SSD-MobileNetV1·

This implements training of SSD-MobileNetV1 on the VOC dataset, mainly modified from pytorch/examples.

Requirements

• Install PyTorch (pytorch.org)
• pip install -r requirements.txt
• Download the VOC dataset from pjreddie
• Download pretrained models (models)

Training

To train a model, run main.py with the path to the VOC dataset:

# a dirctory to save model and label
mkdir models
# training 1p accuracy
bash ./test/train_full_1p.sh --data_path=real_data_path  --validation_data_path=real_validation_path

# training 1p performance
bash ./test/train_performance_1p.sh --data_path=real_data_path  --validation_data_path=real_validation_path

# training 8p accuracy
bash ./test/train_full_8p.sh --data_path=real_data_path --validation_data_path=real_validation_path --loss_scale=128.0

# training 8p performance
bash ./test/train_performance_8p.sh --data_path=real_data_path  --validation_data_path=real_validation_path --loss_scale=128.0

#test 8p accuracy
bash test/train_eval.sh --data_path=real_data_path --pth_path=real_pre_train_model_path

# finetuning 1p 
bash test/train_finetune_1p.sh --data_path=real_data_path --pth_path=real_pre_train_model_path 

Log path: test/output/devie_id/train_${device_id}.log # training detail log test/output/devie_id/SSD-MobileNetV1_bs32_8p_perf.log # 8p training performance result log test/output/devie_id/SSD-MobileNetV1_bs32_8p_acc.log # 8p training accuracy result log

SSD-MobileNetV1 training result

Acc@1	FPS	Npu_nums	Epochs	AMP_Type
0.67662	54	1	240	O1
0.6783	1000	8	240	O2