2. 调整各种Embedding的实现，使得确保来自dev和test的未发现词使用unk的表示
3. 在Embedding中增加dropout_word的选项，使得可以随机drop掉词语
4. 以及其它若干小的bug

2. 重大更新，支持ElmoEmbedding, BertEmbedding

- add Option

1. 完善 trainer 和 tester 部分的文档
2. 研究注释样例与测试

2.Vocabulary.from_dataset(), index_dataset()提供报错时的vocab顺序
3.embedloader在embed读取时遇到不规则的数据跳过这一行.

2. vocabulary增加from_dataset(), index_dataset()函数。避免需要多行写index dataset的问题。
3. 在utils中新增一个cache_result()修饰器，用于cache函数的返回值。
4. callback中新增update_every属性

* 修复fast_param_mapping的一个bug
* Trainer添加自动创建save目录
* Vocabulary的打印，显示内容

* update fastNLP.core.rst
* refine all docstrings in core/

* remove unused codes in metrics.py
* add tests for DataSet
* add tests for FieldArray
* add tests for metrics.py
* fix predictor, add tests for predictor
* fix bucket sampler, add tests for bucket sampler

* 增量添加单词到词典中
* lazy update: 当用到词典的时候才重新build
* 当新添加的词导致词典大小超出限制时，打印一个warning

Update Vocabulary:
* More words can be added after the building.
* Lazy update: rebuild automatically when vocab is used.
* print warning when max size is reached

* remove and fix other unit tests
* add more code comments

* rename preprocess.py into utils.py, because nothing about preprocess in it
* anything in loader/ and saver/ is moved directly into io/
* corresponding unit tests are moved to /test/io
* delete fastnlp.py, because we have new and better APIs
* rename Biaffine_parser/run_test.py to Biaffine_parser/main.py; Otherwise, test will fail.
* A looooooooooot of ancient codes to be refined...........

- add character vocab in preprocessor
- add dataset loader for language model dataset
- other minor adjustments
- preserve only a little example data for language model