PaddleDetection is an end-to-end object-detection development kit based on PaddlePaddle. It implements varied mainstream object detection, instance segmentation, tracking and keypoint detection algorithms in modular design with configurable modules such as network components, data augmentations and losses.
PaddleDetection now comes with a built in W&B integration which logs all your training and validation metrics, as well as your model checkpoints and their corresponding metadata.
Example Blog and Colab
Read our blog here to see how to train a YOLOX model with PaddleDetection on a subset of the COCO2017 dataset. This also comes with a Google Colab and the corresponding live W&B dashboard is available here
The PaddleDetection WandbLogger
The PaddleDetection WandbLogger will log your training and evaluation metrics to Weights & Biases as well as your model checkpoints while training.
Using PaddleDetection with Weights & Biases
Sign up and Login to W&B
Sign up for a free Weights & Biases account, then pip install the wandb library. To login, you'll need to be signed in to you account at www.wandb.ai. Once signed in you will find your API key on the Authorize page.
- Command Line
pip install wandb
!pip install wandb
Activating the WandbLogger in your Training Script
Using the CLI
To use wandb via arguments to
train.py in PaddleDetection:
- Add the
- The first wandb arguments must be preceded by
-o(you only need to pass this once)
- Each individual wandb argument must contain the prefix
wandb-. For example any argument to be passed to
wandb.initwould get the
-c config.yml \
Using a config.yml file
You can also activate wandb via the config file. Add the wandb arguments to the config.yml file under the wandb header like so:
Once you run your
train.py file with Weights & Biases turned on, a link will be generated to bring you to your W&B dashboard: