Ufdjrw

Probability that a draw from a normal distribution is some number greater than another draw from the same distribution

What killed these X2 caps?

Can my sorcerer use a spellbook only to collect spells and scribe scrolls, not cast?

Ambiguity in the definition of entropy

Is it acceptable for a professor to tell male students to not think that they are smarter than female students?

Do UK voters know if their MP will be the Speaker of the House?

What's the in-universe reasoning behind sorcerers needing material components?

If human space travel is limited by the G force vulnerability, is there a way to counter G forces?

Can I run a new neutral wire to repair a broken circuit?

How do conventional missiles fly?

How to tell a function to use the default argument values?

How do I gain back my faith in my PhD degree?

What is the most common color to indicate the input-field is disabled?

Why is it a bad idea to hire a hitman to eliminate most corrupt politicians?

Would Slavery Reparations be considered Bills of Attainder and hence Illegal?

Valid term from quadratic sequence?

Why is this clock signal connected to a capacitor to gnd?

GFCI outlets - can they be repaired? Are they really needed at the end of a circuit?

Mathematica command that allows it to read my intentions

How would I stat a creature to be immune to everything but the Magic Missile spell? (just for fun)

Detention in 1997

What reasons are there for a Capitalist to oppose a 100% inheritance tax?

Is it logically or scientifically possible to artificially send energy to the body?

What does the expression "A Mann!" means

Error in simple example pytorch -> onnx -> tensorflow

Tensorflow: how to save/restore a model?TensorFlow not found using pipNameError: global name 'linear' is not definedMissing Weight Vectors when converting from PyTorch to CoreML via ONNXTrouble using openCV to load a net from ONNX (python/pytorch)Conversion from pytorch to onnx to tensorflow graph definition to tflite - TOCO failed - type check failexporting pytorch model through onnx to caffe2Transferring pretrained pytorch model to onnxError while converting pytorch to tensorflow through onnxPyTorch VAE fails conversion to onnx

0

I'm trying to convert a pytorch model to a tensorflow model for inference using ONNX. Below is a simple example that fails with error:
ValidationError: model with IR version < 3 cannot have opset_import specified. What could cause this error? How is this done correctly?

The code is:

from torch.nn import Linear
from torch.autograd import Variable
from numpy.random import randn
from torch import from_numpy
from torch.onnx import export
from onnx import load
from onnx.checker import check_graph
from onnx_tf.backend import prepare
net = Linear(10, 2)
inp = Variable(from_numpy(randn(16, 10)).float())
export(net, args=inp, f='./test.pb', verbose=True,
 training=False, export_params=True)
model = load('./test.pb')
check_graph(model.graph)
tfmodel = prepare(model)

Here are the versions of the used modules:

• 
  torch : 0.3.1.post2


• 
  numpy : 1.14.3


• 
  onnx : 1.1.1


• 
  onnx_tf : 1.1.1


• 
  tensorflow : 1.5.0

Btw: There's no tag ONNX yet. Is that project dead or still in the womb?

tensorflow pytorch onnx

share|improve this question

edited Mar 7 at 22:16

dashesy

1,52323049

asked May 21 '18 at 21:07

JusJus

1899

add a comment | 

0

I'm trying to convert a pytorch model to a tensorflow model for inference using ONNX. Below is a simple example that fails with error:
ValidationError: model with IR version < 3 cannot have opset_import specified. What could cause this error? How is this done correctly?

The code is:

from torch.nn import Linear
from torch.autograd import Variable
from numpy.random import randn
from torch import from_numpy
from torch.onnx import export
from onnx import load
from onnx.checker import check_graph
from onnx_tf.backend import prepare
net = Linear(10, 2)
inp = Variable(from_numpy(randn(16, 10)).float())
export(net, args=inp, f='./test.pb', verbose=True,
 training=False, export_params=True)
model = load('./test.pb')
check_graph(model.graph)
tfmodel = prepare(model)

Here are the versions of the used modules:

• 
  torch : 0.3.1.post2


• 
  numpy : 1.14.3


• 
  onnx : 1.1.1


• 
  onnx_tf : 1.1.1


• 
  tensorflow : 1.5.0

Btw: There's no tag ONNX yet. Is that project dead or still in the womb?

tensorflow pytorch onnx

share|improve this question

edited Mar 7 at 22:16

dashesy

1,52323049

asked May 21 '18 at 21:07

JusJus

1899

add a comment | 

I'm trying to convert a pytorch model to a tensorflow model for inference using ONNX. Below is a simple example that fails with error:
ValidationError: model with IR version < 3 cannot have opset_import specified. What could cause this error? How is this done correctly?

The code is:

from torch.nn import Linear
from torch.autograd import Variable
from numpy.random import randn
from torch import from_numpy
from torch.onnx import export
from onnx import load
from onnx.checker import check_graph
from onnx_tf.backend import prepare
net = Linear(10, 2)
inp = Variable(from_numpy(randn(16, 10)).float())
export(net, args=inp, f='./test.pb', verbose=True,
 training=False, export_params=True)
model = load('./test.pb')
check_graph(model.graph)
tfmodel = prepare(model)

Here are the versions of the used modules:

• 
  torch : 0.3.1.post2


• 
  numpy : 1.14.3


• 
  onnx : 1.1.1


• 
  onnx_tf : 1.1.1


• 
  tensorflow : 1.5.0

Btw: There's no tag ONNX yet. Is that project dead or still in the womb?

tensorflow pytorch onnx

share|improve this question

edited Mar 7 at 22:16

dashesy

1,52323049

asked May 21 '18 at 21:07

JusJus

1899

from torch.nn import Linear
from torch.autograd import Variable
from numpy.random import randn
from torch import from_numpy
from torch.onnx import export
from onnx import load
from onnx.checker import check_graph
from onnx_tf.backend import prepare
net = Linear(10, 2)
inp = Variable(from_numpy(randn(16, 10)).float())
export(net, args=inp, f='./test.pb', verbose=True,
 training=False, export_params=True)
model = load('./test.pb')
check_graph(model.graph)
tfmodel = prepare(model)

share|improve this question

edited Mar 7 at 22:16

dashesy

1,52323049

asked May 21 '18 at 21:07

JusJus

1899

share|improve this question

edited Mar 7 at 22:16

dashesy

1,52323049

asked May 21 '18 at 21:07

JusJus

1899

edited Mar 7 at 22:16

dashesy

1,52323049

edited Mar 7 at 22:16

dashesy

1,52323049

asked May 21 '18 at 21:07

JusJus

1899

asked May 21 '18 at 21:07

JusJus

1899