[每日一氵] Python 写程序专用 log 分享

众所众知：

写程序不写注释是XX
写程序不缩进是XX
写程序不写log是XX <-------------- 今天说叨说叨这个

之前一直想把，PaddleClas 里边儿那个log 搬下来，自己用用，奈何，那个 logger.py 过于庞大，我把握不住

今天撞大运了，看这个baseline，发现一个简单好用的log，赶紧搬过来，给我用！

原出处在这里：
https://github.com/PerceptionComputingLab/PARSE2022/blob/main/baseline/train.py

import logging

root_nowexp = "test"

def get_logger(filename, verbosity=1, name=None):
    level_dict = {0: logging.DEBUG, 1: logging.INFO, 2: logging.WARNING}
    formatter = logging.Formatter(
        "[%(asctime)s][%(filename)s][line:%(lineno)d][%(levelname)s] %(message)s"
    )
    logger = logging.getLogger(name)
    logger.setLevel(level_dict[verbosity])

    fh = logging.FileHandler(filename, "w")
    fh.setFormatter(formatter)
    logger.addHandler(fh)

    sh = logging.StreamHandler()
    sh.setFormatter(formatter)
    logger.addHandler(sh)

    return logger

logger = get_logger(f'{root_nowexp}exp.log')
logger.info('start training!')
logger.warning('啊哈哈哈哈')
logger.error('你们被拒绝过吗')

info/warning/error 都有

运行一下看看：

[2022-04-14 14:31:54,112][2908748658.py][line:24][INFO] start training!
[2022-04-14 14:31:54,113][2908748658.py][line:25][INFO] 啊哈哈哈哈
[2022-04-14 14:31:54,114][2908748658.py][line:26][INFO] 你们被拒绝过吗

然后会在本地生成一个文件 testexp.log

打开一看，还是这个：

[2022-04-14 14:31:54,112][2908748658.py][line:24][INFO] start training!
[2022-04-14 14:31:54,113][2908748658.py][line:25][INFO] 啊哈哈哈哈
[2022-04-14 14:31:54,114][2908748658.py][line:26][INFO] 你们被拒绝过吗

nice!! 好用的一批好吗！

---

再放一个训练的例子：

logger = get_logger(f'{root_nowexp}exp.log')
logger.info('start training!')

for epoch in range(opt.max_epoch):

    model.train()
    for batch_id, (dcm_image, label_image) in tqdm(enumerate(train_dataLoader),
                                                   total=int(len(train_dataset) / opt.batch_size)):
        XXXXXXXXXXXXXXXX

        optimizer.zero_grad()
        loss.backward()
        optimizer.step()


    Unet.eval()

    logger.info('Epoch:[{}/{}]\t loss={:.5f}\t train dice:{:.1f} \t evalu dice:{:.1f}'.format(epoch, opt.max_epoch,
                                                                                              epoch_loss.avg,
                                                                                              train_dice,
                                                                                              evalu_dice))
    scheduler.step()
    
logger.info('finish training!')

前面是 开始 训练：

logger.info('start training!')

而后是 中间的指标：

logger.info('Epoch:[{}/{}]\t loss={:.5f}\t train dice:{:.1f} \t evalu dice:{:.1f}'.format(epoch, opt.max_epoch,
                                                                                              epoch_loss.avg,
                                                                                              train_dice,
                                                                                              evalu_dice))

最后是训练结束：

logger.info('finish training!')

下一篇：
https://blog.csdn.net/HaoZiHuang/article/details/127127752
来举个例子说下 log 的 level