* sequnce mask support max length tensor input; test=develop

* add rnn_impl.py; test=develop

* add basic gru lstm unittest; test=develop

* fix api spec; test=develop

* fix sequence_mask op bug;
test=develop
test=document_preview

* change +-*x to elmentwise_op; test=develop

* add mkl flag; test=develop

* fix rnn impl bug; test=develop

* update api spec; test=develop

* fix doc bug; test=develop

* fix lstm bugs; test=develop

test=release/1.5

* Fix the GetExpectedKernelType of add_position_encoding_op.

* Fix the doc of lstm_unit outputs in nn.py.

test=release/1.5

* fix default value of fluid.memory_optimize
test=release/1.5

* Use TensorCopySync for sequence_unpad op

* Fix the tensor memory alloc bug

test=release/1.5

add dygraph api doc for fluidDoc and api spec

* fix slice op bug; test=develop

* fix variabel test bug; test=develop

* remove slice while true; test=develop

test=release/1.5

Repair error prompt: Users are prompted to check whether the model or parameter files are damaged when loading parameters are wrong.

* cherry pick 18000, test=release/1.5

* cherry-pick the inference update for win test=develop

* test=develop

* update some op doc, test=release/1.5

* Update backward appending stragety to support double backward and fix some bug. (#18104)

* Update backward.py:
     - If there is no input grad var in all outputs of previous ops, do not append this op into graph.
     - Only apply this stragety when double backward.
* Update some double backward op.
* Update sum_op to judge whether a tensor is empty by numel or IsInitialized().

* add mkldnn Int8v2 slim doc (#17909)

* Change int8v2 CAPI unit test name and add log in the prediction stage (#18200)

* fix issue 18111;test=develop

* fix timer;test=develop

* refine code;test=develop

* test=release/1.5

* add multi process reader
test=release/1.5

* fix save/load in fleet (#17675)

* fix save/load in Fleet
* add UT framework of Fleet (#18058)

* add paddle cloud role maker for customized usage, note this is only for industrial users that have cloud environment pre-configuration (#18121)

add paddle cloud role maker for specific cloud usage. This pr will simplifies user's configuration in distributed training.

* assign role_maker before use (#18137)

* cherry pick refine core warning message (#18063)

* cherry pick core replace x86cpu with py cpuinfo (#18151)

test=release/1.5

Cherry pick retinanet_target_assign_op(#17893), sigmoid_focal_loss_op(#17895) and retinanet_detection_output_op(#17896) for supporting retinanet (#18141)

* test=release/1.5
Fix conflicts in test_layers.py when adding target assign operator for supporting retinanet. Cherry pick #17893

* test=release/1.5
Add sigmoid focal loss operator for supporting retinanet. Cherry pick #17895

* test=release/1.5
Add detection output operator for supporting retinanet. Cherry pick #17896

* test=release/1.5
fix wrong code style in test_layers.py when cherry pick retinanet_target_assign #17893

* test=release/1.5
Fix type error of std::pow in sigmoid_focal_loss. Cherry pick #17895

* update execution_strategy option default value
test=release/1.5

* fix doc error
test=release/1.5

* remove nccl dep when the number of GPU is 1
test=develop

* use multi card run syncBN
test=release/1.5

Add trainer_desc proto DEPS to solve CI random fail.

Add cascade rcnn support.

Cherry pick #18077 and  #18111 unify FP32 vs. INT8 comparison tests output, reuse C-API INT8 unit test application (#18145)

* unify FP32 vs. INT8 comparison tests output (#18111)

test=release/1.5

* reuse C-API INT8 unit test application (#18077)

test=release/1.5

* fix dygraph mem leak, test=release/1.5

* polish msg, test=release/1.5

cherry pick for deform roi pooling

cherry-pick for (https://github.com/PaddlePaddle/Paddle/pull/17402)

Add Pipeline Concurrency Train Mode:
- Cpp: pipeline_trainer & section_worker
- Python: PipelineOptimizer
- Add a new data_feed type: PrivateInstantDataFeed
- Add a test demo of pipeline trainer and the test model is gnn
- Do not support win32 now

* fix multithreading issue

test=develop

* rview fixes

test=develop

* reivew fix: omp->cpu, infernce_api.cc->pybind.cc

test=release/1.5

* fix py_reader iterable bug, test=release/1.5

* move data from buffered_reader,test=release/1.5

* update CPU_NUM config
test=develop

test=release/1.5
cherry-pick from #17952
The scatter op has a calc bug when the indices has same index, the scatter op use overwrite mode to calculate the same index, fix this bug by using the accumulate mode to calculate the same index.At the same time, the gather op has the same bug when the op calc the grad. And we use the lib of open-blas and eigen to optimize the time cost in accumulate mode.

Added unit test for QAT FP32 & INT8 comparison (#17814)
Disable MKLDNN FC in Resnet50 test (#18030)

test=release/1.5

test=release/1.5

Fix #17955
Fix bug in sequence_unpad op, when allocate the output memory do not match actual memory, check memory failed. Fix this bug by allocating the output memeory in correct code position.

test=release/1.5

* support avx and noavx core
* add catch and give some log
* fix build
* add missing package
* fix pybind name
* fix import error
* conbime noavx core
* add requirements
* fix unkown message
* fix api spec
* refine and clean
* update
* follow comments
* refine scripts

* fix: when use the load model from memory mode, the RAM occupy is high

test=release/1.5

fix merge conflict of 'Remove attribute in Allocator::Allocate' and elementwise_add_mkldnn_op (#17949)

test=develop