--- a
+++ b/utils.py
@@ -0,0 +1,88 @@
+""" Utilities """
+import os
+import logging
+import shutil
+import torch
+import torchvision.datasets as dset
+import numpy as np
+import torch.nn as nn
+device = torch.device("cuda")
+
+def get_logger(file_path):
+    """ Make python logger """
+    # [!] Since tensorboardX use default logger (e.g. logging.info()), we should use custom logger
+    logger = logging.getLogger('darts')
+    log_format = '%(asctime)s | %(message)s'
+    formatter = logging.Formatter(log_format, datefmt='%m/%d %I:%M:%S %p')
+    file_handler = logging.FileHandler(file_path)
+    file_handler.setFormatter(formatter)
+    stream_handler = logging.StreamHandler()
+    stream_handler.setFormatter(formatter)
+
+    logger.addHandler(file_handler)
+    logger.addHandler(stream_handler)
+    logger.setLevel(logging.INFO)
+
+    return logger
+
+
+def param_size(model):
+    """ Compute parameter size in MB """
+    n_params = sum(
+        np.prod(v.size()) for k, v in model.named_parameters() if not k.startswith('aux_head'))
+    return n_params / 1024. / 1024.
+
+
+class AverageMeter():
+    """ Computes and stores the average and current value """
+    def __init__(self):
+        self.reset()
+
+    def reset(self):
+        """ Reset all statistics """
+        self.val = 0
+        self.avg = 0
+        self.sum = 0
+        self.count = 1 # avoid the count of some calsses in the first batch is zero
+
+    def update(self, val, n):
+        """ Update statistics """
+        self.val = val
+        self.sum += val * n
+        self.count += n
+        self.avg = self.sum / self.count
+
+def evaluate(logits, label):
+    logits = logits.astype(np.float32)
+    label = label.astype(np.float32)
+    inter = np.dot(logits.flatten(), label.flatten())
+    union = np.sum(logits) + np.sum(label)
+    dice = (2 * inter + 1e-5) / (union + 1e-5)
+    return dice
+
+def save_results(results, path):
+    
+    filename = os.path.join(path, 'final_results.txt')
+    f = open(filename, 'a')
+    f.write('Best dice: {:.5f}\n'.format(results))
+    
+
+def save_checkpoint(state, ckpt_dir, is_best=False):
+    filename = os.path.join(ckpt_dir, 'checkpoint.pth.tar')
+    torch.save(state, filename)
+    if is_best:
+        best_filename = os.path.join(ckpt_dir, 'best.pth.tar')
+        shutil.copyfile(filename, best_filename)
+        
+class log_loss(nn.Module):
+    def __init__(self, w_dice = 0.5, w_cross = 0.5):
+        super(log_loss, self).__init__()
+        self.w_dice = w_dice
+        self.w_cross = w_cross
+    def forward(self, logits, label, smooth = 1.):
+        
+        area_union = torch.sum(logits * label, dim = (0,2,3), keepdim = True)
+        area_logits = torch.sum(logits, dim = (0,2,3), keepdim = True)
+        area_label = torch.sum(label, dim = (0,2,3), keepdim = True)
+        in_dice = torch.mean(torch.pow((-1) * torch.log((2 * area_union + 1e-7)/(area_logits + area_label + smooth)), 0.3))
+        return in_dice