diff --git a/modules/deeplearning/srcnn_l1b_l2.py b/modules/deeplearning/srcnn_l1b_l2.py
index 63d7d474dc2a82400a5b87ac9a57f6466e75e23c..e919b9b50905fa6b446cd49308e31cfb63701658 100644
--- a/modules/deeplearning/srcnn_l1b_l2.py
+++ b/modules/deeplearning/srcnn_l1b_l2.py
@@ -59,12 +59,14 @@ label_param = 'cld_opd_dcomp'
 # label_param = 'cloud_probability'
 
 params = ['temp_11_0um_nom', 'temp_12_0um_nom', 'refl_0_65um_nom', label_param]
-data_params = ['temp_11_0um_nom', 'refl_0_65um_nom']
+data_params_half = ['temp_11_0um_nom']
+data_params_full = ['refl_0_65um_nom']
 # data_params = []
 
 label_idx = params.index(label_param)
 
-print('data_params: ', data_params)
+print('data_params_half: ', data_params_half)
+print('data_params_full: ', data_params_full)
 print('label_param: ', label_param)
 
 KERNEL_SIZE = 3  # target size: (128, 128)
@@ -221,7 +223,7 @@ class SRCNN:
         self.test_data_nda = None
         self.test_label_nda = None
 
-        self.n_chans = len(data_params) + 1
+        self.n_chans = len(data_params_half) + len(data_params_full) + 1
 
         self.X_img = tf.keras.Input(shape=(None, None, self.n_chans))
         # self.X_img = tf.keras.Input(shape=(36, 36, self.n_chans))
@@ -254,7 +256,22 @@ class SRCNN:
             DO_ADD_NOISE = True
 
         data_norm = []
-        for param in data_params:
+        for param in data_params_half:
+            idx = params.index(param)
+            tmp = input_data[:, idx, :, :]
+            tmp = tmp.copy()
+            tmp = np.where(np.isnan(tmp), 0, tmp)
+            # tmp = smooth_2d(tmp, sigma=1.0)
+            # Half res upsampled to full res:
+            tmp = tmp[:, slc_y_2, slc_x_2]
+            tmp = resample_2d_linear(x_2, y_2, tmp, t, s)
+            tmp = tmp[:, y_k, x_k]
+            tmp = normalize(tmp, param, mean_std_dct)
+            if DO_ADD_NOISE:
+                tmp = add_noise(tmp, noise_scale=NOISE_STDDEV)
+            data_norm.append(tmp)
+
+        for param in data_params_full:
             idx = params.index(param)
             tmp = input_data[:, idx, :, :]
             tmp = tmp.copy()
@@ -262,10 +279,6 @@ class SRCNN:
             # tmp = smooth_2d(tmp, sigma=1.0)
             # Full res:
             tmp = tmp[:, slc_y, slc_x]
-            # Half res upsampled to full res:
-            # tmp = tmp[:, slc_y_2, slc_x_2]
-            # tmp = resample_2d_linear(x_2, y_2, tmp, t, s)
-            # tmp = tmp[:, y_k, x_k]
             tmp = normalize(tmp, param, mean_std_dct)
             if DO_ADD_NOISE:
                 tmp = add_noise(tmp, noise_scale=NOISE_STDDEV)