实现pytorch版efficientdet的全过程

一、安装环境

# install requirements
pip install pycocotools numpy opencv-python tqdm tensorboard tensorboardX pyyaml
pip install torch==1.4.0
pip install torchvision==0.5.0

二、下载pytorch版efficientdet源码

git clone https://github.com/zylo117/Yet-Another-EfficientDet-Pytorch.git

源码链接：https://github.com/zylo117/Yet-Another-EfficientDet-Pytorch

三、准备数据集

# your dataset structure should be like this
datasets/
    -your_project_name/
        -train_set_name/
            -*.jpg
        -val_set_name/
            -*.jpg
        -annotations
            -instances_{train_set_name}.json
            -instances_{val_set_name}.json

# for example, coco2017
datasets/
    -coco2017/
        -train2017/
            -000000000001.jpg
            -000000000002.jpg
            -000000000003.jpg
        -val2017/
            -000000000004.jpg
            -000000000005.jpg
            -000000000006.jpg
        -annotations
            -instances_train2017.json
            -instances_val2017.json

四、修改配置文件

# create a yml file {your_project_name}.yml under 'projects'folder 
# modify it following 'coco.yml'
 
# for example
project_name: coco
train_set: train2017
val_set: val2017
num_gpus: 4  # 0 means using cpu, 1-N means using gpus 

# mean and std in RGB order, actually this part should remain unchanged as long as your dataset is similar to coco.
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]

# this is coco anchors, change it if necessary
anchors_scales: '[2 ** 0, 2 ** (1.0 / 3.0), 2 ** (2.0 / 3.0)]'
anchors_ratios: '[(1.0, 1.0), (1.4, 0.7), (0.7, 1.4)]'

# objects from all labels from your dataset with the order from your annotations.
# its index must match your dataset's category_id.
# category_id is one_indexed,
# for example, index of 'car' here is 2, while category_id of is 3
obj_list: ['person', 'bicycle', 'car', ...]

五、训练coco数据集

# train efficientdet-d1 on a custom dataset 
# with batchsize 8 and learning rate 1e-5

python train.py -c 1 -p your_project_name --batch_size 8 --lr 1e-5

六、训练带已训练好的权重的数据集

# train efficientdet-d2 on a custom dataset with pretrained weights
# with batchsize 8 and learning rate 1e-5 for 10 epoches

python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-5 --num_epochs 10 \
 --load_weights /path/to/your/weights/efficientdet-d2.pth

# with a coco-pretrained, you can even freeze the backbone and train heads only
# to speed up training and help convergence.

python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-5 --num_epochs 10 \
 --load_weights /path/to/your/weights/efficientdet-d2.pth \
 --head_only True

权重下载链接：https://github.com/zylo117/Yet-Another-EfficientDet-Pytorch/releases/

七、尽早停止训练

# while training, press Ctrl+c, the program will catch KeyboardInterrupt
# and stop training, save current checkpoint.

八、恢复训练

# let says you started a training session like this.

python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-5 \
 --load_weights /path/to/your/weights/efficientdet-d2.pth \
 --head_only True
 
# then you stopped it with a Ctrl+c, it exited with a checkpoint

# now you want to resume training from the last checkpoint
# simply set load_weights to 'last'

python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-5 \
 --load_weights last \
 --head_only True

九、评估模型性能

# eval on your_project, efficientdet-d5

python coco_eval.py -p your_project_name -c 5 \
 -w /path/to/your/weights

十、调试训练（可选）

# when you get bad result, you need to debug the training result.
python train.py -c 2 -p your_project_name --batch_size 8 --lr 1e-5 --debug True

# then checkout test/ folder, there you can visualize the predicted boxes during training
# don't panic if you see countless of error boxes, it happens when the training is at early stage.
# But if you still can't see a normal box after several epoches, not even one in all image,
# then it's possible that either the anchors config is inappropriate or the ground truth is corrupted.

十一、个人训练总结

最重要的是不放弃！！！

遇到错误，根据错误来源看代码，当然一般按照流程来不会出错。

标签：name,efficientdet,project,python,pytorch,train,weights,全过程,your
来源： https://www.cnblogs.com/cx-99/p/12812613.html