zhengwu

I haven't implement ......., I try to add soon

@mymoongit 昨晚分词方法实现了加载用户词典，目前还在dev1.1 分之

用户词典的格式：`词\t权重` 的格式， 模型结果的权重默认5.0 , 词典中的权重越大出现的可能性越大

线上和发布的模型是同一个

```python if flags_obj.export_dir is not None: # Exports a saved model for the given classifier. dtype=flags_core.get_tf_dtype(flags_obj) input_receiver_fn = export.build_tensor_serving_input_receiver_fn( shape, batch_size=flags_obj.batch_size, dtype=dtype) classifier.export_savedmodel(flags_obj.export_dir, input_receiver_fn) ``` you can set export_dir at...

I load model from ```sys.prefix``` you can copy the model file to /usr/local/Cellar/python3/3.6.3/Frameworks/Python.framework/Versions/3.6/fool/ I will fix this next version ! thanks !!

@liu-nlper you can copy ```fool ``` from ```/usr/local``` to ```/usr/ ```

@liu-nlper 谢谢建议，正在努力中

@ameerhamza6733 Sorry, I know little about tflite

@ShuGao0810 谢谢你的反馈，现在的词典在分词的时候是有效的，analysis 不支持，稍后修改