提交 7e145b7c 编写于 作者: Q Qiao Longfei

optimize test_async_ssa_graph_executor_mnist

上级 9da96aba
...@@ -15,13 +15,13 @@ ...@@ -15,13 +15,13 @@
from __future__ import print_function from __future__ import print_function
import os import os
from PIL import Image import unittest
import numpy import numpy
import paddle import paddle
import paddle.fluid as fluid import paddle.fluid as fluid
BATCH_SIZE = 64 BATCH_SIZE = 64
PASS_NUM = 5
def loss_net(hidden, label): def loss_net(hidden, label):
...@@ -51,11 +51,9 @@ def convolutional_neural_network(img, label): ...@@ -51,11 +51,9 @@ def convolutional_neural_network(img, label):
return loss_net(conv_pool_2, label) return loss_net(conv_pool_2, label)
def train(use_cuda, def train(use_cuda, thread_num, cpu_num):
save_dirname=None,
model_filename=None,
params_filename=None):
if use_cuda and not fluid.core.is_compiled_with_cuda(): if use_cuda and not fluid.core.is_compiled_with_cuda():
print("paddle is not compiled with cuda, exit!")
return return
img = fluid.layers.data(name='img', shape=[1, 28, 28], dtype='float32') img = fluid.layers.data(name='img', shape=[1, 28, 28], dtype='float32')
...@@ -84,8 +82,6 @@ def train(use_cuda, ...@@ -84,8 +82,6 @@ def train(use_cuda,
place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace() place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
exe = fluid.Executor(place)
train_reader = paddle.batch( train_reader = paddle.batch(
paddle.reader.shuffle( paddle.reader.shuffle(
paddle.dataset.mnist.train(), buf_size=500), paddle.dataset.mnist.train(), buf_size=500),
...@@ -94,24 +90,22 @@ def train(use_cuda, ...@@ -94,24 +90,22 @@ def train(use_cuda,
paddle.dataset.mnist.test(), batch_size=BATCH_SIZE) paddle.dataset.mnist.test(), batch_size=BATCH_SIZE)
feeder = fluid.DataFeeder(feed_list=[img, label], place=place) feeder = fluid.DataFeeder(feed_list=[img, label], place=place)
exe = fluid.Executor(place)
exe.run(fluid.default_startup_program()) exe.run(fluid.default_startup_program())
main_program = fluid.default_main_program()
exec_strategy = fluid.ExecutionStrategy() os.environ['CPU_NUM'] = str(cpu_num)
build_strategy = fluid.BuildStrategy()
cpu_num = int(os.environ.get('CPU_NUM'))
thread_num = int(os.getenv("NUM_THREADS"))
print("cpu_num:" + str(cpu_num)) print("cpu_num:" + str(cpu_num))
print("thread_num:" + str(thread_num)) print("thread_num:" + str(thread_num))
build_strategy.async_mode = True build_strategy = fluid.BuildStrategy()
build_strategy.async_mode = True # enable async mode
exec_strategy = fluid.ExecutionStrategy()
exec_strategy.num_threads = thread_num exec_strategy.num_threads = thread_num
exec_strategy.num_iteration_per_drop_scope = 1 exec_strategy.num_iteration_per_run = 2
exec_strategy.num_iteration_per_run = 10
main_program = fluid.default_main_program()
pe = fluid.ParallelExecutor( pe = fluid.ParallelExecutor(
use_cuda=False, use_cuda=False,
loss_name=avg_loss.name, loss_name=avg_loss.name,
...@@ -119,96 +113,26 @@ def train(use_cuda, ...@@ -119,96 +113,26 @@ def train(use_cuda,
build_strategy=build_strategy, build_strategy=build_strategy,
exec_strategy=exec_strategy) exec_strategy=exec_strategy)
lists = []
step = 0 step = 0
for epoch_id in range(PASS_NUM): for step_id, data in enumerate(train_reader()):
for step_id, data in enumerate(train_reader()): loss_val = pe.run(feed=feeder.feed(data), fetch_list=[avg_loss.name])
loss_val, acc_val = pe.run(feed=feeder.feed(data), loss_val = numpy.mean(loss_val)
fetch_list=[avg_loss.name, acc.name]) if step % 100 == 0:
loss_val = numpy.mean(loss_val) print("Batch %d, Cost %f" % (step, loss_val))
acc_val = numpy.mean(acc_val) step += 1
if step % 100 == 0: # test for epoch
print("Pass %d, Batch %d, Cost %f" % (epoch_id, step, loss_val)) avg_loss_val, acc_val = train_test(
step += 1 train_test_program=test_program,
# test for epoch train_test_reader=test_reader,
avg_loss_val, acc_val = train_test( train_test_feed=feeder)
train_test_program=test_program,
train_test_reader=test_reader, print("Test: avg_cost: %s, acc: %s" % (avg_loss_val, acc_val))
train_test_feed=feeder)
print("Test with Epoch %d, avg_cost: %s, acc: %s" % class TestAsyncSSAGraphExecutor(unittest.TestCase):
(epoch_id, avg_loss_val, acc_val)) def test_check_async_ssa_exe_train(self):
lists.append((epoch_id, avg_loss_val, acc_val)) train(use_cuda=False, thread_num=2, cpu_num=2)
if save_dirname is not None:
fluid.io.save_inference_model(
save_dirname, ["img"], [prediction],
exe,
model_filename=model_filename,
params_filename=params_filename)
# find the best pass
best = sorted(lists, key=lambda list: float(list[1]))[0]
print('Best pass is %s, testing Avgcost is %s' % (best[0], best[1]))
print('The classification accuracy is %.2f%%' % (float(best[2]) * 100))
def infer(use_cuda,
save_dirname=None,
model_filename=None,
params_filename=None):
if save_dirname is None:
return
place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
exe = fluid.Executor(place)
def load_image(file): if __name__ == "__main__":
im = Image.open(file).convert('L') unittest.main()
im = im.resize((28, 28), Image.ANTIALIAS)
im = numpy.array(im).reshape(1, 1, 28, 28).astype(numpy.float32)
im = im / 255.0 * 2.0 - 1.0
return im
cur_dir = os.path.dirname(os.path.realpath(__file__))
tensor_img = load_image(cur_dir + '/image/infer_3.png')
inference_scope = fluid.core.Scope()
with fluid.scope_guard(inference_scope):
# Use fluid.io.load_inference_model to obtain the inference program desc,
# the feed_target_names (the names of variables that will be feeded
# data using feed operators), and the fetch_targets (variables that
# we want to obtain data from using fetch operators).
[inference_program, feed_target_names,
fetch_targets] = fluid.io.load_inference_model(
save_dirname, exe, model_filename, params_filename)
# Construct feed as a dictionary of {feed_target_name: feed_target_data}
# and results will contain a list of data corresponding to fetch_targets.
results = exe.run(inference_program,
feed={feed_target_names[0]: tensor_img},
fetch_list=fetch_targets)
lab = numpy.argsort(results)
print("Inference result of image/infer_3.png is: %d" % lab[0][0][-1])
def main(use_cuda):
model_filename = None
params_filename = None
save_dirname = "recognize_digits" + ".inference.model"
# call train() with is_local argument to run distributed train
train(
use_cuda=use_cuda,
save_dirname=save_dirname,
model_filename=model_filename,
params_filename=params_filename)
infer(
use_cuda=use_cuda,
save_dirname=save_dirname,
model_filename=model_filename,
params_filename=params_filename)
if __name__ == '__main__':
use_cuda = False
main(use_cuda=use_cuda)
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册