Python refactor and simplification

export.py

@@ -546,7 +546,7 @@ def export_tfjs(file, int8, prefix=colorstr("TensorFlow.js:")):
         "--quantize_uint8" if int8 else "",
         "--output_node_names=Identity,Identity_1,Identity_2,Identity_3",
         str(f_pb),
-        str(f),
+        f,
     ]
     subprocess.run([arg for arg in args if arg], check=True)
 

models/common.py

@@ -859,11 +859,17 @@ def pandas(self):
     def tolist(self):
         # return a list of Detections objects, i.e. 'for result in results.tolist():'
         r = range(self.n)  # iterable
-        x = [Detections([self.ims[i]], [self.pred[i]], [self.files[i]], self.times, self.names, self.s) for i in r]
-        # for d in x:
-        #    for k in ['ims', 'pred', 'xyxy', 'xyxyn', 'xywh', 'xywhn']:
-        #        setattr(d, k, getattr(d, k)[0])  # pop out of list
-        return x
+        return [
+            Detections(
+                [self.ims[i]],
+                [self.pred[i]],
+                [self.files[i]],
+                self.times,
+                self.names,
+                self.s,
+            )
+            for i in r
+        ]
 
     def print(self):
         LOGGER.info(self.__str__())

train.py

@@ -682,10 +682,7 @@ def main(opt, callbacks=Callbacks()):
             )
 
         # Delete the items in meta dictionary whose first value is False
-        del_ = []
-        for item in meta.keys():
-            if meta[item][0] is False:
-                del_.append(item)
+        del_ = [item for item, value_ in meta.items() if value_[0] is False]
         hyp_GA = hyp.copy()  # Make a copy of hyp dictionary
         for item in del_:
             del meta[item]  # Remove the item from meta dictionary
@@ -696,9 +693,7 @@ def main(opt, callbacks=Callbacks()):
         upper_limit = np.array([meta[k][2] for k in hyp_GA.keys()])
 
         # Create gene_ranges list to hold the range of values for each gene in the population
-        gene_ranges = []
-        for i in range(len(upper_limit)):
-            gene_ranges.append((lower_limit[i], upper_limit[i]))
+        gene_ranges = [(lower_limit[i], upper_limit[i]) for i in range(len(upper_limit))]
 
         # Initialize the population with initial_values or random values
         initial_values = []
@@ -723,25 +718,20 @@ def main(opt, callbacks=Callbacks()):
 
         # Generate random values within the search space for the rest of the population
         if initial_values is None:
-            population = [generate_individual(gene_ranges, len(hyp_GA)) for i in range(pop_size)]
-        else:
-            if pop_size > 1:
-                population = [
-                    generate_individual(gene_ranges, len(hyp_GA)) for i in range(pop_size - len(initial_values))
-                ]
-                for initial_value in initial_values:
-                    population = [initial_value] + population
+            population = [generate_individual(gene_ranges, len(hyp_GA)) for _ in range(pop_size)]
+        elif pop_size > 1:
+            population = [generate_individual(gene_ranges, len(hyp_GA)) for _ in range(pop_size - len(initial_values))]
+            for initial_value in initial_values:
+                population = [initial_value] + population
 
         # Run the genetic algorithm for a fixed number of generations
         list_keys = list(hyp_GA.keys())
         for generation in range(opt.evolve):
             if generation >= 1:
                 save_dict = {}
                 for i in range(len(population)):
-                    little_dict = {}
-                    for j in range(len(population[i])):
-                        little_dict[list_keys[j]] = float(population[i][j])
-                    save_dict["gen" + str(generation) + "number" + str(i)] = little_dict
+                    little_dict = {list_keys[j]: float(population[i][j]) for j in range(len(population[i]))}
+                    save_dict[f"gen{str(generation)}number{str(i)}"] = little_dict
 
                 with open(save_dir / "evolve_population.yaml", "w") as outfile:
                     yaml.dump(save_dict, outfile, default_flow_style=False)
@@ -771,7 +761,7 @@ def main(opt, callbacks=Callbacks()):
 
             # Select the fittest individuals for reproduction using adaptive tournament selection
             selected_indices = []
-            for i in range(pop_size - elite_size):
+            for _ in range(pop_size - elite_size):
                 # Adaptive tournament size
                 tournament_size = max(
                     max(2, tournament_size_min),
@@ -788,7 +778,7 @@ def main(opt, callbacks=Callbacks()):
             selected_indices.extend(elite_indices)
             # Create the next generation through crossover and mutation
             next_generation = []
-            for i in range(pop_size):
+            for _ in range(pop_size):
                 parent1_index = selected_indices[random.randint(0, pop_size - 1)]
                 parent2_index = selected_indices[random.randint(0, pop_size - 1)]
                 # Adaptive crossover rate

utils/downloads.py

@@ -24,9 +24,7 @@ def is_url(url, check=True):
 def gsutil_getsize(url=""):
     # gs://bucket/file size https://cloud.google.com/storage/docs/gsutil/commands/du
     output = subprocess.check_output(["gsutil", "du", url], shell=True, encoding="utf-8")
-    if output:
-        return int(output.split()[0])
-    return 0
+    return int(output.split()[0]) if output else 0
 
 
 def url_getsize(url="https://ultralytics.com/images/bus.jpg"):

utils/general.py

@@ -188,7 +188,7 @@ class Profile(contextlib.ContextDecorator):
     def __init__(self, t=0.0, device: torch.device = None):
         self.t = t
         self.device = device
-        self.cuda = True if (device and str(device)[:4] == "cuda") else False
+        self.cuda = bool(device and str(device).startswith("cuda"))
 
     def __enter__(self):
         self.start = self.time()

utils/loggers/__init__.py

@@ -67,9 +67,7 @@ def _json_default(value):
             value = value.item()
         except ValueError:  # "only one element tensors can be converted to Python scalars"
             pass
-    if isinstance(value, float):
-        return value
-    return str(value)
+    return value if isinstance(value, float) else str(value)
 
 
 class Loggers:
@@ -250,12 +248,12 @@ def on_fit_epoch_end(self, vals, epoch, best_fitness, fi):
                 f.write(s + ("%20.5g," * n % tuple([epoch] + vals)).rstrip(",") + "\n")
         if self.ndjson_console or self.ndjson_file:
             json_data = json.dumps(dict(epoch=epoch, **x), default=_json_default)
-            if self.ndjson_console:
-                print(json_data)
-            if self.ndjson_file:
-                file = self.save_dir / "results.ndjson"
-                with open(file, "a") as f:
-                    print(json_data, file=f)
+        if self.ndjson_console:
+            print(json_data)
+        if self.ndjson_file:
+            file = self.save_dir / "results.ndjson"
+            with open(file, "a") as f:
+                print(json_data, file=f)
 
         if self.tb:
             for k, v in x.items():
@@ -370,10 +368,7 @@ def __init__(self, opt, console_logger, include=("tb", "wandb", "clearml")):
         if clearml and "clearml" in self.include:
             try:
                 # Hyp is not available in classification mode
-                if "hyp" not in opt:
-                    hyp = {}
-                else:
-                    hyp = opt.hyp
+                hyp = {} if "hyp" not in opt else opt.hyp
                 self.clearml = ClearmlLogger(opt, hyp)
             except Exception:
                 self.clearml = None
@@ -427,7 +422,9 @@ def log_graph(self, model, imgsz=(640, 640)):
         if self.tb:
             log_tensorboard_graph(self.tb, model, imgsz)
 
-    def log_model(self, model_path, epoch=0, metadata={}):
+    def log_model(self, model_path, epoch=0, metadata=None):
+        if metadata is None:
+            metadata = {}
         # Log model to all loggers
         if self.wandb:
             art = wandb.Artifact(name=f"run_{wandb.run.id}_model", type="model", metadata=metadata)

utils/loggers/clearml/clearml_utils.py

@@ -31,7 +31,7 @@ def construct_dataset(clearml_info_string):
             "More than one yaml file was found in the dataset root, cannot determine which one contains "
             "the dataset definition this way."
         )
-    elif len(yaml_filenames) == 0:
+    elif not yaml_filenames:
         raise ValueError(
             "No yaml definition found in dataset root path, check that there is a correct yaml file "
             "inside the dataset root path."
@@ -45,7 +45,7 @@ def construct_dataset(clearml_info_string):
         {"train", "test", "val", "nc", "names"}
     ), "The right keys were not found in the yaml file, make sure it at least has the following keys: ('train', 'test', 'val', 'nc', 'names')"
 
-    data_dict = dict()
+    data_dict = {}
     data_dict["train"] = (
         str((dataset_root_path / dataset_definition["train"]).resolve()) if dataset_definition["train"] else None
     )
@@ -96,7 +96,7 @@ def __init__(self, opt, hyp):
         self.data_dict = None
         if self.clearml:
             self.task = Task.init(
-                project_name=opt.project if not str(opt.project).startswith("runs/") else "YOLOv5",
+                project_name="YOLOv5" if str(opt.project).startswith("runs/") else opt.project,
                 task_name=opt.name if opt.name != "exp" else "Training",
                 tags=["YOLOv5"],
                 output_uri=True,
@@ -202,24 +202,26 @@ def log_image_with_boxes(self, image_path, boxes, class_names, image, conf_thres
         class_names (dict): dict containing mapping of class int to class name
         image (Tensor): A torch tensor containing the actual image data
         """
-        if len(self.current_epoch_logged_images) < self.max_imgs_to_log_per_epoch and self.current_epoch >= 0:
-            # Log every bbox_interval times and deduplicate for any intermittend extra eval runs
-            if self.current_epoch % self.bbox_interval == 0 and image_path not in self.current_epoch_logged_images:
-                im = np.ascontiguousarray(np.moveaxis(image.mul(255).clamp(0, 255).byte().cpu().numpy(), 0, 2))
-                annotator = Annotator(im=im, pil=True)
-                for i, (conf, class_nr, box) in enumerate(zip(boxes[:, 4], boxes[:, 5], boxes[:, :4])):
-                    color = colors(i)
-
-                    class_name = class_names[int(class_nr)]
-                    confidence_percentage = round(float(conf) * 100, 2)
-                    label = f"{class_name}: {confidence_percentage}%"
-
-                    if conf > conf_threshold:
-                        annotator.rectangle(box.cpu().numpy(), outline=color)
-                        annotator.box_label(box.cpu().numpy(), label=label, color=color)
-
-                annotated_image = annotator.result()
-                self.task.get_logger().report_image(
-                    title="Bounding Boxes", series=image_path.name, iteration=self.current_epoch, image=annotated_image
-                )
-                self.current_epoch_logged_images.add(image_path)
+        if (
+            len(self.current_epoch_logged_images) < self.max_imgs_to_log_per_epoch
+            and self.current_epoch >= 0
+            and (self.current_epoch % self.bbox_interval == 0 and image_path not in self.current_epoch_logged_images)
+        ):
+            im = np.ascontiguousarray(np.moveaxis(image.mul(255).clamp(0, 255).byte().cpu().numpy(), 0, 2))
+            annotator = Annotator(im=im, pil=True)
+            for i, (conf, class_nr, box) in enumerate(zip(boxes[:, 4], boxes[:, 5], boxes[:, :4])):
+                color = colors(i)
+
+                class_name = class_names[int(class_nr)]
+                confidence_percentage = round(float(conf) * 100, 2)
+                label = f"{class_name}: {confidence_percentage}%"
+
+                if conf > conf_threshold:
+                    annotator.rectangle(box.cpu().numpy(), outline=color)
+                    annotator.box_label(box.cpu().numpy(), label=label, color=color)
+
+            annotated_image = annotator.result()
+            self.task.get_logger().report_image(
+                title="Bounding Boxes", series=image_path.name, iteration=self.current_epoch, image=annotated_image
+            )
+            self.current_epoch_logged_images.add(image_path)