test=develop

* fix quantization doc 

Update PaddleSlim LightNAS

* update

* update

* update

* fix run_ernie.sh

* Update README.md

fix path of dir of files in windows

* PaddleRec xdeepfm add local distributed training

* PaddleRec deepfm dcn add local distributed training

* change PaddleRec deepfm README.md

* add embedding is_sparse flag in argparser

* fix deepfm failed in ver 1.5.2 when use transpiler

* fix xdeepfm failed in ver 1.5.2 when use transpiler

* fix some bugs in dnn deepfm and dcn

* fix embedding size in deepfm

* MRQA2019-D-NET

* delete_data_file

* add distillation demo and doc

* add classification quantization demo

* Disable optimizer fusion for RCNN models

waiting for upstream fix

* Disable completely

* Add support for mixed precision training

* Disable `fuse_all_reduce_ops` when training in mixed precision

waiting for upstream fix

* Add pruning demo for classification

* set encoding=utf8 for open

* set encoding=utf8 for open

* Initial support for distributed training

* Housekeep on rank 0 only

* Conform to models convention

add license for nlp models

* fix model save

* fix doc

* update TensorRT 1.5.2 inference time

* add EfficientNet configuration in README

* add notes for the ported weight of EfficientNet

* fix_infer_pix2pixreader

Add EfficientNet

Fix transformer inference when using ParallelExecutor by setting ExecutionStrategy to clear tensor array after each iteration. (#3374)

* minor fix

* add getting ops from program

* add obtaining op params from paddle program

* add code reusage of light_nas_space

* add comments to get_ops_from_program

* add comments to get_ops_from_program

* fixed bugs

* add get_all_ops

* add get_op_latency

* add platform parameter for get_op_latency

* add get_all_ops

* add get ops from shortcut and se

* bug fixed

* fixed bugs

* add refined code

* bugs fixed

* useless test codes and op files removed

* modifying 7 to length of searched bottlenecks

* bugs fixed

* add softmax op in get_ops_from_program

* remove resize ops in get_ops_from_program

* add get_op_latency from paddle-mobile and modify getting fc op params

* add fc op params

* add batch norm in light_nas_space

* bugs fixed

* update light_nas_space: add get_model_latency

* bug fixed

* bug fixed

* bug fixed

* update get_latency_lookup_table

* add bias of fc layer in get_all_ops

* add docs for light-nas with latency

* update doc for light-nas

* add first part of method descriptions for light-nas based on android and ios

* add descriptions for lightnas basd on android and ios in demo and usage and modify get_latency_lookup_table

* fix errors in usage.md

* fix errors among the contents of android system in usage.md

* fix errors among the contents of ios system in usage.md

* add descriptions in demo.md about lightnas based on android and ios

* add auxillary methods for generating bin files and .a files for op latency test

* add links for latency estimation tools

* minor fix in usage.md

* minor errors fixed in usage.md

* update LightNAS usage

* update docs

* update get_latency_lookup_table.

* add information for get latency lookup table

* add current searched models and their corresponding accuracies for android and ios systems

* bugs in fc_op_prams are fixed in get_ops_from_program.py test=develop

* update docs

* update docs

* update docs

* update PaddleSlim results

* add more lr_strategy

* update docs

* update PaddleSlim results

* update docs

* minor update

* minor update

* modify get_latency_lookup_table.py to make it compatible for python2 and python3; modify usage.md to update the information of PaddleMobile to Paddle-Lite

* minor update to be compatible for py3

* modify get_latency_lookup_table.py to make it compatible for python2 and python3

* convert float number in get_all_ops to int number in order to make it compatible for both python2 and python3

* remove print in get_latency_lookup_table.py

* update docs

* fixed bugs in get_all_ops

* fix merge conflicts

* fix merge conflicts