Step1. Install YOLOX from source.

# clone code
git clone git@github.com:ApolloAuto/apollo-model-yolox.git

cd apollo-model-yolox

# creat conda env
conda create -n apollo_yolox python=3.8
conda activate apollo_yolox

# install requirements
pip3 install -r requirements.txt

Step1. Download a pretrained model from the benchmark table.

Step2. For example, here we use best_L4_ckpt model:

python tools/demo.py image -n yolox-s -c /path/to/your/best_L4_ckpt.pth --path sample/ --conf 0.25 --nms 0.45 --tsize 640 --save_result --device [cpu/gpu]

then you will find result under path YOLOX_outputs/yolox_s/.

Step1. Prepare KITTI dataset

cd <YOLOX_HOME>
ln -s /path/to/your/KITTI ./datasets/KITTI

Step2. Tools for kitti type datasets We provide tools for KITTI type datasets which can help to trans it to VOC type : readme

Step3. change kitti configs

1. change yolox_voc_s

• class number: 8 to 6
• data_dir=os.path.join(get_yolox_datadir(), “CUSTOMER”) to KITTI

2. change voc_classes.py to KITTI class.
3. modify yolox_voc_s.py todo items.
4. modify voc.py line 119 change jpg to png

• self._imgpath = os.path.join(“%s”, “JPEGImages”, “%s.jpg”) # to png

Step4. Reproduce our results on KITTI:

python3 tools/train.py -f exps/example/yolox_voc/yolox_voc_s.py -d 1 -b 16

or resume

python3 tools/train.py -f exps/example/yolox_voc/yolox_voc_s.py -d 1 -b 16 -c /path/to/your/latest_ckpt.pth --resume

• -d: number of gpu devices
• -b: total batch size, the recommended number for -b is num-gpu * 8
• –fp16: mixed precision training
• –cache: caching imgs into RAM to accelarate training, which need large system RAM.
• -c: checkpoint file path

If you want to visuialize your results, please refer to Visualization guides

python tools/export_onnx.py --input data -n yolox-s -c YOLOX_outputs/yolox_voc_s/latest_ckpt.pth  --output-name yolox.onnx

Multi Machine Training

We also support multi-nodes training. Just add the following args:

• –num_machines: num of your total training nodes
• –machine_rank: specify the rank of each node

Suppose you want to train YOLOX on 2 machines, and your master machines’s IP is 123.123.123.123, use port 12312 and TCP.

On master machine, run

python tools/train.py -n yolox-s -b 128 --dist-url tcp://123.123.123.123:12312 --num_machines 2 --machine_rank 0

On the second machine, run

python tools/train.py -n yolox-s -b 128 --dist-url tcp://123.123.123.123:12312 --num_machines 2 --machine_rank 1

Logging to Weights & Biases

To log metrics, predictions and model checkpoints to W&B use the command line argument --logger wandb and use the prefix “wandb-“ to specify arguments for initializing the wandb run.

python tools/train.py -n yolox-s -d 8 -b 64 --fp16 -o [--cache] --logger wandb wandb-project <project name>
                         yolox-m
                         yolox-l
                         yolox-x

An example wandb dashboard is available here

Others

See more information with the following command:

python -m yolox.tools.train --help

We support batch testing for fast evaluation:

python -m yolox.tools.eval -n  yolox-s -c yolox_s.pth -b 64 --exp_file exps/example/yolox_voc/yolox_voc_s.py -d 8 --conf 0.001 [--fp16] [--fuse]
                               yolox-m
                               yolox-l
                               yolox-x

• –fuse: fuse conv and bn
• -d: number of GPUs used for evaluation. DEFAULT: All GPUs available will be used.
• -b: total batch size across on all GPUs

To reproduce speed test, we use the following command:

python -m yolox.tools.eval -n  yolox-s -c yolox_s.pth -b 1 --exp_file exps/example/yolox_voc/yolox_voc_s.py -d 1 --conf 0.001 --fp16 --fuse
                               yolox-m
                               yolox-l
                               yolox-x

• Training on custom data
• Caching for custom data
• Manipulating training image size
• Assignment visualization
• Freezing model