def run(weights='./yolov5s.pt', # weights path
img_size=(640, 640), # image (height, width)
batch_size=1, # batch size
device='cpu', # cuda device, i.e. 0 or 0,1,2,3 or cpu
include=('torchscript', 'onnx', 'coreml'), # include formats
half=False, # FP16 half-precision export
inplace=False, # set YOLOv5 Detect() inplace=True
train=False, # model.train() mode
optimize=False, # TorchScript: optimize for mobile
dynamic=False, # ONNX: dynamic axes
simplify=False, # ONNX: simplify model
opset=12, # ONNX: opset version
):
t = time.time()
include = [x.lower() for x in include]
img_size *= 2 if len(img_size) == 1 else 1 # expand
file = Path(weights)
# Load PyTorch model
device = select_device(device)
assert not (device.type == 'cpu' and half), '--half only compatible with GPU export, i.e. use --device 0'
model = attempt_load(weights, map_location=device) # load FP32 model
names = model.names
# Input
gs = int(max(model.stride)) # grid size (max stride)
img_size = [check_img_size(x, gs) for x in img_size] # verify img_size are gs-multiples
img = torch.zeros(batch_size, 3, *img_size).to(device) # image size(1,3,320,192) iDetection
# Update model
if half:
img, model = img.half(), model.half() # to FP16
model.train() if train else model.eval() # training mode = no Detect() layer grid construction
for k, m in model.named_modules():
if isinstance(m, Conv): # assign export-friendly activations
if isinstance(m.act, nn.Hardswish):
m.act = Hardswish()
elif isinstance(m.act, nn.SiLU):
m.act = SiLU()
elif isinstance(m, Detect):
m.inplace = inplace
m.onnx_dynamic = dynamic
# m.forward = m.forward_export # assign forward (optional)
for _ in range(2):
y = model(img) # dry runs
print(f"\n{colorstr('PyTorch:')} starting from {weights} ({file_size(weights):.1f} MB)")
# Exports
if 'torchscript' in include:
export_torchscript(model, img, file, optimize)
if 'onnx' in include:
export_onnx(model, img, file, opset, train, dynamic, simplify)
if 'coreml' in include:
export_coreml(model, img, file)
# Finish
print(f'\nExport complete ({time.time() - t:.2f}s). Visualize with https://github.com/lutzroeder/netron.')
def yolov5_onnx(pretrained=False, progress=True, num_classes=80, **kwargs):
model = yolov5s(pretrained=pretrained, progress=progress, num_classes=num_classes, **kwargs)
for m in model.modules():
m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatibility
if isinstance(m, Conv) and isinstance(m.act, nn.Hardswish):
m.act = Hardswish() # assign activation
return model
def __init__(self, model_path: str, reso: int = 640, cuda: bool = False):
self.device = torch.device("cuda:0" if cuda else "cpu")
self.half = self.device.type != "cpu"
self.reso = (reso, reso)
self.cuda = cuda
self.model = attempt_load(model_path, map_location=self.device)
# Update model
for k, m in self.model.named_modules():
m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatability
if isinstance(m, models.common.Conv) and isinstance(m.act, nn.Hardswish):
m.act = Hardswish() # assign activation
self.names = (
self.model.module.names
if hasattr(self.model, "module")
else self.model.names
)
if self.half:
self.model.half()
def load_model(weights, device):
# Load model
# model = attempt_load(weights, map_location=device) # load FP32 model
with open('data/coco128.yaml') as f:
data_dict = yaml.load(f, Loader=yaml.FullLoader)
model = Model('models/yolov5s.yaml').to(device)
model.names = data_dict['names']
model = model.fuse().eval()
ckpt = torch.load(weights, map_location=device)
ckpt['model'].float().fuse().eval()
print({
k: (v.shape, model.state_dict()[k].shape)
for k, v in ckpt['model'].float().state_dict().items()
if model.state_dict()[k].shape != v.shape
})
print({
k: (v.shape, ckpt['model'].float().state_dict()[k].shape)
for k, v in model.state_dict().items()
if ckpt['model'].float().state_dict()[k].shape != v.shape
})
ckpt['model'] = {
k: v
for k, v in ckpt['model'].float().state_dict().items()
if model.state_dict()[k].shape == v.shape
}
model.load_state_dict(ckpt['model'], strict=False)
for k, m in model.named_modules():
m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatability
if isinstance(m, Conv) and isinstance(m.act, nn.Hardswish):
m.act = Hardswish()
if isinstance(m, Conv) and isinstance(m.act, nn.SiLU):
m.act = SiLU()
return model
gs = int(max(model.stride)) # grid size (max stride)
opt.img_size = [check_img_size(x, gs)
for x in opt.img_size] # verify img_size are gs-multiples
print('opt:{}'.format(opt))
# Input
img = torch.zeros(opt.batch_size, 3,
*opt.img_size) # image size(1,3,320,192) iDetection
# Update model
for k, m in model.named_modules():
m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatibility
if isinstance(
m, models.common.Conv): # assign export-friendly activations
if isinstance(m.act, nn.Hardswish):
m.act = Hardswish()
elif isinstance(m.act, nn.SiLU):
m.act = SiLU()
# elif isinstance(m, models.yolo.Detect):
# m.forward = m.forward_export # assign forward (optional)
model.model[-1].export = True # set Detect() layer export=True
y = model(img) # dry run
# TorchScript export
try:
print('\nStarting TorchScript export with torch %s...' %
torch.__version__)
f = opt.weights.replace('.pt', '.torchscript.pt') # filename
ts = torch.jit.trace(model, img)
ts.save(f)
print('TorchScript export success, saved as %s' % f)
set_logging()
# Input
img = torch.zeros((opt.batch_size, 3,
*opt.img_size)) # image size(1,3,320,192) iDetection
# Load PyTorch model
model = attempt_load(opt.weights,
map_location=torch.device('cpu')) # load FP32 model
# Update model
for k, m in model.named_modules():
m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatability
if isinstance(m, models.common.Conv) and isinstance(
m.act, nn.Hardswish):
m.act = Hardswish() # assign activation
# if isinstance(m, models.yolo.Detect):
# m.forward = m.forward_export # assign forward (optional)
model.model[-1].export = True # set Detect() layer export=True
y = model(img) # dry run
# TorchScript export
try:
print('\nStarting TorchScript export with torch %s...' %
torch.__version__)
f = opt.weights.replace('.pt', '.torchscript.pt') # filename
ts = torch.jit.trace(model, img)
ts.save(f)
print('TorchScript export success, saved as %s' % f)
except Exception as e:
print('TorchScript export failure: %s' % e)
def run(
weights='./yolov5s.pt', # weights path
img_size=(640, 640), # image (height, width)
batch_size=1, # batch size
device='cpu', # cuda device, i.e. 0 or 0,1,2,3 or cpu
include=('torchscript', 'onnx', 'coreml'), # include formats
half=False, # FP16 half-precision export
inplace=False, # set YOLOv5 Detect() inplace=True
train=False, # model.train() mode
optimize=False, # TorchScript: optimize for mobile
dynamic=False, # ONNX: dynamic axes
simplify=False, # ONNX: simplify model
opset_version=12, # ONNX: opset version
):
t = time.time()
include = [x.lower() for x in include]
img_size *= 2 if len(img_size) == 1 else 1 # expand
# Load PyTorch model
device = select_device(device)
assert not (
device.type == 'cpu' and opt.half
), '--half only compatible with GPU export, i.e. use --device 0'
model = attempt_load(weights, map_location=device) # load FP32 model
labels = model.names
# Input
gs = int(max(model.stride)) # grid size (max stride)
img_size = [check_img_size(x, gs)
for x in img_size] # verify img_size are gs-multiples
img = torch.zeros(batch_size, 3, *img_size).to(
device) # image size(1,3,320,192) iDetection
# Update model
if half:
img, model = img.half(), model.half() # to FP16
model.train() if train else model.eval(
) # training mode = no Detect() layer grid construction
for k, m in model.named_modules():
m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatibility
if isinstance(m, Conv): # assign export-friendly activations
if isinstance(m.act, nn.Hardswish):
m.act = Hardswish()
elif isinstance(m.act, nn.SiLU):
m.act = SiLU()
elif isinstance(m, Detect):
m.inplace = inplace
m.onnx_dynamic = dynamic
# m.forward = m.forward_export # assign forward (optional)
for _ in range(2):
y = model(img) # dry runs
print(
f"\n{colorstr('PyTorch:')} starting from {weights} ({file_size(weights):.1f} MB)"
)
# TorchScript export -----------------------------------------------------------------------------------------------
if 'torchscript' in include or 'coreml' in include:
prefix = colorstr('TorchScript:')
try:
print(
f'\n{prefix} starting export with torch {torch.__version__}...'
)
f = weights.replace('.pt', '.torchscript.pt') # filename
ts = torch.jit.trace(model, img, strict=False)
(optimize_for_mobile(ts) if optimize else ts).save(f)
print(
f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)'
)
except Exception as e:
print(f'{prefix} export failure: {e}')
# ONNX export ------------------------------------------------------------------------------------------------------
if 'onnx' in include:
prefix = colorstr('ONNX:')
try:
import onnx
print(f'{prefix} starting export with onnx {onnx.__version__}...')
f = weights.replace('.pt', '.onnx') # filename
torch.onnx.export(
model,
img,
f,
verbose=False,
opset_version=opset_version,
training=torch.onnx.TrainingMode.TRAINING
if train else torch.onnx.TrainingMode.EVAL,
do_constant_folding=not train,
input_names=['images'],
output_names=['output'],
dynamic_axes={
'images': {
0: 'batch',
2: 'height',
3: 'width'
}, # shape(1,3,640,640)
'output': {
0: 'batch',
1: 'anchors'
} # shape(1,25200,85)
} if dynamic else None)
# Checks
model_onnx = onnx.load(f) # load onnx model
onnx.checker.check_model(model_onnx) # check onnx model
# print(onnx.helper.printable_graph(model_onnx.graph)) # print
# Simplify
if simplify:
try:
check_requirements(['onnx-simplifier'])
import onnxsim
print(
f'{prefix} simplifying with onnx-simplifier {onnxsim.__version__}...'
)
model_onnx, check = onnxsim.simplify(
model_onnx,
dynamic_input_shape=dynamic,
input_shapes={'images': list(img.shape)}
if dynamic else None)
assert check, 'assert check failed'
onnx.save(model_onnx, f)
except Exception as e:
print(f'{prefix} simplifier failure: {e}')
print(
f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)'
)
except Exception as e:
print(f'{prefix} export failure: {e}')
# CoreML export ----------------------------------------------------------------------------------------------------
if 'coreml' in include:
prefix = colorstr('CoreML:')
try:
import coremltools as ct
print(
f'{prefix} starting export with coremltools {ct.__version__}...'
)
assert train, 'CoreML exports should be placed in model.train() mode with `python export.py --train`'
model = ct.convert(ts,
inputs=[
ct.ImageType('image',
shape=img.shape,
scale=1 / 255.0,
bias=[0, 0, 0])
])
f = weights.replace('.pt', '.mlmodel') # filename
model.save(f)
print(
f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)'
)
except Exception as e:
print(f'{prefix} export failure: {e}')
# Finish
print(
f'\nExport complete ({time.time() - t:.2f}s). Visualize with https://github.com/lutzroeder/netron.'
)
def hardswish(input, inplace=False):
return OriginalHardswish.forward(input)