ValueError: slice index 0 of dimension 0 Machine learniçng

Bonjour
j'ai réalisé un modèle d'apprentissage qui marche très bien avec jupyter notebook mais il prends des heures,
du coup je veux le faire avec le TPU de google colab pour accélérer l'apprentissage,j'ai reçu un message d'erreur dès le premier epoch

Code:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
model = Sequential()
model.add(Conv2D(32, (3, 3), input_shape=input_shape))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
 
model.add(Conv2D(32, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
 
model.add(Conv2D(64, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
 
model.add(Flatten())
model.add(Dense(220))
model.add(Activation('relu'))
model.add(Dropout(0.4))
model.add(Dense(120))
model.add(Activation('softmax'))

Optimizer

Code:

1
2
3
4
5
6
opt = tf.train.AdamOptimizer(learning_rate)
 
model.compile(
      optimizer=opt,
      loss='categorical_crossentropy',
      metrics=['acc'])

Convert Keras to TPU

Code:

1
2
3
4
5
6
7
8
9
10
11
try:
    device_name = os.environ['COLAB_TPU_ADDR']
    TPU_ADDRESS = 'grpc://' + device_name
    print('Found TPU at: {}'.format(TPU_ADDRESS))
 
except KeyError:
    print('TPU not found')
tpu_model = tf.contrib.tpu.keras_to_tpu_model(
    model,
    strategy=tf.contrib.tpu.TPUDistributionStrategy(
        tf.contrib.cluster_resolver.TPUClusterResolver(TPU_ADDRESS)))

ImageDataGenerator

Code:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
train_datagen = ImageDataGenerator(
    rescale=1. / 255,
    shear_range=0.2,
    zoom_range=0.2,
    horizontal_flip=True)
 
# this is the augmentation configuration we will use for testing:
# only rescaling
test_datagen = ImageDataGenerator(rescale=1. / 255)
 
train_generator = train_datagen.flow_from_directory(
    train_data_dir,
    target_size=(img_width, img_height),
    batch_size=batch_size,
    class_mode='categorical')#binary ,categorical
 
validation_generator = test_datagen.flow_from_directory(
    validation_data_dir,
    target_size=(img_width, img_height),
    batch_size=batch_size,
    class_mode='categorical')

Model Fit

Code:

1
2
3
4
5
6
7
model_fit=tpu_model.fit_generator(
    train_generator,
 
    epochs=50,
    steps_per_epoch=20, 
 
)

le message d'erreur

Code:

1
2
3
4
5
6
7
8
9
10
11
Epoch 1/50 15/33 [============>.................] - ETA: 8s - loss: 4.7722 - acc: 0.0083INFO:tensorflow:New input shapes; (re-)compiling: mode=train (# of cores 8), [TensorSpec(shape=(0,), dtype=tf.int32, name='core_id_60'), TensorSpec(shape=(0, 128, 128, 3), dtype=tf.float32, name='conv2d_3_input_20'), TensorSpec(shape=(0, 120), dtype=tf.float32, name='activation_13_target_30')] --------------------------------------------------------------------------- InvalidArgumentError Traceback (most recent call last) /usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/ops.py in _create_c_op(graph, node_def, inputs, control_inputs) 1658
try: -> 1659 c_op = c_api.TF_FinishOperation(op_desc) 1660 except errors.InvalidArgumentError as e:
 
InvalidArgumentError: slice index 0 of dimension 0 out of bounds. for 'strided_slice_19' (op: 'StridedSlice') with input shapes: [0], [1], [1], [1] and with computed input tensors: input[1] = <0>, input[2] = <1>, input[3] = <1>.
 
During handling of the above exception, another exception occurred:
 
ValueError Traceback (most recent call last) 17 frames /usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/ops.py in _create_c_op(graph, node_def, inputs, control_inputs) 1660
except errors.InvalidArgumentError as e: 1661 # Convert to ValueError for backwards compatibility. -> 1662 raise ValueError(str(e)) 1663 1664 return c_op
 
ValueError: slice index 0 of dimension 0 out of bounds. for 'strided_slice_19' (op: 'StridedSlice') with input shapes: [0], [1], [1], [1] and with computed input tensors: input[1] = <0>, input[2] = <1>, input[3] = <1>.