List index out of range

[rajendraranabhat]

I am trying to run tflearn for simple mnist classification using MLP. I tried to other example and I am getting this error for almost all. I just started using TFlearn.

IndexError Traceback (most recent call last)
in ()
48 max_checkpoints=10, tensorboard_verbose=0)
49 model.fit(X, Y, n_epoch=100, validation_set=(testX, testY),
---> 50 show_metric=True, batch_size=256, run_id='resnet_mnist')

/usr/local/lib/python2.7/dist-packages/tflearn/models/dnn.pyc in fit(self, X_inputs, Y_targets, n_epoch, validation_set, show_metric, batch_size, shuffle, snapshot_epoch, snapshot_step, excl_trainops, run_id)
155 # TODO: check memory impact for large data and multiple optimizers
156 feed_dict = feed_dict_builder(X_inputs, Y_targets, self.inputs,
--> 157 self.targets)
158 feed_dicts = [feed_dict for i in self.train_ops]
159 val_feed_dicts = None

/usr/local/lib/python2.7/dist-packages/tflearn/utils.pyc in feed_dict_builder(X, Y, net_inputs, net_targets)
265 X = [X]
266 for i, x in enumerate(X):
--> 267 feed_dict[net_inputs[i]] = x
268 else:
269 # If a dict is provided

IndexError: list index out of range

---

Code

from __future__ import division, print_function, absolute_import

import tflearn

# Data loading and preprocessing
import tflearn.datasets.mnist as mnist
X, Y, testX, testY = mnist.load_data(one_hot=True)

# Building deep neural network
input_layer = tflearn.input_data(shape=[None, 784])
dense1 = tflearn.fully_connected(input_layer, 64, activation='tanh',
                                 regularizer='L2', weight_decay=0.001)
dropout1 = tflearn.dropout(dense1, 0.8)
dense2 = tflearn.fully_connected(dropout1, 64, activation='tanh',
                                 regularizer='L2', weight_decay=0.001)
dropout2 = tflearn.dropout(dense2, 0.8)
softmax = tflearn.fully_connected(dropout2, 10, activation='softmax')

# Regression using SGD with learning rate decay and Top-3 accuracy
sgd = tflearn.SGD(learning_rate=0.1, lr_decay=0.96, decay_step=1000)
top_k = tflearn.metrics.Top_k(3)
net = tflearn.regression(softmax, optimizer=sgd, metric=top_k,
                         loss='categorical_crossentropy')

# Training
model = tflearn.DNN(net, tensorboard_verbose=1)
model.fit(X, Y, n_epoch=5, validation_set=(testX, testY), show_metric=True, run_id="dense_model")

[aymericdamien]

This code is working perfectly fine for me. Can you try to re-install tflearn?

[wgottschalk]

I'm getting a similar error. Here's a link to the SO question I asked. http://stackoverflow.com/questions/40200311/tflearn-covnet-example-resulting-in-an-error-w-cifar-10

Are you using python notebooks?

[cemysf]

same error here, while running quickstart tutorial on ipython notebook

Edit: restarting kernel solved the problem

[wgottschalk]

Yeah. I've found when I run the session for the first time it works correctly but I get the index error when I tried to rerun the notebook cell. @cemysf I noticed that in ipython notebooks when using vanilla tensor flow that variables get namespaced to the number of iterations of the cell. My quick fix was to use tf.reset_default_graph() at the top of the cell to clear all of the variables.

Perhaps it's worth submitting a PR to address this issue?

[ShivangiM]

Restarting kernel solved my problem too.

[llealgt]

For me , adding this line before any code do the trick:
tf.reset_default_graph()

[chrisplyn]

For me restarting kernel doesn't work, adding tf.reset_default_graph() does.

[rushatrai]

Restarting the kernel worked. I don't know why this is a recurring problem in Jupyter Notebooks. Some bug I suppose.

[annanurov]

How do you restart a kernel?
Also: is there a function in tflearn that replaces tf.reset_default_graph()?

Thank you

[annanurov]

I used:

from tensorflow import reset_default_graph
reset_default_graph()

but same error is there:
ValueError: Index (16494) out of range (0-7127)

[airportpeople]

Worked for me!
I ran

from tensorflow import reset_default_graph
reset_default_graph()

, then I reran the code that built the network. Then I ran the .fit method.

[anandithaaa]

I still get the same exact error, even with the from tensorflow import reset_default_graph &
reset_default_graph().

I've restarted the kernel as well, no luck.

[uzairali37]

Funzione per me. Grazie!

from tensorflow import reset_default_graph
reset_default_graph()

[ketan4373]

For me, it was working with jupyter-notebook and to tweak a model, I need to reset tensorflow graph by tf.reset_default_graph() every time.

[XushaopengHNU]

I got the same error when using jupyter notebook, and restarting kernel solved the problem.