def run(args):
# Export if we know how, otherwise return
if args.file:
args.copy = False
args.build = False
export.run(args)
else:
if not os.path.exists(os.path.join(args.directory, args.project)):
logging.debug("The project: %s does not exist." %
os.path.join(args.directory, args.project))
return
if args.file:
# known project from records
workspace = Workspace(args.file, os.getcwd())
if args.project:
workspace.build_project(args.project, args.tool)
else:
workspace.build_projects(args.tool)
else:
# not project known by pgen
project_settings = ProjectSettings()
project_files = [os.path.join(args.directory, args.project)]
builder = ToolsSupported().get_value(args.tool, 'builder')
build(builder, args.project, project_files, args.tool,
project_settings)
def test(
weights=ROOT / 'yolov5s.pt', # weights path
imgsz=640, # inference size (pixels)
batch_size=1, # batch size
data=ROOT / 'data/coco128.yaml', # dataset.yaml path
device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu
half=False, # use FP16 half-precision inference
test=False, # test exports only
pt_only=False, # test PyTorch only
):
y, t = [], time.time()
device = select_device(device)
for i, (name, f, suffix, gpu) in export.export_formats().iterrows(
): # index, (name, file, suffix, gpu-capable)
try:
w = weights if f == '-' else \
export.run(weights=weights, imgsz=[imgsz], include=[f], device=device, half=half)[-1] # weights
assert suffix in str(w), 'export failed'
y.append([name, True])
except Exception:
y.append([name, False]) # mAP, t_inference
# Print results
LOGGER.info('\n')
parse_opt()
notebook_init() # print system info
py = pd.DataFrame(y, columns=['Format', 'Export'])
LOGGER.info(f'\nExports complete ({time.time() - t:.2f}s)')
LOGGER.info(str(py))
return py
def run(
weights=ROOT / 'yolov5s.pt', # weights path
imgsz=640, # inference size (pixels)
batch_size=1, # batch size
data=ROOT / 'data/coco128.yaml', # dataset.yaml path
device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu
half=False, # use FP16 half-precision inference
):
y, t = [], time.time()
formats = export.export_formats()
device = select_device(device)
for i, (name, f, suffix, gpu) in formats.iterrows(
): # index, (name, file, suffix, gpu-capable)
try:
if device.type != 'cpu':
assert gpu, f'{name} inference not supported on GPU'
if f == '-':
w = weights # PyTorch format
else:
w = export.run(weights=weights,
imgsz=[imgsz],
include=[f],
device=device,
half=half)[-1] # all others
assert suffix in str(w), 'export failed'
result = val.run(data,
w,
batch_size,
imgsz,
plots=False,
device=device,
task='benchmark',
half=half)
metrics = result[
0] # metrics (mp, mr, map50, map, *losses(box, obj, cls))
speeds = result[2] # times (preprocess, inference, postprocess)
y.append([name, metrics[3], speeds[1]]) # mAP, t_inference
except Exception as e:
LOGGER.warning(f'WARNING: Benchmark failure for {name}: {e}')
y.append([name, None, None]) # mAP, t_inference
# Print results
LOGGER.info('\n')
parse_opt()
notebook_init() # print system info
py = pd.DataFrame(
y, columns=['Format', '[email protected]:0.95', 'Inference time (ms)'])
LOGGER.info(f'\nBenchmarks complete ({time.time() - t:.2f}s)')
LOGGER.info(str(py))
return py
def run(args):
# Export if we know how, otherwise return
if args.file:
args.copy = False
args.build = False
export.run(args)
else:
if not os.path.exists(os.path.join(args.directory, args.project)):
logging.debug("The project: %s does not exist." % os.path.join(args.directory, args.project))
return
if args.file:
# known project from records
workspace = Workspace(args.file, os.getcwd())
if args.project:
workspace.build_project(args.project, args.tool)
else:
workspace.build_projects(args.tool)
else:
# not project known by pgen
project_settings = ProjectSettings()
project_files = [os.path.join(args.directory, args.project)]
builder = ToolsSupported().get_value(args.tool, 'builder')
build(builder, args.project, project_files, args.tool, project_settings)
def run(
weights=ROOT / 'yolov5s.pt', # weights path
imgsz=640, # inference size (pixels)
batch_size=1, # batch size
data=ROOT / 'data/coco128.yaml', # dataset.yaml path
):
y, t = [], time.time()
formats = export.export_formats()
for i, (name, f,
suffix) in formats.iterrows(): # index, (name, file, suffix)
try:
w = weights if f == '-' else export.run(
weights=weights, imgsz=[imgsz], include=[f], device='cpu')[-1]
assert suffix in str(w), 'export failed'
result = val.run(data,
w,
batch_size,
imgsz=imgsz,
plots=False,
device='cpu',
task='benchmark')
metrics = result[
0] # metrics (mp, mr, map50, map, *losses(box, obj, cls))
speeds = result[2] # times (preprocess, inference, postprocess)
y.append([name, metrics[3], speeds[1]]) # mAP, t_inference
except Exception as e:
LOGGER.warning(f'WARNING: Benchmark failure for {name}: {e}')
y.append([name, None, None]) # mAP, t_inference
# Print results
LOGGER.info('\n')
parse_opt()
notebook_init() # print system info
py = pd.DataFrame(
y, columns=['Format', '[email protected]:0.95', 'Inference time (ms)'])
LOGGER.info(f'\nBenchmarks complete ({time.time() - t:.2f}s)')
LOGGER.info(str(py))
return py
def run(
weights=ROOT / 'yolov5s.pt', # weights path
imgsz=640, # inference size (pixels)
batch_size=1, # batch size
data=ROOT / 'data/coco128.yaml', # dataset.yaml path
device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu
half=False, # use FP16 half-precision inference
test=False, # test exports only
pt_only=False, # test PyTorch only
):
y, t = [], time.time()
device = select_device(device)
for i, (name, f, suffix, gpu) in export.export_formats().iterrows(
): # index, (name, file, suffix, gpu-capable)
try:
assert i != 9, 'Edge TPU not supported'
assert i != 10, 'TF.js not supported'
if device.type != 'cpu':
assert gpu, f'{name} inference not supported on GPU'
# Export
if f == '-':
w = weights # PyTorch format
else:
w = export.run(weights=weights,
imgsz=[imgsz],
include=[f],
device=device,
half=half)[-1] # all others
assert suffix in str(w), 'export failed'
# Validate
result = val.run(data,
w,
batch_size,
imgsz,
plots=False,
device=device,
task='benchmark',
half=half)
metrics = result[
0] # metrics (mp, mr, map50, map, *losses(box, obj, cls))
speeds = result[2] # times (preprocess, inference, postprocess)
y.append([
name,
round(file_size(w), 1),
round(metrics[3], 4),
round(speeds[1], 2)
]) # MB, mAP, t_inference
except Exception as e:
LOGGER.warning(f'WARNING: Benchmark failure for {name}: {e}')
y.append([name, None, None, None]) # mAP, t_inference
if pt_only and i == 0:
break # break after PyTorch
# Print results
LOGGER.info('\n')
parse_opt()
notebook_init() # print system info
c = ['Format', 'Size (MB)', '[email protected]:0.95', 'Inference time (ms)'
] if map else ['Format', 'Export', '', '']
py = pd.DataFrame(y, columns=c)
LOGGER.info(f'\nBenchmarks complete ({time.time() - t:.2f}s)')
LOGGER.info(str(py if map else py.iloc[:, :2]))
return py