updates

train.py

@@ -62,12 +62,13 @@ def train():
     epochs = opt.epochs  # 500200 batches at bs 16, 117263 images = 273 epochs
     batch_size = opt.batch_size
     accumulate = opt.accumulate  # effective bs = batch_size * accumulate = 16 * 4 = 64
+    weights = opt.weights  # initial training weights
 
     # Initialize
     init_seeds()
-    weights = 'weights' + os.sep
+    wdir = 'weights' + os.sep  # weights dir
-    last = weights + 'last.pt'
+    last = wdir + 'last.pt'
-    best = weights + 'best.pt'
+    best = wdir + 'best.pt'
     device = torch_utils.select_device(apex=mixed_precision)
     multi_scale = opt.multi_scale
 
@@ -94,26 +95,23 @@ def train():
     cutoff = -1  # backbone reaches to cutoff layer
     start_epoch = 0
     best_fitness = 0.
-    nf = int(model.module_defs[model.yolo_layers[0] - 1]['filters'])  # yolo layer size (i.e. 255)
+    if weights.endswith('.pt'):  # pytorch format
-    if opt.resume or opt.transfer:  # Load previously saved model
+        # possible weights are 'last.pt', 'yolov3-spp.pt', 'yolov3-tiny.pt' etc.
-        if opt.transfer:  # Transfer learning
-            chkpt = torch.load(weights + 'yolov3-spp.pt', map_location=device)
-            model.load_state_dict({k: v for k, v in chkpt['model'].items() if v.numel() > 1 and v.shape[0] != 255},
-                                  strict=False)
-
-            for p in model.parameters():
-                p.requires_grad = True if p.shape[0] == nf else False
-
-        else:  # resume from last.pt
         if opt.bucket:
             os.system('gsutil cp gs://%s/last.pt %s' % (opt.bucket, last))  # download from bucket
-            chkpt = torch.load(last, map_location=device)  # load checkpoint
+        chkpt = torch.load(weights, map_location=device)
-            model.load_state_dict(chkpt['model'])
 
+        # load model
+        if opt.transfer:
+            chkpt['model'] = {k: v for k, v in chkpt['model'].items() if model.state_dict()[k].numel() == v.numel()}
+        model.load_state_dict(chkpt['model'], strict=False)
+
+        # load optimizer
         if chkpt['optimizer'] is not None:
             optimizer.load_state_dict(chkpt['optimizer'])
             best_fitness = chkpt['best_fitness']
 
+        # load results
         if chkpt.get('training_results') is not None:
             with open('results.txt', 'w') as file:
                 file.write(chkpt['training_results'])  # write results.txt
@@ -121,15 +119,14 @@ def train():
         start_epoch = chkpt['epoch'] + 1
         del chkpt
 
-    else:  # Initialize model with backbone (optional)
+    elif weights.endswith('.weights'):  # darknet format
-        if '-tiny.cfg' in cfg:
+        # possible weights are 'yolov3.weights', 'yolov3-tiny.conv.15',  'darknet53.conv.74' etc.
-            cutoff = load_darknet_weights(model, weights + 'yolov3-tiny.conv.15')
+        cutoff = load_darknet_weights(model, weights)
-        else:
-            cutoff = load_darknet_weights(model, weights + 'darknet53.conv.74')
 
-        # Remove old results
+    if opt.transfer:  # transfer learning
-        for f in glob.glob('*_batch*.jpg') + glob.glob('results.txt'):
+        nf = int(model.module_defs[model.yolo_layers[0] - 1]['filters'])  # yolo layer size (i.e. 255)
-            os.remove(f)
+        for p in model.parameters():
+            p.requires_grad = True if p.shape[0] == nf else False
 
     # Scheduler https://github.com/ultralytics/yolov3/issues/238
     # lf = lambda x: 1 - x / epochs  # linear ramp to zero
@@ -181,6 +178,10 @@ def train():
                                              pin_memory=True,
                                              collate_fn=dataset.collate_fn)
 
+    # Remove previous results
+    for f in glob.glob('*_batch*.jpg') + glob.glob('results.txt'):
+        os.remove(f)
+
     # Start training
     model.nc = nc  # attach number of classes to model
     model.hyp = hyp  # attach hyperparameters to model
@@ -327,7 +328,7 @@ def train():
 
             # Save backup every 10 epochs (optional)
             if epoch > 0 and epoch % 10 == 0:
-                torch.save(chkpt, weights + 'backup%g.pt' % epoch)
+                torch.save(chkpt, wdir + 'backup%g.pt' % epoch)
 
             # Delete checkpoint
             del chkpt
@@ -345,7 +346,7 @@ if __name__ == '__main__':
     parser.add_argument('--epochs', type=int, default=273)  # 500200 batches at bs 16, 117263 images = 273 epochs
     parser.add_argument('--batch-size', type=int, default=32)  # effective bs = batch_size * accumulate = 16 * 4 = 64
     parser.add_argument('--accumulate', type=int, default=2, help='batches to accumulate before optimizing')
-    parser.add_argument('--cfg', type=str, default='cfg/yolov3-spp.cfg', help='cfg file path')
+    parser.add_argument('--cfg', type=str, default='cfg/yolov3-spp-1cls.cfg', help='cfg file path')
     parser.add_argument('--data', type=str, default='data/coco.data', help='*.data file path')
     parser.add_argument('--multi-scale', action='store_true', help='adjust (67% - 150%) img_size every 10 batches')
     parser.add_argument('--img-size', type=int, default=416, help='inference size (pixels)')
@@ -358,7 +359,9 @@ if __name__ == '__main__':
     parser.add_argument('--bucket', type=str, default='', help='gsutil bucket')
     parser.add_argument('--img-weights', action='store_true', help='select training images by weight')
     parser.add_argument('--cache-images', action='store_true', help='cache images for faster training')
+    parser.add_argument('--weights', type=str, default='', help='initial weights')  # i.e. weights/darknet.53.conv.74
     opt = parser.parse_args()
+    opt.weights = 'weights/last.pt' if opt.resume else opt.weights
     print(opt)
 
     tb_writer = None