GETTING_STARTED.md

Getting Started with MindYOLO

This document provides a brief introduction to the usage of built-in command-line tools in MindYOLO.

1. Infer a image

• step1, pick a model and its config file from the Model Zoo, such as ./configs/yolov7/yolov7.yaml

• step2, download the corresponding pre-trained checkpoint from the Model Zoo of each model

• step3, run YOLO object detection with the built-in configs:

  # Run with Ascend (By default)
  python demo/predict.py --config ./configs/yolov7/yolov7.yaml --weight=/path_to_ckpt/WEIGHT.ckpt --image_path /path_to_image/IMAGE.jpg

• Notes:

  • The results will be saved in ./detect_results

  • For more details of the command line arguments, see demo/predict.py -h or look at its source code to understand their behavior.

2. Training & Evaluating

• First, Prepare Dataset

  • Prepare your dataset in YOLO format. If trained with COCO (YOLO format), prepare it from yolov5 or the darknet.

    More Details

      coco/
        {train,val}2017.txt
        annotations/
          instances_{train,val}2017.json
        images/
          {train,val}2017/
              00000001.jpg
              ...
              # image files that are mentioned in the corresponding train/val2017.txt
        labels/
          {train,val}2017/
              00000001.txt
              ...
              # label files that are mentioned in the corresponding train/val2017.txt
    

• Second, Training

  • 1 NPU/CPU:

    python train.py --config ./configs/yolov7/yolov7.yaml 

  • To train a model on 8 NPUs:

    msrun --worker_num=8 --local_worker_num=8 --bind_core=True --log_dir=./yolov7_log python train.py --config ./configs/yolov7/yolov7.yaml  --is_parallel True

• Finally, Evaluating

  • To evaluate a model's performance on 1 NPU/CPU:

    python test.py --config ./configs/yolov7/yolov7.yaml --weight /path_to_ckpt/WEIGHT.ckpt

  • To evaluate a model's performance 8 NPUs:

    msrun --worker_num=8 --local_worker_num=8 --bind_core=True --log_dir=./yolov7_log python test.py --config ./configs/yolov7/yolov7.yaml --weight /path_to_ckpt/WEIGHT.ckpt --is_parallel True

• Notes:

  • The default hyper-parameter is used for 8-card training, and some parameters need to be adjusted in the case of a single card.

  • The default device is Ascend, and you can modify it by specifying 'device_target' as Ascend/CPU, as these are currently supported.

  • For more options, see train/test.py -h.

  • Notice that if you are using msrun startup with 2 devices, please add --bind_core=True to improve performance. For example:

    msrun --bind_core=True --worker_num=2 --local_worker_num=2 --master_port=8118 \
          --log_dir=msrun_log --join=True --cluster_time_out=300 \
          python train.py --config ./configs/yolov7/yolov7.yaml  --is_parallel True

    For more usage of msrun, please reference to MindSpore Docs.

Deployment

See depoly readme.

To use MindYOLO APIs in Your Code

To be supplemented.