def get_predictions_board(X_pairs,
model_filepath,
imgs_per_cluster=1000,
sw=False):
phi = []
eta = []
pt = []
iterations = int(len(X_pairs) / imgs_per_cluster)
#ie = microndla.MDLA()
ie = fwdnxt.FWDNXT()
ie.SetFlag('debug', 'wb')
ie.SetFlag("options", "V")
ie.SetFlag('imgs_per_cluster', str(imgs_per_cluster))
swnresults = ie.Compile('4x1x1', model_filepath, 'save.bin', 1, 1)
nresults = ie.Init('save.bin', '')
result = np.ndarray(swnresults, dtype=np.float32)
start = 0
end = 0
for iter_i in range(0, len(X_pairs), imgs_per_cluster):
start = iter_i
end = start + imgs_per_cluster
print(iter_i)
print(start)
print(end)
current_batch = np.ascontiguousarray(np.swapaxes(
np.expand_dims(np.expand_dims(X_pairs[start:end, :], 0), 0), 0, 2),
dtype=np.float32())
print(current_batch.shape)
print(current_batch[0][0][0])
result = np.ndarray(swnresults, dtype=np.float32)
ie.Run_sw(current_batch, result)
cur_phi_pred, cur_eta_pred, cur_pt_pred = parse_board_results(
result, imgs_per_cluster, outputs=3)
phi.extend(cur_phi_pred)
eta.extend(cur_eta_pred)
pt.extend(cur_pt_pred)
ie.Free()
return phi, eta, pt
def ieprocess(image_file, network_file):
# load image and resize it:
img = Image.open(image_file)
#Resize it to the size expected by the network
img = img.resize((224, 224), resample=Image.BILINEAR)
#Convert to numpy float
img = np.array(img).astype(np.float32) / 255
#Transpose to plane-major, as required by our API
img = np.ascontiguousarray(img.transpose(2, 0, 1))
# print(img)
print('Image shape:', img.shape)
#Normalize images
stat_mean = list([0.485, 0.456, 0.406])
stat_std = list([0.229, 0.224, 0.225])
for i in range(3):
img[i] = (img[i] - stat_mean[i]) / stat_std[i]
#Create and initialize the Inference Engine object
ie = fwdnxt.FWDNXT()
#Compile to a file
swnresults = ie.Compile("{:d}x{:d}x{:d}".format(224, 224, 3), network_file,
'save.bin')
#Init fpga
nresults = ie.Init('save.bin', '')
#Create the storage for the result and run one inference
result = np.ndarray(swnresults, dtype=np.float32)
ie.Run(img, result)
#Convert to numpy and print top-5
idxs = (-result).argsort()
rstring = []
with open("categories.txt") as f:
categories = f.read().splitlines()
for i in range(5):
rstring.append(
str(categories[idxs[i]]) + ', ' + str(result[idxs[i]]))
#Free
ie.Free()
return rstring
class VectorAdd(torch.nn.Module):
def __init__(self):
super(VectorAdd, self).__init__()
def forward(self, x1, x2):
x3 = x1 + x2
return x3
D = args.d
inVec1 = torch.randn(1, D, 1, 1, dtype=torch.float32)
inVec2 = torch.randn(1, D, 1, 1, dtype=torch.float32)
modelProd = VectorAdd()
torch.onnx.export(modelProd, (inVec1, inVec2), "net_vector_add.onnx")
sf = fwdnxt.FWDNXT()
if args.verbose:
sf.SetFlag('debug', 'b') #debug options
# Compile to generate binary
snwresults = sf.Compile(
'{:d}x{:d}x{:d};{:d}x{:d}x{:d}'.format(1, 1, D, 1, 1, D),
'net_vector_add.onnx', 'net_vector_add.bin', 1, 1)
sf.Init("./net_vector_add.bin", "")
in_1 = np.ascontiguousarray(inVec1)
in_2 = np.ascontiguousarray(inVec2)
result = np.ascontiguousarray(np.ndarray((1, 1, snwresults), dtype=np.float32))
sf.Run((in_1, in_2), result)
outhw = modelProd(inVec1, inVec2)
for i in range(2):
#Convert to numpy float
img[i] = np.array(img[i]).astype(np.float32) / 255
#Transpose to plane-major, as required by our API
img[i] = np.ascontiguousarray(img[i].transpose(2, 0, 1))
#Normalize images
stat_mean = list([0.485, 0.456, 0.406])
stat_std = list([0.229, 0.224, 0.225])
for j in range(3):
img[i][j] = (img[i][j] - stat_mean[j]) / stat_std[j]
#Create and initialize the Inference Engine object
ie = fwdnxt.FWDNXT()
#ie.SetFlag('hwlinear','0')
#ie.SetFlag('debug','bw')
#Compile to a file
ie.Compile("{:d}x{:d}x{:d}".format(args.res1[2], args.res1[1], args.res1[0]),
args.modelpath1, 'save1.bin')
ie.Compile("{:d}x{:d}x{:d}".format(args.res2[2], args.res2[1], args.res2[0]),
args.modelpath2, 'save2.bin')
ie.Loadmulti(('save1.bin', 'save2.bin'))
#Init fpga
nresults = ie.Init('', args.load)
#Create the storage for the result and run one inference
result1 = np.ndarray(nresults[0], dtype=np.float32)
