KylinA1

Word2Vec

3

It seems this code is written in python2. It may come across several problems. 1. for reading data part. using `text_words = (f.read(f.namelist()[0]).lower().decode('utf-8')).split()` or `text_words =tf.compat.as_str(f.read(f.namelist()[0])).split()` convert b string to...

Got a different huber_loss result

1

when I just use the huber loss in file "03_linreg_placeholder" in line 36, I get different results from the notes. w=-4.2249994 b=78.485054

userid and itemid start from 1

3

Hello, I just notice that you start count id from 1, which lead to one more dimension abuse in both user and item. For example, the number of user and...

what is the purpose to define the same two functions ?

Hi, I notice when you generate train batch, you didn't go through all the positive rating, just random pick some. Thus I suggest go through all the positive ratings and...