notebook.community

Edit and run



In [12]:

    
import warnings
warnings.filterwarnings('ignore')
from ds.models.model import Model
from ds.models.Explore import TextInterpret



In [13]:

    
sentiment = Model()
sentiment.load('sentiment')









    



transferring from s3



In [14]:

    
TextInterpret('i love this product', sentiment.predict)









    





        
        
        
        
        
        






    Out[14]:





<lime.explanation.Explanation at 0x7f6d27442350>



In [17]:

    
sentiment.comments.display_history()









    Out[17]:






  
    
      
      ID
      comment
      name
      ts
    
  
  
    
      0
      0
      Created Thread
      SYSTEM
      2017-01-06 23:37:37.551286



In [9]:

    
new_comment = """
Looks decent, remind me again though what the class definitions are?
"""
sentiment.comments.add_comment(new_comment, 'Pramit')



In [18]:

    
sentiment.save()









    



saving to s3



In [24]:

    
from ds import datasets
reviews = datasets.load('movie_reviews')
sentence = " ".join(reviews.sents()[0])
TextInterpret(sentence, sentiment.predict)









    





        
        
        
        
        
        






    Out[24]:





<lime.explanation.Explanation at 0x7f6d2744e590>

What else we've learned we need as candidates for feature additions:

1: Additional types of explanations/interpretations

2: Plotting/Textual explanations

3: Implemetation details (open source/ in house / combo)

4: Platform integration: abstractions around models (particularly within platform)



In [ ]: