I am trying to replicate the results from: https://github.com/raghakot/keras-vis/blob/master/examples/mnist/attention.ipynb to produce a saliency map of zero. I put in the exact same code, and as suggested, used:
from vis.visualization import visualize_saliency
from vis.utils import utils
from keras import activations
# Utility to search for layer index by name.
# Alternatively we can specify this as -1 since it corresponds to the last layer.
layer_idx = utils.find_layer_idx(model, 'preds')
# Swap softmax with linear
model.layers[layer_idx].activation = activations.linear
model = utils.apply_modifications(model)
grads = visualize_saliency(model, layer_idx, filter_indices=class_idx,seed_input=x_test[idx])
# Plot with 'jet' colormap to visualize as a heatmap.
plt.imshow(grads, cmap='jet')
However, I keep getting the following error:
InvalidArgumentError: conv2d_1_input_3:0 is both fed and fetched.
I looked elsewhere and have seen suggestions to upgrade keras-vis, I have done this but the same error shows up. The error appears to be in
grads = visualize_saliency(model, layer_idx, filter_indices=class_idx,seed_input=x_test[idx])
as when I comment this line out, no error shows.
How can I solve this?
It's solved! If anyone has this problem, use:
pip install git+git://github.com/raghakot/keras-vis.git --upgrade --no-deps
Related
I am currently trying to implement Mask RCNN by following Matterport repo. I have a doubt regarding implementation of tensorboard.
The dataset is similar to coco dataset. Inside model.py under def train, tensorboard is mentioned as
callbacks = [ keras.callbacks.TensorBoard(log_dir=self.log_dir,histogram_freq=0, write_graph=True, write_images=False)
But What else I should mention for using tensorboard? When I try to run the tensorboard, it say log file not found. I know that there is something I am missing some where!!. Please help me out !
In your model.train() ensure you set custom_callbacks = callbacks parameter.
If you specified these parameters exactly like this, then it means that your issue is that you do not properly open the logs directory.
Open (inside Anaconda/PyCharm) or a separate Python terminal and put the absolute path(to make sure it works):
tensorboard --logdir = my_absolute_path/logs/
I train a boject detection model on pytorch, and I have exported to onnx file.
And I want to convert it to caffe2 model :
import onnx
import caffe2.python.onnx.backend as onnx_caffe2_backend
# Load the ONNX ModelProto object. model is a standard Python protobuf object
model = onnx.load("CPU4export.onnx")
# prepare the caffe2 backend for executing the model this converts the ONNX model into a
# Caffe2 NetDef that can execute it. Other ONNX backends, like one for CNTK will be
# availiable soon.
prepared_backend = onnx_caffe2_backend.prepare(model)
# run the model in Caffe2
# Construct a map from input names to Tensor data.
# The graph of the model itself contains inputs for all weight parameters, after the input image.
# Since the weights are already embedded, we just need to pass the input image.
# Set the first input.
W = {model.graph.input[0].name: x.data.numpy()}
# Run the Caffe2 net:
c2_out = prepared_backend.run(W)[0]
# Verify the numerical correctness upto 3 decimal places
np.testing.assert_almost_equal(torch_out.data.cpu().numpy(), c2_out, decimal=3)
print("Exported model has been executed on Caffe2 backend, and the result looks good!")
I always got this error :
RuntimeError: ONNX conversion failed, encountered 1 errors:
Error while processing node: input: "90"
input: "91"
output: "92"
op_type: "Resize"
attribute {
name: "mode"
s: "nearest"
type: STRING
}
. Exception: Don't know how to translate op Resize
How can I solve it ?
The problem is that the Caffe2 ONNX backend does not yet support the export of the Resize operator.
Please raise an issue on the Caffe2 / PyTorch github -- there's an active community of developers who should be able to address this use case.
I'm using CNTK as the backend for Keras. I'm trying to use my model which I have trained using Keras in C++.
I have trained and saved my model using Keras which is in HDF5. How do I now use CNTK API to save it in their model-v2 format?
I tried this:
model = load_model('model2.h5')
cntk.ops.functions.Function.save(model, 'CNTK_model2.pb')
but i got the following error:
TypeError: save() missing 1 required positional argument: 'filename'
If tensorflow were the backend I would have done this:
model = load_model('model2.h5')
sess = K.get_session()
tf_saver = tf.train.Saver()
tf_saver.save(sess=sess, save_path=checkpoint_path)
How can I achieve the same thing?
As per the comments here, I was able to use this:
import cntk as C
import keras.backend as K
keras_model = K.load_model('my_keras_model.h5')
C.combine(keras_model.model.outputs).save('my_cntk_model')
cntk_model = C.load_model('my_cntk_model')
You can do something like this
model.outputs[0].save('CNTK_model2.pb')
I'm assuming here you have called model.compile (i.e. that's the only case I have tried :-)
The reason you see this error is because keras' cntk backend use a user defined function to do reshape on batch axis, which can't been serialized. We have fixed this issue in CNTK v2.2. Please upgrade your cntk to v2.2, and upgrade keras to last master.
Please see this pull request:
https://github.com/fchollet/keras/pull/7907
I have some experience with writing machine learning programs in python, but I'm new to TensorFlow and am checking it out. My dev environment is a lubuntu 14.04 64-bit virtual machine. I've created a python 3.5 conda environment from miniconda and installed TensorFlow 0.12 and its dependencies. I began trying to run some example code from TensorFlow's tutorials and encountered this warning when calling fit() in the boston.py example for input functions: source.
WARNING:tensorflow:Rank of input Tensor (1) should be the same as
output_rank (2) for column. Will attempt to expand dims. It is highly
recommended that you resize your input, as this behavior may change.
After some searching in Google, I found other people encountered this same warning:
https://github.com/tensorflow/tensorflow/issues/6184
https://github.com/tensorflow/tensorflow/issues/5098
Tensorflow - Boston Housing Data Tutorial Errors
However, they also experienced errors which prevent code execution from completing. In my case, the code executes with the above warning. Unfortunately, I couldn't find a single answer in those links regarding what caused the warning and how to fix the warning. They all focused on the error. How does one remove the warning? Or is the warning safe to ignore?
Cheers!
Extra info, I also see the following warnings when running the aforementioned boston.py example.
WARNING:tensorflow:*******************************************************
WARNING:tensorflow:TensorFlow's V1 checkpoint format has been
deprecated. WARNING:tensorflow:Consider switching to the more
efficient V2 format: WARNING:tensorflow:
'tf.train.Saver(write_version=tf.train.SaverDef.V2)'
WARNING:tensorflow:now on by default.
WARNING:tensorflow:*******************************************************
and
WARNING:tensorflow:From
/home/kade/miniconda3/envs/tensorflow/lib/python3.5/site-packages/tensorflow/contrib/learn/python/learn/estimators/dnn_linear_combined.py:1053
in predict.: calling BaseEstimator.predict (from
tensorflow.contrib.learn.python.learn.estimators.estimator) with x is
deprecated and will be removed after 2016-12-01. Instructions for
updating: Estimator is decoupled from Scikit Learn interface by moving
into separate class SKCompat. Arguments x, y and batch_size are only
available in the SKCompat class, Estimator will only accept input_fn.
Example conversion: est = Estimator(...) -> est =
SKCompat(Estimator(...))
UPDATE (2016-12-22):
I've tracked the warning to this file:
https://github.com/tensorflow/tensorflow/blob/master/tensorflow/contrib/layers/python/layers/feature_column_ops.py
and this code block:
except NotImplementedError:
with variable_scope.variable_scope(
None,
default_name=column.name,
values=columns_to_tensors.values()):
tensor = column._to_dense_tensor(transformed_tensor)
tensor = fc._reshape_real_valued_tensor(tensor, 2, column.name)
variable = [
contrib_variables.model_variable(
name='weight',
shape=[tensor.get_shape()[1], num_outputs],
initializer=init_ops.zeros_initializer(),
trainable=trainable,
collections=weight_collections)
]
predictions = math_ops.matmul(tensor, variable[0], name='matmul')
Note the line: tensor = fc._reshape_real_valued_tensor(tensor, 2, column.name)
The method signature is: _reshape_real_valued_tensor(input_tensor, output_rank, column_name=None)
The value 2 is hardcoded as the value of output_rank, but the boston.py example is passing in an input_tensor of rank 1. I will continue to investigate.
If you specify the shape of your tensor explicitly:
tf.constant(df[k].values, shape=[df[k].size, 1])
the warning should go away.
After I specify the shape of the tensor explicitly.
continuous_cols = {k: tf.constant(df[k].values, shape=[df[k].size, 1]) for k in CONTINUOUS_COLUMNS}
It works!
I am following a tutorial to create convolution neural network with Theano. Although, I got a problem in a piece of code:
>> x = theano.floatX.xmatrix(theano.config.floatX) # rasterized images
AttributeError: 'module' object has no attribute 'floatX'
I loaded floatX with:
>> from theano import config
and checked with:
>> print(theano.config.floatX)
float 32
But still cannot load the module xmatrix, which should be in theano.config.floatX, judging from documentation. Does somebody know where can I find it?
Thank you in advance!
This sections of the convnet tutorial has a bug or is very outdated. Symbolic variables in Theano are located in theano.tensor package. This package theano.floatX even doesn't exist!
The current version in the tutorial github repository works fine. They allocate the symbolic variable the right way:
# allocate symbolic variables for the data
index = T.lscalar() # index to a [mini]batch
x = T.matrix('x') # the data is presented as rasterized images
y = T.ivector('y') # the labels are presented as 1D vector of
# [int] labels
Browsing the tutorial repository I found the revision where this bug was corrected.
They seem to have forgotten to update the tutorial text with this fix.