I want to use pretrained weights for 2 parts of my model. I have 2 checkpoints from different models, from which I can load only one into my main model with tf.estimator.WarmStart as I'm using the estimator architecture.
tf.WarmStartSettings(ckpt_to_initialize_from=X)
from the doc:
Either the directory or a specific checkpoint can be provided (in the case of the former, the latest checkpoint will be used).
I can't see how I can add an additional checkpoint. Maybe there is a way to load the weights from both checkpoint into one and load that one?
You can use init_from_checkpoint.
First, define assignment map:
dir = 'path_to_checkpoint_files'
vars_to_load = [i[0] for i in tf.train.list_variables(dir)]
This creates a list of all variables in checkpoints
assignment_map = {variable.op.name: variable for variable in tf.global_variables() if variable.op.name in vars_to_load}
And this creates a dict that has variables from current graph as key and variables from checkpoint as values
tf.train.init_from_checkpoint(dir, assignment_map)
This function is placed inside estimator's model_fn. It will override standard variable initialization.
Related
I have this config:
network = {"source_embed_raw": {"class": "linear", ...}}
I want to load the params for layer source_embed_raw from some existing checkpoint.
In that checkpoint, param is called differently (output/rec/target_embed_raw/W).
I understand, that I can load parameters with preload_from_files, but I am not sure about the exact way to do that in my case, because the names of the layers differ, thus simply adding a prefix does not do the job.
This is currently not possible with preload_from_files in this way.
So I currently see these possible options:
We could extend the logic of preload_from_files (and CustomCheckpointLoader) to allow for sth like that (some generic variable/layer name mapping).
Or you could rename your layer from source_embed_raw to e.g. old_model__target_embed_raw and then use preload_from_files with the prefix option. If you do not want to rename it, you could still add a layer like old_model__target_embed_raw and then use parameter sharing in source_embed_raw.
If the parameter in the checkpoint is actually called sth like output/rec/target_embed_raw/..., you could create a SubnetworkLayer named old_model__output, in that another SubnetworkLayer with name rec, and in that a layer named target_embed_raw.
You could write a script to simply load the existing checkpoint, and store is as a new checkpoint but with renamed variable names (this is also totally independent from RETURNN).
LinearLayer (and most other layers) allows to specify exactly how the parameters are initialized (forward_weights_init and bias_init). The parameter initialization is quite flexible. E.g. there is sth like load_txt_file_initializer which can be used. Currently there is no such function to directly load it from an existing checkpoint but we could add that. Or you could simply implement the logic inside your config (it will only be sth like 5 lines of code or so).
Instead of using preload_from_files, you could also use SubnetworkLayer and the load_on_init option. And then a similar logic as in option 2.
I am sorry if this question seems pretty straight forward. But reading the Keras save and restore help page :
https://www.tensorflow.org/beta/tutorials/keras/save_and_restore_models
I do not understand how to use the "ModelCheckpoint" for saving during training. The help file mentions it should give 3 files, I see only one, MODEL.ckpt.
Here is my code:
checkpoint_dir = FolderName + "/tmp/model.ckpt"
cp_callback = k.callbacks.ModelCheckpoint(checkpoint_dir,verbose=1,save_weights_only=True)
parallel_model.compile(optimizer=tf.keras.optimizers.Adam(lr=learning_rate),loss=my_cost_MSE, metrics=['accuracy])
parallel _model.fit(image, annotation, epochs=epoch,
batch_size=batch_size, steps_per_epoch=10,
validation_data=(image_val,annotation_val),validation_steps=num_batch_val,callbacks=callbacks_list)
Also, when I want to load the weights after training with:
model = k.models.load_model(file_checkpoint)
I get the error:
"raise ValueError('Unknown ' + printable_module_name + ':' + object_name)
ValueError: Unknown loss function:my_cost_MSE"
my-cost_MSE is my cost function that is used in the training.
First of all, it looks like you are using the tf.keras (from tensorflow) implementation rather than keras (from the keras-team/keras repo). In this case, as stated in the tf.keras guide :
When saving a model's weights, tf.keras defaults to the checkpoint
format. Pass save_format='h5' to use HDF5.
On the other hand, note that adding the callback ModelCheckpoint is, usually, roughly equivalent to call model.save(...) at the end of each epoch, so that's why you should expect three files to be saved (according to the checkpoint format).
The reason it's not doing so is because, by using the option save_weights_only=True, you are saving just the weights. Roughly equivalent to replace the call to model.save for model.save_weights at the end of each epoch. Hence, the only file that's being saved is the one with the weights.
From here, you can proceed in two different ways:
Storing just the weights
You need your model (the structure, let's say) to be loaded beforehand and then call model.load_weights instead of keras.models.load_model:
model = MyModel(...) # Your model definition as used in training
model.load_weights(file_checkpoint)
Note that in this case, you won't have problems with custom definitions (my_cost_MSE) since you are just loading model weights.
Storing the whole model
Another way to proceed is to store the whole model and load it accordingly:
cp_callback = k.callbacks.ModelCheckpoint(
checkpoint_dir,verbose=1,
save_weights_only=False
)
parallel_model.compile(
optimizer=tf.keras.optimizers.Adam(lr=learning_rate),
loss=my_cost_MSE,
metrics=['accuracy']
)
model.fit(..., callbacks=[cp_callback])
Then you could load it by:
model = k.models.load_model(file_checkpoint, custom_objects={"my_cost_MSE": my_cost_MSE})
Note that in this latter case, you need to specify custom_objects since its definition is needed to deserialize the model.
keras has a save command. It saves all the details needed to rebuild the model.
(from the keras docs)
from keras.models import load_model
model.save('my_model.h5') # creates a HDF5 file 'my_model.h5'
del model # deletes the existing model
# returns am identical compiled model
model = load_model('my_model.h5')
Let's say I have a model with Y layers.
I am trying to restore the model with setting Y-1 layers to trainable=False, so I insert all Y-1 layers(variable names) into var_list when defining tf.train.Saver(var_list=list_of_Y-1_layers) so they can be restored.
I would like to not restore the last layer, which I would like to train myself, so if I put it var_list it gets restored and if I don't put it there, it doesn't save inside the checkpoint during training.
Does this variable gets saved elsewhere ? Or am I doing something wrong for saving/restoring?
Side note:
To check if a trainable variable is saved or not, I use the function inspect_checkpoint(), which is defined in tensorflow/tensorflow/python/tools/inspect_checkpoint.py
You can create two objects, one for saving, the other for restoring:
#used to restore:
saver_restore = tf.train.Saver(var_list=list_of_Y-1_layers)
#used to save, will save all variables
saver_save = tf.train.Saver()
You can save your entire model without specifying a var_list. This will save all of the variables in a checkpoint. Then when you restore, you can specify var_list to the restore Saver to only restore your desired subset of layers.
Sources:
https://www.tensorflow.org/programmers_guide/saved_model#choosing_which_variables_to_save_and_restore
I am training a Generative Adversarial Network (GAN) in tensorflow, where basically we have two different networks each one with its own optimizer.
self.G, self.layer = self.generator(self.inputCT,batch_size_tf)
self.D, self.D_logits = self.discriminator(self.GT_1hot)
...
self.g_optim = tf.train.MomentumOptimizer(self.learning_rate_tensor, 0.9).minimize(self.g_loss, global_step=self.global_step)
self.d_optim = tf.train.AdamOptimizer(self.learning_rate, beta1=0.5) \
.minimize(self.d_loss, var_list=self.d_vars)
The problem is that I train one of the networks (g) first, and then, I want to train g and d together. However, when I call the load function:
self.sess.run(tf.initialize_all_variables())
self.sess.graph.finalize()
self.load(self.checkpoint_dir)
def load(self, checkpoint_dir):
print(" [*] Reading checkpoints...")
ckpt = tf.train.get_checkpoint_state(checkpoint_dir)
if ckpt and ckpt.model_checkpoint_path:
ckpt_name = os.path.basename(ckpt.model_checkpoint_path)
self.saver.restore(self.sess, ckpt.model_checkpoint_path)
return True
else:
return False
I have an error like this (with a lot more traceback):
Tensor name "beta2_power" not found in checkpoint files checkpoint/MR2CT.model-96000
I can restore the g network and keep training with that function, but when I want to star d from scratch, and g from the the stored model I have that error.
To restore a subset of variables, you must create a new tf.train.Saver and pass it a specific list of variables to restore in the optional var_list argument.
By default, a tf.train.Saver will create ops that (i) save every variable in your graph when you call saver.save() and (ii) lookup (by name) every variable in the given checkpoint when you call saver.restore(). While this works for most common scenarios, you have to provide more information to work with specific subsets of the variables:
If you only want to restore a subset of the variables, you can get a list of these variables by calling tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES, scope=G_NETWORK_PREFIX), assuming that you put the "g" network in a common with tf.name_scope(G_NETWORK_PREFIX): or tf.variable_scope(G_NETWORK_PREFIX): block. You can then pass this list to the tf.train.Saver constructor.
If you want to restore a subset of the variable and/or they variables in the checkpoint have different names, you can pass a dictionary as the var_list argument. By default, each variable in a checkpoint is associated with a key, which is the value of its tf.Variable.name property. If the name is different in the target graph (e.g. because you added a scope prefix), you can specify a dictionary that maps string keys (in the checkpoint file) to tf.Variable objects (in the target graph).
I had a similar problem when restoring only part of my variables from a checkpoint and some of the saved variables did not exist in the new model.
Inspired by #Lidong answer I modified a little the reading function:
def get_tensors_in_checkpoint_file(file_name,all_tensors=True,tensor_name=None):
varlist=[]
var_value =[]
reader = pywrap_tensorflow.NewCheckpointReader(file_name)
if all_tensors:
var_to_shape_map = reader.get_variable_to_shape_map()
for key in sorted(var_to_shape_map):
varlist.append(key)
var_value.append(reader.get_tensor(key))
else:
varlist.append(tensor_name)
var_value.append(reader.get_tensor(tensor_name))
return (varlist, var_value)
and added a loading function:
def build_tensors_in_checkpoint_file(loaded_tensors):
full_var_list = list()
# Loop all loaded tensors
for i, tensor_name in enumerate(loaded_tensors[0]):
# Extract tensor
try:
tensor_aux = tf.get_default_graph().get_tensor_by_name(tensor_name+":0")
except:
print('Not found: '+tensor_name)
full_var_list.append(tensor_aux)
return full_var_list
Then you can simply load all common variables using:
CHECKPOINT_NAME = path to save file
restored_vars = get_tensors_in_checkpoint_file(file_name=CHECKPOINT_NAME)
tensors_to_load = build_tensors_in_checkpoint_file(restored_vars)
loader = tf.train.Saver(tensors_to_load)
loader.restore(sess, CHECKPOINT_NAME)
Edit: I am using tensorflow 1.2
Inspired by #mrry, I propose a solution for this problem.
To make it clear, I formulate the problem as restoring a subset of the variable from the checkpoint, when the model is built on a pre-trained model.
First, we should use print_tensors_in_checkpoint_file function from the library inspect_checkpoint or just simply extract this function by:
from tensorflow.python import pywrap_tensorflow
def print_tensors_in_checkpoint_file(file_name, tensor_name, all_tensors):
varlist=[]
reader = pywrap_tensorflow.NewCheckpointReader(file_name)
if all_tensors:
var_to_shape_map = reader.get_variable_to_shape_map()
for key in sorted(var_to_shape_map):
varlist.append(key)
return varlist
varlist=print_tensors_in_checkpoint_file(file_name=the path of the ckpt file,all_tensors=True,tensor_name=None)
Then we use tf.get_collection() just like #mrry saied:
variables = tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES)
Finally, we can initialize the saver by:
saver = tf.train.Saver(variable[:len(varlist)])
The complete version can be found at my github: https://github.com/pobingwanghai/tensorflow_trick/blob/master/restore_from_checkpoint.py
In my situation, the new variables are added at the end of the model, so I can simply use [:length()] to identify the needed variables, for a more complex situation, you might have to do some hand-alignment work or write a simple string matching function to determine the required variables.
You can create a separate instance of tf.train.Saver() with the var_list argument set to the variables you want to restore.
And create a separate instance to save the variables
I'm trying to train an LSTM in Tensorflow using minibatches, but after training is complete I would like to use the model by submitting one example at a time to it. I can set up the graph within Tensorflow to train my LSTM network, but I can't use the trained result afterward in the way I want.
The setup code looks something like this:
#Build the LSTM model.
cellRaw = rnn_cell.BasicLSTMCell(LAYER_SIZE)
cellRaw = rnn_cell.MultiRNNCell([cellRaw] * NUM_LAYERS)
cell = rnn_cell.DropoutWrapper(cellRaw, output_keep_prob = 0.25)
input_data = tf.placeholder(dtype=tf.float32, shape=[SEQ_LENGTH, None, 3])
target_data = tf.placeholder(dtype=tf.float32, shape=[SEQ_LENGTH, None])
initial_state = cell.zero_state(batch_size=BATCH_SIZE, dtype=tf.float32)
with tf.variable_scope('rnnlm'):
output_w = tf.get_variable("output_w", [LAYER_SIZE, 6])
output_b = tf.get_variable("output_b", [6])
outputs, final_state = seq2seq.rnn_decoder(input_list, initial_state, cell, loop_function=None, scope='rnnlm')
output = tf.reshape(tf.concat(1, outputs), [-1, LAYER_SIZE])
output = tf.nn.xw_plus_b(output, output_w, output_b)
...Note the two placeholders, input_data and target_data. I haven't bothered including the optimizer setup. After training is complete and the training session closed, I would like to set up a new session that uses the trained LSTM network whose input is provided by a completely different placeholder, something like:
with tf.Session() as sess:
with tf.variable_scope("simulation", reuse=None):
cellSim = cellRaw
input_data_sim = tf.placeholder(dtype=tf.float32, shape=[1, 1, 3])
initial_state_sim = cell.zero_state(batch_size=1, dtype=tf.float32)
input_list_sim = tf.unpack(input_data_sim)
outputsSim, final_state_sim = seq2seq.rnn_decoder(input_list_sim, initial_state_sim, cellSim, loop_function=None, scope='rnnlm')
outputSim = tf.reshape(tf.concat(1, outputsSim), [-1, LAYER_SIZE])
with tf.variable_scope('rnnlm'):
output_w = tf.get_variable("output_w", [LAYER_SIZE, nOut])
output_b = tf.get_variable("output_b", [nOut])
outputSim = tf.nn.xw_plus_b(outputSim, output_w, output_b)
This second part returns the following error:
tensorflow.python.framework.errors.InvalidArgumentError: You must feed a value for placeholder tensor 'Placeholder' with dtype float
[[Node: Placeholder = Placeholder[dtype=DT_FLOAT, shape=[], _device="/job:localhost/replica:0/task:0/cpu:0"]()]]
...Presumably because the graph I'm using still has the old training placeholders attached to the trained LSTM nodes. What's the right way to 'extract' the trained LSTM and put it into a new, different graph that has a different style of inputs? The Varible scoping features that Tensorflow has seem to address something like this, but the examples in the documentation all talk about using variable scope as a way of managing variable names so that the same piece of code will generate similar subgraphs within the same graph. The 'reuse' feature seems to be close to what I want, but I don't find the Tensorflow documentation linked above to be clear at all on what it does. The cells themselves cannot be given a name (in other words,
cellRaw = rnn_cell.MultiRNNCell([cellRaw] * NUM_LAYERS, name="multicell")
is not valid), and while I can give a name to a seq2seq.rnn_decoder(), I presumably wouldn't be able to remove the rnn_cell.DropoutWrapper() if I used that node unchanged.
Questions:
What is the proper way to move trained LSTM weights from one graph to another?
Is it correct to say that starting a new session "releases resources", but doesn't erase the graph built in memory?
It seems to me like the 'reuse' feature allows Tensorflow to search outside of the current variable scope for variables with the same name (existing in a different scope), and use them in the current scope. Is this correct? If it is, what happens to all of the graph edges from the non-current scope that link to that variable? If it isn't, why does Tensorflow throw an error if you try to have the same variable name within two different scopes? It seems perfectly reasonable to define two variables with identical names in two different scopes, e.g. conv1/sum1 and conv2/sum1.
In my code I'm working within a new scope but the graph won't run without data to be fed into a placeholder from the initial, default scope. Is the default scope always 'in-scope' for some reason?
If graph edges can span different scopes, and names in different scopes can't be shared unless they refer to the exact same node, then that would seem to defeat the purpose of having different scopes in the first place. What am I misunderstanding here?
Thanks!
What is the proper way to move trained LSTM weights from one graph to another?
You can create your decoding graph first (with a saver object to save the parameters) and create a GraphDef object that you can import in your bigger training graph:
basegraph = tf.Graph()
with basegraph.as_default():
***your graph***
traingraph = tf.Graph()
with traingraph.as_default():
tf.import_graph_def(basegraph.as_graph_def())
***your training graph***
make sure you load your variables when you start a session for a new graph.
I don't have experience with this functionality so you may have to look into it a bit more
Is it correct to say that starting a new session "releases resources", but doesn't erase the graph built in memory?
yep, the graph object still hold it
It seems to me like the 'reuse' feature allows Tensorflow to search outside of the current variable scope for variables with the same name (existing in a different scope), and use them in the current scope. Is this correct? If it is, what happens to all of the graph edges from the non-current scope that link to that variable? If it isn't, why does Tensorflow throw an error if you try to have the same variable name within two different scopes? It seems perfectly reasonable to define two variables with identical names in two different scopes, e.g. conv1/sum1 and conv2/sum1.
No, reuse is to determine the behaviour when you use get_variable on an existing name, when it is true it will return the existing variable, otherwise it will return a new one. Normally tensorflow should not throw an error. Are you sure your using tf.get_variable and not just tf.Variable?
In my code I'm working within a new scope but the graph won't run without data to be fed into a placeholder from the initial, default scope. Is the default scope always 'in-scope' for some reason?
I don't really see what you mean. The do not always have to be used. If a placeholder is not required for running an operation you don't have to define it.
If graph edges can span different scopes, and names in different scopes can't be shared unless they refer to the exact same node, then that would seem to defeat the purpose of having different scopes in the first place. What am I misunderstanding here?
I think your understanding or usage of scopes is flawed, see above