YOLOX

Integrate W&B with YOLOX to track object detection model training, log metrics, and visualize detection results.

YOLOX is an anchor-free version of YOLO for object detection. You can use the YOLOX W&B integration to turn on logging of metrics related to training, validation, and the system, and to interactively validate predictions with a single command-line argument.

This guide shows you how to authenticate with W&B, install the integration, and enable W&B logging when you train a YOLOX object detection model so you can track metrics and inspect predictions in the W&B UI.

Sign up and create an API key

An API key authenticates your machine to W&B. You can generate an API key from your user profile.

For a more streamlined approach, go to User Settings and create an API key. Copy the API key immediately and save it in a secure location such as a password manager.

1. Click your user profile icon in the upper right corner.
2. Select User Settings, then scroll to the API Keys section.

Install the wandb library and log in

To install the wandb library locally and log in:

Command line

1. Set the WANDB_API_KEY environment variable to your API key. Replace values enclosed in <> with your own:

   export WANDB_API_KEY=<your_api_key>
   

2. Install the wandb library and log in.

   pip install wandb
   
   wandb login
   

Python

pip install wandb

import wandb
wandb.login()

Python notebook

!pip install wandb

import wandb
wandb.login()

Log metrics

With the wandb library installed and your machine authenticated, you can enable W&B logging from the YOLOX training script.

Use the --logger wandb command-line argument to turn on logging with wandb. Optionally, you can also pass all of the arguments that wandb.init() expects. Prepend each argument with wandb-.

num_eval_imges controls the number of validation set images and predictions that W&B logs to tables for model evaluation.

Replace the following placeholders before you run the command:

• <project-name>: The name of your W&B project.
• <entity>: Your W&B entity (username or team name).
• <run-name>: A name for this training run.
• <run-id>: A unique identifier for this run.
• <save-dir>: The directory where YOLOX saves checkpoints and logs.
• <num-images>: The number of validation images to log.
• <bool>: Whether to log checkpoints (true or false).

# Log in to W&B
wandb login

# Call your YOLOX training script with the wandb logger argument
python tools/train.py .... --logger wandb \
                wandb-project <project-name> \
                wandb-entity <entity> \
                wandb-name <run-name> \
                wandb-id <run-id> \
                wandb-save_dir <save-dir> \
                wandb-num_eval_imges <num-images> \
                wandb-log_checkpoints <bool>

Example

After your training run starts, YOLOX streams training, validation, and system metrics to your W&B project, where you can compare runs and inspect predictions. See the following example for what a populated dashboard looks like.

See the example dashboard with YOLOX training and validation metrics.

If you have questions or issues about this W&B integration, open an issue in the YOLOX repository.