def experiment()

in data_augmentation/hydra_app_local.py [0:0]

• 40 lines of code
• 7 McCabe index (conditional complexity)

def experiment(cfg):
    import data_augmentation.my_training as my_training 
    try:
        cfg.classifier.gpu
    except:
        cfg.classifier.gpu = None
    try:
        cfg.classifier.seed
    except:
        cfg.classifier.seed = None

    args = cfg.classifier
    checkpoint_path = os.path.join(args.result_dir, 'checkpoints')
    if os.path.exists(checkpoint_path):
        checkpoint_folders = os.listdir(checkpoint_path)
    else:
        checkpoint_folders = []
    job_id = ""

    args = set_submitit_vars(args)

    checkpoint_folder = []
    for f in checkpoint_folders:
        f_id = ('_').join(f.split('_')[4:])
        if job_id == f_id:
            checkpoint_folder.append(f)
    if len(checkpoint_folder) > 1:
        print(checkpoint_folder)
        raise Exception("More than one matching folder found.")
    elif len(checkpoint_folder)==0:
        print("No matching folder found.")
        my_training.run_model(cfg.classifier, ckpt_path=None, repo=None)
    else:
        checkpoint_folder = checkpoint_folder[0]
        print("One folder found from {}".format(checkpoint_folder))
        ckpt_path = os.path.join(checkpoint_path,'{0}/checkpoint.pth'.format(checkpoint_folder))
        repo_path=os.path.join(checkpoint_path,checkpoint_folder)
        if os.path.exists(ckpt_path):
            print("It has a checkpoint.pth, resuming from there")
            my_training.run_model(cfg.classifier, ckpt_path, repo=repo_path)
        else:
            print("But has no checkpoint.pth")
            my_training.run_model(cfg.classifier, ckpt_path=None, repo=repo_path)