning rate, and mAP metrics.

```bash
tensorboard --logdir=work_dirs/yolov5_s-v61_syncbn_fast_1xb32-100e_cat

```

--------------------------------

### Install Labelme Annotation Tool

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/labeling_to_deployment_tutorials.html

Set up a conda environment and install the labelme tool. It is recommended to use Python 3.8 and a specific version of labelme (5.1.1) for compatibility.

```bash
conda create -n labelme python=3.8
conda activate labelme
pip install labelme==5.1.1

```

--------------------------------

### Start Labelme Annotation Tool

Source: https://mmyolo.readthedocs.io/en/latest/recommended_topics/labeling_to_deployment_tutorials.html

Launch the labelme application with the specified image directory and output directory for label files. The `--autosave` flag enables automatic saving, and `--nodata` reduces file size by not encoding images.

```bash
labelme ${image dir path (same as the previous step)} \
        --output ${the dir path of label file（same as --out-dir）} \
        --autosave \
        --nodata

```

--------------------------------

### Install MMEngine, MMCV, MMDetection, and MMYOLO on Google Colab

Source: https://mmyolo.readthedocs.io/en/latest/tutorials/custom_installation.html

Install necessary OpenMMLab packages on Google Colab using MIM. This includes MMEngine, MMCV, and MMDetection.

```bash
!pip3 install openmim
!mim install "mmengine>=0.6.0"
!mim install "mmcv>=2.0.0rc4,<2.1.0"
!mim install "mmdet>=3.0.0,<4.0.0"
```
Tokens:
Raw