Classify training cleanup (#33)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

ultralytics/yolo/utils/configs/default.yml

@@ -1,25 +1,27 @@
-model: null
-data: null
+# YOLO 🚀 by Ultralytics, GPL-3.0 license
+# Default training settings and hyperparameters for medium-augmentation COCO training
 
-# Training options
+
+# Train settings -------------------------------------------------------------------------------------------------------
+model: null  # i.e. yolov5s.pt
+data: null  # i.e. coco128.yaml
 epochs: 300
 batch_size: 16
 img_size: 640
 nosave: False
-cache: False # True/ram for ram, or disc
-device: '' # cuda device, i.e. 0 or 0,1,2,3 or cpu
+cache: False  # True/ram, disk or False
+device: ''  # cuda device, i.e. 0 or 0,1,2,3 or cpu
 workers: 8
-project: "ultralytics-yolo"
-name: "exp" # TODO: make this informative, maybe exp{#number}_{datetime} ?
+project: 'runs'
+name: 'exp'
 exist_ok: False
 pretrained: False
-optimizer: "Adam" # choices=['SGD', 'Adam', 'AdamW', 'RMSProp']
+optimizer: 'SGD'  # choices=['SGD', 'Adam', 'AdamW', 'RMSProp']
 verbose: False
 seed: 0
 local_rank: -1
-#-----------------------------------#
 
-# Hyper-parameters
+# Hyperparameters ------------------------------------------------------------------------------------------------------
 lr0: 0.001  # initial learning rate (SGD=1E-2, Adam=1E-3)
 lrf: 0.01  # final OneCycleLR learning rate (lr0 * lrf)
 momentum: 0.937  # SGD momentum/Adam beta1
@@ -50,9 +52,8 @@ mosaic: 1.0  # image mosaic (probability)
 mixup: 0.0  # image mixup (probability)
 copy_paste: 0.0  # segment copy-paste (probability)
 
-# Hydra configs -------------------------------------
-# to disable hydra directory creation
+# Hydra configs --------------------------------------------------------------------------------------------------------
 hydra:
-  output_subdir: null
+  output_subdir: null  # disable hydra directory creation
   run:
     dir: .

ultralytics/yolo/utils/modeling/__init__.py

@@ -107,18 +107,17 @@ def parse_model(d, ch):  # model_dict, input_channels(3)
     return nn.Sequential(*layers), sorted(save)
 
 
-def get_model(model: str):
+def get_model(model='s.pt', pretrained=True):
+    # Load a YOLO model locally, from torchvision, or from Ultralytics assets
     if model.endswith(".pt"):
         model = model.split(".")[0]
 
-    if Path(model + ".pt").is_file():
-        trained_model = torch.load(model + ".pt", map_location='cpu')
-    elif model in torchvision.models.__dict__:  # try torch hub classifier models
-        trained_model = torch.hub.load("pytorch/vision", model, pretrained=True)
-    else:
-        model_ckpt = attempt_download(model + ".pt")  # try ultralytics assets
-        trained_model = torch.load(model_ckpt, map_location='cpu')
-    return trained_model
+    if Path(f"{model}.pt").is_file():  # local file
+        return torch.load(f"{model}.pt", map_location='cpu')
+    elif model in torchvision.models.__dict__:  # TorchVision models i.e. resnet50, efficientnet_b0
+        return torchvision.models.__dict__[model](weights='IMAGENET1K_V1' if pretrained else None)
+    else:  # Ultralytics assets
+        return torch.load(attempt_download(f"{model}.pt"), map_location='cpu')
 
 
 def yaml_load(file='data.yaml'):