Cleanup argument handling in Model class (#4614)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Sebastian Stapf <42514241+Wiqzard@users.noreply.github.com>

docs/reference/cfg/__init__.md

@@ -17,6 +17,10 @@ keywords: Ultralytics, YOLO, Configuration, cfg2dict, handle_deprecation, merge_
 ## ::: ultralytics.cfg.get_cfg
 <br><br>
 
+---
+## ::: ultralytics.cfg.get_save_dir
+<br><br>
+
 ---
 ## ::: ultralytics.cfg._handle_deprecation
 <br><br>

tests/test_python.py

@@ -182,9 +182,11 @@ def test_export_openvino():
 
 def test_export_coreml():
     if not WINDOWS:  # RuntimeError: BlobWriter not loaded with coremltools 7.0 on windows
-        f = YOLO(MODEL).export(format='coreml', nms=True)
         if MACOS:
-            YOLO(f)(SOURCE)  # model prediction only supported on macOS
+            f = YOLO(MODEL).export(format='coreml')
+            YOLO(f)(SOURCE)  # model prediction only supported on macOS for nms=False models
+        else:
+            YOLO(MODEL).export(format='coreml', nms=True)
 
 
 def test_export_tflite(enabled=False):

ultralytics/engine/model.py

@@ -5,12 +5,10 @@ import sys
 from pathlib import Path
 from typing import Union
 
-from ultralytics.cfg import get_cfg, get_save_dir
+from ultralytics.cfg import TASK2DATA, get_cfg, get_save_dir
-from ultralytics.engine.exporter import Exporter
 from ultralytics.hub.utils import HUB_WEB_ROOT
 from ultralytics.nn.tasks import attempt_load_one_weight, guess_model_task, nn, yaml_model_load
-from ultralytics.utils import (ASSETS, DEFAULT_CFG, DEFAULT_CFG_DICT, DEFAULT_CFG_KEYS, LOGGER, RANK, callbacks, emojis,
+from ultralytics.utils import ASSETS, DEFAULT_CFG_DICT, DEFAULT_CFG_KEYS, LOGGER, RANK, callbacks, emojis, yaml_load
-                               yaml_load)
 from ultralytics.utils.checks import check_file, check_imgsz, check_pip_update_available, check_yaml
 from ultralytics.utils.downloads import GITHUB_ASSETS_STEMS
 from ultralytics.utils.torch_utils import smart_inference_mode
@@ -118,9 +116,9 @@ class Model:
         cfg_dict = yaml_model_load(cfg)
         self.cfg = cfg
         self.task = task or guess_model_task(cfg_dict)
-        model = model or self.smart_load('model')
+        self.model = (model or self.smart_load('model'))(cfg_dict, verbose=verbose and RANK == -1)  # build model
-        self.model = model(cfg_dict, verbose=verbose and RANK == -1)  # build model
         self.overrides['model'] = self.cfg
+        self.overrides['task'] = self.task
 
         # Below added to allow export from YAMLs
         args = {**DEFAULT_CFG_DICT, **self.overrides}  # combine model and default args, preferring model args
@@ -220,28 +218,22 @@ class Model:
         if source is None:
             source = ASSETS
             LOGGER.warning(f"WARNING ⚠️ 'source' is missing. Using 'source={source}'.")
+
         is_cli = (sys.argv[0].endswith('yolo') or sys.argv[0].endswith('ultralytics')) and any(
             x in sys.argv for x in ('predict', 'track', 'mode=predict', 'mode=track'))
-        # Check prompts for SAM/FastSAM
+
-        prompts = kwargs.pop('prompts', None)
+        custom = {'conf': 0.25, 'save': is_cli}  # method defaults
-        overrides = self.overrides.copy()
+        args = {**self.overrides, **custom, **kwargs, 'mode': 'predict'}  # highest priority args on the right
-        overrides['conf'] = 0.25
+        prompts = args.pop('prompts', None)  # for SAM-type models
-        overrides.update(kwargs)  # prefer kwargs
+
-        overrides['mode'] = kwargs.get('mode', 'predict')
-        assert overrides['mode'] in ['track', 'predict']
-        if not is_cli:
-            overrides['save'] = kwargs.get('save', False)  # do not save by default if called in Python
         if not self.predictor:
-            self.task = overrides.get('task') or self.task
+            self.predictor = (predictor or self.smart_load('predictor'))(overrides=args, _callbacks=self.callbacks)
-            predictor = predictor or self.smart_load('predictor')
-            self.predictor = predictor(overrides=overrides, _callbacks=self.callbacks)
             self.predictor.setup_model(model=self.model, verbose=is_cli)
         else:  # only update args if predictor is already setup
-            self.predictor.args = get_cfg(self.predictor.args, overrides)
+            self.predictor.args = get_cfg(self.predictor.args, args)
-            if 'project' in overrides or 'name' in overrides:
+            if 'project' in args or 'name' in args:
                 self.predictor.save_dir = get_save_dir(self.predictor.args)
-        # Set prompts for SAM/FastSAM
+        if prompts and hasattr(self.predictor, 'set_prompts'):  # for SAM-type models
-        if len and hasattr(self.predictor, 'set_prompts'):
             self.predictor.set_prompts(prompts)
         return self.predictor.predict_cli(source=source) if is_cli else self.predictor(source=source, stream=stream)
 
@@ -257,46 +249,31 @@ class Model:
 
         Returns:
             (List[ultralytics.engine.results.Results]): The tracking results.
-
         """
         if not hasattr(self.predictor, 'trackers'):
             from ultralytics.trackers import register_tracker
             register_tracker(self, persist)
         # ByteTrack-based method needs low confidence predictions as input
-        conf = kwargs.get('conf') or 0.1
+        kwargs['conf'] = kwargs.get('conf') or 0.1
-        kwargs['conf'] = conf
         kwargs['mode'] = 'track'
         return self.predict(source=source, stream=stream, **kwargs)
 
     @smart_inference_mode()
-    def val(self, data=None, validator=None, **kwargs):
+    def val(self, validator=None, **kwargs):
         """
         Validate a model on a given dataset.
 
         Args:
-            data (str): The dataset to validate on. Accepts all formats accepted by yolo
             validator (BaseValidator): Customized validator.
             **kwargs : Any other args accepted by the validators. To see all args check 'configuration' section in docs
         """
-        overrides = self.overrides.copy()
+        custom = {'rect': True}  # method defaults
-        overrides['rect'] = True  # rect batches as default
+        args = {**self.overrides, **custom, **kwargs, 'mode': 'val'}  # highest priority args on the right
-        overrides.update(kwargs)
+        args['imgsz'] = check_imgsz(args['imgsz'], max_dim=1)
-        overrides['mode'] = 'val'
-        if overrides.get('imgsz') is None:
-            overrides['imgsz'] = self.model.args['imgsz']  # use trained imgsz unless custom value is passed
-        args = get_cfg(cfg=DEFAULT_CFG, overrides=overrides)
-        args.data = data or args.data
-        if 'task' in overrides:
-            self.task = args.task
-        else:
-            args.task = self.task
-        validator = validator or self.smart_load('validator')
-        args.imgsz = check_imgsz(args.imgsz, max_dim=1)
 
-        validator = validator(args=args, _callbacks=self.callbacks)
+        validator = (validator or self.smart_load('validator'))(args=args, _callbacks=self.callbacks)
         validator(model=self.model)
         self.metrics = validator.metrics
-
         return validator.metrics
 
     @smart_inference_mode()
@@ -309,17 +286,16 @@ class Model:
         """
         self._check_is_pytorch_model()
         from ultralytics.utils.benchmarks import benchmark
-        overrides = self.model.args.copy()
+
-        overrides.update(kwargs)
+        custom = {'verbose': False}  # method defaults
-        overrides['mode'] = 'benchmark'
+        args = {**DEFAULT_CFG_DICT, **self.model.args, **custom, **kwargs, 'mode': 'benchmark'}
-        overrides = {**DEFAULT_CFG_DICT, **overrides}  # fill in missing overrides keys with defaults
         return benchmark(
             model=self,
             data=kwargs.get('data'),  # if no 'data' argument passed set data=None for default datasets
-            imgsz=overrides['imgsz'],
+            imgsz=args['imgsz'],
-            half=overrides['half'],
+            half=args['half'],
-            int8=overrides['int8'],
+            int8=args['int8'],
-            device=overrides['device'],
+            device=args['device'],
             verbose=kwargs.get('verbose'))
 
     def export(self, **kwargs):
@@ -327,22 +303,13 @@ class Model:
         Export model.
 
         Args:
-            **kwargs : Any other args accepted by the predictors. To see all args check 'configuration' section in docs
+            **kwargs : Any other args accepted by the Exporter. To see all args check 'configuration' section in docs.
         """
         self._check_is_pytorch_model()
-        overrides = self.overrides.copy()
+        from .exporter import Exporter
-        overrides.update(kwargs)
+
-        overrides['mode'] = 'export'
+        custom = {'imgsz': self.model.args['imgsz'], 'batch': 1, 'data': None, 'verbose': False}  # method defaults
-        if overrides.get('imgsz') is None:
+        args = {**self.overrides, **custom, **kwargs, 'mode': 'export'}  # highest priority args on the right
-            overrides['imgsz'] = self.model.args['imgsz']  # use trained imgsz unless custom value is passed
-        if 'batch' not in kwargs:
-            overrides['batch'] = 1  # default to 1 if not modified
-        if 'data' not in kwargs:
-            overrides['data'] = None  # default to None if not modified (avoid int8 calibration with coco.yaml)
-        if 'verbose' not in kwargs:
-            overrides['verbose'] = False
-        args = get_cfg(cfg=DEFAULT_CFG, overrides=overrides)
-        args.task = self.task
         return Exporter(overrides=args, _callbacks=self.callbacks)(model=self.model)
 
     def train(self, trainer=None, **kwargs):
@@ -359,20 +326,15 @@ class Model:
                 LOGGER.warning('WARNING ⚠️ using HUB training arguments, ignoring local training arguments.')
             kwargs = self.session.train_args
         check_pip_update_available()
-        overrides = self.overrides.copy()
+
-        if kwargs.get('cfg'):
+        overrides = yaml_load(check_yaml(kwargs['cfg'])) if kwargs.get('cfg') else self.overrides
-            LOGGER.info(f"cfg file passed. Overriding default params with {kwargs['cfg']}.")
+        custom = {'data': TASK2DATA[self.task]}  # method defaults
-            overrides = yaml_load(check_yaml(kwargs['cfg']))
+        args = {**overrides, **custom, **kwargs, 'mode': 'train'}  # highest priority args on the right
-        overrides.update(kwargs)
+        if args.get('resume'):
-        overrides['mode'] = 'train'
+            args['resume'] = self.ckpt_path
-        if not overrides.get('data'):
+
-            raise AttributeError("Dataset required but missing, i.e. pass 'data=coco128.yaml'")
+        self.trainer = (trainer or self.smart_load('trainer'))(overrides=args, _callbacks=self.callbacks)
-        if overrides.get('resume'):
+        if not args.get('resume'):  # manually set model only if not resuming
-            overrides['resume'] = self.ckpt_path
-        self.task = overrides.get('task') or self.task
-        trainer = trainer or self.smart_load('trainer')
-        self.trainer = trainer(overrides=overrides, _callbacks=self.callbacks)
-        if not overrides.get('resume'):  # manually set model only if not resuming
             self.trainer.model = self.trainer.get_model(weights=self.model if self.ckpt else None, cfg=self.model.yaml)
             self.model = self.trainer.model
         self.trainer.hub_session = self.session  # attach optional HUB session
@@ -455,7 +417,7 @@ class Model:
             name = self.__class__.__name__
             mode = inspect.stack()[1][3]  # get the function name.
             raise NotImplementedError(
-                emojis(f'WARNING ⚠️ `{name}` model does not support `{mode}` mode for `{self.task}` task yet.')) from e
+                emojis(f"WARNING ⚠️ '{name}' model does not support '{mode}' mode for '{self.task}' task yet.")) from e
 
     @property
     def task_map(self):

ultralytics/engine/trainer.py

@@ -175,8 +175,13 @@ class BaseTrainer:
         if world_size > 1 and 'LOCAL_RANK' not in os.environ:
             # Argument checks
             if self.args.rect:
-                LOGGER.warning("WARNING ⚠️ 'rect=True' is incompatible with Multi-GPU training, setting rect=False")
+                LOGGER.warning("WARNING ⚠️ 'rect=True' is incompatible with Multi-GPU training, setting 'rect=False'")
                 self.args.rect = False
+            if self.args.batch == -1:
+                LOGGER.warning("WARNING ⚠️ 'batch=-1' for AutoBatch is incompatible with Multi-GPU training, setting "
+                               "default 'batch=16'")
+                self.args.batch = 16
+
             # Command
             cmd, file = generate_ddp_command(world_size, self)
             try:
@@ -186,6 +191,7 @@ class BaseTrainer:
                 raise e
             finally:
                 ddp_cleanup(self, str(file))
+
         else:
             self._do_train(world_size)
 
@@ -248,9 +254,6 @@ class BaseTrainer:
         if self.batch_size == -1:
             if RANK == -1:  # single-GPU only, estimate best batch size
                 self.args.batch = self.batch_size = check_train_batch_size(self.model, self.args.imgsz, self.amp)
-            else:
-                SyntaxError('batch=-1 to use AutoBatch is only available in Single-GPU training. '
-                            'Please pass a valid batch size value for Multi-GPU DDP training, i.e. batch=16')
 
         # Dataloaders
         batch_size = self.batch_size // max(world_size, 1)

ultralytics/nn/autobackend.py

@@ -18,7 +18,6 @@ from PIL import Image
 from ultralytics.utils import ARM64, LINUX, LOGGER, ROOT, yaml_load
 from ultralytics.utils.checks import check_requirements, check_suffix, check_version, check_yaml
 from ultralytics.utils.downloads import attempt_download_asset, is_url
-from ultralytics.utils.ops import xywh2xyxy
 
 
 def check_class_names(names):
@@ -363,9 +362,13 @@ class AutoBackend(nn.Module):
             # im = im.resize((192, 320), Image.BILINEAR)
             y = self.model.predict({'image': im_pil})  # coordinates are xywh normalized
             if 'confidence' in y:
-                box = xywh2xyxy(y['coordinates'] * [[w, h, w, h]])  # xyxy pixels
+                raise TypeError('Ultralytics only supports inference of non-pipelined CoreML models exported with '
-                conf, cls = y['confidence'].max(1), y['confidence'].argmax(1).astype(np.float)
+                                f"'nms=False', but 'model={w}' has an NMS pipeline created by an 'nms=True' export.")
-                y = np.concatenate((box, conf.reshape(-1, 1), cls.reshape(-1, 1)), 1)
+                # TODO: CoreML NMS inference handling
+                # from ultralytics.utils.ops import xywh2xyxy
+                # box = xywh2xyxy(y['coordinates'] * [[w, h, w, h]])  # xyxy pixels
+                # conf, cls = y['confidence'].max(1), y['confidence'].argmax(1).astype(np.float32)
+                # y = np.concatenate((box, conf.reshape(-1, 1), cls.reshape(-1, 1)), 1)
             elif len(y) == 1:  # classification model
                 y = list(y.values())
             elif len(y) == 2:  # segmentation model

ultralytics/nn/modules/transformer.py

@@ -355,15 +355,15 @@ class DeformableTransformerDecoder(nn.Module):
         for i, layer in enumerate(self.layers):
             output = layer(output, refer_bbox, feats, shapes, padding_mask, attn_mask, pos_mlp(refer_bbox))
 
-            # refine bboxes, (bs, num_queries+num_denoising, 4)
+            bbox = bbox_head[i](output)
-            refined_bbox = torch.sigmoid(bbox_head[i](output) + inverse_sigmoid(refer_bbox))
+            refined_bbox = torch.sigmoid(bbox + inverse_sigmoid(refer_bbox))
 
             if self.training:
                 dec_cls.append(score_head[i](output))
                 if i == 0:
                     dec_bboxes.append(refined_bbox)
                 else:
-                    dec_bboxes.append(torch.sigmoid(bbox_head[i](output) + inverse_sigmoid(last_refined_bbox)))
+                    dec_bboxes.append(torch.sigmoid(bbox + inverse_sigmoid(last_refined_bbox)))
             elif i == self.eval_idx:
                 dec_cls.append(score_head[i](output))
                 dec_bboxes.append(refined_bbox)