Created by: huayong

出现的问题如title， 主要目的是实现分割任务，所以自己写了个reader生成器来读取自己的训练数据，代码如下：

def load_image(file, image_size=224, mean_file=None): im = cv2.imread(file) im = cv2.resize(im, (image_size, image_size)) im = np.array(im).astype(np.float32) mean = np.array([103.94, 116.78, 123.68]) im = im / 255.0 - mean / 255.0 im = im / 255.0 im = im.transpose((2, 0, 1)) im = im.astype(np.float32) return im

def load_label(file, label_size=224): label_data = cv2.imread(file, cv2.IMREAD_GRAYSCALE) label_data = cv2.resize(label_data, (label_size, label_size)) label_data = np.array(label_data) label_data = label_data.astype(int) return label_data

def reader_creator(sub_name):

def reader():
    set_file = SET_FILE.format(sub_name)
    sets = [line.strip() for line in open(set_file, 'r')]
    for line in sets:
        line = line.strip()
        data_file = DATA_FILE.format(line)
        label_file = LABEL_FILE.format(line)
        data = load_image(data_file)
        label = load_label(label_file)
        data = np.array(data)
        label = np.array(label)
        yield data, label
return reader

但是在train过程中出现错误如下： Traceback (most recent call last): File "/Applications/PyCharm CE.app/Contents/helpers/pydev/pydevd.py", line 1599, in globals = debugger.run(setup['file'], None, None, is_module) File "/Applications/PyCharm CE.app/Contents/helpers/pydev/pydevd.py", line 1026, in run pydev_imports.execfile(file, globals, locals) # execute the script File "/Volumes/Huayong/baidu/paddlepaddle/Seg/seg_train.py", line 96, in train(net, class_num) File "/Volumes/Huayong/baidu/paddlepaddle/Seg/seg_train.py", line 79, in train 'label': 1}) File "/Library/Python/2.7/site-packages/paddle/v2/trainer.py", line 169, in train in_args = feeder(data_batch) File "/Library/Python/2.7/site-packages/py_paddle/dataprovider_converter.py", line 282, in call return self.convert(dat, argument) File "/Library/Python/2.7/site-packages/paddle/v2/data_feeder.py", line 133, in convert return DataProviderConverter.convert(self, reorder_data(dat), argument) File "/Library/Python/2.7/site-packages/py_paddle/dataprovider_converter.py", line 277, in convert scanner.finish_scan(argument) File "/Library/Python/2.7/site-packages/py_paddle/dataprovider_converter.py", line 211, in finish_scan ids = swig_paddle.IVector.create(self.ids, self.data_in_gpu) File "/Library/Python/2.7/site-packages/py_paddle/swig_paddle.py", line 1344, in create return _swig_paddle.IVector_create(args) NotImplementedError: Wrong number or type of arguments for overloaded function 'IVector_create'. Possible C/C++ prototypes are: IVector::create(std::vector< int,std::allocator< int > > const &,bool) IVector::create(std::vector< int,std::allocator< int > > const &) train代码如下： def train(net, class_num): lbl = paddle.layer.data( name="label", type=paddle.data_type.integer_value(224224))

cost = paddle.layer.cross_entropy_cost(input=net,
                                       label=lbl)


# Create parameters
parameters = paddle.parameters.create(cost)

# Create optimizer
# momentum_optimizer = paddle.optimizer.Momentum(
#     momentum=0.9,
#     regularization=paddle.optimizer.L2Regularization(rate=0.0002 * 128),
#     learning_rate=0.1 / 128.0,
#     learning_rate_decay_a=0.1,
#     learning_rate_decay_b=50000 * 100,
#     learning_rate_schedule='discexp')

# create optimizer
adam_optimizer = paddle.optimizer.Adam(
    learning_rate=2e-3,
    regularization=paddle.optimizer.L2Regularization(rate=8e-4),
    model_average=paddle.optimizer.ModelAverage(average_window=0.5))

# Create trainer
trainer = paddle.trainer.SGD(
    cost=cost, parameters=parameters, update_equation=adam_optimizer)

# End batch and end pass event handler
def event_handler(event):
    if isinstance(event, paddle.event.EndIteration):
        if event.batch_id % 2 == 0:
            print "\nPass %d, Batch %d, Cost %f, %s" % (
                event.pass_id, event.batch_id, event.cost, event.metrics)
        else:
            sys.stdout.write('.')
            sys.stdout.flush()
    if isinstance(event, paddle.event.EndPass):
        # save parameters
        with open('params_pass_%d.tar' % event.pass_id, 'w') as f:
            trainer.save_parameter_to_tar(f)

        result = trainer.test(
            reader=paddle.batch(
                seg_reader.train(), batch_size=1),
            feeding={'image': 0,
                     'label': 1})
        print "\nTest with Pass %d, %s" % (event.pass_id, result.metrics)

trainer.train(
    reader=paddle.batch(
        paddle.reader.shuffle(
            seg_reader.train(), buf_size=50000),
        batch_size=1),
    num_passes=1,
    event_handler=event_handler,
    feeding={'image': 0,
             'label': 1})
return

pycharm debug模式下查看读取的数据应该没有问题