We Keep Coding

Error in TF variable singleton variable creation in updating dynamic model

Following the papers Progressive Gans (https://arxiv.org/abs/1710.10196), I implement keras.Model that needs to grow in size (layers). I first initialize the full model. But when the time making an inference, I will use only partial of model but the same trainable_variables, e.g. 4x4 then 8x8. So that the trainable_variables passing to train_step which decorated with tf.function will be different. This work properly for computing gradient etc. but not optimizer.apply_gradients.
The code look something like this:
strategy = tf.distribute.MirroredStrategy()
G = Generator()
with strategy.scope():
Opt = keras.optimizers.Adam()
G.initialize_model() # initialize full model
#tf.function
def train_step(optimizer, model, var_to_train):
with tf.GradientTape() as tape:
Loss = loss(model(datasets))
grads = tape.gradient(Loss, variables)
opt.apply_gradients(zip(grads, variables)) # this will raise ValueError
res = 4 # resolution of 4x4
for ep in range(epochs):
if ep % 100 == 0:
res = res * 2
cur_model = G.forward(output_shape=(res, res, 3)) # output for given image resolution
var = cur_model.trainable_variables # this variables will be increasing as we grow model
strategy.run(train_step, args=(Opt, cur_model, var))
Note, however, that this will work fine when train_step is not used in the context of tf.function or in MirroredStrategy. From last section of seem not to solve the problem. I tried tf.distribute.ReplicaContext.all_reduce or any equivalent method for obtaining local results from all replica but it won't work since the trainable_variables are created inside the strategy.scope() so every update must be in the context of Replica.
The only naive solution I could is to train, for example 4x4 model and save it. Then use transfer learning load it back to 8x8 model.
I want to use usual keras optimizer which support any dynamic trainable_variables passed through tf.function context.
[1]: https://www.tensorflow.org/guide/function#creating_tfvariables:~:text=shape%3D()%2C%20dtype%3Dfloat32)

First call optimizer._create_all_weights(var) where the argument var is the full model. This will make the optimizer create all variables and must be done at the beginning before making any updates. In updating gradient it won't create again but still can provide a subset of it. This work in the context of tf.function too.

Access output of intermediate layers in Tensor-flow 2.0 in eager mode

I have CNN that I have built using on Tensor-flow 2.0. I need to access outputs of the intermediate layers. I was going over other stackoverflow questions that were similar but all had solutions involving Keras sequential model.
I have tried using model.layers[index].output but I get
Layer conv2d has no inbound nodes.
I can post my code here (which is super long) but I am sure even without that someone can point to me how it can be done using just Tensorflow 2.0 in eager mode.

I stumbled onto this question while looking for an answer and it took me some time to figure out as I use the model subclassing API in TF 2.0 by default (as in here https://www.tensorflow.org/tutorials/quickstart/advanced).
If somebody is in a similar situation, all you need to do is assign the intermediate output you want, as an attribute of the class. Then keep the test_step without the #tf.function decorator and create its decorated copy, say val_step, for efficient internal computation of validation performance during training. As a short example, I have modified a few functions of the tutorial from the link accordingly. I'm assuming we need to access the output after flattening.
def call(self, x):
x = self.conv1(x)
x = self.flatten(x)
self.intermediate=x #assign it as an object attribute for accessing later
x = self.d1(x)
return self.d2(x)
#Remove #tf.function decorator from test_step for prediction
def test_step(images, labels):
predictions = model(images, training=False)
t_loss = loss_object(labels, predictions)
test_loss(t_loss)
test_accuracy(labels, predictions)
return
#Create a decorated val_step for object's internal use during training
#tf.function
def val_step(images, labels):
return test_step(images, labels)
Now when you run model.predict() after training, using the un-decorated test step, you can access the intermediate output using model.intermediate which would be an EagerTensor whose value is obtained simply by model.intermediate.numpy(). However, if you don't remove the #tf_function decorator from test_step, this would return a Tensor whose value is not so straightforward to obtain.

Thanks for answering my earlier question. I wrote this simple example to illustrate how what you're trying to do might be done in TensorFlow 2.x, using the MNIST dataset as the example problem.
The gist of the approach:
Build an auxiliary model (aux_model in the example below), which is so-called "functional model" with multiple outputs. The first output is the output of the original model and will be used for loss calculation and backprop, while the remaining output(s) are the intermediate-layer outputs that you want to access.
Use tf.GradientTape() to write a custom training loop and expose the detailed gradient values on each individual variable of the model. Then you can pick out the gradients that are of interest to you. This requires that you know the ordering of the model's variables. But that should be relatively easy for a sequential model.
import tensorflow as tf
(x_train, y_train), (_, _) = tf.keras.datasets.mnist.load_data()
# This is the original model.
model = tf.keras.Sequential([
tf.keras.layers.Flatten(input_shape=[28, 28, 1]),
tf.keras.layers.Dense(100, activation="relu"),
tf.keras.layers.Dense(10, activation="softmax")])
# Make an auxiliary model that exposes the output from the intermediate layer
# of interest, which is the first Dense layer in this case.
aux_model = tf.keras.Model(inputs=model.inputs,
outputs=model.outputs + [model.layers[1].output])
# Define a custom training loop using `tf.GradientTape()`, to make it easier
# to access gradients on specific variables (the kernel and bias of the first
# Dense layer in this case).
cce = tf.keras.losses.CategoricalCrossentropy()
optimizer = tf.optimizers.Adam()
with tf.GradientTape() as tape:
# Do a forward pass on the model, retrieving the intermediate layer's output.
y_pred, intermediate_output = aux_model(x_train)
print(intermediate_output) # Now you can access the intermediate layer's output.
# Compute loss, to enable backprop.
loss = cce(tf.one_hot(y_train, 10), y_pred)
# Do backprop. `gradients` here are for all variables of the model.
# But we know we want the gradients on the kernel and bias of the first
# Dense layer, which happens to be the first two variables of the model.
gradients = tape.gradient(loss, aux_model.variables)
# This is the gradient on the first Dense layer's kernel.
intermediate_layer_kerenl_gradients = gradients[0]
print(intermediate_layer_kerenl_gradients)
# This is the gradient on the first Dense layer's bias.
intermediate_layer_bias_gradients = gradients[1]
print(intermediate_layer_bias_gradients)
# Update the variables of the model.
optimizer.apply_gradients(zip(gradients, aux_model.variables))

The most straightforward solution would go like this:
mid_layer = model.get_layer("layer_name")
you can now treat the "mid_layer" as a model, and for instance:
mid_layer.predict(X)
Oh, also, to get the name of a hidden layer, you can use this:
model.summary()
this will give you some insights about the layer input/output as well.

TensorFlow on multiple GPU

Recently, I try to learn how to use Tensorflow on multiple GPU by reading the official tutorial. However, there is something that I am confused about. The following code is part of the official tutorial, which calculates the loss on single GPU.
def tower_loss(scope, images, labels):
# Build inference Graph.
logits = cifar10.inference(images)
# Build the portion of the Graph calculating the losses. Note that we will
# assemble the total_loss using a custom function below.
_ = cifar10.loss(logits, labels)
# Assemble all of the losses for the current tower only.
losses = tf.get_collection('losses', scope)
# Calculate the total loss for the current tower.
total_loss = tf.add_n(losses, name='total_loss')
# Attach a scalar summary to all individual losses and the total loss; do the
# same for the averaged version of the losses.
for l in losses + [total_loss]:
# Remove 'tower_[0-9]/' from the name in case this is a multi-GPU training
# session. This helps the clarity of presentation on tensorboard.
loss_name = re.sub('%s_[0-9]*/' % cifar10.TOWER_NAME, '', l.op.name)
tf.summary.scalar(loss_name, l)
return total_loss
The training process is as the following.
def train():
with tf.device('/cpu:0'):
# Create a variable to count the number of train() calls. This equals the
# number of batches processed * FLAGS.num_gpus.
global_step = tf.get_variable(
'global_step', [],
initializer=tf.constant_initializer(0), trainable=False)
# Calculate the learning rate schedule.
num_batches_per_epoch = (cifar10.NUM_EXAMPLES_PER_EPOCH_FOR_TRAIN /
FLAGS.batch_size / FLAGS.num_gpus)
decay_steps = int(num_batches_per_epoch * cifar10.NUM_EPOCHS_PER_DECAY)
# Decay the learning rate exponentially based on the number of steps.
lr = tf.train.exponential_decay(cifar10.INITIAL_LEARNING_RATE,
global_step,
decay_steps,
cifar10.LEARNING_RATE_DECAY_FACTOR,
staircase=True)
# Create an optimizer that performs gradient descent.
opt = tf.train.GradientDescentOptimizer(lr)
# Get images and labels for CIFAR-10.
images, labels = cifar10.distorted_inputs()
batch_queue = tf.contrib.slim.prefetch_queue.prefetch_queue(
[images, labels], capacity=2 * FLAGS.num_gpus)
# Calculate the gradients for each model tower.
tower_grads = []
with tf.variable_scope(tf.get_variable_scope()):
for i in xrange(FLAGS.num_gpus):
with tf.device('/gpu:%d' % i):
with tf.name_scope('%s_%d' % (cifar10.TOWER_NAME, i)) as scope:
# Dequeues one batch for the GPU
image_batch, label_batch = batch_queue.dequeue()
# Calculate the loss for one tower of the CIFAR model. This function
# constructs the entire CIFAR model but shares the variables across
# all towers.
loss = tower_loss(scope, image_batch, label_batch)
# Reuse variables for the next tower.
tf.get_variable_scope().reuse_variables()
# Retain the summaries from the final tower.
summaries = tf.get_collection(tf.GraphKeys.SUMMARIES, scope)
However, I am confused about the for loop about 'for i in xrange(FLAGS.num_gpus)'. It seems that I have to get a new batch image from batch_queue and calculate every gradient. I think this process is serialized instead of parallel. If there anything wrong with my own understanding? By the way, I can also use the iterator to feed image to my model rather than the dequeue right?
Thank you everybody!

This is a common misconception with Tensorflow's coding model.
What you are showing here is the computation graph's construction, NOT the actual execution.
The block:
for i in xrange(FLAGS.num_gpus):
with tf.device('/gpu:%d' % i):
with tf.name_scope('%s_%d' % (cifar10.TOWER_NAME, i)) as scope:
# Dequeues one batch for the GPU
image_batch, label_batch = batch_queue.dequeue()
# Calculate the loss for one tower of the CIFAR model. This function
# constructs the entire CIFAR model but shares the variables across
# all towers.
loss = tower_loss(scope, image_batch, label_batch)
translates to:
For each GPU device (`for i in range..` & `with device...`):
- build operations needed to dequeue a batch
- build operations needed to run the batch through the network and compute the loss
Note how via tf.get_variable_scope().reuse_variables() you're telling the graph that the variables used for the graph GPU must be shared among all (i.e., all graphs on the multiple devices "reuse" the same variables).
None of this actually runs the network once (note how there is no sess.run()): you're just giving instructions on how data must flow.
Then, when you'll start the actual training (I guess you missed that piece of the code when copying it here) each GPU will pull its own batch and produce the per-tower loss. I guess these losses are averaged somewhere in the subsequent code and the average is the loss passed to the optimizer.
Up until the point where the tower losses are averaged together, everything is independent from the other devices, so getting the batch and computing the loss can be done in parallel. Then the gradients and parameter update is done only once, variables are updated and the cycle repeats.
So, to answer your question, no, per-batch loss computation is not serialized, but since this is synchronous distributed computation you need to collect all losses from all GPUs before being allowed to continue with gradients computation and parameters update, so you still have some part of the graph that cannot be independent.

How to gradually train with more and more classes?

I'm trying to create an incremental classifier that will get trained on data containing n classes for some set number of epochs, then n+m classes for a set number of epochs, then n+m+k, etc, where each successive set of classes contains the previous set as a subset.
In order to do this without having to train the model, save it, manually edit the graph, re-train, repeat, I'm simply defining all the weights I will need to classify the entire set of classes, but keeping the weights corresponding to unseen classes frozen at 0 until the classifier is introduced to those classes.
My strategy for this is to define a placeholder that is fed in an array of Boolean values defining whether or not some given set of weights are trainable.
Relevant code below:
output_train = tf.placeholder(tf.int32, shape = (num_incremental_grps), name = "output_train")
.
.
.
weights = []
biases = []
for i in range(num_incremental_grps):
W = tf.Variable(tf.zeros([batch_size, classes_per_grp]),
trainable=tf.cond(tf.equal(output_train[i], tf.constant(1)),lambda: tf.constant(True), lambda: tf.constant(False)))
weights.append(W)
b = tf.Variable(tf.zeros([classes_per_grp]), trainable=tf.cond(tf.equal(output_train[i],
tf.constant(1)), lambda:tf.constant(True), lambda: tf.constant(False)))
biases.append(b)
out_weights = tf.stack(weights, axis=1).reshape((batch_size, -1))
out_biases = tf.stack(biases, axis=1).reshape((batch_size, -1))
outputs = tf.identity(tf.matmul(inputs, out_weights) + out_biases, name='values')
.
.
.
# Will change this to an array that progressively updates as classes are added.
output_trainable = np.ones(num_incremental_grps, dtype=bool)
.
.
.
with tf.Session() as sess:
init.run()
for epoch in range(epochs):
for iteration in range(iterations):
X_batch, y_batch = batch.getBatch()
fd={X: X_batch, y: y_batch, training: True, output_train: output_trainable}
_, loss_val = sess.run([training_op, loss], feed_dict=fd)
This returns the error message
Using a 'tf.Tensor' as a Python `bool` is not allowed. Use `if t is not None:` instead of
`if t:` to test if a tensor is defined,and use TensorFlow ops such as tf.cond to execute
subgraphs conditioned on the value of a tensor.
I've tried tinkering around with this, like making the initial placeholder datatype tf.bool instead of tf.int32. I've also tried just feeding in a slice of the tensor into the 'trainable' argument in the weights/biases like this
W = tf.Variable(tf.zeros([batch_size, classes_per_grp]), trainable=output_variable[i])
but I get the same error message. I'm not sure how to proceed from here, aside from trying a completely different approach to updating the number of predictable classes. Any help would be much appreciated.

The error occurs because tf.cond takes a decision based on a single boolean — much like an if statement. What you want here is to make a choice per element of your tensor.
You could use tf.where to fix that problem, but then you will run into another one, which is that trainable is not a property that you can fix at runtime, it is part of the definition of a variable. If a variable will be trained at some point, perhaps not at the beginning but definitely later, then it must be trainable.
I would suggest to take a much simpler route: define output_train to be an array of tf.float32
output_train = tf.placeholder(tf.float32, shape=(num_incremental_grps), name="output_train")
then later simply multiply your weights and variables with this vector.
W = tf.Variable(...)
W = W * output_train
...
Provide values of 1 to output_train where you want training to happen, 0 otherwise.
Be careful to also mask your loss to ignore output from unwanted channels, because event though they now always output 0, that may still affect your loss. For example,
logits = ...
logits = tf.matrix_transpose(tf.boolean_mask(
tf.matrix_transpose(logits ),
output_train == 1))
loss = tf.nn.softmax_cross_entropy_with_logits_v2(logits=logits, labels=labels)

How to reuse RNN in TensorFlow

I want to implement a model like DSSM (Deep Semantic Similarity Model).
I want to train one RNN model and use this model to get three hidden vector for three different inputs, and use these hidden vector to compute loss function.
I try to code in a variable scope with reuse=None like:
gru_cell = tf.nn.rnn_cell.GRUCell(size)
gru_cell = tf.nn.rnn_cell.DropoutWrapper(gru_cell,output_keep_prob=0.5)
cell = tf.nn.rnn_cell.MultiRNNCell([gru_cell] * 2, state_is_tuple=True)
embedding = tf.get_variable("embedding", [vocab_size, wordvec_size])
inputs = tf.nn.embedding_lookup(embedding, self._input_data)
inputs = tf.nn.dropout(inputs, 0.5)
with tf.variable_scope("rnn"):
_, self._states_2 = rnn_states_2[config.num_layers-1] = tf.nn.dynamic_rnn(cell, inputs, sequence_length=self.lengths, dtype=tf.float32)
self._states_1 = rnn_states_1[config.num_layers-1]
with tf.variable_scope("rnn", reuse=True):
_, rnn_states_2 = tf.nn.dynamic_rnn(cell,inputs,sequence_length=self.lengths,dtype=tf.float32)
self._states_2 = rnn_states_2[config.num_layers-1]
I use the same inputs and reuse the RNN model, but when I print 'self_states_1' and 'self_states_2', these two vectors are different.
I use with tf.variable_scope("rnn", reuse=True): to compute 'rnn_states_2' because I want to use the same RNN model like 'rnn_states_1'.
But why I get different hidden vectors with the same inputs and the same model?
Where did i go wrong?
Thanks for your answering.
Update:
I find the reason may be the 'tf.nn.rnn_cell.DropoutWrapper' , when I remove the drop out wrapper, the hidden vectors are same, when I add the drop out wrapper, these vector become different.
So, the new question is :
How to fix the part of vector which be 'dropped out' ? By setting the 'seed' parameter ?
When training a DSSM, should I fix the drop out action ?

If you structure your code to use tf.contrib.rnn.DropoutWrapper, you can set variational_recurrent=True in your wrapper, which causes the same dropout mask to be used at all steps, i.e. the dropout mask will be constant. Is that what you want?
Setting the seed parameter in tf.nn.dropout will just make sure that you get the same sequence of dropout masks every time you run with that seed. That does not mean the dropout mask will be constant, just that you'll always see the same dropout mask at a particular iteration. The mask will be different for every iteration.