Unnecesary arguments removed

create_model.py

@@ -7,7 +7,7 @@ def create_model():
 
     parser.add_argument("--input-data-path", "-i", help="The path to a data file.", type=str, required=True)
     parser.add_argument("--output-model-folder", "-o", help="Prefix to the folder to save output model.", type=str)
-    parser.add_argument("--latent_dim", "-ld", help="dimensionality of the latent space", type=int)
+    parser.add_argument("--latent_dim", "-ld", help="dimensionality of the noise", type=int)
     args = {k: v for k, v in vars(parser.parse_args()).items() if v is not None}
 
     runner = CreateModelRunner(**args)

runners/TrainModelRunner.py

@@ -18,7 +18,7 @@ class TrainModelRunner:
     lambda_gp = 10
 
     def __init__(self, input_data_path, output_model_folder, decode_mols_save_path='', n_epochs=200, starting_epoch=1,
-                 batch_size=64, lr=0.0002, b1=0.5, b2=0.999, n_cpu=4, n_critic=5, clip_value=0.01, sample_interval=10,
+                 batch_size=64, lr=0.0002, b1=0.5, b2=0.999,  n_critic=5, sample_interval=10,
                  save_interval=100, sample_after_training=100, message=""):
         self.message = message
 
@@ -31,9 +31,7 @@ def __init__(self, input_data_path, output_model_folder, decode_mols_save_path='
         self.lr = lr
         self.b1 = b1
         self.b2 = b2
-        self.n_cpu = n_cpu
         self.n_critic = n_critic
-        self.clip_value = clip_value
         self.sample_interval = sample_interval
         self.save_interval = save_interval
         self.sample_after_training = sample_after_training
@@ -160,19 +158,18 @@ def run(self):
         with open(os.path.join(self.output_model_folder, 'gen_loss.json'), 'w') as json_file:
             json.dump(g_loss_log, json_file)
 
+        # Sampling after training
         if self.sample_after_training > 0:
             # sampling mode
             torch.no_grad()
             self.G.eval()
 
             S = Sampler(generator=self.G)
             latent = S.sample(self.sample_after_training)
-            # latent = ((latent + 1) * self.peak / 2) + self.min
             latent = latent.detach().cpu().numpy().tolist()
 
             sampled_mols_save_path = os.path.join(self.output_model_folder, 'sampled.json')
             with open(sampled_mols_save_path, 'w') as json_file:
-                # array_fake_mols = fake_mols.data
                 json.dump(latent, json_file)
 
             # decoding sampled mols

train_model.py

@@ -14,16 +14,12 @@ def train_model():
     parser.add_argument("--lr", type=float, help="adam: learning rate")
     parser.add_argument("--b1", type=float, help="adam: decay of first order momentum of gradient")
     parser.add_argument("--b2", type=float, help="adam: decay of first order momentum of gradient")
-    parser.add_argument("--n-cpu", type=int, help="number of cpu threads to use during batch generation")
     parser.add_argument("--latent-dim", type=int, help="dimensionality of the latent space")
-    parser.add_argument("--img-size", type=int, help="size of each image dimension")
-    parser.add_argument("--channels", type=int, help="number of image channels")
     parser.add_argument("--n-critic", type=int, help="number of training steps for discriminator per iter")
-    parser.add_argument("--clip-value", type=float, help="lower and upper clip value for disc. weights")
     parser.add_argument("--sample-interval", type=int, help="interval between samples")
     parser.add_argument("--save-interval", type=int, help="interval between saving the model")
     parser.add_argument("--sample-after-training", type=int, help="Number of molecules to sample after training")
-    parser.add_argument("--message", "-m", type=str, help="Number of molecules to sample after training")
+    parser.add_argument("--message", "-m", type=str, help="The message to print before the training starts")
 
     args = {k: v for k, v in vars(parser.parse_args()).items() if v is not None}