diff --git a/.travis.yml b/.travis.yml index 4fb2ca938795bb6a69f7d7991aee9f7386947bf2..31d76ecb876f117217566359e69a6bd7fdb2c45e 100644 --- a/.travis.yml +++ b/.travis.yml @@ -57,7 +57,7 @@ before_install: - if [[ "$JOB" == "PRE_COMMIT" ]]; then sudo ln -s /usr/bin/clang-format-3.8 /usr/bin/clang-format; fi # Paddle is using protobuf 3.1 currently. Protobuf 3.2 breaks the compatibility. So we specify the python # protobuf version. - - pip install numpy wheel 'protobuf==3.1' sphinx recommonmark sphinx_rtd_theme virtualenv pre-commit requests==2.9.2 LinkChecker + - pip install numpy wheel 'protobuf==3.1' sphinx recommonmark sphinx_rtd_theme virtualenv pre-commit requests==2.9.2 LinkChecker 'scikit-learn>=0.18.0' 'scipy>=0.18.0' script: - paddle/scripts/travis/main.sh notifications: diff --git a/demo/mnist/api_train_v2.py b/demo/mnist/api_train_v2.py index 767dffdad82f59d73b1505586260a61f5008f1f8..45a70bc84afa29c5f12d2a8dddf17b8034e1c541 100644 --- a/demo/mnist/api_train_v2.py +++ b/demo/mnist/api_train_v2.py @@ -1,15 +1,6 @@ import numpy import paddle.v2 as paddle -import mnist_util - - -def train_reader(): - train_file = './data/raw_data/train' - generator = mnist_util.read_from_mnist(train_file) - for item in generator: - yield item - def main(): paddle.init(use_gpu=False, trainer_count=1) @@ -30,27 +21,29 @@ def main(): adam_optimizer = paddle.optimizer.Adam(learning_rate=0.01) + trainer = paddle.trainer.SGD(topology=cost, + parameters=parameters, + update_equation=adam_optimizer) + def event_handler(event): if isinstance(event, paddle.event.EndIteration): - if event.batch_id % 100 == 0: - print "Pass %d, Batch %d, Cost %f, %s" % ( - event.pass_id, event.batch_id, event.cost, event.metrics) + if event.batch_id % 1000 == 0: + result = trainer.test(reader=paddle.reader.batched( + paddle.dataset.mnist.test_creator(), batch_size=256)) + + print "Pass %d, Batch %d, Cost %f, %s, Testing metrics %s" % ( + event.pass_id, event.batch_id, event.cost, event.metrics, + result.metrics) + else: pass - trainer = paddle.trainer.SGD(update_equation=adam_optimizer) - - trainer.train(train_data_reader=train_reader, - topology=cost, - parameters=parameters, - event_handler=event_handler, - batch_size=32, # batch size should be refactor in Data reader - data_types=[ # data_types will be removed, It should be in - # network topology - ('pixel', images.type), - ('label', label.type)], - reader_dict={'pixel':0, 'label':1} - ) + trainer.train( + reader=paddle.reader.batched( + paddle.reader.shuffle( + paddle.dataset.mnist.train_creator(), buf_size=8192), + batch_size=32), + event_handler=event_handler) if __name__ == '__main__': diff --git a/python/CMakeLists.txt b/python/CMakeLists.txt index 71af50a9a4ed835635362c24d6c1ae5e92de050b..48e0a1993d07f801e65dfa54a991995c593fe475 100644 --- a/python/CMakeLists.txt +++ b/python/CMakeLists.txt @@ -4,7 +4,7 @@ set(OUTPUT_DIR file(GLOB TRAINER_PY_FILES . ./paddle/trainer/*.py) file(GLOB HELPERS_PY_FILES . ./paddle/trainer_config_helpers/*.py) file(GLOB UTILS_PY_FILES . ./paddle/utils/*.py) -file(GLOB V2_PY_FILES . ./paddle/v2/*.py) +file(GLOB_RECURSE V2_PY_FILES ./paddle/v2/ *.py) set(PY_FILES paddle/__init__.py ${TRAINER_PY_FILES} @@ -24,7 +24,7 @@ add_custom_target(paddle_python ALL DEPENDS ${OUTPUT_DIR}/.timestamp) add_subdirectory(paddle/trainer_config_helpers/tests) -add_subdirectory(paddle/reader/tests) +add_subdirectory(paddle/v2/reader/tests) add_subdirectory(paddle/v2/tests) install(DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}/dist/ diff --git a/python/paddle/reader/tests/CMakeLists.txt b/python/paddle/reader/tests/CMakeLists.txt deleted file mode 100644 index da072fb3dbeed5bee15fa1f64372ad8dec497070..0000000000000000000000000000000000000000 --- a/python/paddle/reader/tests/CMakeLists.txt +++ /dev/null @@ -1,9 +0,0 @@ -add_test(NAME reader_decorator_test - COMMAND ${PROJ_ROOT}/paddle/.set_python_path.sh -d ${PROJ_ROOT}/python/ - ${PYTHON_EXECUTABLE} ${PROJ_ROOT}/python/paddle/reader/tests/decorator_test.py - WORKING_DIRECTORY ${PROJ_ROOT}/python/paddle) - -add_test(NAME reader_creator_test - COMMAND ${PROJ_ROOT}/paddle/.set_python_path.sh -d ${PROJ_ROOT}/python/ - ${PYTHON_EXECUTABLE} ${PROJ_ROOT}/python/paddle/reader/tests/creator_test.py - WORKING_DIRECTORY ${PROJ_ROOT}/python/paddle) diff --git a/python/paddle/v2/__init__.py b/python/paddle/v2/__init__.py index 1dff754edf1faafcabd3bcea733970235964d344..cd39191713048bb0bc53b95e8cd2b2d448073bee 100644 --- a/python/paddle/v2/__init__.py +++ b/python/paddle/v2/__init__.py @@ -19,13 +19,15 @@ import trainer import event import data_type import data_feeder +from . import dataset +from . import reader import attr import pooling import py_paddle.swig_paddle as api __all__ = [ 'optimizer', 'layer', 'activation', 'parameters', 'init', 'trainer', - 'event', 'data_type', 'attr', 'pooling', 'data_feeder' + 'event', 'data_type', 'attr', 'pooling', 'data_feeder', 'dataset', 'reader' ] diff --git a/python/paddle/v2/dataset/__init__.py b/python/paddle/v2/dataset/__init__.py index e69de29bb2d1d6434b8b29ae775ad8c2e48c5391..9647e9850332b795f6480d26e4f9c736129782ae 100644 --- a/python/paddle/v2/dataset/__init__.py +++ b/python/paddle/v2/dataset/__init__.py @@ -0,0 +1,3 @@ +import mnist + +__all__ = ['mnist'] diff --git a/python/paddle/v2/event.py b/python/paddle/v2/event.py index 835e28e6218df22e1cad7f7bb31c3c9941657252..a78bcf076cc65e0dfdfc5760e099900418162f35 100644 --- a/python/paddle/v2/event.py +++ b/python/paddle/v2/event.py @@ -11,7 +11,10 @@ There are: TODO(yuyang18): Complete it! """ import py_paddle.swig_paddle as api -__all__ = ['EndIteration', 'BeginIteration', 'BeginPass', 'EndPass'] + +__all__ = [ + 'EndIteration', 'BeginIteration', 'BeginPass', 'EndPass', 'TestResult' +] class WithMetric(object): @@ -30,6 +33,11 @@ class WithMetric(object): return retv +class TestResult(WithMetric): + def __init__(self, evaluator): + super(TestResult, self).__init__(evaluator) + + class BeginPass(object): """ Event On One Pass Training Start. diff --git a/python/paddle/reader/__init__.py b/python/paddle/v2/reader/__init__.py similarity index 100% rename from python/paddle/reader/__init__.py rename to python/paddle/v2/reader/__init__.py diff --git a/python/paddle/reader/creator.py b/python/paddle/v2/reader/creator.py similarity index 100% rename from python/paddle/reader/creator.py rename to python/paddle/v2/reader/creator.py diff --git a/python/paddle/reader/decorator.py b/python/paddle/v2/reader/decorator.py similarity index 90% rename from python/paddle/reader/decorator.py rename to python/paddle/v2/reader/decorator.py index 9f4234358f469b9741dfef96aab3853d5972b5f4..d83e2f4577f20bcd2220a1a29b394d925126922e 100644 --- a/python/paddle/reader/decorator.py +++ b/python/paddle/v2/reader/decorator.py @@ -14,7 +14,7 @@ __all__ = [ 'map_readers', 'buffered', 'compose', 'chain', 'shuffle', - 'ComposeNotAligned' + 'ComposeNotAligned', 'batched' ] from Queue import Queue @@ -191,3 +191,25 @@ def buffered(reader, size): e = q.get() return data_reader + + +def batched(reader, batch_size): + """ + Create a batched reader. + :param reader: the data reader to read from. + :param batch_size: batch_size + :return: the batched reader. + """ + + def __impl__(): + r = reader() + batch = [] + for instance in r: + batch.append(instance) + if len(batch) == batch_size: + yield batch + batch = [] + if batch: + yield batch + + return __impl__ diff --git a/python/paddle/v2/reader/tests/CMakeLists.txt b/python/paddle/v2/reader/tests/CMakeLists.txt new file mode 100644 index 0000000000000000000000000000000000000000..a51f700406b48f8186e45f1ced94765e343a8b5e --- /dev/null +++ b/python/paddle/v2/reader/tests/CMakeLists.txt @@ -0,0 +1,3 @@ +add_test(NAME reader_tests + COMMAND bash ${PROJ_ROOT}/python/paddle/v2/reader/tests/run_tests.sh + ${PYTHON_EXECUTABLE}) diff --git a/python/paddle/v2/reader/tests/__init__.py b/python/paddle/v2/reader/tests/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/python/paddle/reader/tests/creator_test.py b/python/paddle/v2/reader/tests/creator_test.py similarity index 89% rename from python/paddle/reader/tests/creator_test.py rename to python/paddle/v2/reader/tests/creator_test.py index eda8ab6715b2be0c9cb6163adf60d8fbdf2d7e8c..9f8d7133b8694aae5541eff9576eaba8a31e77dc 100644 --- a/python/paddle/reader/tests/creator_test.py +++ b/python/paddle/v2/reader/tests/creator_test.py @@ -11,17 +11,19 @@ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. +import os import unittest -import paddle.reader.creator + import numpy as np -import os + +import paddle.v2.reader.creator class TestNumpyArray(unittest.TestCase): def test_numpy_array(self): l = [[1, 2, 3], [4, 5, 6]] x = np.array(l, np.int32) - reader = paddle.reader.creator.np_array(x) + reader = paddle.v2.reader.creator.np_array(x) for idx, e in enumerate(reader()): self.assertItemsEqual(e, l[idx]) @@ -29,7 +31,7 @@ class TestNumpyArray(unittest.TestCase): class TestTextFile(unittest.TestCase): def test_text_file(self): path = os.path.join(os.path.dirname(__file__), "test_data_creator.txt") - reader = paddle.reader.creator.text_file(path) + reader = paddle.v2.reader.creator.text_file(path) for idx, e in enumerate(reader()): self.assertEqual(e, str(idx * 2) + " " + str(idx * 2 + 1)) diff --git a/python/paddle/reader/tests/decorator_test.py b/python/paddle/v2/reader/tests/decorator_test.py similarity index 81% rename from python/paddle/reader/tests/decorator_test.py rename to python/paddle/v2/reader/tests/decorator_test.py index 0396a61786539bf57be4cab9ebbd108ade9e7c83..734154b9790a4dc118d11992343648364c907305 100644 --- a/python/paddle/reader/tests/decorator_test.py +++ b/python/paddle/v2/reader/tests/decorator_test.py @@ -11,9 +11,10 @@ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. -import unittest -import paddle.reader import time +import unittest + +import paddle.v2.reader def reader_creator_10(dur): @@ -37,7 +38,7 @@ class TestMap(unittest.TestCase): yield "h" yield "i" - r = paddle.reader.map_readers(tokenize, read) + r = paddle.v2.reader.map_readers(tokenize, read) for i, e in enumerate(r()): self.assertEqual(e, i) @@ -45,7 +46,7 @@ class TestMap(unittest.TestCase): class TestBuffered(unittest.TestCase): def test_read(self): for size in range(20): - b = paddle.reader.buffered(reader_creator_10(0), size) + b = paddle.v2.reader.buffered(reader_creator_10(0), size) c = 0 for i in b(): self.assertEqual(i, c) @@ -54,7 +55,7 @@ class TestBuffered(unittest.TestCase): def test_buffering(self): # read have 30ms delay. - b = paddle.reader.buffered(reader_creator_10(0.03), 10) + b = paddle.v2.reader.buffered(reader_creator_10(0.03), 10) last_time = time.time() for idx, i in enumerate(b()): elapsed_time = time.time() - last_time @@ -68,17 +69,17 @@ class TestBuffered(unittest.TestCase): class TestCompose(unittest.TestCase): def test_compse(self): - reader = paddle.reader.compose( + reader = paddle.v2.reader.compose( reader_creator_10(0), reader_creator_10(0)) for idx, e in enumerate(reader()): self.assertEqual(e, (idx, idx)) def test_compose_not_aligned(self): total = 0 - reader = paddle.reader.compose( - paddle.reader.chain(reader_creator_10(0), reader_creator_10(0)), + reader = paddle.v2.reader.compose( + paddle.v2.reader.chain(reader_creator_10(0), reader_creator_10(0)), reader_creator_10(0)) - with self.assertRaises(paddle.reader.ComposeNotAligned): + with self.assertRaises(paddle.v2.reader.ComposeNotAligned): for e in reader(): total += 1 # expecting 10, not 20 @@ -86,8 +87,8 @@ class TestCompose(unittest.TestCase): def test_compose_not_aligned_no_check(self): total = 0 - reader = paddle.reader.compose( - paddle.reader.chain(reader_creator_10(0), reader_creator_10(0)), + reader = paddle.v2.reader.compose( + paddle.v2.reader.chain(reader_creator_10(0), reader_creator_10(0)), reader_creator_10(0), check_alignment=False) for e in reader(): @@ -98,7 +99,7 @@ class TestCompose(unittest.TestCase): class TestChain(unittest.TestCase): def test_chain(self): - c = paddle.reader.chain(reader_creator_10(0), reader_creator_10(0)) + c = paddle.v2.reader.chain(reader_creator_10(0), reader_creator_10(0)) idx = 0 for e in c(): self.assertEqual(e, idx % 10) @@ -111,7 +112,7 @@ class TestShuffle(unittest.TestCase): case = [(0, True), (1, True), (10, False), (100, False)] a = reader_creator_10(0) for size, checkEq in case: - s = paddle.reader.shuffle(a, size) + s = paddle.v2.reader.shuffle(a, size) total = 0 for idx, e in enumerate(s()): if checkEq: diff --git a/python/paddle/v2/reader/tests/run_tests.sh b/python/paddle/v2/reader/tests/run_tests.sh new file mode 100755 index 0000000000000000000000000000000000000000..a544a5636021bcf8bd9a35966c91ae343c149d14 --- /dev/null +++ b/python/paddle/v2/reader/tests/run_tests.sh @@ -0,0 +1,35 @@ +#!/bin/bash +# Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + + +pushd `dirname $0` > /dev/null +SCRIPTPATH=$PWD +popd > /dev/null + +cd $SCRIPTPATH +$1 -m pip install ../../../../../paddle/dist/*.whl + +test_list="creator_test.py decorator_test.py" + +export PYTHONPATH=$PWD/../../../../../python/ + +for fn in $test_list +do + echo "test $fn" + $1 $fn + if [ $? -ne 0 ]; then + exit 1 + fi +done diff --git a/python/paddle/reader/tests/test_data_creator.txt b/python/paddle/v2/reader/tests/test_data_creator.txt similarity index 100% rename from python/paddle/reader/tests/test_data_creator.txt rename to python/paddle/v2/reader/tests/test_data_creator.txt diff --git a/python/paddle/v2/trainer.py b/python/paddle/v2/trainer.py index 097814d2f4619797470668cbd0ea95f112a1fde6..77232d5ac5e411e7cb6c44236c7d1a7e9341af05 100644 --- a/python/paddle/v2/trainer.py +++ b/python/paddle/v2/trainer.py @@ -23,24 +23,25 @@ def default_event_handler(event): pass +def __bfs_travel_topology__(callback, *topologies): + for each_layer in topologies: + callback(each_layer) + __bfs_travel_topology__(callback, + *each_layer.__parent_layers__.values()) + + class ITrainer(object): """ The interface of Trainer. The only exposed method is `train`. """ - def train(self, - train_data_reader, - topology, - parameters, - test_data_reader=None, - event_handler=None): + def train(self, reader, topology, parameters, event_handler=None): """ train method. - :param train_data_reader: + :param reader: :param topology: :param parameters: - :param test_data_reader: :param event_handler: :return: """ @@ -49,83 +50,99 @@ class ITrainer(object): class SGD(ITrainer): - def __init__(self, update_equation): + def __init__(self, topology, parameters, update_equation): """ Simple SGD Trainer. :param update_equation: The optimizer object. :type update_equation: v2_optimizer.Optimizer """ + if not isinstance(parameters, v2_parameters.Parameters): + raise TypeError('parameters should be parameters') + if not isinstance(update_equation, v2_optimizer.Optimizer): - raise ValueError("update equation parameter must be " - "paddle.v2.optimizer.Optimizer") + raise TypeError("update equation parameter must be " + "paddle.v2.optimizer.Optimizer") self.__optimizer__ = update_equation + self.__topology__ = topology + self.__parameters__ = parameters + self.__topology_in_proto__ = v2_layer.parse_network(topology) + data_types = dict() + + def __travel__(l): + if hasattr(l, 'type'): + data_types[l.name] = l.type + + if not isinstance(topology, collections.Sequence): + topology = [topology] + __bfs_travel_topology__(__travel__, *topology) + self.__data_types__ = [ + (iname, data_types[iname]) + for iname in self.__topology_in_proto__.input_layer_names + ] + + if not isinstance(self.__topology_in_proto__, ModelConfig): + raise TypeError('topology should be a model config') + + gm = api.GradientMachine.createFromConfigProto( + self.__topology_in_proto__, api.CREATE_MODE_NORMAL, + self.__optimizer__.enable_types()) + assert isinstance(gm, api.GradientMachine) + parameters.append_gradient_machine(gm) + self.__gradient_machine__ = gm + self.__gradient_machine__.randParameters() - def train(self, - train_data_reader, - topology, - parameters, - num_passes=1, - test_data_reader=None, - event_handler=None, - batch_size=32, - data_types=None, - reader_dict=None): + def train(self, reader, num_passes=1, event_handler=None, reader_dict=None): """ Training method. Will train num_passes of input data. - :param train_data_reader: + :param reader: :param topology: Network Topology, use one or more Layers to represent it. :param parameters: The parameter pools. :param num_passes: The total train passes. - :param test_data_reader: :param event_handler: Event handler. A method will be invoked when event occurred. :type event_handler: (BaseEvent) => None - :param batch_size: Not important, will be removed after data refactor. :param data_types: Not important, will be removed after data refactor. :return: """ if event_handler is None: event_handler = default_event_handler - topology = v2_layer.parse_network(topology) + if reader_dict is None: + reader_dict = self.default_reader_dict() __check_train_args__(**locals()) - - gm = api.GradientMachine.createFromConfigProto( - topology, api.CREATE_MODE_NORMAL, self.__optimizer__.enable_types()) - assert isinstance(gm, api.GradientMachine) - parameters.append_gradient_machine(gm) - gm.randParameters() updater = self.__optimizer__.create_local_updater() - updater.init(gm) + updater.init(self.__gradient_machine__) - gm.start() - batch_evaluator = gm.makeEvaluator() + self.__gradient_machine__.start() + batch_evaluator = self.__gradient_machine__.makeEvaluator() assert isinstance(batch_evaluator, api.Evaluator) - pass_evaluator = gm.makeEvaluator() + pass_evaluator = self.__gradient_machine__.makeEvaluator() assert isinstance(pass_evaluator, api.Evaluator) out_args = api.Arguments.createArguments(0) - feeder = DataFeeder(data_types, reader_dict) + feeder = DataFeeder(self.__data_types__, reader_dict) for pass_id in xrange(num_passes): event_handler(v2_event.BeginPass(pass_id)) pass_evaluator.start() updater.startPass() - for batch_id, data_batch in enumerate( - __data_reader_to_batch__(train_data_reader, batch_size, - topology)): + for batch_id, data_batch in enumerate(reader()): + pass_type = updater.startBatch(len(data_batch)) + self.__gradient_machine__.forwardBackward( + feeder(data_batch), out_args, pass_type) batch_evaluator.start() event_handler( v2_event.BeginIteration( pass_id=pass_id, batch_id=batch_id)) pass_type = updater.startBatch(len(data_batch)) - gm.forwardBackward(feeder(data_batch), out_args, pass_type) - gm.eval(pass_evaluator) - gm.eval(batch_evaluator) - for each_param in gm.getParameters(): + self.__gradient_machine__.forwardBackward( + feeder(data_batch), out_args, pass_type) + self.__gradient_machine__.eval(pass_evaluator) + self.__gradient_machine__.eval(batch_evaluator) + for each_param in self.__gradient_machine__.getParameters(): updater.update(each_param) # Get cost. We use numpy to calculate total cost for this batch. cost_vec = out_args.getSlotValue(0) @@ -143,59 +160,38 @@ class SGD(ITrainer): updater.finishPass() pass_evaluator.finish() event_handler(v2_event.EndPass(pass_id, evaluator=pass_evaluator)) - gm.finish() + self.__gradient_machine__.finish() + def default_reader_dict(self): + reader_dict = dict() + for i, tp in enumerate(self.__data_types__): + reader_dict[tp[0]] = i + return reader_dict -def __data_reader_to_batch__(reader, batch_size, topology): - """ - This function is not important, and will be removed when data refactored. - """ + def test(self, reader, reader_dict=None): + if reader_dict is None: + reader_dict = self.default_reader_dict() - def input_reorder(func): - for item in func(): - retv = [] - for __layer_name__ in topology.input_layer_names: - retv.append(item[__layer_name__]) - yield retv + feeder = DataFeeder(self.__data_types__, reader_dict) + evaluator = self.__gradient_machine__.makeEvaluator() + out_args = api.Arguments.createArguments(0) + evaluator.start() + for data_batch in reader(): + self.__gradient_machine__.forward( + feeder(data_batch), out_args, api.PASS_TEST) + self.__gradient_machine__.eval(evaluator) - return __generator_to_batch__(input_reorder(reader), batch_size=batch_size) + evaluator.finish() + return v2_event.TestResult(evaluator=evaluator) -def __generator_to_batch__(generator, batch_size): - """ - This function is not important, and will be removed when data refactored. - """ - ret_val = list() - for each_item in generator: - ret_val.append(each_item) - if len(ret_val) == batch_size: - yield ret_val - ret_val = list() - if len(ret_val) != 0: - yield ret_val - - -def __check_train_args__(train_data_reader, topology, parameters, - test_data_reader, event_handler, **kwargs): +def __check_train_args__(reader, event_handler, **kwargs): """ Check train function's argument types """ - if not callable(train_data_reader) or not isinstance(train_data_reader(), - collections.Iterator): - raise ValueError('train_data_reader should be a function, ' - 'which can return a iterator') - - if test_data_reader is not None: - if not callable(test_data_reader) or not isinstance( - test_data_reader(), collections.Iterator): - raise ValueError('test_data_reader should be a function, which can ' - 'return a iterator') - - if not isinstance(topology, ModelConfig): - raise ValueError('topology should be a model config') - - if not isinstance(parameters, v2_parameters.Parameters): - raise ValueError('parameters should be a parameter pool') + if not callable(reader) or not isinstance(reader(), collections.Iterator): + raise TypeError('train_data_reader should be a function, ' + 'which can return a iterator') if not callable(event_handler): - raise ValueError('event handler should be a function') + raise TypeError('event handler should be a function') diff --git a/python/setup.py.in b/python/setup.py.in index 1e1324eea825ab1945a38cb43eceec29a4ebb1a1..c7e46293b79fe0f8e92f7425247ddefef6777a62 100644 --- a/python/setup.py.in +++ b/python/setup.py.in @@ -5,7 +5,9 @@ packages=['paddle', 'paddle.trainer', 'paddle.trainer_config_helpers', 'paddle.utils', - 'paddle.v2'] + 'paddle.v2', + 'paddle.v2.dataset', + 'paddle.v2.reader'] setup(name='paddle', version='${PADDLE_VERSION}', @@ -13,5 +15,9 @@ setup(name='paddle', packages=packages, package_dir={ '': '${CMAKE_CURRENT_SOURCE_DIR}' - } + }, + install_requires = [ + 'scikit-learn>=0.18.0', + 'scipy>=0.18.0', + ] )