Fix FleetWrapper:
1. fix shrink dense: just scale show
2. add datanorm scale: divide datanorm's gradient by batch_size

* change INT8 to template so that checking dst_dt with if-else could be removed. CI will be enabled after fixing reviews

* reverse user_residual_memory_p and user_bias_memory_p declaration scope
test=develop

* split test_dist_se_resnext.py into 4 testcases

* extend matmul op to support multiple head multiplication

With the support of multiple head, the multiplication of two big matrixes is
split into multiplication of several (head_number) small matrixes. e.g. if
Mat A is [3, 24] and Mat B is [24, 4], when multiple A and B with head_number
as 4, Mat A will be split as 4 matrix of [3, 6] and Mat B will be 4 matrix of
[6, 4]. The result of final matrix will be 4 matrix of [3, 4], i.e. [3, 16].

* Make lod reset op support for append lod level.

* Fix API.spec
test=develop

* Fix unitest.
test=develop

* Add python api for lod append.
test=develop

* Fix API.spec
test=develop

* Fix format of doc.
test=develop

* Fix unitest.
test=develop

* Fix doc.
test=develop

test=develop

* prun backward ops
test=develop

Modify auc doc. Add output variable description, previously was the scalar type, now changed to the tuple type.test=develop (#18771)

* update paddle-trt for:
    1. fix bug: when batch > 2, core in split plugin.
    2. add leaky_relu trt5.0 support (yolov3 from 65ms to 42ms.)
    3. add new attr to dropout.
    4. shuffle channel, swish, relu6 support
    test=develop

* 1. fix ci
test=develop

The change includes 2 things:

1. save delta model and shrink table are control by the same parameter before, now add delete_after_unseen_days to control shrink table.
2. value in sparse table has no slot before, now add slot in sparse table, and add DownpureCtrAccessor to support the new meta.
test=develop

* modify install GPU97

* modify install GPU97

test=develop

- compileation fix

- Yet another compilation fix

- Even yet another compilation fix

- Surprise! Again compilation fix

- lint fixes

test=develop

- Fix to workspace acquire of LRN

test=develop

- Fix to hash of BWD LRN

test=develop

- fix to lrn BWD PD acquire

test=develop

- Fixing LRN PD creation

test=develop

- cosmetic fix in comment

test=develop

- Fixes after review

test=develop

test=develop

（1）support patch data （merge slots of instances of same line id, modify dense layer which
changes its size）
（2）add fleet load_one_table interface, support load from paddle model and load from pslib model
（3）fix push sparse bug which cause push sparse cost more time（about 10% in my testcase）
（4）when some slots are not in one of your network (join/update, etc.)，data feed、collect label info、push/pull sparse will skip these slots， instead of throw error.
（5）add more debug info in TrainFilesWithProfiler

* support sparse gradients
test=develop

* rewrite the conv_op using cudnn_conv_helper

* add workspace limit for v7 test=develop

* fix test=develop

* add half float test=develop

* fix test=develop

* fix test=develop

* revise code style test=develop

* fix test=develop

* supports distributed classification training
* update API.spec
* fix evenly division in python3
* change "index_range" to "index_num" in shard_index operator
test=document_preview
test=develop

add parameter server launch so that a user can easily launch parameter server

The change includes 3 things:

1. Set CPU_NUM to 1 in the tests because the ParallelExecutor will print warning that CPU_NUM is not set and use default 1.

2. Old tests compare two RNNs, hand written simple RNN and same RNN built by Paddle, but initialized RNN weights in numpy random and Paddle random separately. Fixed it by setting weights and bias values.

3. Also set numpy random seed in the tests. Now the two RNNs diff can be smaller (rtol from 0.1, 0.2 to. 0.01) in the tests.

test=develop

test=develop

test=develop

test=develop

do some odd jobs, test=develop

* split different comm method for mnist distributed training

add license

* fix clip_by_norm doc, test=develop

Test PaddingRNN on V100 GPU device.

Test configuration: large model, padding mode (which is the mode using recurrentOp), one GPU.
                   
GPU memory (MiB):   6414 (this PR)     vs   6837 (without this PR)
Speed (steps/s):         10.28 (this PR)    vs    9.89 (without this PR)
 

test=develop

test=develop

* add check of executor, test=develop

Optimize the content of error reporting information, print error code and official document web sites (#18671)

 optimize the error reporting information of cuda related API
 index on develop: 130ac177 Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop

* feature/auto_growth_allocator, test=develop

* add unittest of AlignedAllocator, test=develop

* try to turn on auto_growth to test on CI, test=develop

* fix segmentation fault in mixed_vector.h, test=develop

* add unittests, test=develop

* hash_op support int64 hash_size
* add corresponding UT