Update README.md

Author: bedapudi6788

README.md

@@ -1,40 +1,25 @@
-# Deep-Segmentation
-A sentence segmenter that actually works! Now for English, French and Italian.
+# DeepSegment: A sentence segmenter that actually works!
+# For the original implementation please use the "master" branch of this repo.
 
 The Demo is available at http://bpraneeth.com/projects
 
-The code and pre-trained models for "DeepCorrection 1: Sentence Segmentation of unpunctuated text." as explained in the medium posts at https://medium.com/@praneethbedapudi/deepcorrection-1-sentence-segmentation-of-unpunctuated-text-a1dbc0db4e98 and
-https://medium.com/@praneethbedapudi/deepsegment-2-0-multilingual-text-segmentation-with-vector-alignment-fd76ce62194f
-
-
-The pre-trained models are available at https://github.com/bedapudi6788/DeepSegment-Models
-
+# Installation:
+```
+pip install --upgrade deepsegment
+# please install tensorflow or tensorflow-gpu separately. Tested with tf and tf-gpu versions 1.8 to 2.0
+```
 
-# Requirements:
-seqtag
+# Usage:
 
 ```
-# if you are using gpu for prediction, please see https://stackoverflow.com/questions/34199233/how-to-prevent-tensorflow-from-allocating-the-totality-of-a-gpu-memory for restricting memory usage
-
 from deepsegment import DeepSegment
-# the config file can be found at in the pre-trained model zip. Change the model paths in the config file before loading. 
-# Since the complete glove embeddings are not needed for predictions, "glove_path" can be left empty in config file
-segmenter = DeepSegment('path_to_config')
+# The default language is 'en'
+segmenter = DeepSegment('en')
 segmenter.segment('I am Batman i live in gotham')
-['I am Batman', 'i live in gotham']
+# ['I am Batman', 'i live in gotham']
+
 ```
 
 # To Do:
-Add a sliding window for processing very long texts.
-
-Update the seqtag model to work with tf 2.0 (Change to tf.data may be).
-
-Update to add Indic languages.
-
-
-# Notes:
-Of all the sentence segmentation models I evaluated, without doubt deepsegment is the best in terms of accuracy in real word (bad punctuation, wrong punctuation)
-
-I trained flair's ner model on the same data and flair has better results but, it's miniscule (0.3% absolute accuracy increase).
-
-Since I want to keep using tf and keras for now, and since flair embeddings are not available for all the languages I want deepsegment to work on, I am going to keep using seqtag for this project.
+1. Add a sliding window for processing very long texts.
+2. Publish docker tf-serving image and deepsegment-client.