Skip to content

glee1228/landmark_dacon

Repository files navigation

Dacon: Landmark classification

Private score 2nd 0.9971


Usage

  1. git clone https://github.com/glee1228/landmark_dacon.git

  2. edit docker-compose.yml

    services:
      main:
        container_name: landmark
        ...
        ports:
          - "{host ssh}:22"
          - "{host tensorboard}:6006"
        ipc: host
        stdin_open: true
    
  3. docker-compose up -d

  4. Train main.py

    #/workspace
    python main.py --ckpt_dir {ckpt directory} --epoch 100 --batch_size 64 --lr 0.001 --weight_decay 1e-5 -step 10 --gamma 0.8 

    Tensorboard

    # host에서 
    docker exec landmark tensorboard --logdir={ckpt directory} --bind_all
  5. submit {ckpt directory}/submission.csv

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages