def __init__(
self,
c1,
c2,
k=1,
s=1,
p=None,
g=1,
act=True,
): # ch_in, ch_out, kernel, stride, padding, groups
assert isinstance(
self.use_v3,
bool), "You need to decide whether use_yolov3 is True or False"
super(Conv, self).__init__()
if isinstance(k, list):
assert len(k) <= 2 and k[0] == k[-1]
k = k[0]
if isinstance(s, list):
assert len(s) <= 2 and s[0] == s[-1]
s = s[0]
self.conv = nn.Conv(c1, c2, k, s, autopad(k, p), groups=g, bias=False)
self.bn = nn.BatchNorm(c2)
self.act = (
nn.LeakyReLU(0.1) if self.use_v3 else SiLU()) if act is True else (
act if isinstance(act, nn.Module) else nn.Identity())
def __init__(self):
super(SpatialAttentionModule, self).__init__()
self.conv2d = nn.Conv2d(in_channels=2,
out_channels=1,
kernel_size=7,
stride=1,
padding=3)
self.act = SiLU()
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 __init__(self, c1, reduction=16):
super(ChannelAttentionModule, self).__init__()
mid_channel = c1 // reduction
self.avg_pool = nn.AdaptiveAvgPool2d(1)
self.max_pool = nn.AdaptiveMaxPool2d(1)
self.shared_MLP = nn.Sequential(
nn.Linear(in_features=c1, out_features=mid_channel),
nn.LeakyReLU(0.1, inplace=True),
nn.Linear(in_features=mid_channel, out_features=c1))
# self.sigmoid = nn.Sigmoid()
self.act = SiLU()
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
def run(data=ROOT / 'data/coco128.yaml', # 'dataset.yaml path'
weights=ROOT / 'yolov5s.pt', # weights path
imgsz=(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
int8=False, # CoreML/TF INT8 quantization
dynamic=False, # ONNX/TF: dynamic axes
simplify=False, # ONNX: simplify model
opset=12, # ONNX: opset version
verbose=False, # TensorRT: verbose log
workspace=4, # TensorRT: workspace size (GB)
topk_per_class=100, # TF.js NMS: topk per class to keep
topk_all=100, # TF.js NMS: topk for all classes to keep
iou_thres=0.45, # TF.js NMS: IoU threshold
conf_thres=0.25 # TF.js NMS: confidence threshold
):
t = time.time()
include = [x.lower() for x in include]
tf_exports = list(x in include for x in ('saved_model', 'pb', 'tflite', 'tfjs')) # TensorFlow exports
imgsz *= 2 if len(imgsz) == 1 else 1 # expand
file = Path(url2file(weights) if str(weights).startswith(('http:/', 'https:/')) else 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, inplace=True, fuse=True) # load FP32 model
nc, names = model.nc, model.names # number of classes, class names
# Input
gs = int(max(model.stride)) # grid size (max stride)
imgsz = [check_img_size(x, gs) for x in imgsz] # verify img_size are gs-multiples
im = torch.zeros(batch_size, 3, *imgsz).to(device) # image size(1,3,320,192) BCHW iDetection
# Update model
if half:
im, model = im.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.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(im) # dry runs
LOGGER.info(f"\n{colorstr('PyTorch:')} starting from {file} ({file_size(file):.1f} MB)")
# Exports
if 'torchscript' in include:
export_torchscript(model, im, file, optimize)
if 'onnx' in include:
export_onnx(model, im, file, opset, train, dynamic, simplify)
if 'engine' in include:
export_engine(model, im, file, train, half, simplify, workspace, verbose)
if 'coreml' in include:
export_coreml(model, im, file)
# TensorFlow Exports
if any(tf_exports):
pb, tflite, tfjs = tf_exports[1:]
assert not (tflite and tfjs), 'TFLite and TF.js models must be exported separately, please pass only one type.'
model = export_saved_model(model, im, file, dynamic, tf_nms=tfjs, agnostic_nms=tfjs,
topk_per_class=topk_per_class, topk_all=topk_all, conf_thres=conf_thres,
iou_thres=iou_thres) # keras model
if pb or tfjs: # pb prerequisite to tfjs
export_pb(model, im, file)
if tflite:
export_tflite(model, im, file, int8=int8, data=data, ncalib=100)
if tfjs:
export_tfjs(model, im, file)
# Finish
LOGGER.info(f'\nExport complete ({time.time() - t:.2f}s)'
f"\nResults saved to {colorstr('bold', file.parent.resolve())}"
f'\nVisualize with https://netron.app')
def run(data=ROOT / 'data/coco128.yaml', # 'dataset.yaml path'
weights=ROOT / 'yolov5s.pt', # weights path
imgsz=(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'), # 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
int8=False, # CoreML/TF INT8 quantization
dynamic=False, # ONNX/TF: dynamic axes
simplify=False, # ONNX: simplify model
opset=12, # ONNX: opset version
verbose=False, # TensorRT: verbose log
workspace=4, # TensorRT: workspace size (GB)
nms=False, # TF: add NMS to model
agnostic_nms=False, # TF: add agnostic NMS to model
topk_per_class=100, # TF.js NMS: topk per class to keep
topk_all=100, # TF.js NMS: topk for all classes to keep
iou_thres=0.45, # TF.js NMS: IoU threshold
conf_thres=0.25 # TF.js NMS: confidence threshold
):
t = time.time()
include = [x.lower() for x in include]
tf_exports = list(x in include for x in ('saved_model', 'pb', 'tflite', 'edgetpu', 'tfjs')) # TensorFlow exports
file = Path(url2file(weights) if str(weights).startswith(('http:/', 'https:/')) else 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, inplace=True, fuse=True) # load FP32 model
nc, names = model.nc, model.names # number of classes, class names
# Checks
imgsz *= 2 if len(imgsz) == 1 else 1 # expand
opset = 12 if ('openvino' in include) else opset # OpenVINO requires opset <= 12
assert nc == len(names), f'Model class count {nc} != len(names) {len(names)}'
# Input
gs = int(max(model.stride)) # grid size (max stride)
imgsz = [check_img_size(x, gs) for x in imgsz] # verify img_size are gs-multiples
im = torch.zeros(batch_size, 3, *imgsz).to(device) # image size(1,3,320,192) BCHW iDetection
# Update model
if half:
im, model = im.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.SiLU):
m.act = SiLU()
elif isinstance(m, Detect):
m.inplace = inplace
m.onnx_dynamic = dynamic
if hasattr(m, 'forward_export'):
m.forward = m.forward_export # assign custom forward (optional)
for _ in range(2):
y = model(im) # dry runs
shape = tuple(y[0].shape) # model output shape
LOGGER.info(f"\n{colorstr('PyTorch:')} starting from {file} with output shape {shape} ({file_size(file):.1f} MB)")
# Exports
f = [''] * 10 # exported filenames
warnings.filterwarnings(action='ignore', category=torch.jit.TracerWarning) # suppress TracerWarning
if 'torchscript' in include:
f[0] = export_torchscript(model, im, file, optimize)
if 'engine' in include: # TensorRT required before ONNX
f[1] = export_engine(model, im, file, train, half, simplify, workspace, verbose)
if ('onnx' in include) or ('openvino' in include): # OpenVINO requires ONNX
f[2] = export_onnx(model, im, file, opset, train, dynamic, simplify)
if 'openvino' in include:
f[3] = export_openvino(model, im, file)
if 'coreml' in include:
_, f[4] = export_coreml(model, im, file)
# TensorFlow Exports
if any(tf_exports):
pb, tflite, edgetpu, tfjs = tf_exports[1:]
if int8 or edgetpu: # TFLite --int8 bug https://github.com/ultralytics/yolov5/issues/5707
check_requirements(('flatbuffers==1.12',)) # required before `import tensorflow`
assert not (tflite and tfjs), 'TFLite and TF.js models must be exported separately, please pass only one type.'
model, f[5] = export_saved_model(model, im, file, dynamic, tf_nms=nms or agnostic_nms or tfjs,
agnostic_nms=agnostic_nms or tfjs, topk_per_class=topk_per_class,
topk_all=topk_all, conf_thres=conf_thres, iou_thres=iou_thres) # keras model
if pb or tfjs: # pb prerequisite to tfjs
f[6] = export_pb(model, im, file)
if tflite or edgetpu:
f[7] = export_tflite(model, im, file, int8=int8 or edgetpu, data=data, ncalib=100)
if edgetpu:
f[8] = export_edgetpu(model, im, file)
if tfjs:
f[9] = export_tfjs(model, im, file)
# Finish
f = [str(x) for x in f if x] # filter out '' and None
if any(f):
LOGGER.info(f'\nExport complete ({time.time() - t:.2f}s)'
f"\nResults saved to {colorstr('bold', file.parent.resolve())}"
f"\nDetect: python detect.py --weights {f[-1]}"
f"\nPyTorch Hub: model = torch.hub.load('ultralytics/yolov5', 'custom', '{f[-1]}')"
f"\nValidate: python val.py --weights {f[-1]}"
f"\nVisualize: https://netron.app")
return f # return list of exported files/dirs
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)
except Exception as e:
print('TorchScript export failure: %s' % e)
opt.img_size = [check_img_size(x, gs)
for x in opt.img_size] # verify img_size are gs-multiples
# 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)
if isinstance(m,
models.common.ShuffleV2Block): #shufflenet block nn.SiLU
for i in range(len(m.branch1)):
if isinstance(m.branch1[i], nn.SiLU):
m.branch1[i] = SiLU()
for i in range(len(m.branch2)):
if isinstance(m.branch2[i], nn.SiLU):
m.branch2[i] = SiLU()
model.model[-1].export = True # set Detect() layer export=True
y = model(img) # dry run
# ONNX export
print('\nStarting ONNX export with onnx %s...' % onnx.__version__)
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.'
)
