xdobro23/YOLOv8-16bit
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

YOLOv5 updates (#90)

Browse Source 
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
This commit is contained in:
Glenn Jocher
2022-12-25 14:33:18 +01:00
committed by GitHub
parent ebd3cfb2fd
commit 98815d560f
27 changed files with 281 additions and 161 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
ultralytics/yolo/v8/classify/predict.py
View File

@ -52,7 +52,7 @@ class ClassificationPredictor(BasePredictor):
return log_string
@hydra.main(version_base=None, config_path=DEFAULT_CONFIG.parent, config_name=DEFAULT_CONFIG.name)
@hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
def predict(cfg):
cfg.model = cfg.model or "squeezenet1_0"
sz = cfg.imgsz

2
ultralytics/yolo/v8/classify/train.py
View File

@ -59,7 +59,7 @@ class ClassificationTrainer(BaseTrainer):
pass
@hydra.main(version_base=None, config_path=DEFAULT_CONFIG.parent, config_name=DEFAULT_CONFIG.name)
@hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
def train(cfg):
cfg.model = cfg.model or "resnet18"
cfg.data = cfg.data or "imagenette160" # or yolo.ClassificationDataset("mnist")

2
ultralytics/yolo/v8/classify/val.py
View File

@ -35,7 +35,7 @@ class ClassificationValidator(BaseValidator):
return ["top1", "top5"]
@hydra.main(version_base=None, config_path=DEFAULT_CONFIG.parent, config_name=DEFAULT_CONFIG.name)
@hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
def val(cfg):
cfg.data = cfg.data or "imagenette160"
cfg.model = cfg.model or "resnet18"

2
ultralytics/yolo/v8/detect/predict.py
View File

@ -81,7 +81,7 @@ class DetectionPredictor(BasePredictor):
return log_string
@hydra.main(version_base=None, config_path=DEFAULT_CONFIG.parent, config_name=DEFAULT_CONFIG.name)
@hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
def predict(cfg):
cfg.model = cfg.model or "n.pt"
sz = cfg.imgsz

18
ultralytics/yolo/v8/detect/train.py
View File

@ -53,7 +53,9 @@ class DetectionTrainer(BaseTrainer):
args=self.args)
def criterion(self, preds, batch):
return Loss(self.model)(preds, batch)
if not hasattr(self, 'compute_loss'):
self.compute_loss = Loss(de_parallel(self.model))
return self.compute_loss(preds, batch)
def label_loss_items(self, loss_items=None, prefix="train"):
# We should just use named tensors here in future
@ -61,8 +63,8 @@ class DetectionTrainer(BaseTrainer):
return dict(zip(keys, loss_items)) if loss_items is not None else keys
def progress_string(self):
return ('\n' + '%11s' * 6) % \
('Epoch', 'GPU_mem', *self.loss_names, 'Size')
return ('\n' + '%11s' * 7) % \
('Epoch', 'GPU_mem', *self.loss_names, 'Instances', 'Size')
def plot_training_samples(self, batch, ni):
images = batch["img"]
@ -79,7 +81,7 @@ class DetectionTrainer(BaseTrainer):
# Criterion class for computing training losses
class Loss:
def __init__(self, model):
def __init__(self, model): # model must be de-paralleled
device = next(model.parameters()).device # get model device
h = model.args # hyperparameters
@ -90,7 +92,7 @@ class Loss:
# Class label smoothing https://arxiv.org/pdf/1902.04103.pdf eqn 3
self.cp, self.cn = smooth_BCE(eps=h.get("label_smoothing", 0.0)) # positive, negative BCE targets
m = de_parallel(model).model[-1] # Detect() module
m = model.model[-1] # Detect() module
self.BCEcls = BCEcls
self.hyp = h
self.stride = m.stride # model strides
@ -169,12 +171,12 @@ class Loss:
return loss.sum() * batch_size, loss.detach() # loss(box, cls, dfl)
@hydra.main(version_base=None, config_path=DEFAULT_CONFIG.parent, config_name=DEFAULT_CONFIG.name)
@hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
def train(cfg):
cfg.model = cfg.model or "models/yolov8n.yaml"
cfg.data = cfg.data or "coco128.yaml" # or yolo.ClassificationDataset("mnist")
cfg.imgsz = 160
cfg.epochs = 5
# cfg.imgsz = 160
# cfg.epochs = 5
trainer = DetectionTrainer(cfg)
trainer.train()

8
ultralytics/yolo/v8/detect/val.py
View File

@ -119,9 +119,9 @@ class DetectionValidator(BaseValidator):
if len(stats) and stats[0].any():
self.metrics.process(*stats)
self.nt_per_class = np.bincount(stats[-1].astype(int), minlength=self.nc) # number of targets per class
metrics = {"fitness": self.metrics.fitness()}
metrics |= zip(self.metric_keys, self.metrics.mean_results())
return metrics
fitness = {"fitness": self.metrics.fitness()}
metrics = dict(zip(self.metric_keys, self.metrics.mean_results()))
return {**metrics, **fitness}
def print_results(self):
pf = '%22s' + '%11i' * 2 + '%11.3g' * len(self.metric_keys) # print format
@ -198,7 +198,7 @@ class DetectionValidator(BaseValidator):
names=self.names) # pred
@hydra.main(version_base=None, config_path=DEFAULT_CONFIG.parent, config_name=DEFAULT_CONFIG.name)
@hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
def val(cfg):
cfg.data = cfg.data or "coco128.yaml"
validator = DetectionValidator(args=cfg)

2
ultralytics/yolo/v8/segment/predict.py
View File

@ -99,7 +99,7 @@ class SegmentationPredictor(DetectionPredictor):
return log_string
@hydra.main(version_base=None, config_path=DEFAULT_CONFIG.parent, config_name=DEFAULT_CONFIG.name)
@hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
def predict(cfg):
cfg.model = cfg.model or "n.pt"
sz = cfg.imgsz

6
ultralytics/yolo/v8/segment/train.py
View File

@ -214,8 +214,8 @@ class SegmentationTrainer(DetectionTrainer):
return dict(zip(keys, loss_items)) if loss_items is not None else keys
def progress_string(self):
return ('\n' + '%11s' * 7) % \
('Epoch', 'GPU_mem', *self.loss_names, 'Size')
return ('\n' + '%11s' * 8) % \
('Epoch', 'GPU_mem', *self.loss_names, 'Instances', 'Size')
def plot_training_samples(self, batch, ni):
images = batch["img"]
@ -230,7 +230,7 @@ class SegmentationTrainer(DetectionTrainer):
plot_results(file=self.csv, segment=True) # save results.png
@hydra.main(version_base=None, config_path=DEFAULT_CONFIG.parent, config_name=DEFAULT_CONFIG.name)
@hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
def train(cfg):
cfg.model = cfg.model or "models/yolov8n-seg.yaml"
cfg.data = cfg.data or "coco128-seg.yaml" # or yolo.ClassificationDataset("mnist")

2
ultralytics/yolo/v8/segment/val.py
View File

@ -211,7 +211,7 @@ class SegmentationValidator(DetectionValidator):
self.plot_masks.clear()
@hydra.main(version_base=None, config_path=DEFAULT_CONFIG.parent, config_name=DEFAULT_CONFIG.name)
@hydra.main(version_base=None, config_path=str(DEFAULT_CONFIG.parent), config_name=DEFAULT_CONFIG.name)
def val(cfg):
cfg.data = cfg.data or "coco128-seg.yaml"
validator = SegmentationValidator(args=cfg)