Preparing for transfer learning on Cloud 95 dataset

main/config.py

@@ -28,12 +28,12 @@ from pytorch.models import UNet
 wd = '/mnt/CEPH_PROJECTS/cci_snow/dfrisinghelli/'
 
 # define which dataset to train on
-dataset_name = 'Sparcs'
-# dataset_name = 'Cloud95'
+# dataset_name = 'Sparcs'
+dataset_name = 'Cloud95'
 
 # path to the dataset
-dataset_path = os.path.join(wd, '_Datasets/Sparcs')
-# dataset_path = os.path.join(wd, '_Datasets/Cloud95/Training')
+# dataset_path = os.path.join(wd, '_Datasets/Sparcs')
+dataset_path = os.path.join(wd, '_Datasets/Cloud95/Training')
 
 # the csv file containing the names of the informative patches of the
 # Cloud95 dataset
@@ -46,7 +46,7 @@ bands = ['red', 'green', 'blue', 'nir']
 
 # define the size of the network input
 # if None, the size will default to the size of a scene
-tile_size = 125
+tile_size = 192
 
 # -----------------------------------------------------------------------------
 # -----------------------------------------------------------------------------
@@ -83,10 +83,10 @@ kwargs = {'kernel_size': 3,  # the size of the convolving kernel
 state_path = os.path.join(wd, 'git/deep-learning/main/_models/')
 
 # whether to use a pretrained model
-pretrained = False
+pretrained = True
 
 # name of the pretrained model
-pretrained_model = 'UNet_SparcsDataset_t125_b64_rgbn.pt'
+pretrained_model = 'UNet_SparcsDataset_t125_b128_rgbn.pt'
 
 
 # Dataset split ---------------------------------------------------------------
@@ -100,12 +100,12 @@ ttratio = 1
 
 # (ttratio * tvratio) * 100 % will be used as the training dataset
 # (1 - ttratio * tvratio) * 100 % will be used as the validation dataset
-tvratio = 0.8
+tvratio = 0.05
 
 # define the batch size
 # determines how many samples of the dataset are processed until the weights
 # of the network are updated
-batch_size = 128
+batch_size = 64
 
 # Training configuration ------------------------------------------------------
 
@@ -114,14 +114,14 @@ checkpoint = False
 
 # whether to early stop training if the accuracy (loss) on the validation set
 # does not increase (decrease) more than delta over patience epochs
-early_stop = True
+early_stop = False
 mode = 'max'
 delta = 0
 patience = 10
 
 # define the number of epochs: the number of maximum iterations over the whole
 # training dataset
-epochs = 200
+epochs = 10
 
 # define the number of threads
 nthreads = os.cpu_count()
@@ -149,7 +149,7 @@ plot_cm = False
 # whether to save plots of (input, ground truth, prediction) of the validation
 # dataset to disk
 # output path is: current_working_directory/_samples/
-plot_samples = True
+plot_samples = False
 
 # number of samples to plot
 # if nsamples = -1, all samples are plotted