boostcampaitech2/model-optimization-level3-nlp-03

model-optimization-level3-nlp-03 created by GitHub Classroom

boostcamp_pstage10

Environment

1. Docker

docker run -it --gpus all --ipc=host -v ${path_to_code}:/opt/ml/code -v ${path_to_dataset}:/opt/ml/data placidus36/pstage4_lightweight:v0.4 /bin/bash

2. Install dependencies

pip install -r requirements.txt

Run

1. train

python train.py --model ${path_to_model_config} --data ${path_to_data_config}

Ex) python train.py --model configs/model/mobilenetv3.yaml --data configs/data/taco.yaml

2. inference(submission.csv)

python inference.py --model_dir ${저장된 모델 root 경로} --weight_name ${모델 weight 이름} --img_root ${test 데이터 경로}

EX) python inference.py --model_dir /opt/ml/code/exp/latest --weight best.pt --img_root /opt/ml/data/test/

Reference

Our basic structure is based on Kindle(by JeiKeiLim)