Count consecutive rows with same value for a column in a database using django?

Let my model Notification model be:

Class Notification(models.Model):
    user = models.ForeignKey(
        settings.AUTH_USER_MODEL,
        related_name='notifications',
        on_delete=models.CASCADE)
    notif = models.CharField(max_length=255)
    date_created = models.DateTimeField(auto_now_add=True)

The database rows are as:

id    |   user    |     notif       |
------------------------------------
1       john doe     liked your pic 
2       john doe     commented on your pic
3       james        liked your pic
4       james        commented on your pic
5       john doe     pinged you
6       john doe     showed interest
7       john doe     is busy

Basically, I am trying to join consecutive rows by user

The above notifications then are to be shown as:

2 notifications from john doe
2 notification from james
3 notofications from john doe

instead of

5 notifications from john doe
2 notification from james

or

1 notifications from john doe
1 notifications from john doe
1 notification from james
1 notification from james
1 notofications from john doe
1 notofications from john doe
1 notofications from john doe

In order to achieve this, we are looking for a dictionary like:

{
"john doe": ["notif1", "notif2"],
"james": ["notif1", "notif2"],
"john doe": ["notif1", "notif2", "notif3"] #duplicate key.
}

But, that's not possible as duplicate keys are not allowed. Hence I am going with array of tuples instead.

[
  ('john doe', ['notif1', 'notif2']),
  ('james', ['notif1', 'notif2']),
  ('john doe', ['notif1', 'notif2', 'notif3']),
]

So, we first sort the Notifications by date_created. Then we use itertools.groupby to make groups per user.

from itertools import groupby
from operator import attrgetter

qs = Notification.objects.select_related('user').order_by('date_created')
notifs= [(u, list(nf)) for (u, nf) in groupby(qs, attrgetter('user'))]

You have everything sorted as needed in notifs.

Done!