>

aspp.py

@@ -1,4 +1,4 @@
-# camera-ready if everythng works (need to modify paths)
+# camera-ready if everythng works
 
 import torch
 import torch.nn as nn

deeplabv3.py

@@ -1,4 +1,4 @@
-# camera-ready if everything works (need to modify paths)
+# camera-ready if everything works
 
 import torch
 import torch.nn as nn

preprocess_data.py

@@ -1,4 +1,4 @@
-# camera-ready if everything works (need to modify paths)
+# camera-ready if everything works
 
 import pickle
 import numpy as np
@@ -94,8 +94,8 @@
 val_dirs = ["frankfurt/", "munster/", "lindau/"]
 test_dirs = ["berlin", "bielefeld", "bonn", "leverkusen", "mainz", "munich"]
 
-cityscapes_data_path = "/datasets/cityscapes"
-cityscapes_meta_path = "/staging/frexgus/cityscapes/meta"
+cityscapes_data_path = "/root/deeplabv3/data/cityscapes"
+cityscapes_meta_path = "/root/deeplabv3/data/cityscapes/meta"
 
 if not os.path.exists(cityscapes_meta_path + "/label_imgs"):
     os.makedirs(cityscapes_meta_path + "/label_imgs")

train.py

@@ -1,4 +1,4 @@
-# camera-ready if everything works (need to modify paths)
+# camera-ready if everything works
 
 from datasets import DatasetTrain, DatasetVal # (this needs to be imported before torch, because cv2 needs to be imported before torch for some reason)
 from deeplabv3 import DeepLabV3
@@ -28,12 +28,12 @@
 batch_size = 16
 learning_rate = 0.0001
 
-network = DeepLabV3(model_id, project_dir="/staging/frexgus/multitask").cuda()
+network = DeepLabV3(model_id, project_dir="/root/deeplabv3").cuda()
 
-train_dataset = DatasetTrain(cityscapes_data_path="/datasets/cityscapes",
-                             cityscapes_meta_path="/staging/frexgus/cityscapes/meta")
-val_dataset = DatasetVal(cityscapes_data_path="/datasets/cityscapes",
-                         cityscapes_meta_path="/staging/frexgus/cityscapes/meta")
+train_dataset = DatasetTrain(cityscapes_data_path="/root/deeplabv3/data/cityscapes",
+                             cityscapes_meta_path="/root/deeplabv3/data/cityscapes/meta")
+val_dataset = DatasetVal(cityscapes_data_path="/root/deeplabv3/data/cityscapes",
+                         cityscapes_meta_path="/root/deeplabv3/data/cityscapes/meta")
 
 num_train_batches = int(len(train_dataset)/batch_size)
 num_val_batches = int(len(val_dataset)/batch_size)
@@ -50,7 +50,7 @@
 params = add_weight_decay(network, l2_value=0.0001)
 optimizer = torch.optim.Adam(params, lr=learning_rate)
 
-with open("/staging/frexgus/cityscapes/meta/class_weights.pkl", "rb") as file: # (needed for python3)
+with open("/root/deeplabv3/data/cityscapes/meta/class_weights.pkl", "rb") as file: # (needed for python3)
     class_weights = np.array(pickle.load(file))
 class_weights = torch.from_numpy(class_weights)
 class_weights = Variable(class_weights.type(torch.FloatTensor)).cuda()