I read the data and processed it using the following code :
data = pd.read_csv('Step1_output.csv')
data = data.sample(frac=1).reset_index(drop=True)
data1 = pd.DataFrame(data, columns=['Res_pair'])
# creating instance of labelencoder
labelencoder = LabelEncoder()
# Assigning numerical values and storing in another column
data1['Res_pair_ID'] = labelencoder.fit_transform(data1['Res_pair'])
data['Res_pair'] = data1['Res_pair_ID']
data = data.to_numpy()
train_X = data[0:data.shape[0],0:566]
train_y = data[0:data.shape[0],566:data.shape[1]]
train_X = train_X.reshape((train_X.shape[0], train_X.shape[1], 1))
I build the model using following code where I have tried to distribute the dataset using mirrored strategy of Tensorflow :
print("Hyper-parameter values:\n")
print('Momentum Rate =',momentum_rate,'\n')
print('learning rate =',learning_rate,'\n')
print('Number of neurons =',neurons,'\n')
strategy = tensorflow.distribute.MirroredStrategy()
with strategy.scope():
model = tf.keras.Sequential([
tf.keras.layers.Conv1D(64,kernel_size = 3,activation='relu',input_shape=train_X.shape[1:]),
tf.keras.layers.Flatten(),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(10, activation='softmax'),])
sgd = optimizers.SGD(lr=learning_rate, decay=1e-6, momentum=momentum_rate, nesterov=True)
model.compile(loss='categorical_crossentropy',optimizer=sgd,metrics=['accuracy',tensorflow.keras.metrics.Precision()])
results = model.fit(train_X,train_y,validation_split = 0.2,epochs=10,batch_size = 100)
print(results)
path = 'saved_model/'
model.save(path, save_format='tf')
for k in range(100):
momentum_rate = random.random()
learning_rate = random.uniform(0,0.2)
neurons = random.randint(10,50)
I tried to run the code on GPU but it runs for some time and then throws this error :
Hyper-parameter values:
Momentum Rate = 0.6477407029392913
learning rate = 0.03988890117492503
Number of neurons = 35
Epoch 1/10
1/270110 [..............................] - ETA: 28s - loss: nan - accuracy: 0.0100 - precision: 0.0100Traceback (most recent call last):
File "parallelised_script_realdata2.py", line 56, in <module>
results = model.fit(train_X,train_y,validation_split = 0.2,epochs=10,batch_size = 100)
File "/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/engine/training.py", line 108, in _method_wrapper
return method(self, *args, **kwargs)
File "/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/engine/training.py", line 1098, in fit
tmp_logs = train_function(iterator)
File "/usr/local/lib64/python3.6/site-packages/tensorflow/python/eager/def_function.py", line 780, in __call__
result = self._call(*args, **kwds)
File "/usr/local/lib64/python3.6/site-packages/tensorflow/python/eager/def_function.py", line 807, in _call
return self._stateless_fn(*args, **kwds) # pylint: disable=not-callable
File "/usr/local/lib64/python3.6/site-packages/tensorflow/python/eager/function.py", line 2829, in __call__
return graph_function._filtered_call(args, kwargs) # pylint: disable=protected-access
File "/usr/local/lib64/python3.6/site-packages/tensorflow/python/eager/function.py", line 1848, in _filtered_call
cancellation_manager=cancellation_manager)
File "/usr/local/lib64/python3.6/site-packages/tensorflow/python/eager/function.py", line 1924, in _call_flat
ctx, args, cancellation_manager=cancellation_manager))
File "/usr/local/lib64/python3.6/site-packages/tensorflow/python/eager/function.py", line 550, in call
ctx=ctx)
File "/usr/local/lib64/python3.6/site-packages/tensorflow/python/eager/execute.py", line 60, in quick_execute
inputs, attrs, num_outputs)
tensorflow.python.framework.errors_impl.InvalidArgumentError: assertion failed: [predictions must be >= 0] [Condition x >= y did not hold element-wise:] [x (sequential/dense_4/Softmax:0) = ] [[nan nan nan...]...] [y (Cast_6/x:0) = ] [0]
[[{{node assert_greater_equal/Assert/AssertGuard/else/_21/assert_greater_equal/Assert/AssertGuard/Assert}}]] [Op:__inference_train_function_1270]
Function call stack:
train_function
Update: The code works well if I don't use strategy = tensorflow.distribute.MirroredStrategy(). Like the code below (but will fail for larger datasets for memory shortage):
def convolutional_neural_network(x, y):
print("Hyper-parameter values:\n")
print('Momentum Rate =',momentum_rate,'\n')
print('learning rate =',learning_rate,'\n')
print('Number of neurons =',neurons,'\n')
model = Sequential()
model.add(Conv1D(filters=64,input_shape=train_X.shape[1:],activation='relu',kernel_size = 3))
model.add(Flatten())
model.add(Dense(neurons,activation='relu')) # first hidden layer
model.add(Dense(neurons, activation='relu')) # second hidden layer
model.add(Dense(neurons, activation='relu'))
model.add(Dense(neurons, activation='relu'))
model.add(Dense(10, activation='softmax'))
sgd = optimizers.SGD(lr=learning_rate, decay=1e-6, momentum=momentum_rate, nesterov=True)
model.compile(loss='categorical_crossentropy', optimizer=sgd, metrics=['accuracy',tensorflow.keras.metrics.Precision()])
history = model.fit(train_X, train_y, validation_split=0.2, epochs=10, batch_size=100)
momentum_rate = 0.09
learning_rate = 0.01
neurons = 40
print(convolutional_neural_network(train_X, train_y))
Update 2: Still facing a similar issue with smaller dataset
_________________________________________________________________
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv1d (Conv1D) (None, 564, 64) 256
_________________________________________________________________
flatten (Flatten) (None, 36096) 0
_________________________________________________________________
dense (Dense) (None, 50) 1804850
_________________________________________________________________
dense_1 (Dense) (None, 50) 2550
_________________________________________________________________
dense_2 (Dense) (None, 50) 2550
_________________________________________________________________
dense_3 (Dense) (None, 50) 2550
_________________________________________________________________
dense_4 (Dense) (None, 10) 510
=================================================================
Total params: 1,813,266
Trainable params: 1,813,266
Non-trainable params: 0
The model definition seems fine, so does the strategy.
Can you just verify train_Y for sanity check? Mostly I'm sure the error lies there.
If that's not the case, try running model.fit and latter ones outside the scope.
Related
I'm creating a Bidirectional LSTM but I faced following error
ValueError: Dimensions must be equal, but are 5 and 250 for '{{node Equal}} = Equal[T=DT_INT64, incompatible_shape_error=true](ArgMax, ArgMax_1)' with input shapes: [?,5], [?,250]
I have no idea what is wrong and how to fix it!
I have a text dataset with 59k row for train the model and i would divid them into 15 classes which then I would use for text similarity base on classes for the received new text.
Based on the other post I played with loss but still it doesn't solve the issue.
Here is the model plot:
Also sequential model would be as follow:
model_lstm = Sequential()
model_lstm.add(InputLayer(250,))
model_lstm.add(Embedding(input_dim=max_words+1, output_dim=200, weights=[embedding_matrix],
mask_zero=True, trainable= True, name='corpus_embed'))
enc_lstm = Bidirectional(LSTM(128, activation='sigmoid', return_sequences=True, name='LSTM_Encod'))
model_lstm.add(enc_lstm)
model_lstm.add(Dropout(0.25))
model_lstm.add(Bidirectional(LSTM( 128, activation='sigmoid',dropout=0.25, return_sequences=True, name='LSTM_Decod')))
model_lstm.add(Dropout(0.25))
model_lstm.add(Dense(15, activation='softmax'))
model_lstm.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['Accuracy'])
## Feed the model
history = model_lstm.fit(x=corpus_seq_train,
y=target_seq_train,
batch_size=128,
epochs=50,
validation_data=(corpus_seq_test,target_seq_test),
callbacks=[tensorboard],
sample_weight= sample_wt_mat)
This is the model summary:
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
corpus_embed (Embedding) (None, 250, 200) 4000200
bidirectional (Bidirectiona (None, 250, 256) 336896
l)
dropout (Dropout) (None, 250, 256) 0
bidirectional_1 (Bidirectio (None, 250, 256) 394240
nal)
dropout_1 (Dropout) (None, 250, 256) 0
dense (Dense) (None, 250, 15) 3855
=================================================================
Total params: 4,735,191
Trainable params: 4,735,191
Non-trainable params: 0
_________________________________
and dataset shape:
corpus_seq_train.shape, target_seq_train.shape
((59597, 250), (59597, 5, 8205))
Finally, here is the error:
Epoch 1/50
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
C:\Users\AMIRSH~1\AppData\Local\Temp/ipykernel_10004/3838451254.py in <module>
9 ## Feed the model
10
---> 11 history = model_lstm.fit(x=corpus_seq_train,
12 y=target_seq_train,
13 batch_size=128,
C:\ProgramData\Anaconda3\lib\site-packages\keras\utils\traceback_utils.py in error_handler(*args, **kwargs)
65 except Exception as e: # pylint: disable=broad-except
66 filtered_tb = _process_traceback_frames(e.__traceback__)
---> 67 raise e.with_traceback(filtered_tb) from None
68 finally:
69 del filtered_tb
C:\ProgramData\Anaconda3\lib\site-packages\keras\engine\training.py in tf__train_function(iterator)
13 try:
14 do_return = True
---> 15 retval_ = ag__.converted_call(ag__.ld(step_function), (ag__.ld(self), ag__.ld(iterator)), None, fscope)
16 except:
17 do_return = False
ValueError: in user code:
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\engine\training.py", line 1051, in train_function *
return step_function(self, iterator)
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\engine\training.py", line 1040, in step_function **
outputs = model.distribute_strategy.run(run_step, args=(data,))
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\engine\training.py", line 1030, in run_step **
outputs = model.train_step(data)
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\engine\training.py", line 894, in train_step
return self.compute_metrics(x, y, y_pred, sample_weight)
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\engine\training.py", line 987, in compute_metrics
self.compiled_metrics.update_state(y, y_pred, sample_weight)
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\engine\compile_utils.py", line 501, in update_state
metric_obj.update_state(y_t, y_p, sample_weight=mask)
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\utils\metrics_utils.py", line 70, in decorated
update_op = update_state_fn(*args, **kwargs)
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\metrics\base_metric.py", line 140, in update_state_fn
return ag_update_state(*args, **kwargs)
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\metrics\base_metric.py", line 646, in update_state **
matches = ag_fn(y_true, y_pred, **self._fn_kwargs)
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\metrics\metrics.py", line 3295, in categorical_accuracy
return metrics_utils.sparse_categorical_matches(
File "C:\ProgramData\Anaconda3\lib\site-packages\keras\utils\metrics_utils.py", line 893, in sparse_categorical_matches
matches = tf.cast(tf.equal(y_true, y_pred), backend.floatx())
ValueError: Dimensions must be equal, but are 5 and 250 for '{{node Equal}} = Equal[T=DT_INT64, incompatible_shape_error=true](ArgMax, ArgMax_1)' with input shapes: [?,5], [?,250].
the problem is because of the Loss function and y-label shape.
we should not pad y_label and it should fit the model directly without any other process
Code:
Y = Y.to_numpy()
X = X.to_numpy()
X.reshape((1, 478405, 33))
opt = tf.keras.optimizers.Adam(lr=0.001, decay=1e-6)
model = Sequential()
model.add(LSTM(33, return_sequences=True, input_shape=(X.shape[1], X.shape[0]), activation='sigmoid'))
model.add(Dropout(0.2))
model.add(LSTM(33, return_sequences=True))
model.add(Dropout(0.2))
model.add(Dense(1, activation = "sigmoid"))
model.compile(optimizer='adam', loss='mse', metrics=['accuracy'])
filepath = "RNN_Final-{epoch:02d}-{val_acc:.3f}" # unique file name that will include the epoch and the validation acc for that epoch
checkpoint = ModelCheckpoint("models/{}.model".format(filepath, monitor='val_acc', verbose=1, save_best_only=True, mode='max')) # saves only the best ones
history = model.fit(X, Y, epochs=35, batch_size=1, shuffle=False)
scores = model.evaluate(X, Y)
Error:
WARNING:tensorflow:Model was constructed with shape (None, 33, 478405) for input KerasTensor(type_spec=TensorSpec(shape=(None, 33, 478405), dtype=tf.float32, name='lstm_input'), name='lstm_input', description="created by layer 'lstm_input'"), but it was called on an input with incompatible shape (1, 33).
Traceback (most recent call last):
File "C:\Users\W10\PycharmProjects\TheCryptoBot\cryptobot\app\ai-model -2.py", line 84, in <module>
history = model.fit(X, Y, epochs=35, batch_size=1, shuffle=False)
File "C:\Users\W10\PycharmProjects\TheCryptoBot\venv\lib\site-packages\keras\utils\traceback_utils.py", line 67, in error_handler
raise e.with_traceback(filtered_tb) from None
File "C:\Users\W10\PycharmProjects\TheCryptoBot\venv\lib\site-packages\tensorflow\python\framework\func_graph.py", line 1129, in autograph_handler
raise e.ag_error_metadata.to_exception(e)
ValueError: in user code:
File "C:\Users\W10\PycharmProjects\TheCryptoBot\venv\lib\site-packages\keras\engine\training.py", line 878, in train_function *
return step_function(self, iterator)
File "C:\Users\W10\PycharmProjects\TheCryptoBot\venv\lib\site-packages\keras\engine\training.py", line 867, in step_function **
outputs = model.distribute_strategy.run(run_step, args=(data,))
File "C:\Users\W10\PycharmProjects\TheCryptoBot\venv\lib\site-packages\keras\engine\training.py", line 860, in run_step **
outputs = model.train_step(data)
File "C:\Users\W10\PycharmProjects\TheCryptoBot\venv\lib\site-packages\keras\engine\training.py", line 808, in train_step
y_pred = self(x, training=True)
File "C:\Users\W10\PycharmProjects\TheCryptoBot\venv\lib\site-packages\keras\utils\traceback_utils.py", line 67, in error_handler
raise e.with_traceback(filtered_tb) from None
File "C:\Users\W10\PycharmProjects\TheCryptoBot\venv\lib\site-packages\keras\engine\input_spec.py", line 213, in assert_input_compatibility
raise ValueError(f'Input {input_index} of layer "{layer_name}" '
ValueError: Exception encountered when calling layer "sequential" (type Sequential).
Input 0 of layer "lstm" is incompatible with the layer: expected ndim=3, found ndim=2. Full shape received: (1, 33)
Call arguments received:
• inputs=tf.Tensor(shape=(1, 33), dtype=float32)
• training=True
• mask=None
Process finished with exit code 1
Model:
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
lstm (LSTM) (None, 478405, 33) 63153948
dropout (Dropout) (None, 478405, 33) 0
lstm_1 (LSTM) (None, 478405, 33) 8844
dropout_1 (Dropout) (None, 478405, 33) 0
dense (Dense) (None, 478405, 1) 34
=================================================================
Total params: 63,162,826
Trainable params: 63,162,826
Non-trainable params: 0
_________________________________________________________________
I think the problem is that you are reshaping the variable X like so X.reshape((1, 478405, 33)), however, this does not change the shape of X on its own. You need to set the result to X, like this X = X.reshape((1, 478405, 33)).
For time series you must use a TimeseriesGenerator
generator = TimeseriesGenerator(X, Y, length=478404, batch_size=100)
# print each sample
#for i in range(len(generator)):
#x, y = generator[i]
#print('%s => %s' % (x, y))
opt = tf.keras.optimizers.Adam(learning_rate=0.001, decay=1e-6)
print("Adding layer 1...")
model = Sequential()
model.add(LSTM(33, return_sequences=True, input_shape=(478404, 33), activation='sigmoid'))
print("Adding layer 2...")
model.add(Dropout(0.2))
print("Adding layer 3...")
model.add(LSTM(33, return_sequences=True))
print("Adding layer 4...")
model.add(Dropout(0.2))
print("Adding layer 5...")
model.add(Dense(1, activation="sigmoid"))
print("Adding layer 6...")
model.compile(optimizer='adam', loss='mse', metrics=['accuracy'])
print ('model compiled')
print (model.summary())
# Compile model
filepath = "RNN_Final-{epoch:02d}-{val_acc:.3f}" # unique file name that will include the epoch and the validation acc for that epoch
checkpoint = ModelCheckpoint("models/{}.model".format(filepath, monitor='val_acc', verbose=1, save_best_only=True, mode='max')) # saves only the best ones
history = model.fit(generator, steps_per_epoch=1, epochs=30, verbose=0)
print("Fit DOne")
print(history.history.keys())
# evaluate the model
scores = model.evaluate(generator)
I am trying to implement an image classifier using "The Street View House Numbers (SVHN) Dataset" from this link. I am using format 2 which contains 32x32 RGB centered digit images from 0 to 9. When I try to compile and fit the model I get the following error:
Epoch 1/10
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-37-31870b6986af> in <module>()
3
4 model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
----> 5 model.fit(trainX, trainY, validation_data=(validX, validY), batch_size=128, epochs=10)
9 frames
/usr/local/lib/python3.7/dist-packages/tensorflow/python/framework/func_graph.py in wrapper(*args, **kwargs)
975 except Exception as e: # pylint:disable=broad-except
976 if hasattr(e, "ag_error_metadata"):
--> 977 raise e.ag_error_metadata.to_exception(e)
978 else:
979 raise
ValueError: in user code:
/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/training.py:805 train_function *
return step_function(self, iterator)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/training.py:795 step_function **
outputs = model.distribute_strategy.run(run_step, args=(data,))
/usr/local/lib/python3.7/dist-packages/tensorflow/python/distribute/distribute_lib.py:1259 run
return self._extended.call_for_each_replica(fn, args=args, kwargs=kwargs)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/distribute/distribute_lib.py:2730 call_for_each_replica
return self._call_for_each_replica(fn, args, kwargs)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/distribute/distribute_lib.py:3417 _call_for_each_replica
return fn(*args, **kwargs)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/training.py:788 run_step **
outputs = model.train_step(data)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/training.py:756 train_step
y, y_pred, sample_weight, regularization_losses=self.losses)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/compile_utils.py:203 __call__
loss_value = loss_obj(y_t, y_p, sample_weight=sw)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/losses.py:152 __call__
losses = call_fn(y_true, y_pred)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/losses.py:256 call **
return ag_fn(y_true, y_pred, **self._fn_kwargs)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/util/dispatch.py:201 wrapper
return target(*args, **kwargs)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/losses.py:1537 categorical_crossentropy
return K.categorical_crossentropy(y_true, y_pred, from_logits=from_logits)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/util/dispatch.py:201 wrapper
return target(*args, **kwargs)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/backend.py:4833 categorical_crossentropy
target.shape.assert_is_compatible_with(output.shape)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/framework/tensor_shape.py:1134 assert_is_compatible_with
raise ValueError("Shapes %s and %s are incompatible" % (self, other))
ValueError: Shapes (None, 1) and (None, 10) are incompatible
The code is:
model = Sequential([
Conv2D(filters=64, kernel_size=3, strides=2, activation='relu', input_shape=(32,32,3)),
MaxPooling2D(pool_size=(2, 2), strides=1, padding='same'),
Conv2D(filters=32, kernel_size=3, strides=1, activation='relu'),
MaxPooling2D(pool_size=(2, 2), strides=1, padding='same'),
Flatten(),
Dense(10, activation='softmax')
])
model.summary()
Model: "sequential_10"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv2d_23 (Conv2D) (None, 15, 15, 64) 1792
_________________________________________________________________
max_pooling2d_23 (MaxPooling (None, 15, 15, 64) 0
_________________________________________________________________
conv2d_24 (Conv2D) (None, 13, 13, 32) 18464
_________________________________________________________________
max_pooling2d_24 (MaxPooling (None, 13, 13, 32) 0
_________________________________________________________________
flatten_10 (Flatten) (None, 5408) 0
_________________________________________________________________
dense_13 (Dense) (None, 10) 54090
=================================================================
Total params: 74,346
Trainable params: 74,346
Non-trainable params: 0
_________________________________________________________________
model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
model.fit(trainX, trainY, validation_data=(validX, validY), batch_size=128, epochs=10)
I was unable to solve the error, does anyone have any ideas on how to fix it?
As i could not see your coding for trainY; seems like - your trainY has only one column and your model output have 10 neurons, so Shapes (None, 1) and (None, 10) are incompatible. you can try this on your trainY(i.e one-hot encoding)
from sklearn.preprocessing import LabelBinarizer
label_as_binary = LabelBinarizer()
train__y_labels = label_as_binary.fit_transform(trainY)
and compile will look like as (look for train__y_labels)
model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
model.fit(train_X_input, train__y_labels, batch_size=128, epochs=1)
note: if your valid also throws the error, same would be needed on all y(s).
Change the compile statement so that
loss = 'sparse_categorical_cross_entropy'
The "sparse" indicates that the y values are numeric rather than one-hot
I have a dataset with 565 features and 10 different columns on the prediction site for predicting labels in the training model.Here is the model summary dimensions :
_________________________________________________________________
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv1d (Conv1D) (None, 564, 64) 256
_________________________________________________________________
flatten (Flatten) (None, 36096) 0
_________________________________________________________________
dense (Dense) (None, 50) 1804850
_________________________________________________________________
dense_1 (Dense) (None, 50) 2550
_________________________________________________________________
dense_2 (Dense) (None, 50) 2550
_________________________________________________________________
dense_3 (Dense) (None, 50) 2550
_________________________________________________________________
dense_4 (Dense) (None, 10) 510
=================================================================
Total params: 1,813,266
Trainable params: 1,813,266
Non-trainable params: 0
_________________________________________________________________
Here is the code I have used :
import pandas as pd
import numpy as np
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Conv1D, Flatten
from tensorflow.keras import optimizers
from sklearn.metrics import confusion_matrix
import tensorflow as tf
import tensorflow.keras.metrics
data = pd.read_csv('Step1_reducedfile.csv',skiprows = 1,header = None)
data = data.sample(frac=1).reset_index(drop=True)
train_X = data[0:data.shape[0],0:566]
train_y = data[0:data.shape[0],566:data.shape[1]]
train_X = train_X.reshape((train_X.shape[0], train_X.shape[1], 1))
import random
neurons = 50
strategy = tensorflow.distribute.MirroredStrategy()
with strategy.scope():
model = tf.keras.Sequential([
tf.keras.layers.Conv1D(64,kernel_size = 3,activation='relu',input_shape=train_X.shape[1:]),
tf.keras.layers.Flatten(),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(10, activation='softmax'),])
model.summary()
sgd = optimizers.SGD(lr=0.05, decay=1e-6, momentum=0.24, nesterov=True)
model.compile(loss='categorical_crossentropy',optimizer=sgd,metrics=['accuracy',tensorflow.keras.metrics.Precision()])
model.summary()
results = model.fit(train_X,train_y,validation_split = 0.2,epochs=10,batch_size = 100)
print(results)
I am getting the following error :
ValueError: in user code:
/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/engine/training.py:806 train_function *
return step_function(self, iterator)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/engine/training.py:796 step_function **
outputs = model.distribute_strategy.run(run_step, args=(data,))
/usr/local/lib64/python3.6/site-packages/tensorflow/python/distribute/distribute_lib.py:1211 run
return self._extended.call_for_each_replica(fn, args=args, kwargs=kwargs)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/distribute/distribute_lib.py:2585 call_for_each_replica
return self._call_for_each_replica(fn, args, kwargs)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/distribute/mirrored_strategy.py:585 _call_for_each_replica
self._container_strategy(), fn, args, kwargs)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/distribute/mirrored_run.py:96 call_for_each_replica
return _call_for_each_replica(strategy, fn, args, kwargs)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/distribute/mirrored_run.py:237 _call_for_each_replica
coord.join(threads)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/training/coordinator.py:389 join
six.reraise(*self._exc_info_to_raise)
/usr/local/lib/python3.6/site-packages/six.py:703 reraise
raise value
/usr/local/lib64/python3.6/site-packages/tensorflow/python/training/coordinator.py:297 stop_on_exception
yield
/usr/local/lib64/python3.6/site-packages/tensorflow/python/distribute/mirrored_run.py:323 run
self.main_result = self.main_fn(*self.main_args, **self.main_kwargs)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/engine/training.py:789 run_step **
outputs = model.train_step(data)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/engine/training.py:749 train_step
y, y_pred, sample_weight, regularization_losses=self.losses)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/engine/compile_utils.py:204 __call__
loss_value = loss_obj(y_t, y_p, sample_weight=sw)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/losses.py:149 __call__
losses = ag_call(y_true, y_pred)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/losses.py:253 call **
return ag_fn(y_true, y_pred, **self._fn_kwargs)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/util/dispatch.py:201 wrapper
return target(*args, **kwargs)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/losses.py:1535 categorical_crossentropy
return K.categorical_crossentropy(y_true, y_pred, from_logits=from_logits)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/util/dispatch.py:201 wrapper
return target(*args, **kwargs)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/keras/backend.py:4687 categorical_crossentropy
target.shape.assert_is_compatible_with(output.shape)
/usr/local/lib64/python3.6/site-packages/tensorflow/python/framework/tensor_shape.py:1134 assert_is_compatible_with
raise ValueError("Shapes %s and %s are incompatible" % (self, other))
ValueError: Shapes (None, 9) and (None, 10) are incompatible
That error shows that you are giving a wrong shape of label array to your model. It is s expecting an array of shape (None, 9), while you are giving an array of shape (None, 10). This may be because your dataset has 9 classes as rightly mentioned by Dr.Snoopy.
For the benefit of community here i am providing complete working code.
import pandas as pd
import numpy as np
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Conv1D, Flatten
from tensorflow.keras import optimizers
from sklearn.metrics import confusion_matrix
import tensorflow as tf
import tensorflow.keras.metrics
data = pd.read_csv('Step1_reducedfile.csv',skiprows = 1,header = None)
data = data.sample(frac=1).reset_index(drop=True)
train_X = data[0:data.shape[0],0:566]
train_y = data[0:data.shape[0],566:data.shape[1]]
train_X = train_X.reshape((train_X.shape[0], train_X.shape[1], 1))
import random
neurons = 50
strategy = tensorflow.distribute.MirroredStrategy()
with strategy.scope():
model = tf.keras.Sequential([
tf.keras.layers.Conv1D(64,kernel_size = 3,activation='relu',input_shape=train_X.shape[1:]),
tf.keras.layers.Flatten(),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(neurons,activation='relu'),
tf.keras.layers.Dense(9, activation='softmax'),])
model.summary()
sgd = optimizers.SGD(lr=0.05, decay=1e-6, momentum=0.24, nesterov=True)
model.compile(loss='categorical_crossentropy',optimizer=sgd,metrics=['accuracy',tensorflow.keras.metrics.Precision()])
model.summary()
results = model.fit(train_X,train_y,validation_split = 0.2,epochs=10,batch_size = 100)
print(results)
I am trying to build a tweet generator with keras using RNN. I came across this problem and I can not figure it where it comes from. I also searched the internet for hours but found nothing. I am sure it is something small, but I can't get it...
Here is the code(from https://github.com/schuyler-jackson/RNN_tweet_generation/blob/master/final_model.ipynb):
data = pd.read_csv('data/election2020.csv', usecols=[0, 4], names=['id', 'text'], encoding="latin-1")
# all tweets into one string
tweet_txt = data['text'][:].str.cat(sep=' ')
print(f'total characters in our dataset: {len(tweet_txt)}')
# get unique chars and make character mapping
chars = list(set(tweet_txt))
chars.sort()
char_to_index = dict((c,i) for i,c in enumerate(chars))
index_to_char = np.array(chars)
print(f"unique characters: {len(chars)}")
maxlen = 100
tweet_int = np.array([char_to_index[char] for char in tweet_txt])
seq_length = 100
examples_per_epoch = len(tweet_txt)//seq_length
char_dataset = tf.data.Dataset.from_tensor_slices(tweet_int)
sequences = char_dataset.batch(seq_length+1, drop_remainder=True)
def split_input_target(chunk):
input_text = chunk[:-1]
target_text = chunk[1:]
return input_text, target_text
dataset = sequences.map(split_input_target)
BATCH_SIZE = 64
steps_per_epoch = examples_per_epoch//BATCH_SIZE
BUFFER_SIZE = 10000
dataset.shuffle(BUFFER_SIZE).batch(BATCH_SIZE, drop_remainder=True)
print(dataset)
# Here is a model using the Keras Functional Api.
import functools
rnn = functools.partial(keras.layers.GRU, recurrent_activation='sigmoid')
def build_model(vocab_size, embedding_dim, rnn_units, batch_size):
model = Sequential()
model.add(Embedding(vocab_size, embedding_dim, batch_input_shape=[batch_size, None]))
model.add(rnn(rnn_units, return_sequences=True, recurrent_initializer='glorot_uniform', stateful=True))
model.add(Dropout(rate=0.2, noise_shape=(batch_size, 1, rnn_units)))
model.add(rnn(rnn_units, return_sequences=True, recurrent_initializer='glorot_uniform', stateful=True))
model.add(Dense(vocab_size))
return model
vocab_size = len(chars)
embedding_dim = 256
rnn_units = 256
batch_size = BATCH_SIZE
model = build_model(vocab_size=vocab_size, embedding_dim=embedding_dim, rnn_units=rnn_units, batch_size=batch_size)
model.summary()
def loss(labels, logits):
return sparse_categorical_crossentropy(labels, logits, from_logits=True)
model.compile(optimizer= Adam(), loss=loss)
checkpoint_dir = "model_gen/checkpoints"
checkpoint_prefix = os.path.join(checkpoint_dir, "ckpt_{epoch}.hdf5")
checkpoint_callback = ModelCheckpoint(filepath=checkpoint_prefix, save_weights_only=True)
EPOCHS = 5
early_stopping = EarlyStopping(monitor='val_loss', patience=3, restore_best_weights=True)
dataset2 = dataset
history = model.fit(np.array(dataset2), validation_data=dataset, validation_steps=30, epochs=EPOCHS, steps_per_epoch=steps_per_epoch, callbacks=[checkpoint_callback])
data looks like this:
id text
0 1204000574099857409 Democrats launch impeachment endgame with risi...
1 1203998807928823809 ***********************#biden2020 #Election202...
2 1203998376376832000 Any congressional representation doing this sh...
3 1203997840718086144 I"m glad to see this. #Booker deserves to be s...
4 1203997705938362368 #realDonaldTrump #AmericaFirst #KAG2020 #Trump...
and the outpus is this:
Using TensorFlow backend.
total characters in our dataset: 4786659
unique characters: 186
<MapDataset shapes: ((100,), (100,)), types: (tf.int32, tf.int32)>
Model: "sequential_1"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
embedding_1 (Embedding) (64, None, 256) 47616
_________________________________________________________________
gru_1 (GRU) (64, None, 256) 393984
_________________________________________________________________
dropout_1 (Dropout) (64, None, 256) 0
_________________________________________________________________
gru_2 (GRU) (64, None, 256) 393984
_________________________________________________________________
dense_1 (Dense) (64, None, 186) 47802
=================================================================
Total params: 883,386
Trainable params: 883,386
Non-trainable params: 0
_________________________________________________________________
Traceback (most recent call last):
File ".../src/tweet_generator_2.py", line 97, in <module>
history = model.fit(np.array(dataset2), validation_data=dataset, validation_steps=30, epochs=EPOCHS, steps_per_epoch=steps_per_epoch, callbacks=[checkpoint_callback])
File "...\Anaconda\envs\gputest\lib\site-packages\keras\engine\training.py", line 1154, in fit
batch_size=batch_size)
File "...\Anaconda\envs\gputest\lib\site-packages\keras\engine\training.py", line 579, in _standardize_user_data
exception_prefix='input')
File "...\Anaconda\envs\gputest\lib\site-packages\keras\engine\training_utils.py", line 135, in standardize_input_data
'with shape ' + str(data_shape))
ValueError: Error when checking input: expected embedding_1_input to have 2 dimensions, but got array with shape ()
Process finished with exit code 1
Does anyone know how can I solve the issue? I do not understand where does shape () come from.
Thank you!
I have reproduced your error, the issue was with the data you are providing when you fit the model.
In the code, you are generating data using tf.data which converts into tensorflow.python.data.ops.dataset_ops.DatasetV1Adapter type. But while .fit you are converting that to np.array(dataset2) of type numpy.ndarray which does not hold any data for input.
You need to assign to dataset when you are doing shuffle which was missing in your code. If you don't assign it to dataset your DatasetV1Adapter will have different shape.
I have modified your code and was able to run without any issues.
data = pd.read_csv('data/election2020.csv', usecols=[0, 4], names=['id', 'text'], encoding="latin-1")
# all tweets into one string
tweet_txt = data['text'][:].str.cat(sep=' ')
print(f'total characters in our dataset: {len(tweet_txt)}')
# get unique chars and make character mapping
chars = list(set(tweet_txt))
chars.sort()
char_to_index = dict((c,i) for i,c in enumerate(chars))
index_to_char = np.array(chars)
print(f"unique characters: {len(chars)}")
maxlen = 100
tweet_int = np.array([char_to_index[char] for char in tweet_txt])
seq_length = 100
examples_per_epoch = len(tweet_txt)//seq_length
char_dataset = tf.data.Dataset.from_tensor_slices(tweet_int)
sequences = char_dataset.batch(seq_length+1, drop_remainder=True)
def split_input_target(chunk):
input_text = chunk[:-1]
target_text = chunk[1:]
return input_text, target_text
dataset = sequences.map(split_input_target)
BATCH_SIZE = 64
steps_per_epoch = examples_per_epoch//BATCH_SIZE
BUFFER_SIZE = 10000
dataset = dataset.shuffle(BUFFER_SIZE).batch(BATCH_SIZE, drop_remainder=True)
print(dataset)
# Here is a model using the Keras Functional Api.
import functools
rnn = functools.partial(tf.keras.layers.GRU, recurrent_activation='sigmoid')
def build_model(vocab_size, embedding_dim, rnn_units, batch_size):
model = Sequential()
model.add(Embedding(vocab_size, embedding_dim, batch_input_shape=[batch_size, None]))
model.add(rnn(rnn_units, return_sequences=True, recurrent_initializer='glorot_uniform', stateful=True))
model.add(Dropout(rate=0.2, noise_shape=(batch_size, 1, rnn_units)))
model.add(rnn(rnn_units, return_sequences=True, recurrent_initializer='glorot_uniform', stateful=True))
model.add(Dense(vocab_size))
return model
vocab_size = len(chars)
embedding_dim = 256
rnn_units = 256
batch_size = BATCH_SIZE
model = build_model(vocab_size=vocab_size, embedding_dim=embedding_dim, rnn_units=rnn_units, batch_size=batch_size)
model.summary()
def loss(labels, logits):
return tf.keras.losses.sparse_categorical_crossentropy(labels, logits, from_logits=True)
model.compile(optimizer= tf.train.AdamOptimizer(), loss=loss)
checkpoint_dir = "model_gen/checkpoints"
checkpoint_prefix = os.path.join(checkpoint_dir, "ckpt_{epoch}.hdf5")
checkpoint_callback = tf.keras.callbacks.ModelCheckpoint(filepath=checkpoint_prefix, save_weights_only=True)
EPOCHS = 5
early_stopping = tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=3, restore_best_weights=True)
history = model.fit(dataset.repeat(), validation_data=dataset, validation_steps=30, epochs=EPOCHS, steps_per_epoch=steps_per_epoch, callbacks=[checkpoint_callback])
Model Summary:
Layer (type) Output Shape Param #
=================================================================
embedding (Embedding) (64, None, 256) 123648
_________________________________________________________________
gru (GRU) (64, None, 256) 393984
_________________________________________________________________
dropout (Dropout) (64, None, 256) 0
_________________________________________________________________
gru_1 (GRU) (64, None, 256) 393984
_________________________________________________________________
dense (Dense) (64, None, 483) 124131
=================================================================
Total params: 1,035,747
Trainable params: 1,035,747
Non-trainable params: 0
Training in progress:
Epoch 1/5
180/710 [======>.......................] - ETA: 19:52 - loss: 3.5081