def make_network(minibatch_size = 128):
pre_net = load_network("trained.data")
inp = pre_net.outputs[-1]
visitor = NetworkVisitor(inp).all_oprs
for i in visitor:
if isinstance(i, BN):
i.set_freezed()
if isinstance(i, ParamProvider):
i.set_freezed()
if isinstance(i, DataProvider):
dp = i
lay = O.ZeroGrad(inp)
chl = inp.partial_shape[1]
p = []
for tt in range(3):
lay = deconv_bn_relu("encoder_deconv_{}0".format(tt), lay, kernel_shape = 3, stride = 1, padding = 1, output_nr_channel = chl)
lay = deconv_bn_relu("encoder_deconv_{}1".format(tt), lay, kernel_shape = 3, stride = 1, padding = 1, output_nr_channel = chl)
p.append(lay)
if tt != 2:
lay = deconv_bn_relu("encoder_deconv_{}2".format(tt), lay, kernel_shape = 2, stride = 2, padding = 0, output_nr_channel = chl // 2)
chl = chl // 2
lay = deconv_bn_relu("outputs", lay, kernel_shape = 3, stride = 1, padding = 1, output_nr_channel = 3, isbnrelu = False)
loss = ((lay - dp)**2).sum(axis = 3).sum(axis = 2).sum(axis = 1).mean()
network = Network(outputs = [lay, inp] + p)
network.loss_var = loss
return network
def make_network(minibatch_size = 128):
pre_net = load_network("rand.data")
inp = pre_net.outputs[-1]
visitor = NetworkVisitor(inp).all_oprs
for i in visitor:
#if isinstance(i, BN):
# i.set_freezed()
if isinstance(i, ParamProvider):
i.set_freezed()
if isinstance(i, DataProvider):
dp = i
lay = O.ZeroGrad(inp)
chl = inp.partial_shape[1]
p = []
num_blocks = 1
for tt in range(num_blocks):
for j in range(2):
lay = deconv_bn_relu("encoder_deconv_{}{}".format(tt, j), lay, kernel_shape = 3, stride = 1, padding = 0, output_nr_channel = chl)
#lay = deconv_bn_relu("encoder_deconv_{}1".format(tt), lay, kernel_shape = 3, stride = 1, padding = 1, output_nr_channel = chl)
p.append(lay)
if tt != num_blocks:
lay = deconv_bn_relu("encoder_deconv_{}{}".format(tt, j + 1), lay, kernel_shape = 2, stride = 2, padding = 0, output_nr_channel = chl // 2)
chl = chl // 2
lay = deconv_bn_relu("outputs", lay, kernel_shape = 3, stride = 1, padding = 1, output_nr_channel = 3, isbnrelu = False)
mid = lay.partial_shape[2] // 2
lay = lay[:, :, mid-16:mid+16, mid-16:mid+16]
print(lay.partial_shape)
loss = ((lay - dp)**2).sum(axis = 3).sum(axis = 2).sum(axis = 1).mean()
network = Network(outputs = [lay, inp] + p)
network.loss_var = loss
return network
def __init__(self, name):
with open("{}".format(name), "rb") as f:
self.net = load_network(f)
self.test_func = Function().compile([self.net.outputs[0], self.net.loss_var])
with open("/unsullied/sharefs/liuyanyi02/lyy/CIFAR/meanstd.data", "rb") as f:
self.mean, self.std = pickle.load(f)
self.mean = np.array(self.mean).reshape(3, 32, 32)
self.std = np.array(self.std).reshape(3, 32, 32)
def __init__(self, model_file, env):
self.env = env
self.net = load_network(model_file)
self.N = 16000
self.batch_size = 64
self.nperimage = 20
self.new_net = load_network(model_file)
print("finished loading weights")
self.oprs_dict = self.net.loss_visitor.all_oprs_dict
self.new_oprs_dict = self.new_net.loss_visitor.all_oprs_dict
self.convs = self.get_convs_from_net()
print(self.convs)
data_func = self.get_data(is_val=False)
data_func_val = self.get_data(is_val=True)
self.data_iter = data_func()
self.data_iter_val = data_func_val()
def bn_post_process(model_file: str, save_model_file: str, data):
with TrainingEnv(name=model_file + "bn_post_proc", part_count=2) as env:
net = load_network(open(model_file, "rb"))
#loss_func = env.make_func_from_loss_var(net.loss_var, "val", train_state = False)
bn_oprs = [
opr for opr in net.loss_visitor.all_oprs
if isinstance(opr, BatchNormalization)
]
bn_inputs = [opr.inputs[0] for opr in bn_oprs]
mean_Esqr_nodes = []
for i in bn_inputs:
if i.partial_shape.ndim == 2:
mean = i.mean(axis=0).reshape((1, -1))
mean.vflags.data_parallel_reduce_method = 'sum'
Esqr = (i**2).mean(axis=0).reshape((1, -1))
Esqr.vflags.data_parallel_reduce_method = 'sum'
if i.partial_shape.ndim == 4:
mean = i.mean(axis=3).mean(axis=2).mean(axis=0).reshape(
(1, -1))
mean.vflags.data_parallel_reduce_method = 'sum'
Esqr = (i**2).mean(axis=3).mean(axis=2).mean(axis=0).reshape(
(1, -1))
Esqr.vflags.data_parallel_reduce_method = 'sum'
mean_Esqr_nodes.append(mean)
mean_Esqr_nodes.append(Esqr)
func = Function().compile(mean_Esqr_nodes)
for i in range(len(bn_oprs)):
opr = bn_oprs[i]
layer_mean, layer_var = _get_dataset_mean_var(data, func, i)
if layer_mean.ndim == 0:
layer_mean = layer_mean.reshape((1, ))
if layer_var.ndim == 0:
layer_var = layer_var.reshape((1, ))
state = opr.State(channels=layer_mean.shape[0],
val=[layer_mean, layer_var, 1])
state.owner_opr_type = type(opr)
opr.set_opr_state(state)
opr.freezed = True
env.register_checkpoint_component("network", net)
env.save_checkpoint(save_model_file)
def test(net=None):
if net == None:
net = load_network(open("./data/resnet110.data", "rb"))
test_func = Function().compile(net.outputs[0])
"""
from megskull.network import NetworkVisitor
visitor = NetworkVisitor(net.loss_var)
for i in visitor.all_oprs:
print(i)
print("input = ", i.inputs)
print("output = ", i.outputs)
a = np.array(i.outputs[0].get_value())
print(a)
input()
a = np.array(visitor.all_oprs_dict["conv1:W"].get_value())
print(a)
print("mean = ", np.mean(a, axis = 0))
print("std = ", np.std(a, axis = 0))
exit()
"""
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2
for i in range(10):
img = data[i].transpose(1, 2, 0)
img = img[:,::-1,:]
cv2.imshow('x', img)
cv2.waitKey(0)
"""
#data = data.astype(np.float32)
#data = (data - 128) / 256
pred = test_func(data=data)
print(np.array(pred).shape)
pred = np.argmax(np.array(pred), axis=1)
acc = (np.array(pred) == np.array(label)).mean()
print(acc)
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
net = load_network(open("./data/resnet20.data_bned", "rb"))
test_func = Function().compile(net.outputs[0])
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2
for i in range(10):
img = data[i].transpose(1, 2, 0)
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
from numpy.linalg import svd
net = load_network(open("./data/plain30_dp_sup.data", "rb"))
func = Function().compile(net.outputs)
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
idx = np.random.randint(data.shape[0], size=3000)
data = [data[i] for i in idx]
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
from numpy.linalg import svd
net = load_network(open("./data/plain30_orth.data", "rb"))
func = Function().compile(net.outputs)
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
idx = np.random.randint(data.shape[0], size=3000)
data = [data[i] for i in idx]
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
net = load_network(open("./data/wc0.data", "rb"))
test_func = Function().compile(net.outputs[0])
lis = net.loss_var.owner_opr._param_weights
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("/home/liuyanyi02/CIFAR/meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
net = load_network(open("./data/resnet110.data", "rb"))
test_func = Function().compile(net.outputs[0])
from megskull.network import NetworkVisitor
visitor = NetworkVisitor(net.loss_var)
#print(visitor.all_oprs)
"""
for i in visitor.all_oprs:
print(i)
print("input = ", i.inputs)
print("output = ", i.outputs)
a = np.array(i.outputs[0].get_value())
print(a)
input()
"""
a = np.array(visitor.all_oprs_dict["conv1:W"].get_value())
print(a)
print("mean = ", np.mean(a, axis = 0))
print("std = ", np.std(a, axis = 0))
exit()
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding = "bytes")
return dic
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
import cv2
net = load_network(open("./data/resnet20.data_acc91.14", "rb"))
test_func = Function().compile(net.outputs[0])
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
raw_data = data.copy()
raw_data = np.resize(raw_data, (10000, 3, 32, 32))
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
idx = 2
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
net = load_network(open("./data/Res110Mixup.data", "rb"))
test_func = Function().compile(net.outputs[0])
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("/home/liuyanyi02/CIFAR/meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2
for i in range(10):
img = data[i].transpose(1, 2, 0)
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
net = load_network(open("./data/fixedfc.data", "rb"))
test_func = Function().compile(net.outputs[0])
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("/home/liuyanyi02/CIFAR/meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2
for i in range(10):
img = data[i].transpose(1, 2, 0)
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
net = load_network(open("./data/fixed_res110.data", "rb"))
test_func = Function().compile(net.outputs[0])
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("/home/liuyanyi02/CIFAR/meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2
for i in range(10):
img = data[i].transpose(1, 2, 0)
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
net = load_network(open("./data/densenetl100k24.data", "rb"))
test_func = Function().compile(net.outputs[0])
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding = "bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2
for i in range(10):
img = data[i].transpose(1, 2, 0)
img = img[:,::-1,:]
cv2.imshow('x', img)
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
import cv2
net = load_network(open("./data/multiscal_v2.data", "rb"))
test_func = Function().compile(net.outputs[0])
from megskull.network import NetworkVisitor
visitor = NetworkVisitor(net.loss_var)
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
net = load_network(open("./data/plain30_dropout.data", "rb"))
test_func = Function().compile(net.outputs[0])
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding = "bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2
for i in range(10):
img = data[i].transpose(1, 2, 0)
img = img[:,::-1,:]
cv2.imshow('x', img)
def main():
parser = argparse.ArgumentParser(
description='Dump the Python Megbrain model to C++ model, by the way '
'optimizing for inference',
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('input', help='input pkl model file ')
parser.add_argument('-o', '--output', help='output file', required=True)
parser.add_argument('--init-bn',
action='store_true',
help='initialize untrained batch-normalization, to '
'avoid NaN or Inf results')
parser.add_argument('--silent',
action='store_true',
help='set verbose to False in AssertEqual opr')
parser.add_argument('--optimize-for-inference',
action='store_true',
help='enbale optimization for inference')
parser.add_argument('--discard-var-name',
action='store_true',
help='discard variable and param names in the '
'generated output')
parser.add_argument('--output-strip-info',
action='store_true',
help='output code strip information')
parser.add_argument(
'--enable-io16xc32',
action='store_true',
help='transform the mode to float16 io float32 compute')
parser.add_argument('--enable-ioc16',
action='store_true',
help='transform the dtype of the model to float16 io '
'and compute')
parser.add_argument('--enable-fuse-conv-bias-nonlinearity',
action='store_true',
help='fuse convolution bias and nonlinearity opr to a '
'conv_bias opr and compute')
parser.add_argument('--enable-hwcd4',
action='store_true',
help='transform the model format from NCHW to NHWCD4 '
'for inference; you may need to disable CUDA and set '
'MGB_USE_MEGDNN_DBG=2')
parser.add_argument('--enable-nchw4',
action='store_true',
help='transform the model format from NCHW to NCHW4 '
'for inference')
parser.add_argument('--enable-nchw88',
action='store_true',
help='transform the model format from NCHW to NCHW88 '
'for inference')
parser.add_argument('--enable-nchw44',
action='store_true',
help='transform the model format from NCHW to NCHW44 '
'for inference')
parser.add_argument(
'--enable-nchw44-dot',
action='store_true',
help='transform the model format from NCHW to NCHW44_DOT '
'for optimizing armv8.2 dot in inference')
parser.add_argument('--enable-chwn4',
action='store_true',
help='transform the model format to CHWN4 '
'for inference, mainly used for nvidia tensorcore')
parser.add_argument('--enable-nchw32',
action='store_true',
help='transform the model format from NCHW4 to NCHW32 '
'for inference on nvidia TensoCore')
parser.add_argument(
'--enable-fuse-conv-bias-with-z',
action='store_true',
help='fuse conv_bias with z input for inference on '
'nvidia GPU (this optimization pass will result in mismatch '
'of the precision of output of training and inference)')
args = parser.parse_args()
env = FpropEnv(verbose_fprop=False)
outputs = io.load_network(args.input).outputs
output_mgbvars = list(map(env.get_mgbvar, outputs))
output_mgbvars = optimize_for_inference(args, output_mgbvars)
if args.discard_var_name:
sereg_kwargs = dict(keep_var_name=0, keep_param_name=False)
else:
sereg_kwargs = dict(keep_var_name=2, keep_param_name=True)
stat = mgb.serialize_comp_graph_to_file(
args.output,
output_mgbvars,
append=False,
output_strip_info=args.output_strip_info,
**sereg_kwargs)
logger.info(
'graph dump sizes: tot_size={:.3f}KiB overhead={:.3f}KiB'.format(
stat.tot_bytes / 1024,
(stat.tot_bytes - stat.tensor_value_bytes) / 1024))
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
from numpy.linalg import svd
net = load_network(open("./data/plain30_xcep.data", "rb"))
func = Function().compile(net.outputs)
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding = "bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
idx = np.random.randint(data.shape[0], size = 3000)
data = [data[i] for i in idx]
def pca_ener(A):
A = A.T
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
net = load_network(open("./data/fixed_den12.data", "rb"))
test_func = Function().compile(net.outputs[0])
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("/home/liuyanyi02/CIFAR/meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2
for i in range(10):
img = data[i].transpose(1, 2, 0)
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
from megskull.opr.all import Grad
import numpy as np
import cv2
from train import MyMomentum
net = load_network(open("./data/plain30_dfconv.data", "rb"))
#test_func = Function().compile(net.outputs[0])
from megskull.network import NetworkVisitor
visitor = NetworkVisitor(net.loss_var)
offsets = []
offsets_name = []
locs = []
for i in visitor.all_oprs:
print(i, i.name)
if "Astype" in i.name:
locs.append(i)
if i.name in ["pooling32offsetx"]:
offsets.append(i)
offsets_name.append(i.name)
print("A")
"""
for j in i.inputs:
if j.name == "Mul630":
print(i.name, "***")
offsets.append(i)
offsets_name.append(i.name)
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
net = load_network(open("./data/densenetl40k12.data", "rb"))
test_func = Function().compile(net.outputs[0])
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding="bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2
for i in range(10):
img = data[i].transpose(1, 2, 0)
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
import cv2
from train import MyMomentum
net = load_network(open("./data/dfconv.data", "rb"))
#test_func = Function().compile(net.outputs[0])
from megskull.network import NetworkVisitor
visitor = NetworkVisitor(net.loss_var)
offsets = []
locs = []
for i in visitor.all_oprs:
print(i, i.name)
if "Astype" in i.name:
locs.append(i)
if i.name == "conv2offsetx":
offsets.append(i)
print("A")
print(len(locs))
locs = locs[::4]
outs = [net.outputs[0]] + locs
test_func = Function().compile(outs)
outs1 = offsets
offs_func = Function().compile(outs1)
def load_data(name):
import pickle
from megskull.graph import FpropEnv
from meghair.utils.io import load_network
from megskull.graph import Function
import numpy as np
import cv2
net = load_network(open("./data/comp.data", "rb"))
test_func = Function().compile(net.outputs[0])
from megskull.network import NetworkVisitor
visitor = NetworkVisitor(net.loss_var)
def load_data(name):
import pickle
with open(name, "rb") as fo:
dic = pickle.load(fo, encoding = "bytes")
return dic
dic = load_data("/home/liuyanyi02/CIFAR/cifar-10-batches-py/test_batch")
data = dic[b'data']
label = dic[b'labels']
data = data.astype(np.float32)
import pickle
with open("meanstd.data", "rb") as f:
mean, std = pickle.load(f)
data = (data - mean) / std
data = np.resize(data, (10000, 3, 32, 32))
data = data.astype(np.float32)
"""
import cv2