在几个时期之后，Finetuning停止了。主要在时代5或8中。上一个时期的数量在不同的运行中是不同的。
错误：
File "/opt/keras-python3.4/lib/python3.4/site-packages/PIL/JpegImagePlugin.py", line 126, in APP
dpi = x_resolution[0] / x_resolution[1]
ZeroDivisionError: division by zero
我的配置：
有什么问题？为什么它会在几个时代之后出现？ 是否有可能一个损坏的图像文件出现此问题？为什么在第一个时代它没有发生？
代码：
from keras.applications.inception_v3 import InceptionV3
from keras.models import Model
from keras.layers import Dense, GlobalAveragePooling2D
from keras.callbacks import ModelCheckpoint, TensorBoard, CSVLogger, Callback
from keras.optimizers import SGD
# create the base pre-trained model
from keras.preprocessing.image import ImageDataGenerator
base_model = InceptionV3(weights='imagenet', include_top=False)
# add a global spatial average pooling layer
x = base_model.output
x = GlobalAveragePooling2D()(x)
# let's add a fully-connected layer
x = Dense(1024, activation='relu')(x)
# and a logistic layer -- let's say we have 2 classes
# predictions = Dense(1, activation='softmax')(x) #A
predictions = Dense(2, activation='softmax')(x) #B
# this is the model we will train
model = Model(input=base_model.input, output=predictions)
# first: train only the top layers (which were randomly initialized)
# i.e. freeze all convolutional InceptionV3 layers
for layer in base_model.layers:
# layer.trainable = False
layer.trainable = True
model.compile(optimizer=SGD(lr=0.001, momentum=0.9), loss='sparse_categorical_crossentropy') #B
# train the model on the new data for a few epochs
train_datagen = ImageDataGenerator(
rescale=1. / 255,
shear_range=0.2,
zoom_range=0.2,
horizontal_flip=True)
# I changed flow_from_directory() a bit/
train_generator = train_datagen.flow_from_directory(
'.../train/',
_mode='b-w',
classes=['white','black'],
follow_links=True,
shuffle=True,
target_size=(299, 299),
batch_size=16,
class_mode='binary')
test_datagen = ImageDataGenerator(rescale=1./255)
# I changed flow_from_directory() a bit/
validation_generator = test_datagen.flow_from_directory(
'.../val/',
_mode='train-val-test_b-w', _set='val',
classes=['white', 'black'],
target_size=(299, 299),
batch_size=16,
follow_links=True,
class_mode='binary')
class LossHistory(Callback):
def on_train_begin(self, logs={}):
self.f = open('./history-log/log.txt', 'w')
self.f.write('batch' + ' , ' + 'loss\n')
def on_batch_end(self, batch, logs={}):
self.f.write(str(logs.get('batch')) + ' , ' + str(logs.get('loss')) + '\n')
model_checkpoint = ModelCheckpoint('./saved_models/{epoch:02d}-{val_loss:.2f}.hdf5', monitor='val_loss', verbose=0, save_best_only=False, save_weights_only=False, mode='auto', period=1)
csv_logger = CSVLogger('./csv-log/log.csv', separator=',', append=False)
history = LossHistory()
from PIL import ImageFile
ImageFile.LOAD_TRUNCATED_IMAGES = True
model.fit_generator(
train_generator,
steps_per_epoch=733,
epochs=1000,
callbacks=[model_checkpoint, csv_logger, history],
validation_data=validation_generator,
verbose=1,
validation_steps=706)
输出：
Using TensorFlow backend.
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcublas.so.8.0 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcudnn.so.5 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcufft.so.8.0 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcurand.so.8.0 locally
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE3 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:910] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
I tensorflow/core/common_runtime/gpu/gpu_device.cc:885] Found device 0 with properties:
name: GeForce GTX TITAN
major: 3 minor: 5 memoryClockRate (GHz) 0.8755
pciBusID 0000:01:00.0
Total memory: 5.94GiB
Free memory: 5.75GiB
I tensorflow/core/common_runtime/gpu/gpu_device.cc:906] DMA: 0
I tensorflow/core/common_runtime/gpu/gpu_device.cc:916] 0: Y
I tensorflow/core/common_runtime/gpu/gpu_device.cc:975] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX TITAN, pci bus id: 0000:01:00.0)
/patna/patna-codes/python/tensorlow-keras-test/finetune.py:27: UserWarning: Update your Model call to the Keras 2 API: Model(inputs=Tensor("in..., outputs=Tensor("de...)
model = Model(input=base_model.input, output=predictions)
Found 93792 images belonging to 2 classes.
Found 90260 images belonging to 2 classes.
Epoch 1/1000
I tensorflow/core/common_runtime/gpu/pool_allocator.cc:247] PoolAllocator: After 2278 get requests, put_count=2134 evicted_count=1000 eviction_rate=0.468604 and unsatisfied allocation rate=0.546093
I tensorflow/core/common_runtime/gpu/pool_allocator.cc:259] Raising pool_size_limit_ from 100 to 110
9/733 [..............................] - ETA: 869s - loss: 0.7533I tensorflow/core/common_runtime/gpu/pool_allocator.cc:247] PoolAllocator: After 2286 get requests, put_count=2222 evicted_count=1000 eviction_rate=0.450045 and unsatisfied allocation rate=0.474628
I tensorflow/core/common_runtime/gpu/pool_allocator.cc:259] Raising pool_size_limit_ from 233 to 256
21/733 [..............................] - ETA: 646s - loss: 0.6912I tensorflow/core/common_runtime/gpu/pool_allocator.cc:247] PoolAllocator: After 2419 get requests, put_count=2718 evicted_count=1000 eviction_rate=0.367918 and unsatisfied allocation rate=0.312112
I tensorflow/core/common_runtime/gpu/pool_allocator.cc:259] Raising pool_size_limit_ from 596 to 655
732/733 [============================>.] - ETA: 0s - loss: 0.4559/opt/keras-python3.4/lib/python3.4/site-packages/PIL/TiffImagePlugin.py:709: UserWarning: Corrupt EXIF data. Expecting to read 4 bytes but only got 0.
warnings.warn(str(msg))
/opt/keras-python3.4/lib/python3.4/site-packages/PIL/Image.py:885: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images
'to RGBA images')
733/733 [==============================] - 846s - loss: 0.4558 - val_loss: 0.2498
Epoch 2/1000
732/733 [============================>.] - ETA: 0s - loss: 0.3979/opt/keras-python3.4/lib/python3.4/site-packages/PIL/TiffImagePlugin.py:709: UserWarning: Corrupt EXIF data. Expecting to read 4 bytes but only got 2.
warnings.warn(str(msg))
733/733 [==============================] - 844s - loss: 0.3977 - val_loss: 0.1956
Epoch 3/1000
733/733 [==============================] - 820s - loss: 0.3665 - val_loss: 0.2093
Epoch 4/1000
733/733 [==============================] - 819s - loss: 0.3549 - val_loss: 0.1918
Epoch 5/1000
732/733 [============================>.] - ETA: 0s - loss: 0.3427Exception in thread Thread-6:
Traceback (most recent call last):
File "/usr/lib/python3.4/threading.py", line 920, in _bootstrap_inner
self.run()
File "/usr/lib/python3.4/threading.py", line 868, in run
self._target(*self._args, **self._kwargs)
File "/opt/keras-python3.4/lib/python3.4/site-packages/keras/engine/training.py", line 606, in data_generator_task
generator_output = next(self._generator)
File "/opt/keras-python3.4/lib/python3.4/site-packages/keras/preprocessing/image.py", line 756, in next
return self.next(*args, **kwargs)
File "/opt/keras-python3.4/lib/python3.4/site-packages/keras/preprocessing/image.py", line 1328, in next
target_size=self.target_size)
File "/opt/keras-python3.4/lib/python3.4/site-packages/keras/preprocessing/image.py", line 320, in load_img
img = pil_image.open(path)
File "/opt/keras-python3.4/lib/python3.4/site-packages/PIL/Image.py", line 2439, in open
im = _open_core(fp, filename, prefix)
File "/opt/keras-python3.4/lib/python3.4/site-packages/PIL/Image.py", line 2429, in _open_core
im = factory(fp, filename)
File "/opt/keras-python3.4/lib/python3.4/site-packages/PIL/JpegImagePlugin.py", line 761, in jpeg_factory
im = JpegImageFile(fp, filename)
File "/opt/keras-python3.4/lib/python3.4/site-packages/PIL/ImageFile.py", line 100, in init
self._open()
File "/opt/keras-python3.4/lib/python3.4/site-packages/PIL/JpegImagePlugin.py", line 332, in _open
handler(self, i)
File "/opt/keras-python3.4/lib/python3.4/site-packages/PIL/JpegImagePlugin.py", line 126, in APP
dpi = x_resolution[0] / x_resolution[1]
ZeroDivisionError: division by zero
Traceback (most recent call last):
File "/patna/patna-codes/python/tensorlow-keras-test/finetune.py", line 105, in
validation_steps=706)
File "/opt/keras-python3.4/lib/python3.4/site-packages/keras/legacy/interfaces.py", line 88, in wrapper
return func(*args, **kwargs)
File "/opt/keras-python3.4/lib/python3.4/site-packages/keras/engine/training.py", line 1899, in fit_generator
pickle_safe=pickle_safe)
File "/opt/keras-python3.4/lib/python3.4/site-packages/keras/legacy/interfaces.py", line 88, in wrapper
return func(*args, **kwargs)
File "/opt/keras-python3.4/lib/python3.4/site-packages/keras/engine/training.py", line 1985, in evaluate_generator
str(generator_output))
ValueError: output of generator should be a tuple (x, y, sample_weight) or (x, y). Found: None
Process finished with exit code 1
答案 0 :(得分：0)
我在数据集中发现了一些损坏的图像。
我注意到，对于批量大小128而不是16，设置了steps_per_epoch = 733和validation_steps = 706，因此所有图像都没有提供给网络，因此在第一个时期不会发生错误。