I tried generating a saliency map using the following code:

backprop = Backprop(net)
target_class = 4
backprop.visualize(image, target_class, guided=True)

My net is a standard PyTorch neural network, and the image is a (1,3,84,84) image. However, I get the following error:

File "visualizations.py", line 89, in test backprop.visualize(image, target_class, guided=True) File "/home/user/.local/lib/python3.6/site-packages/flashtorch/saliency/backprop.py", line 168, in visualize guided=guided) File "/home/user/.local/lib/python3.6/site-packages/flashtorch/saliency/backprop.py", line 120, in calculate_gradients output.backward(gradient=target) File "/home/user/.local/lib/python3.6/site-packages/torch/tensor.py", line 150, in backward torch.autograd.backward(self, gradient, retain_graph, create_graph) File "/home/user/.local/lib/python3.6/site-packages/torch/autograd/init.py", line 99, in backward allow_unreachable=True) # allow_unreachable flag File "/home/user/.local/lib/python3.6/site-packages/flashtorch/saliency/backprop.py", line 214, in _record_gradients if self.gradients.shape == grad_in[0].shape: AttributeError: 'NoneType' object has no attribute 'shape'

Do I need to turn the image into a parameter with requires_grad = True? Is this usually done by the load_image function? I found that this function was re-sizing my image to 224x224. Thanks for your help!

hello, does flashtorch work on ensemble model, I passed output of one neural network to another, and got error.

I used g_ascent.visualize(model.encoder[0], title='conv');

I followed the tutorial on the README page, and the visualizations are not displaying. I am using a pretrained model, and I use a tensor with requires_grad=True as the image. My code doesn't run in to any errors. My code is below:

backprop = Backprop(net) backprop.visualize(image, target_class, guided = True)

Hi, I was interested in visualizing a CNN I created. I was following the example provided but when I try to use my model instead of a pretrained model from torchvision I get:

File "cnn_layer_visualization.py", line 82, in <module>
    gradients = backprop.calculate_gradients(input_, target_class)
  File "C:\Users\Mike\Anaconda3\lib\site-packages\flashtorch\saliency\backprop.py", line 106, in calculate_gradients
    output.backward(gradient=target)
  File "C:\Users\Mike\Anaconda3\lib\site-packages\torch\tensor.py", line 107, in backward
    torch.autograd.backward(self, gradient, retain_graph, create_graph)
  File "C:\Users\Mike\Anaconda3\lib\site-packages\torch\autograd\__init__.py", line 93, in backward
    allow_unreachable=True)  # allow_unreachable flag
RuntimeError: invalid gradient at index 0 - expected type torch.cuda.FloatTensor but got torch.FloatTensor

I looked through the source code a bit and saw that Backprop moves the network to cuda (if available) regardless if I move it there myself or not. Is this possibly a bug? I've had a bit of trouble with CUDA in the past so it might be a problem on my end.

Thanks for your time, if there's any other information I can provide please let me know. Cheers.

Problem Faced: Backprop visualize doesn't consider models with grayscale input.

Description: Refer to this line: https://github.com/MisaOgura/flashtorch/blob/ba5e3db3911afee583180bead9150ce4c392e09e/flashtorch/saliency/backprop.py#L222 In this line we add a check for 3 channels which results in skipping grayscale inputs to the conv layer

My current workaround: I have a copy of the backprop class with the in_channels set to 1 for my grayscale image use case.

Possible Fix: Would it make sense to remove the in_channels == 3 check since the first conv would be where the input image is supplied and should be good to register the hook at or am I missing something important here?

Hello, The current example for guided backpropogation only supports Alexnet which is a quite old. I've added an example to allow loading any supported model from the pytorch zoo and visualizing the output of guided backpropagation on it. Please review and merge it. Specific enhancements:

• The visualize method assumes normalization parameters of imagenet & doesn't support changing this. Edited the class interface to allow these to be specified. However it defaults to imagenet values so it does not affect past behavior.
• The examples queries the tochvision.models module for supported models and allows user to visualize backdrop output on the pre-trained models from pytorch repo.

enjoy, nav

My network does binary classifcation for detection of 11 different classes. E.G. It predicts if or if not there are apples, oranges, pineapple, and pears in the image I give it. So the output is a binary vector of length 11.

Can I use this project with out modifying it? I have checked and top_label will always be 9, so what I put for target_label will be ignored. I get the error

The predicted class index 9 does notequal the target class index 3. Calculatingthe gradient w.r.t. the predicted class. 'the gradient w.r.t. the predicted class.'

So in my example, if the network outputs a binary vector of length 4 corresponding to whether or not apple, orange, pineapple, and pear are in the image, how can I make it so that when I set target = 3 the code will show the gradient corresponding to the task of detecting pineapples?

Also, I am using modified ResNet-18 (transfer learned).

Describe the bug When testing the visualization tool using a custom pre-trained CNN that performs Binary Classification, the error "AttributeError: 'NoneType' object has no attribute 'shape'" crops up when trying to visualize an image that from the dataset used to train the model.

To Reproduce Steps to reproduce the behavior: Load image using custom transformation: transform = transforms.Compose([ transforms.RandomHorizontalFlip(), transforms.RandomRotation(20), transforms.ToTensor(), transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)) ]) that pertains to the CNN model developed. Load the image and apply the transformation: from flashtorch.utils import load_image from flashtorch.saliency import Backprop backprop=Backprop(n) image =load_image("img0042.png") i = transform(image) i = torch.unsqueeze(i,0) target_class=0 gradients=backprop.calculate_gradients(i,0) (general summary of the code)

Error message

AttributeError Traceback (most recent call last) in 1 from torch.autograd import Variable ----> 2 gradients=backprop.calculate_gradients(i,0)

~\Anaconda3\lib\site-packages\flashtorch\saliency\backprop.py in calculate_gradients(self, input_, target_class, take_max, guided, use_gpu) 118 # Calculate gradients of the target class output w.r.t. input_ 119 --> 120 output.backward(gradient=target) 121 122 # Detach the gradients from the graph and move to cpu

~\Anaconda3\lib\site-packages\torch\tensor.py in backward(self, gradient, retain_graph, create_graph) 164 products. Defaults to False. 165 """ --> 166 torch.autograd.backward(self, gradient, retain_graph, create_graph) 167 168 def register_hook(self, hook):

~\Anaconda3\lib\site-packages\torch\autograd_init_.py in backward(tensors, grad_tensors, retain_graph, create_graph, grad_variables) 97 Variable._execution_engine.run_backward( 98 tensors, grad_tensors, retain_graph, create_graph, ---> 99 allow_unreachable=True) # allow_unreachable flag 100 101

~\Anaconda3\lib\site-packages\flashtorch\saliency\backprop.py in _record_gradients(module, grad_in, grad_out) 212 def _register_conv_hook(self): 213 def _record_gradients(module, grad_in, grad_out): --> 214 if self.gradients.shape == grad_in[0].shape: 215 self.gradients = grad_in[0] 216

AttributeError: 'NoneType' object has no attribute 'shape'

Environment (please complete the following information):

• OS: [Windows 10 1909 version: 18363.476]
• Python version: [3.7..4]
• FlashTorch version: [0.1.1]

Describe the bug Complete output from command python setup.py egg_info: Traceback (most recent call last): File "", line 1, in File "C:\Users\HUAJI0~1\AppData\Local\Temp\pip-install-uc9zsivx\flashtorch\setup.py", line 5, in long_description = fh.read()

UnicodeDecodeError: 'gbk' codec can't decode byte 0x9d in position 5424: illegal multibyte sequence

To Reproduce pip install flashtorch

Environment (please complete the following information):

• OS: windows10
• Python version: 3.7
• FlashTorch version: 0.1.0

fix readme.md line 101 “see” fix to "see"

Hi, Thanks for your work. I am trying to use "Activation maximization" notebook with my own custom model. The difference is that my model takes an input of (56,56,24). The gradient ascent function accepts my model without an error. I perform my own transformation on a random generated input and the input shape is torch.Size([1, 24, 56, 56]).

When I call visualize function giving it any intermediate layer and filter, it throws below error:

TypeError Traceback (most recent call last) in ----> 1 g_ascent.visualize(layer2_0_conv2, title='layer1_0_conv2')

~/.local/lib/python3.6/site-packages/flashtorch/activmax/gradient_ascent.py in visualize(self, layer, filter_idxs, lr, num_iter, num_subplots, figsize, title, return_output) 210 num_iter, 211 len(filter_idxs), --> 212 title=title) 213 214 if return_output:

~/.local/lib/python3.6/site-packages/flashtorch/activmax/gradient_ascent.py in _visualize_filters(self, layer, filter_idxs, num_iter, num_subplots, title) 347 standardize_and_clip(output[-1], 348 saturation=0.15, --> 349 brightness=0.7))) 350 351 plt.subplots_adjust(wspace=0, hspace=0); # noqa

~/.local/lib/python3.6/site-packages/matplotlib/init.py in inner(ax, data, *args, **kwargs) 1563 def inner(ax, *args, data=None, **kwargs): 1564 if data is None: -> 1565 return func(ax, *map(sanitize_sequence, args), **kwargs) 1566 1567 bound = new_sig.bind(ax, *args, **kwargs)

~/.local/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py in wrapper(*args, **kwargs) 356 f"%(removal)s. If any parameter follows {name!r}, they " 357 f"should be pass as keyword, not positionally.") --> 358 return func(*args, **kwargs) 359 360 return wrapper

~/.local/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py in wrapper(*args, **kwargs) 356 f"%(removal)s. If any parameter follows {name!r}, they " 357 f"should be pass as keyword, not positionally.") --> 358 return func(*args, **kwargs) 359 360 return wrapper

~/.local/lib/python3.6/site-packages/matplotlib/axes/_axes.py in imshow(self, X, cmap, norm, aspect, interpolation, alpha, vmin, vmax, origin, extent, shape, filternorm, filterrad, imlim, resample, url, **kwargs) 5624 resample=resample, **kwargs) 5625 -> 5626 im.set_data(X) 5627 im.set_alpha(alpha) 5628 if im.get_clip_path() is None:

~/.local/lib/python3.6/site-packages/matplotlib/image.py in set_data(self, A) 697 or self._A.ndim == 3 and self._A.shape[-1] in [3, 4]): 698 raise TypeError("Invalid shape {} for image data" --> 699 .format(self._A.shape)) 700 701 if self._A.ndim == 3:

TypeError: Invalid shape (56, 56, 24) for image data

Can anyone help me with this please.

I am using a custom pretrained fully connected Pytorch model which was trained on CIFAR10 using GPU. While trying to use the backprop.visualize, I am getting the following error.

---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
<ipython-input-24-554f7ff5427b> in <module>()
      8 # Ready to roll!
      9 
---> 10 backprop.visualize(owl, target_class, guided=True, use_gpu=True)

8 frames
/usr/local/lib/python3.6/dist-packages/flashtorch/saliency/backprop.py in visualize(self, input_, target_class, guided, use_gpu, figsize, cmap, alpha, return_output)
    166         gradients = self.calculate_gradients(input_,
    167                                              target_class,
--> 168                                              guided=guided)
    169         max_gradients = self.calculate_gradients(input_,
    170                                                  target_class,

/usr/local/lib/python3.6/dist-packages/flashtorch/saliency/backprop.py in calculate_gradients(self, input_, target_class, take_max, guided, use_gpu)
     87         # Get a raw prediction value (logit) from the last linear layer
     88 
---> 89         output = self.model(input_)
     90 
     91         # Don't set the gradient target if the model is a binary classifier

/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py in __call__(self, *input, **kwargs)
    539             result = self._slow_forward(*input, **kwargs)
    540         else:
--> 541             result = self.forward(*input, **kwargs)
    542         for hook in self._forward_hooks.values():
    543             hook_result = hook(self, input, result)

/content/archs/cifar10/fc1.py in forward(self, x)
     18     def forward(self, x):
     19         x = torch.flatten(x, 1)
---> 20         x = self.classifier(x)
     21         return x
     22 #====================================

/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py in __call__(self, *input, **kwargs)
    539             result = self._slow_forward(*input, **kwargs)
    540         else:
--> 541             result = self.forward(*input, **kwargs)
    542         for hook in self._forward_hooks.values():
    543             hook_result = hook(self, input, result)

/usr/local/lib/python3.6/dist-packages/torch/nn/modules/container.py in forward(self, input)
     90     def forward(self, input):
     91         for module in self._modules.values():
---> 92             input = module(input)
     93         return input
     94 

/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py in __call__(self, *input, **kwargs)
    539             result = self._slow_forward(*input, **kwargs)
    540         else:
--> 541             result = self.forward(*input, **kwargs)
    542         for hook in self._forward_hooks.values():
    543             hook_result = hook(self, input, result)

/usr/local/lib/python3.6/dist-packages/torch/nn/modules/linear.py in forward(self, input)
     85 
     86     def forward(self, input):
---> 87         return F.linear(input, self.weight, self.bias)
     88 
     89     def extra_repr(self):

/usr/local/lib/python3.6/dist-packages/torch/nn/functional.py in linear(input, weight, bias)
   1368     if input.dim() == 2 and bias is not None:
   1369         # fused op is marginally faster
-> 1370         ret = torch.addmm(bias, input, weight.t())
   1371     else:
   1372         output = input.matmul(weight.t())

RuntimeError: Expected object of device type cuda but got device type cpu for argument #2 'mat1' in call to _th_addmm

My model :

class fc1(nn.Module):

    def __init__(self, num_classes=10, hidden=300):
        super(fc1, self).__init__()
        self.classifier = nn.Sequential(
            nn.Linear(3*32*32, hidden),
            nn.ReLU(inplace=True),
            nn.Linear(hidden, 100),
            nn.ReLU(inplace=True),
            nn.Linear(100, num_classes),
        )

    def forward(self, x):
        x = torch.flatten(x, 1)
        x = self.classifier(x)
        return x

Loaded the model using the following lines :

from archs.cifar10 import fc1

import torch
model = torch.load("/content/epoch_198.pth.tar")
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

model = model.to(device)

Thanks a lot for developing fantastic package FlashTorch!

I have a question about the inputs. My inputs contains two parts:

• the image
• additional information of image (one node)

The additional node is inputed in the middle layers with torch.cat.

In such case, can I use FlashTorch to obtain my saliency maps? Thanks a lot!

RuntimeError Traceback (most recent call last) in 9 # Ready to roll! 10 ---> 11 backprop.visualize(owl, target_class, guided=True)

~\Downloads\flashtorch-master\flashtorch\saliency\backprop.py in visualize(self, input_, target_class, guided, use_gpu, figsize, cmap, alpha, return_output) 180 # (title, [(image1, cmap, alpha), (image2, cmap, alpha)]) 181 ('Input image', --> 182 [(format_for_plotting(denormalize(input_)), None, None)]), 183 ('Gradients across RGB channels', 184 [(format_for_plotting(standardize_and_clip(gradients)),

~\Downloads\flashtorch-master\flashtorch\utils_init_.py in denormalize(tensor) 117 118 for channel, mean, std in zip(denormalized[0], means, stds): --> 119 channel.mul_(std).add_(mean) 120 121 return denormalized

RuntimeError: Output 0 of UnbindBackward is a view and is being modified inplace. This view is the output of a function that returns multiple views. Such functions do not allow the output views to be modified inplace. You should replace the inplace operation by an out-of-place one.

Describe the bug When running the google colab version Part 4. Visualize Saliency Maps throws up the following error

RuntimeError Traceback (most recent call last) in () 9 # Ready to roll! 10 ---> 11 backprop.visualize(owl, target_class, guided=True, use_gpu=True)

1 frames /usr/local/lib/python3.6/dist-packages/flashtorch/utils/init.py in denormalize(tensor) 117 118 for channel, mean, std in zip(denormalized[0], means, stds): --> 119 channel.mul_(std).add_(mean) 120 121 return denormalized

RuntimeError: Output 0 of UnbindBackward is a view and is being modified inplace. This view is the output of a function that returns multiple views. Such functions do not allow the output views to be modified inplace. You should replace the inplace operation by an out-of-place one.

To Reproduce Steps to reproduce the behavior:

1. Go to https://colab.research.google.com/github/MisaOgura/flashtorch/blob/master/examples/visualize_saliency_with_backprop_colab.ipynb
2. Click on 'Run all'
3. Scroll down to 'Section 4'
4. See error

Expected behavior A clear and concise description of what you expected to happen.

Screenshots If applicable, add screenshots to help explain your problem.

Environment (please complete the following information):

• OS: [e.g. iOS]
• Python version: [e.g. 3.6]
• FlashTorch version: [e.g. 0.1.0]

Additional context Add any other context about the problem here.

Thanks so much for sharing your work and make it so easy...in theory...to use. Although I haven't been able to get it to work by running your Colab example(s).

Describe the bug The two middle images involving gradients for each example, "Gradients across RGB channels" and "Max Gradients", appear as uniform color. The RGB is all grey and the Max is all purple.

To Reproduce Steps to reproduce the behavior:

1. Go to the README page for this repo, scroll down to "Want to try?"
2. Next to "Saliency maps:", click on the "Open in Colab" badge.
3. Save a copy of the notebook to Drive (or not; this step is not essential to produce the error)
4. Edit > Notebook Settings, select GPU
5. Runtime > Restart and run all
6. Scroll down and observe that only the leftmost and right most images for owl, peacock and toucan are visible, the middle two images are "blank"

Expected behavior Gradient images should show content, such as colored pixels around the owl's eyes as in your Medium post. That is not what running the Colab demo gives me. See my sample screenshot below.

Screenshots

Environment (please complete the following information): Colab. Whatever OS that's running. Brand new installation of FlashTorch via the !pip install flashtorch in the notebook. Looks like (0.1.3) Looks like Torch 1.6

Additional context From pip (after I re-ran it a second time):

Collecting flashtorch
  Downloading https://files.pythonhosted.org/packages/de/cb/482274e95812c9a17bd156956bef80a8e2683a2b198a505fb922f1c01a71/flashtorch-0.1.3.tar.gz
Requirement already satisfied: matplotlib in /usr/local/lib/python3.6/dist-packages (from flashtorch) (3.2.2)
Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from flashtorch) (1.18.5)
Requirement already satisfied: Pillow in /usr/local/lib/python3.6/dist-packages (from flashtorch) (7.0.0)
Requirement already satisfied: torch in /usr/local/lib/python3.6/dist-packages (from flashtorch) (1.6.0+cu101)
Requirement already satisfied: torchvision in /usr/local/lib/python3.6/dist-packages (from flashtorch) (0.7.0+cu101)
Collecting importlib_resources
  Downloading https://files.pythonhosted.org/packages/ba/03/0f9595c0c2ef12590877f3c47e5f579759ce5caf817f8256d5dcbd8a1177/importlib_resources-3.0.0-py2.py3-none-any.whl
Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.6/dist-packages (from matplotlib->flashtorch) (0.10.0)
Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->flashtorch) (2.4.7)
Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->flashtorch) (2.8.1)
Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->flashtorch) (1.2.0)
Requirement already satisfied: future in /usr/local/lib/python3.6/dist-packages (from torch->flashtorch) (0.16.0)
Requirement already satisfied: zipp>=0.4; python_version < "3.8" in /usr/local/lib/python3.6/dist-packages (from importlib_resources->flashtorch) (3.1.0)
Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from cycler>=0.10->matplotlib->flashtorch) (1.15.0)
Building wheels for collected packages: flashtorch
  Building wheel for flashtorch (setup.py) ... done
  Created wheel for flashtorch: filename=flashtorch-0.1.3-cp36-none-any.whl size=26248 sha256=95cdabc7c3dbc87e25d0795eb47d01971668600bdc58cd558b670c5e7ce0b725
  Stored in directory: /root/.cache/pip/wheels/03/6d/b1/2d3c5987b69e900fcceceeef39d3ed92dfe46ba1359b9c79f8
Successfully built flashtorch
Installing collected packages: importlib-resources, flashtorch
Successfully installed flashtorch-0.1.3 importlib-resources-3.0.0

AttributeError Traceback (most recent call last) in () 4 input_=apply_transforms(image) 5 target_class=24 ----> 6 backprop.visualize(input,target_class,guided=True)

1 frames /usr/local/lib/python3.6/dist-packages/flashtorch/saliency/backprop.py in calculate_gradients(self, input_, target_class, take_max, guided, use_gpu) 84 self.model.zero_grad() 85 ---> 86 self.gradients = torch.zeros(input_.shape) 87 88 # Get a raw prediction value (logit) from the last linear layer

AttributeError: 'function' object has no attribute 'shape'

how to fix it?