add train_on_cloud.py in fluid/language_model

add cloud

* revise se_resnext for imagenet classification

* Add the method to train a SE-ResNeXt model

The current code can successfully implement the result of SE-ResNeXt-50.

* Update ImageNet2012 URL

* update

* update readme

* Update readme with download URL

* add unzip

* update se_resnext.py with parallel_exe

* delete train function in se_resnext.py

* add eval.py and infer.py

* move cosine_decay from se_resnext.py to train.py

增加在云平台上运行的中文序列标注任务

* ready to coco_reader

* complete coco_reader.py & coco_train.py

* complete coco reader

* rename file

* use argparse instead of explicit assignment

* fix

* fix reader bug for some gray image in coco data

* ready to train coco

* fix bug in test()

* fix bug in test()

* change coco dataset to coco2017 dataset

* change dataset from coco to coco2017

* change learning rate

* fix bug in gt label (category id 2 label)

* fix bug in background label

* save model when train finished

* use coco map

* adding coco year version args: 2014 or 2017

* add coco dataset download, and README.md

* fix

* fix image truncted IOError, map version error

* add test config

* add eval.py for evaluate trained model 

* fix

* fix bug when cocoMAP

* updata READEME.md

* fix cocoMAP bug

* find strange with test_program = fluid.default_main_program().clone(for_test=True)

* add inference and visualize, awa, README.md

* upload infer&visual example image

* refine image

* refine

* fix bug after merge

* follow yapf

* follow comments

* fix bug after separate eval and eval_cocoMAP

* follow yapf

* follow comments

* follow yapf

* follow yapf

* ready to coco_reader

* complete coco_reader.py & coco_train.py

* complete coco reader

* rename file

* use argparse instead of explicit assignment

* fix

* fix reader bug for some gray image in coco data

* ready to train coco

* fix bug in test()

* fix bug in test()

* change coco dataset to coco2017 dataset

* change dataset from coco to coco2017

* change learning rate

* fix bug in gt label (category id 2 label)

* fix bug in background label

* save model when train finished

* use coco map

* adding coco year version args: 2014 or 2017

* add coco dataset download, and README.md

* fix

* fix image truncted IOError, map version error

* add test config

* add eval.py for evaluate trained model 

* fix

* fix bug when cocoMAP

* updata READEME.md

* fix cocoMAP bug

* find strange with test_program = fluid.default_main_program().clone(for_test=True)

* add inference and visualize, awa, README.md

* upload infer&visual example image

* refine image

* refine

* fix bug after merge

* follow yapf

* follow comments

* fix bug after separate eval and eval_cocoMAP

* follow yapf

* follow comments

* follow yapf

* follow yapf

Enable the parallel training of mobilenet

caffe2fluid: support Reshape and Scale convertion; fix bug in dropout; add more tools for comparing results (#880)

Replace feed_dict by feed

Change parallel_do to parallel_executor.

Fix params grad

* add a language model sample in fluid

* minor update

* fix code style

Add error rate measuring script

Add eval.py and fix bug for MobileNet-SSD.

Add the aishell example & release data