fastNLP V0.3.1

* 升级parser API和模型
* update docs: add new pages for tutorials
* upgrade CWS api download source
* add a new method for dataset field access
* add introduction for bert
* add more unit tests for api/processor
* remove unused test data. Add new test data.

* load pre-trained BERT weights from local binary
* add tests

* rename callback methods. Use fastai's notation.
* add a new callback method - on_valid_begin

* 添加两类Callback
* 完善Trainer对error的捕捉

* 添加测试：FieldArray的初始化

* 重构dtype的检测代码，在FieldArray的初始化和append两处，达到更好的代码复用
* 类型检测的责任完全落在FieldArray，DataSet与之配合
测试：
* 整理dtype相关的测试代码
* 给所有tutorial添加测试
其他：
* 完善一个完整的Conll dataset loader
* 升级POS tag model训练脚本

2. FieldArray默认使用AutoPadder, AutoPadder的行为与之前不使用padder是一致的的
3. 为了解决二维padding的问题，引入了EngChar2dPadder用于对character进行padding
4. 增加一份padding的tutorial。

* 将dataset.py中的assert改为raise error
* 给trainer添加try-except,捕捉EarlyStopError
* 优化trainer代码
* 给callbacks添加测试

* 改名: chinese_word_segment ---> Chinese_word_segmentation
* 改名: pos_tag_model ---> POS_tagging
* 添加4个对Batch的测试
* 删除无用的chinese_word_segment/run.py

* move used readers from reproduction to io/dataset_loader.py
(API shall not call anything from reproduction/)

- refine & fix Transformer Encoder
- refine & speed up biaffine parser

fix tutorial link

fastNLP V0.3.0

* pass all tests
* prepare CWS & POS API
* update tutorials
* add README.md in tutorials/ & api/

# Conflicts:
#	fastNLP/api/api.py
#	fastNLP/modules/encoder/variational_rnn.py

add parser API

# Conflicts solved:
#	fastNLP/io/dataset_loader.py
#	fastNLP/io/embed_loader.py
#	fastNLP/modules/aggregator/attention.py
#	fastNLP/modules/decoder/CRF.py

* update README