KeyError in name = param_map[id(u)]

[hughperkins]

When I run:

import torch
from torch import nn
import torchviz

h1 = nn.Linear(1, 1)

inputs = torch.zeros(1, 1, requires_grad=True)
inputs.data = torch.rand(1, 1)
print('inputs', inputs)
outputs = h1(inputs)
print('outputs', outputs)
torchviz.make_dot(outputs, params={'inputs': inputs})

I get:

inputs tensor([[0.3559]], requires_grad=True)
outputs tensor([[0.4822]], grad_fn=<ThAddmmBackward>)
---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
<ipython-input-4-a36a1735c26c> in <module>()
     10 outputs = h1(inputs)
     11 print('outputs', outputs)
---> 12 torchviz.make_dot(outputs, params={'inputs': inputs})

/toknas/hugh/condap4/lib/python3.6/site-packages/torchviz/dot.py in make_dot(var, params)
     68             add_nodes(v.grad_fn)
     69     else:
---> 70         add_nodes(var.grad_fn)
     71 
     72     resize_graph(dot)

/toknas/hugh/condap4/lib/python3.6/site-packages/torchviz/dot.py in add_nodes(var)
     57                     if u[0] is not None:
     58                         dot.edge(str(id(u[0])), str(id(var)))
---> 59                         add_nodes(u[0])
     60             if hasattr(var, 'saved_tensors'):
     61                 for t in var.saved_tensors:

/toknas/hugh/condap4/lib/python3.6/site-packages/torchviz/dot.py in add_nodes(var)
     57                     if u[0] is not None:
     58                         dot.edge(str(id(u[0])), str(id(var)))
---> 59                         add_nodes(u[0])
     60             if hasattr(var, 'saved_tensors'):
     61                 for t in var.saved_tensors:

/toknas/hugh/condap4/lib/python3.6/site-packages/torchviz/dot.py in add_nodes(var)
     45             elif hasattr(var, 'variable'):
     46                 u = var.variable
---> 47                 name = param_map[id(u)] if params is not None else ''
     48                 node_name = '%s\n %s' % (name, size_to_str(u.size()))
     49                 dot.node(str(id(var)), node_name, fillcolor='lightblue')

KeyError: 140051140404712

collectenv output:

Collecting environment information...
PyTorch version: 0.4.1
Is debug build: No
CUDA used to build PyTorch: 9.0.176

OS: Ubuntu 16.04.5 LTS
GCC version: (Ubuntu 5.4.0-6ubuntu1~16.04.10) 5.4.0 20160609
CMake version: version 3.5.1

Python version: 3.6
Is CUDA available: Yes
CUDA runtime version: 9.0.176
GPU models and configuration: GPU 0: Tesla M60
Nvidia driver version: 384.130
cuDNN version: Probably one of the following:
/usr/lib/x86_64-linux-gnu/libcudnn.so.5.1.10
/usr/lib/x86_64-linux-gnu/libcudnn.so.6.0.21
/usr/lib/x86_64-linux-gnu/libcudnn.so.7.0.3
/usr/lib/x86_64-linux-gnu/libcudnn_static_v5.a
/usr/lib/x86_64-linux-gnu/libcudnn_static_v6.a
/usr/lib/x86_64-linux-gnu/libcudnn_static_v7.a

Versions of relevant libraries:
[pip] numpy (1.14.3)
[pip] pytorchviz (0.0.1)
[pip] torch (0.4.1)
[pip] torchtext (0.2.3)
[pip] torchvision (0.2.1)
[conda] pytorch                   0.4.1           py36_cuda9.0.176_cudnn7.1.2_1    pytorch
[conda] pytorchviz                0.0.1                     <pip>
[conda] torchtext                 0.2.3                     <pip>
[conda] torchvision               0.2.1                    py36_1    pytorch

[szagoruyko]

Hi @hughperkins , it's been a while! you need to provide either None parameters, or * all * tensors that require grad in the graph, this will fix your example:

torchviz.make_dot(outputs, params={**{'inputs': inputs}, **dict(h1.named_parameters())})

we should be either giving a better error message, or allow missing named nodes.

[hughperkins]

it's been a while!

Yes :)

you need to provide either None parameters, or * all * tensors that require grad in the graph

Ah, ok. Awesome, thanks! Will give this a try :)

[aaronlelevier]

I am getting this issue as well. Here is the final stacktrace line:

~/Documents/github/ssdmultibox/venv/lib/python3.7/site-packages/torchviz/dot.py in add_nodes(var)
     45             elif hasattr(var, 'variable'):
     46                 u = var.variable
---> 47                 name = param_map[id(u)] if params is not None else ''
     48                 node_name = '%s\n %s' % (name, size_to_str(u.size()))
     49                 dot.node(str(id(var)), node_name, fillcolor='lightblue')

KeyError: 4871725920

I tried:

torchviz.make_dot(loss, params={**{'inputs': inputs}, **dict(model.named_parameters())})

And also:

torchviz.make_dot(loss, params=dict(model.named_parameters()))

But, I still got the same error. Is this user error on my part? My model is an SSD Multibox model where the predictions are a list of tensors of different sizes. I am wondering if that might be the issue and it doesn't work with make_dot. Thanks

[Jacoobr]

Hi @szagoruyko ，I met the same error when i train my network using 'make_dot' function to visulize my nerual network structure. There are the errors as follow:

output = self.lstmlayer(output)[0]
Traceback (most recent call last):
  File "/data/xiaojianli/sorftWare/pycharm-2017.3.1/helpers/pydev/pydevd.py", line 1668, in <module>
    main()
  File "/data/xiaojianli/sorftWare/pycharm-2017.3.1/helpers/pydev/pydevd.py", line 1662, in main
    globals = debugger.run(setup['file'], None, None, is_module)
  File "/data/xiaojianli/sorftWare/pycharm-2017.3.1/helpers/pydev/pydevd.py", line 1072, in run
    pydev_imports.execfile(file, globals, locals)  # execute the script
  File "/mnt/data/polygonRNN_data/pytorch-polygon-rnn-master-pmri/train.py", line 113, in <module>
    make_dot(r, params=dict(list(PolygonNet().named_parameters())+[('x',x),('x1',x1),('x2',x2),('x3',x3)]))
  File "/usr/local/lib/python2.7/dist-packages/torchviz/dot.py", line 70, in make_dot
    add_nodes(var.grad_fn)
  File "/usr/local/lib/python2.7/dist-packages/torchviz/dot.py", line 59, in add_nodes
    add_nodes(u[0])
  File "/usr/local/lib/python2.7/dist-packages/torchviz/dot.py", line 59, in add_nodes
    add_nodes(u[0])
  File "/usr/local/lib/python2.7/dist-packages/torchviz/dot.py", line 59, in add_nodes
    add_nodes(u[0])
  File "/usr/local/lib/python2.7/dist-packages/torchviz/dot.py", line 59, in add_nodes
    add_nodes(u[0])
  File "/usr/local/lib/python2.7/dist-packages/torchviz/dot.py", line 59, in add_nodes
    add_nodes(u[0])
  File "/usr/local/lib/python2.7/dist-packages/torchviz/dot.py", line 47, in add_nodes
    name = param_map[id(u)] if params is not None else ''
KeyError: 140707794969536

there are the 'make_dot' function in my code:

        x = Variable(data[0].type(dtype), requires_grad=True)
        x1 = Variable(data[1].type(dtype), requires_grad=True)
        x2 = Variable(data[2].type(dtype), requires_grad=True)
        x3 = Variable(data[3].type(dtype), requires_grad=True)
        ta = Variable(data[4].type(dtype_t))
        optimizer.zero_grad()   #make the grad be zero
        r = net(x,x1,x2,x3)     # go to forward func in model.py script
        make_dot(r, params=dict(list(PolygonNet().named_parameters())+[('x',x),('x1',x1),('x2',x2),('x3',x3)]))

when i debuge my code, i found this line code output_o = self.lstmlayer(output_i)[0] that call lstmlayer casued this Error. Why when i call lstmlayer() the error occured? What's more, when i uncomment make_dot(r, params=dict(list(PolygonNet().named_parameters())+[('x',x),('x1',x1),('x2',x2),('x3',x3)])) code, the training code goes well. So the error only occured when i try to visulize the network structure. Waiting for your appreciative apply. Thank U~

[chensonglu]

@szagoruyko set None parameters works for me, thanks

[vcharvet]

There's a hacky way to fix it
Change line 47 in torchviz/dot.py to
name = param_map.get(id(u), ' ') if params is not None else ''