1
00:00:00,470 --> 00:00:04,610
In the last video we saw LSTMs and
how they work with cell state

2
00:00:04,610 --> 00:00:08,710
to help keep context in a way that
helps with understanding language.

3
00:00:08,710 --> 00:00:12,200
Well, words that aren't immediate
neighbors can affect each other's context.

4
00:00:13,400 --> 00:00:18,250
In this video, you'll see some other
options of RNN including convolutions,

5
00:00:18,250 --> 00:00:21,700
Gated Recurrent Units also called GRUs,
and

6
00:00:21,700 --> 00:00:24,690
more on how you can write the code for
them.

7
00:00:24,690 --> 00:00:27,580
You'll investigate the impact
that they have on training.

8
00:00:27,580 --> 00:00:29,950
I'm not going to go into
depth on how they work,

9
00:00:29,950 --> 00:00:33,570
and that information is available in the
deep learning specialization from Andrew.

10
00:00:33,570 --> 00:00:34,470
So do check it out there.