Fix model re-fuse() in inference loops (#466)

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

ultralytics/__init__.py

@@ -1,6 +1,6 @@
 # Ultralytics YOLO 🚀, GPL-3.0 license
 
-__version__ = "8.0.8"
+__version__ = "8.0.9"
 
 from ultralytics.yolo.engine.model import YOLO
 from ultralytics.yolo.utils import ops

ultralytics/nn/tasks.py

@@ -63,7 +63,8 @@ class BaseModel(nn.Module):
 
     def _profile_one_layer(self, m, x, dt):
         """
-        Profile the computation time and FLOPs of a single layer of the model on a given input. Appends the results to the provided list.
+        Profile the computation time and FLOPs of a single layer of the model on a given input.
+        Appends the results to the provided list.
 
         Args:
             m (nn.Module): The layer to be profiled.
@@ -74,10 +75,10 @@ class BaseModel(nn.Module):
             None
         """
         c = m == self.model[-1]  # is final layer, copy input as inplace fix
-        o = thop.profile(m, inputs=(x.copy() if c else x,), verbose=False)[0] / 1E9 * 2 if thop else 0  # FLOPs
+        o = thop.profile(m, inputs=(x.clone() if c else x,), verbose=False)[0] / 1E9 * 2 if thop else 0  # FLOPs
         t = time_sync()
         for _ in range(10):
-            m(x.copy() if c else x)
+            m(x.clone() if c else x)
         dt.append((time_sync() - t) * 100)
         if m == self.model[0]:
             LOGGER.info(f"{'time (ms)':>10s} {'GFLOPs':>10s} {'params':>10s}  module")
@@ -87,20 +88,36 @@ class BaseModel(nn.Module):
 
     def fuse(self):
         """
-        Fuse the `Conv2d()` and `BatchNorm2d()` layers of the model into a single layer, in order to improve the computation efficiency.
+        Fuse the `Conv2d()` and `BatchNorm2d()` layers of the model into a single layer, in order to improve the
+        computation efficiency.
 
         Returns:
             (nn.Module): The fused model is returned.
         """
-        LOGGER.info('Fusing layers... ')
+        if not self.is_fused():
+            LOGGER.info('Fusing... ')
             for m in self.model.modules():
                 if isinstance(m, (Conv, DWConv)) and hasattr(m, 'bn'):
                     m.conv = fuse_conv_and_bn(m.conv, m.bn)  # update conv
                     delattr(m, 'bn')  # remove batchnorm
                     m.forward = m.forward_fuse  # update forward
             self.info()
+
         return self
 
+    def is_fused(self, thresh=10):
+        """
+        Check if the model has less than a certain threshold of BatchNorm layers.
+
+        Args:
+            thresh (int, optional): The threshold number of BatchNorm layers. Default is 10.
+
+        Returns:
+            bool: True if the number of BatchNorm layers in the model is less than the threshold, False otherwise.
+        """
+        bn = tuple(v for k, v in nn.__dict__.items() if 'Norm' in k)  # normalization layers, i.e. BatchNorm2d()
+        return sum(isinstance(v, bn) for v in self.modules()) < thresh  # True if < 'thresh' BatchNorm layers in model
+
     def info(self, verbose=False, imgsz=640):
         """
         Prints model information

ultralytics/yolo/cli.py

@@ -1,7 +1,9 @@
 # Ultralytics YOLO 🚀, GPL-3.0 license
 
 import argparse
+import re
 import shutil
+import sys
 from pathlib import Path
 
 from ultralytics import __version__, yolo
@@ -17,7 +19,7 @@ CLI_HELP_MSG = \
 
         pip install ultralytics
 
-    2. Train, Val, Predict and Export using 'yolo' commands of the form:
+    2. Train, Val, Predict and Export using 'yolo' commands:
 
             yolo TASK MODE ARGS
 
@@ -97,9 +99,14 @@ def entrypoint():
     It uses the package's default config and initializes it using the passed overrides.
     Then it calls the CLI function with the composed config
     """
+    if len(sys.argv) == 1:  # no arguments passed
+        LOGGER.info(CLI_HELP_MSG)
+        return
+
     parser = argparse.ArgumentParser(description='YOLO parser')
     parser.add_argument('args', type=str, nargs='+', help='YOLO args')
     args = parser.parse_args().args
+    args = re.sub(r'\s*=\s*', '=', ' '.join(args)).split(' ')  # remove whitespaces around = sign
 
     tasks = 'detect', 'segment', 'classify'
     modes = 'train', 'val', 'predict', 'export'

ultralytics/yolo/configs/default.yaml

@@ -8,7 +8,7 @@ mode: "train" # choices=['train', 'val', 'predict'] # mode to run task in.
 model: null # i.e. yolov8n.pt, yolov8n.yaml. Path to model file
 data: null # i.e. coco128.yaml. Path to data file
 epochs: 100 # number of epochs to train for
-patience: 50  # TODO: epochs to wait for no observable improvement for early stopping of training
+patience: 50  # epochs to wait for no observable improvement for early stopping of training
 batch: 16 # number of images per batch
 imgsz: 640 # size of input images
 save: True # save checkpoints

ultralytics/yolo/data/datasets/Argoverse.yaml

@@ -28,10 +28,9 @@ names:
 # Download script/URL (optional) ---------------------------------------------------------------------------------------
 download: |
   import json
-
   from tqdm import tqdm
-  from utils.general import download, Path
+  from ultralytics.yolo.utils.downloads import download
-
+  from pathlib import Path
 
   def argoverse2yolo(set):
       labels = {}

ultralytics/yolo/data/datasets/GlobalWheat2020.yaml

@@ -32,8 +32,8 @@ names:
 
 # Download script/URL (optional) ---------------------------------------------------------------------------------------
 download: |
-  from utils.general import download, Path
+  from ultralytics.yolo.utils.downloads import download
-
+  from pathlib import Path
 
   # Download
   dir = Path(yaml['path'])  # dataset root dir

ultralytics/yolo/data/datasets/Objects365.yaml

@@ -386,7 +386,12 @@ names:
 download: |
   from tqdm import tqdm
 
-  from utils.general import Path, check_requirements, download, np, xyxy2xywhn
+  from ultralytics.yolo.utils.checks import check_requirements
+  from ultralytics.yolo.utils.downloads import download
+  from ultralytics.yolo.utils.ops import xyxy2xywhn
+
+  import numpy as np
+  from pathlib import Path
 
   check_requirements(('pycocotools>=2.0',))
   from pycocotools.coco import COCO

ultralytics/yolo/data/datasets/SKU-110K.yaml

@@ -21,9 +21,14 @@ names:
 # Download script/URL (optional) ---------------------------------------------------------------------------------------
 download: |
   import shutil
+  from pathlib import Path
+
+  import numpy as np
+  import pandas as pd
   from tqdm import tqdm
-  from utils.general import np, pd, Path, download, xyxy2xywh
 
+  from ultralytics.yolo.utils.downloads import download
+  from ultralytics.yolo.utils.ops import xyxy2xywh
 
   # Download
   dir = Path(yaml['path'])  # dataset root dir

ultralytics/yolo/data/datasets/VOC.yaml

@@ -48,8 +48,8 @@ download: |
   import xml.etree.ElementTree as ET
 
   from tqdm import tqdm
-  from utils.general import download, Path
+  from ultralytics.yolo.utils.downloads import download
-
+  from pathlib import Path
 
   def convert_label(path, lb_path, year, image_id):
       def convert_box(size, box):

ultralytics/yolo/data/datasets/VisDrone.yaml

@@ -29,7 +29,10 @@ names:
 
 # Download script/URL (optional) ---------------------------------------------------------------------------------------
 download: |
-  from utils.general import download, os, Path
+  import os
+  from pathlib import Path
+
+  from ultralytics.yolo.utils.downloads import download
 
   def visdrone2yolo(dir):
       from PIL import Image

ultralytics/yolo/data/datasets/coco.yaml

@@ -99,7 +99,9 @@ names:
 
 # Download script/URL (optional)
 download: |
-  from utils.general import download, Path
+  from ultralytics.yoloutils.downloads import download
+  from pathlib import Path
+
   # Download labels
   segments = True  # segment or box labels
   dir = Path(yaml['path'])  # dataset root dir

ultralytics/yolo/data/datasets/xView.yaml

@@ -87,8 +87,8 @@ download: |
   from PIL import Image
   from tqdm import tqdm
 
-  from utils.dataloaders import autosplit
+  from ultralytics.yolo.data.dataloaders.v5loader import autosplit
-  from utils.general import download, xyxy2xywhn
+  from ultralytics.yolo.utils.ops import xyxy2xywhn
 
 
   def convert_labels(fname=Path('xView/xView_train.geojson')):

ultralytics/yolo/engine/model.py

@@ -7,7 +7,7 @@ from ultralytics.nn.tasks import ClassificationModel, DetectionModel, Segmentati
 from ultralytics.yolo.configs import get_config
 from ultralytics.yolo.engine.exporter import Exporter
 from ultralytics.yolo.utils import DEFAULT_CONFIG, LOGGER, yaml_load
-from ultralytics.yolo.utils.checks import check_imgsz, check_yaml
+from ultralytics.yolo.utils.checks import check_yaml
 from ultralytics.yolo.utils.torch_utils import guess_task_from_head, smart_inference_mode
 
 # Map head to model, trainer, validator, and predictor classes
@@ -43,6 +43,7 @@ class YOLO:
         self.TrainerClass = None  # trainer class
         self.ValidatorClass = None  # validator class
         self.PredictorClass = None  # predictor class
+        self.predictor = None  # reuse predictor
         self.model = None  # model object
         self.trainer = None  # trainer object
         self.task = None  # task type
@@ -131,11 +132,12 @@ class YOLO:
         overrides.update(kwargs)
         overrides["mode"] = "predict"
         overrides["save"] = kwargs.get("save", False)  # not save files by default
-        predictor = self.PredictorClass(overrides=overrides)
+        if not self.predictor:
-
+            self.predictor = self.PredictorClass(overrides=overrides)
-        predictor.args.imgsz = check_imgsz(predictor.args.imgsz, min_dim=2)  # check image size
+            self.predictor.setup_model(model=self.model)
-        predictor.setup(model=self.model, source=source)
+        else:  # only update args if predictor is already setup
-        return predictor(stream=stream, verbose=verbose)
+            self.predictor.args = get_config(self.predictor.args, overrides)
+        return self.predictor(source=source, stream=stream, verbose=verbose)
 
     @smart_inference_mode()
     def val(self, data=None, **kwargs):
@@ -170,6 +172,7 @@ class YOLO:
         args = get_config(config=DEFAULT_CONFIG, overrides=overrides)
         args.task = self.task
 
+        print(args)
         exporter = Exporter(overrides=args)
         exporter(model=self.model)
 
@@ -224,10 +227,14 @@ class YOLO:
     def _reset_ckpt_args(args):
         args.pop("project", None)
         args.pop("name", None)
+        args.pop("exist_ok", None)
+        args.pop("resume", None)
         args.pop("batch", None)
         args.pop("epochs", None)
         args.pop("cache", None)
         args.pop("save_json", None)
+        args.pop("half", None)
+        args.pop("v5loader", None)
 
         # set device to '' to prevent from auto DDP usage
         args["device"] = ''

ultralytics/yolo/engine/predictor.py

@@ -76,15 +76,15 @@ class BasePredictor:
         project = self.args.project or Path(SETTINGS['runs_dir']) / self.args.task
         name = self.args.name or f"{self.args.mode}"
         self.save_dir = increment_path(Path(project) / name, exist_ok=self.args.exist_ok)
-        if self.args.save:
-            (self.save_dir / 'labels' if self.args.save_txt else self.save_dir).mkdir(parents=True, exist_ok=True)
         if self.args.conf is None:
             self.args.conf = 0.25  # default conf=0.25
-        self.done_setup = False
+        self.done_warmup = False
 
         # Usable if setup is done
         self.model = None
         self.data = self.args.data  # data_dict
+        self.bs = None
+        self.imgsz = None
         self.device = None
         self.dataset = None
         self.vid_path, self.vid_writer = None, None
@@ -105,11 +105,13 @@ class BasePredictor:
     def postprocess(self, preds, img, orig_img):
         return preds
 
-    def setup(self, source=None, model=None):
+    def setup_source(self, source=None):
+        if not self.model:
+            raise Exception("setup model before setting up source!")
         # source
         source, webcam, screenshot, from_img = self.check_source(source)
         # model
-        stride, pt = self.setup_model(model)
+        stride, pt = self.model.stride, self.model.pt
         imgsz = check_imgsz(self.args.imgsz, stride=stride, min_dim=2)  # check image size
 
         # Dataloader
@@ -143,14 +145,12 @@ class BasePredictor:
                                       transforms=getattr(self.model.model, 'transforms', None),
                                       vid_stride=self.args.vid_stride)
         self.vid_path, self.vid_writer = [None] * bs, [None] * bs
-        self.model.warmup(imgsz=(1 if pt or self.model.triton else bs, 3, *imgsz))  # warmup
 
         self.webcam = webcam
         self.screenshot = screenshot
         self.from_img = from_img
         self.imgsz = imgsz
-        self.done_setup = True
+        self.bs = bs
-        return model
 
     @smart_inference_mode()
     def __call__(self, source=None, model=None, verbose=False, stream=False):
@@ -167,8 +167,20 @@ class BasePredictor:
 
     def stream_inference(self, source=None, model=None, verbose=False):
         self.run_callbacks("on_predict_start")
-        if not self.done_setup:
+
-            self.setup(source, model)
+        # setup model
+        if not self.model:
+            self.setup_model(model)
+        # setup source. Run every time predict is called
+        self.setup_source(source)
+        # check if save_dir/ label file exists
+        if self.args.save:
+            (self.save_dir / 'labels' if self.args.save_txt else self.save_dir).mkdir(parents=True, exist_ok=True)
+        # warmup model
+        if not self.done_warmup:
+            self.model.warmup(imgsz=(1 if self.model.pt or self.model.triton else self.bs, 3, *self.imgsz))
+            self.done_warmup = True
+
         self.seen, self.windows, self.dt = 0, [], (ops.Profile(), ops.Profile(), ops.Profile())
         for batch in self.dataset:
             self.run_callbacks("on_predict_batch_start")
@@ -223,11 +235,9 @@ class BasePredictor:
         device = select_device(self.args.device)
         model = model or self.args.model
         self.args.half &= device.type != 'cpu'  # half precision only supported on CUDA
-        model = AutoBackend(model, device=device, dnn=self.args.dnn, fp16=self.args.half)
+        self.model = AutoBackend(model, device=device, dnn=self.args.dnn, fp16=self.args.half)
-        self.model = model
         self.device = device
         self.model.eval()
-        return model.stride, model.pt
 
     def check_source(self, source):
         source = source if source is not None else self.args.source

ultralytics/yolo/engine/results.py

@@ -85,11 +85,11 @@ class Results:
 
     def __repr__(self):
         s = f'Ultralytics YOLO {self.__class__} instance\n'  # string
-        if self.boxes:
+        if self.boxes is not None:
             s = s + self.boxes.__repr__() + '\n'
-        if self.masks:
+        if self.masks is not None:
             s = s + self.masks.__repr__() + '\n'
-        if self.probs:
+        if self.probs is not None:
             s = s + self.probs.__repr__()
         s += f'original size: {self.orig_shape}\n'
 

ultralytics/yolo/engine/trainer.py

@@ -205,7 +205,7 @@ class BaseTrainer:
             self.model = DDP(self.model, device_ids=[rank])
         # Check imgsz
         gs = max(int(self.model.stride.max() if hasattr(self.model, 'stride') else 32), 32)  # grid size (max stride)
-        self.args.imgsz = check_imgsz(self.args.imgsz, stride=gs, floor=gs * 2)
+        self.args.imgsz = check_imgsz(self.args.imgsz, stride=gs, floor=gs)
         # Batch size
         if self.batch_size == -1:
             if RANK == -1:  # single-GPU only, estimate best batch size

ultralytics/yolo/utils/__init__.py

@@ -372,7 +372,14 @@ def set_sentry(dsn=None):
         import sentry_sdk  # noqa
 
         import ultralytics
-        sentry_sdk.init(dsn=dsn, traces_sample_rate=1.0, release=ultralytics.__version__, debug=False)
+        sentry_sdk.init(
+            dsn=dsn,
+            debug=False,
+            traces_sample_rate=1.0,
+            release=ultralytics.__version__,
+            send_default_pii=True,
+            environment='production',  # 'dev' or 'production'
+            ignore_errors=[KeyboardInterrupt, torch.cuda.OutOfMemoryError])
 
 
 def get_settings(file=USER_CONFIG_DIR / 'settings.yaml', version='0.0.1'):

ultralytics/yolo/v8/classify/predict.py

@@ -5,7 +5,7 @@ import torch
 
 from ultralytics.yolo.engine.predictor import BasePredictor
 from ultralytics.yolo.engine.results import Results
-from ultralytics.yolo.utils import DEFAULT_CONFIG, ROOT
+from ultralytics.yolo.utils import DEFAULT_CONFIG, ROOT, is_git_directory
 from ultralytics.yolo.utils.plotting import Annotator
 
 
@@ -67,7 +67,8 @@ class ClassificationPredictor(BasePredictor):
 @hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
 def predict(cfg):
     cfg.model = cfg.model or "yolov8n-cls.pt"  # or "resnet18"
-    cfg.source = cfg.source if cfg.source is not None else ROOT / "assets"
+    cfg.source = cfg.source if cfg.source is not None else ROOT / "assets" if is_git_directory() \
+        else "https://ultralytics.com/images/bus.jpg"
     predictor = ClassificationPredictor(cfg)
     predictor.predict_cli()
 

ultralytics/yolo/v8/classify/train.py

@@ -140,10 +140,13 @@ class ClassificationTrainer(BaseTrainer):
 def train(cfg):
     cfg.model = cfg.model or "yolov8n-cls.pt"  # or "resnet18"
     cfg.data = cfg.data or "mnist160"  # or yolo.ClassificationDataset("mnist")
-    cfg.lr0 = 0.1
+
-    cfg.weight_decay = 5e-5
+    # Reproduce ImageNet results
-    cfg.label_smoothing = 0.1
+    # cfg.lr0 = 0.1
-    cfg.warmup_epochs = 0.0
+    # cfg.weight_decay = 5e-5
+    # cfg.label_smoothing = 0.1
+    # cfg.warmup_epochs = 0.0
+
     cfg.device = cfg.device if cfg.device is not None else ''
     # trainer = ClassificationTrainer(cfg)
     # trainer.train()

ultralytics/yolo/v8/detect/predict.py

@@ -5,7 +5,7 @@ import torch
 
 from ultralytics.yolo.engine.predictor import BasePredictor
 from ultralytics.yolo.engine.results import Results
-from ultralytics.yolo.utils import DEFAULT_CONFIG, ROOT, ops
+from ultralytics.yolo.utils import DEFAULT_CONFIG, ROOT, is_git_directory, ops
 from ultralytics.yolo.utils.plotting import Annotator, colors, save_one_box
 
 
@@ -84,7 +84,8 @@ class DetectionPredictor(BasePredictor):
 @hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
 def predict(cfg):
     cfg.model = cfg.model or "yolov8n.pt"
-    cfg.source = cfg.source if cfg.source is not None else ROOT / "assets"
+    cfg.source = cfg.source if cfg.source is not None else ROOT / "assets" if is_git_directory() \
+        else "https://ultralytics.com/images/bus.jpg"
     predictor = DetectionPredictor(cfg)
     predictor.predict_cli()
 

ultralytics/yolo/v8/segment/predict.py

@@ -4,7 +4,7 @@ import hydra
 import torch
 
 from ultralytics.yolo.engine.results import Results
-from ultralytics.yolo.utils import DEFAULT_CONFIG, ROOT, ops
+from ultralytics.yolo.utils import DEFAULT_CONFIG, ROOT, is_git_directory, ops
 from ultralytics.yolo.utils.plotting import colors, save_one_box
 from ultralytics.yolo.v8.detect.predict import DetectionPredictor
 
@@ -101,8 +101,8 @@ class SegmentationPredictor(DetectionPredictor):
 @hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
 def predict(cfg):
     cfg.model = cfg.model or "yolov8n-seg.pt"
-    cfg.source = cfg.source if cfg.source is not None else ROOT / "assets"
+    cfg.source = cfg.source if cfg.source is not None else ROOT / "assets" if is_git_directory() \
-
+        else "https://ultralytics.com/images/bus.jpg"
     predictor = SegmentationPredictor(cfg)
     predictor.predict_cli()
 

ultralytics/yolo/v8/segment/val.py

@@ -45,6 +45,7 @@ class SegmentationValidator(DetectionValidator):
         self.jdict = []
         self.stats = []
         if self.args.save_json:
+            check_requirements('pycocotools>=2.0.6')
             self.process = ops.process_mask_upsample  # more accurate
         else:
             self.process = ops.process_mask  # faster
@@ -189,8 +190,9 @@ class SegmentationValidator(DetectionValidator):
         self.plot_masks.clear()
 
     def pred_to_json(self, predn, filename, pred_masks):
-        # Save one JSON result {"image_id": 42, "category_id": 18, "bbox": [258.15, 41.29, 348.26, 243.78], "score": 0.236}
+        # Save one JSON result
-        from pycocotools.mask import encode
+        # Example result = {"image_id": 42, "category_id": 18, "bbox": [258.15, 41.29, 348.26, 243.78], "score": 0.236}
+        from pycocotools.mask import encode  # noqa
 
         def single_encode(x):
             rle = encode(np.asarray(x[:, :, None], order="F", dtype="uint8"))[0]