Unified model loading with backwards compatibility (#132)

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

docs/quickstart.md

@@ -42,9 +42,9 @@ Ultralytics YOLO comes with pythonic Model and Trainer interface.
     import ultralytics
     from ultralytics import YOLO
 
-    model = YOLO("s-seg.yaml") # automatically detects task type
-    model = YOLO("s-seg.pt") # load checkpoint
-    model.train(data="coco128-segments", epochs=1, lr0=0.01, ...)
-    model.train(data="coco128-segments", epochs=1, lr0=0.01, device="0,1,2,3") # DDP mode
+    model = YOLO("yolov8n-seg.yaml") # automatically detects task type
+    model = YOLO("yolov8n.pt") # load checkpoint
+    model.train(data="coco128-seg.yaml", epochs=1, lr0=0.01, ...)
+    model.train(data="coco128-seg.yaml", epochs=1, lr0=0.01, device="0,1,2,3") # DDP mode
     ```
 [API Guide](sdk.md){ .md-button .md-button--primary}

tests/test_model.py

@@ -1,11 +1,7 @@
 import torch
 
 from ultralytics import YOLO
-
-
-def test_model_init():
-    model = YOLO("yolov8n.yaml")
-    model.info()
+from ultralytics.yolo.utils import ROOT
 
 
 def test_model_forward():
@@ -29,9 +25,9 @@ def test_model_fuse():
     model.fuse()
 
 
-def test_visualize_preds():
+def test_predict_dir():
     model = YOLO("yolov8n.pt")
-    model.predict(source="ultralytics/assets")
+    model.predict(source=ROOT / "assets")
 
 
 def test_val():
@@ -39,7 +35,7 @@ def test_val():
     model.val(data="coco128.yaml", imgsz=32)
 
 
-def test_model_resume():
+def test_train_resume():
     model = YOLO("yolov8n.yaml")
     model.train(epochs=1, imgsz=32, data="coco128.yaml")
     try:
@@ -48,16 +44,21 @@ def test_model_resume():
         print("Successfully caught resume assert!")
 
 
-def test_model_train_pretrained():
-    model = YOLO("yolov8n.pt")
-    model.train(data="coco128.yaml", epochs=1, imgsz=32)
+def test_train_scratch():
     model = YOLO("yolov8n.yaml")
     model.train(data="coco128.yaml", epochs=1, imgsz=32)
     img = torch.rand(1, 3, 320, 320)
     model(img)
 
 
-def test_exports():
+def test_train_pretrained():
+    model = YOLO("yolov8n.pt")
+    model.train(data="coco128.yaml", epochs=1, imgsz=32)
+    img = torch.rand(1, 3, 320, 320)
+    model(img)
+
+
+def test_export_torchscript():
     """
                        Format     Argument           Suffix    CPU    GPU
     0                 PyTorch            -              .pt   True   True
@@ -74,26 +75,35 @@ def test_exports():
     11           PaddlePaddle       paddle    _paddle_model   True   True
     """
     from ultralytics.yolo.engine.exporter import export_formats
-
     print(export_formats())
 
     model = YOLO("yolov8n.yaml")
     model.export(format='torchscript')
+
+
+def test_export_onnx():
+    model = YOLO("yolov8n.yaml")
     model.export(format='onnx')
+
+
+def test_export_openvino():
+    model = YOLO("yolov8n.yaml")
     model.export(format='openvino')
+
+
+def test_export_coreml():
+    model = YOLO("yolov8n.yaml")
     model.export(format='coreml')
-    model.export(format='paddle')
 
 
-def test():
-    test_model_forward()
-    test_model_info()
-    test_model_fuse()
-    test_visualize_preds()
-    test_val()
-    test_model_resume()
-    test_model_train_pretrained()
+def test_export_paddle():
+    model = YOLO("yolov8n.yaml")
+    model.export(format='paddle')
 
 
-if __name__ == "__main__":
-    test()
+# def run_all_tests():  # do not name function test_...
+#     pass
+#
+#
+# if __name__ == "__main__":
+#     run_all_tests()

ultralytics/hub/utils.py

@@ -124,7 +124,7 @@ def smart_request(*args, retry=3, timeout=30, thread=True, code=-1, method="post
         return func(*args, **kwargs)
 
 
-def sync_analytics(cfg, all_keys=False, enabled=True):
+def sync_analytics(cfg, all_keys=False, enabled=False):
     """
    Sync analytics data if enabled in the global settings
 

ultralytics/nn/tasks.py

@@ -10,11 +10,13 @@ import torchvision
 from ultralytics.nn.modules import (C1, C2, C3, C3TR, SPP, SPPF, Bottleneck, BottleneckCSP, C2f, C3Ghost, C3x, Classify,
                                     Concat, Conv, ConvTranspose, Detect, DWConv, DWConvTranspose2d, Ensemble, Focus,
                                     GhostBottleneck, GhostConv, Segment)
-from ultralytics.yolo.utils import LOGGER, colorstr, yaml_load
+from ultralytics.yolo.utils import DEFAULT_CONFIG, LOGGER, colorstr, yaml_load
 from ultralytics.yolo.utils.checks import check_yaml
 from ultralytics.yolo.utils.torch_utils import (fuse_conv_and_bn, initialize_weights, intersect_dicts, make_divisible,
                                                 model_info, scale_img, time_sync)
 
+DEFAULT_CONFIG_DICT = yaml_load(DEFAULT_CONFIG, append_filename=False)
+
 
 class BaseModel(nn.Module):
     '''
@@ -211,7 +213,7 @@ class DetectionModel(BaseModel):
         return y
 
     def load(self, weights, verbose=True):
-        csd = weights['model'].float().state_dict()  # checkpoint state_dict as FP32
+        csd = weights.float().state_dict()  # checkpoint state_dict as FP32
         csd = intersect_dicts(csd, self.state_dict())  # intersect
         self.load_state_dict(csd, strict=False)  # load
         if verbose:
@@ -281,21 +283,21 @@ class ClassificationModel(BaseModel):
 # Functions ------------------------------------------------------------------------------------------------------------
 
 
-def attempt_load_weights(weights, device=None, inplace=True, fuse=True):
+def attempt_load_weights(weights, device=None, inplace=True, fuse=False):
     # Loads an ensemble of models weights=[a,b,c] or a single model weights=[a] or weights=a
     from ultralytics.yolo.utils.downloads import attempt_download
+    default_keys = DEFAULT_CONFIG_DICT.keys()
 
     model = Ensemble()
     for w in weights if isinstance(weights, list) else [weights]:
         ckpt = torch.load(attempt_download(w), map_location='cpu')  # load
+        args = {**DEFAULT_CONFIG_DICT, **ckpt['train_args']}
         ckpt = (ckpt.get('ema') or ckpt['model']).to(device).float()  # FP32 model
 
         # Model compatibility updates
-        if not hasattr(ckpt, 'stride'):
-            ckpt.stride = torch.tensor([32.])
-        if hasattr(ckpt, 'names') and isinstance(ckpt.names, (list, tuple)):
-            ckpt.names = dict(enumerate(ckpt.names))  # convert to dict
+        ckpt.args = {k: v for k, v in args.items() if k in default_keys}
 
+        # Append
         model.append(ckpt.fuse().eval() if fuse and hasattr(ckpt, 'fuse') else ckpt.eval())  # model in eval mode
 
     # Module compatibility updates
@@ -310,7 +312,7 @@ def attempt_load_weights(weights, device=None, inplace=True, fuse=True):
     if len(model) == 1:
         return model[-1]
 
-    # Return detection ensemble
+    # Return ensemble
     print(f'Ensemble created with {weights}\n')
     for k in 'names', 'nc', 'yaml':
         setattr(model, k, getattr(model[0], k))

ultralytics/yolo/engine/exporter.py

@@ -164,8 +164,8 @@ class Exporter:
             assert not self.args.dynamic, '--half not compatible with --dynamic, i.e. use either --half or --dynamic'
 
         # Checks
-        if self.args.batch_size == 16:
-            self.args.batch_size = 1  # TODO: resolve batch_size 16 default in config.yaml
+        # if self.args.batch_size == model.args['batch_size']:  # user has not modified training batch_size
+        self.args.batch_size = 1
         self.imgsz = check_imgsz(self.args.imgsz, stride=model.stride, min_dim=2)  # check image size
         if self.args.optimize:
             assert self.device.type == 'cpu', '--optimize not compatible with cuda devices, i.e. use --device cpu'
@@ -778,7 +778,7 @@ def export(cfg):
         if Path(cfg.model).suffix == '.yaml':
             model = DetectionModel(cfg.model)
         elif Path(cfg.model).suffix == '.pt':
-            model = attempt_load_weights(cfg.model)
+            model = attempt_load_weights(cfg.model, fuse=True)
         else:
             TypeError(f'Unsupported model type {cfg.model}')
     exporter(model=model)

ultralytics/yolo/engine/model.py

@@ -77,13 +77,12 @@ class YOLO:
         Args:
             weights (str): model checkpoint to be loaded
         """
-        self.ckpt = torch.load(weights, map_location="cpu")
-        self.task = self.ckpt["train_args"]["task"]
-        self.overrides = dict(self.ckpt["train_args"])
+        self.model = attempt_load_weights(weights)
+        self.task = self.model.args["task"]
+        self.overrides = self.model.args
         self.overrides["device"] = ''  # reset device
         self.ModelClass, self.TrainerClass, self.ValidatorClass, self.PredictorClass = \
             self._guess_ops_from_task(self.task)
-        self.model = attempt_load_weights(weights, fuse=False)
 
     def reset(self):
         """
@@ -189,7 +188,7 @@ class YOLO:
             raise AttributeError("dataset not provided! Please define `data` in config.yaml or pass as an argument.")
 
         self.trainer = self.TrainerClass(overrides=overrides)
-        self.trainer.model = self.trainer.load_model(weights=self.ckpt,
+        self.trainer.model = self.trainer.load_model(weights=self.model,
                                                      model_cfg=self.model.yaml if self.task != "classify" else None)
         self.model = self.trainer.model  # override here to save memory
 

ultralytics/yolo/engine/validator.py

@@ -106,6 +106,9 @@ class BaseValidator:
                 data = check_dataset_yaml(self.args.data)
             else:
                 data = check_dataset(self.args.data)
+
+            if self.device.type == 'cpu':
+                self.args.workers = 0  # faster CPU val as time dominated by inference, not dataloading
             self.dataloader = self.dataloader or \
                               self.get_dataloader(data.get("val") or data.set("test"), self.args.batch_size)
 

ultralytics/yolo/utils/__init__.py

@@ -271,19 +271,20 @@ def yaml_save(file='data.yaml', data=None):
         yaml.safe_dump({k: str(v) if isinstance(v, Path) else v for k, v in data.items()}, f, sort_keys=False)
 
 
-def yaml_load(file='data.yaml'):
+def yaml_load(file='data.yaml', append_filename=True):
     """
     Load YAML data from a file.
 
     Args:
         file (str, optional): File name. Default is 'data.yaml'.
+        append_filename (bool): Add the YAML filename to the YAML dictionary. Default is True.
 
     Returns:
         dict: YAML data and file name.
     """
     with open(file, errors='ignore') as f:
         # Add YAML filename to dict and return
-        return {**yaml.safe_load(f), 'yaml_file': str(file)}
+        return {**yaml.safe_load(f), 'yaml_file': str(file)} if append_filename else yaml.safe_load(f)
 
 
 def get_settings(file=USER_CONFIG_DIR / 'settings.yaml'):

ultralytics/yolo/v8/detect/train.py

@@ -54,7 +54,7 @@ class DetectionTrainer(BaseTrainer):
         self.model.names = self.data["names"]
 
     def load_model(self, model_cfg=None, weights=None, verbose=True):
-        model = DetectionModel(model_cfg or weights["model"].yaml, ch=3, nc=self.data["nc"], verbose=verbose)
+        model = DetectionModel(model_cfg or weights.yaml, ch=3, nc=self.data["nc"], verbose=verbose)
         if weights:
             model.load(weights, verbose)
         return model

ultralytics/yolo/v8/segment/train.py

@@ -17,7 +17,7 @@ from ultralytics.yolo.utils.torch_utils import de_parallel
 class SegmentationTrainer(v8.detect.DetectionTrainer):
 
     def load_model(self, model_cfg=None, weights=None, verbose=True):
-        model = SegmentationModel(model_cfg or weights["model"].yaml, ch=3, nc=self.data["nc"], verbose=verbose)
+        model = SegmentationModel(model_cfg or weights.yaml, ch=3, nc=self.data["nc"], verbose=verbose)
         if weights:
             model.load(weights, verbose)
         return model