Fix issue shivamswarnkar#2

* images are generated and saved to different files (not one with all generated images)

DCGAN/generate.py

@@ -3,25 +3,32 @@
 import numpy as np
 from generator import Generator 
 import matplotlib.pyplot as plt
+from tqdm import tqdm
+import os
 
 def generate_images(args):
+    # set up device
+    device = torch.device('cuda:0' 
+        if (torch.cuda.is_available() and args.ngpu>0)  
+        else 'cpu')
 
-	# set up device
-	device = torch.device('cuda:0' 
-		if (torch.cuda.is_available() and args.ngpu>0)  
-		else 'cpu')
+    # load generator model
+    netG = Generator(args).to(device)
+    netG.load_state_dict(torch.load(args.netG))
 
-	# load generator model
-	netG = Generator(args).to(device)
-	netG.load_state_dict(torch.load(args.netG))
+    filename, ext = os.path.splitext(os.path.basename(args.output_path))
+
+    for i in tqdm(range(args.n)):
+        # create random noise
+        noise = torch.randn(1, args.nz, 1, 1, device=device)
 
-	# create random noise
-	noise = torch.randn(args.n, args.nz, 1, 1, device=device)
-	fake = netG(noise).detach().cpu()
-	img = vutils.make_grid(fake, padding=2, normalize=True)
-
-	# save image
-	plt.axis("off")
-	plt.imshow(np.transpose(img,(1,2,0)))
-	plt.savefig(args.output_path)
+        with torch.no_grad():
+            fake = netG(noise).detach().cpu()
 
+        img_np = np.transpose(vutils.make_grid(fake, padding=2, normalize=True).numpy(), (1, 2, 0))
+
+        new_filename = f"{filename}_{str(i).zfill(3)}{ext}"
+        new_output_path = os.path.join(os.path.dirname(args.output_path), new_filename)
+
+        # save image
+        plt.imsave(new_output_path, img_np)