반응형
wandb로 이미지 로딩하면 생각보다 메모리를 많이 차지 한다
root@129922e51c90:~# df -H
Filesystem Size Used Avail Use% Mounted on
overlay 106G 17G 84G 17% /
tmpfs 68M 0 68M 0% /dev
tmpfs 48G 0 48G 0% /sys/fs/cgroup
shm 4.3G 0 4.3G 0% /dev/shm
/dev/xvdb1 106G 17G 84G 17% /etc/hosts
tmpfs 48G 13k 48G 1% /proc/driver/nvidia
/dev/xvda1 51G 13G 36G 27% /usr/bin/nvidia-smi
udev 48G 0 48G 0% /dev/nvidia0
tmpfs 48G 0 48G 0% /proc/acpi
tmpfs 48G 0 48G 0% /proc/scsi
tmpfs 48G 0 48G 0% /sys/firmware
root@129922e51c90:~# ls
code input
root@129922e51c90:~# cd code
root@129922e51c90:~/code# ls
OCR_EDA.ipynb detect.py inference.py pths trained_models
__pycache__ deteval.py loss.py requirements.txt urls.txt
convert_mlt.py download.sh model.py train-Copy1.py wandb
dataset.py east_dataset.py nohup.out train.py
root@129922e51c90:~/code# rm -rf wandb
root@129922e51c90:~/code# df -H
Filesystem Size Used Avail Use% Mounted on
overlay 106G 9.8G 91G 10% /
tmpfs 68M 0 68M 0% /dev
tmpfs 48G 0 48G 0% /sys/fs/cgroup
shm 4.3G 0 4.3G 0% /dev/shm
/dev/xvdb1 106G 9.8G 91G 10% /etc/hosts
tmpfs 48G 13k 48G 1% /proc/driver/nvidia
/dev/xvda1 51G 13G 36G 27% /usr/bin/nvidia-smi
udev 48G 0 48G 0% /dev/nvidia0
tmpfs 48G 0 48G 0% /proc/acpi
tmpfs 48G 0 48G 0% /proc/scsi
tmpfs 48G 0 48G 0% /sys/firmware
root@129922e51c90:~/code#반응형
'Boostcamp AI tech 3기' 카테고리의 다른 글
| opencv video mp4로 저장하기 (0) | 2022.05.19 |
|---|---|
| 부스트 캠프 끝나고 해볼만한 ai 관련 공고/프로그램 (스마일게이트, ssafy 8기) (0) | 2022.05.15 |
| 파이널 프로젝트 때 참고할 사이트들 (0) | 2022.05.02 |
| vgg16_bn-6c64b313.pth 다운받기 (0) | 2022.04.19 |
| ImportError: libGL.so.1: cannot open shared object file: No such file or directory 에러 해결 (0) | 2022.04.14 |
| mmcv, nvcc 설치 (0) | 2022.03.30 |
| conda install error in pstage (0) | 2022.03.30 |
| global /io/opencv/modules/imgcodecs/src/loadsave.cpp (239) findDecoder imreaead_(~~) : can't open/read file: check file path/integrity 확장자 조심 (0) | 2022.03.27 |