►
From YouTube: Applied ML weekly team meeting - July 22, 2021
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
B
A
Yeah
we
can
start
from
the
beginning
yeah.
So
five
out
of
five
apache
beam
pipelines
are
complete
and
can
be
run
locally,
and
since
we
have
the
google
cloud
project
like
I
can,
I
can
test
these
pipelines
with
the
google
dataflow
backend,
so
it
will
take
like
one
and
two
days.
I
I
think
that
starting
starting
from
monday,
I
can
I
can
help
or
I
can
start
another
task.
B
And
overall
great
progress
on
the
the
apache
beam
related
work.
A
Oh
yeah,
I
forgot
to
say
yes,
sorry,
that
right
now,
these
pipelines,
they
work
by
reading
data
from
local
files
or
from
files
stored
on
google's
storage,
but
for
the
pork
purposes
it
will
be
enough.
It's
exactly
the
same
functionality
that
we
that
that
I
had
before.
A
A
B
So
I
know
you
know
alexandria.
We
just
had
our
separate
one-on-ones
this
morning.
If
I
make
it
a
different
day
than
the
day
we
have
our
group
meeting,
but
that's
okay.
So,
but
so
we
talked
about
this.
I
talked
about
this
with
both
of
you
so
who's
doing.
B
What
on
the
proof
of
concept
the
poc
environment
what's
left
is
anyone's
stuck
with
not
having
the
access
or
knowledge
to
do
what
is
needed
and
I
think
miho
and
I
talked
about
like
do
we
want
a
list
of
who's
exactly
doing
what
and
what
are
the
steps
are
and
we
decided
we
came
no,
that's
probably
too
structured
for
right.
Now
we
can
just
kind
of
figure
it
out
as
we
go
and
be
less
structured.
C
C
There
is
a
cluster
with
one
node,
which
is
exactly
what
we
had
in
the
issue
to
begin
with,
and
I
think
that's
enough
and
we
have
one
virtual
machine
with
gpu
enabled.
If
we
don't
want
that,
then
we
can
easily
destroy
it
using
the
terraform
scripts.
I
terraforms
not
scripts
exactly,
but
there
are
from
files
I
wrote.
Currently
they
are
in
my
own
private
namespace,
since
I
created
an
a
project
in
the
ammo
orbs
group
under
gitlab.org,
but
it
turns
out.
C
C
Nice,
I
think,
stan
stan
who
has
owner
to
that
group
to
ammo
ops
group.
So
he
probably
needs
to
add
me
as
a
owner
to
the
project
itself,
so
I
can
configure
it
it's
in
my
own
private
namespace
for
now.
If
anyone
wants
access.
C
C
B
A
A
We
can,
we
can
reuse,
docker
images.
I
think.
C
A
I
forget
their
name
just
a
second.
C
B
Could
not
be
google
to
be
cloud
vendor
agnostic
on
it,
yeah.
A
So
we
can
check
confluence
right,
that's
the
name
of
the
company,
so
they
provide
kafka
and,
as
I
understood
we
can,
they
have
some
templates
to
install
it
on
top
of
the
google
cloud
platform,
so
I
will
I'll
put
a
link
inside
the
doc.
B
Yeah,
I
know
you
know-
we've
talked
about
this
before
it's
great
to
talk
about
it
here
to
kind
of
level
set.
Let's
also
continue
to
communicate
and
coordinate
on
this
async
via
the
issue
issue
comments
in
the
poc
issue.
I
think
it's
a
great
thing
to
keep
things
transparent
and
async,
so
we
can
move
quickly,
yeah,
okay,
I
don't
want
to
cut
off
conversation
now,
though,
that's
just
kind
of
a
general
comment.
Anything
else
we
want
to
discuss
on
this
today.
A
Yeah
I
see
that
we
have
a
male
ops
namespace
right
now.
Yes,
so
if
we
don't
mean
that
I
can
yeah
doesn't
mean
that
they
can
move
their
repository
with
the
pipelines.
There.
C
B
C
Access
to
that
namespace
is
stan.
Currently,
I
think,
with
access
level
higher
than
developer
and
stan
is
the
owner
of
that
namespace,
so
we'd
have
to
ask
stan
or
we'd
actually
have
to
file
an
issue
to
get
a
request
for
this.
I'm
not
sure
what
what's
the
procedure
here.
What's
the
correct
procedure
here.
C
B
B
C
I'm
probably
going
to
use
it
only
for
the
ci
ci
configuration
I'm
not
gonna
mess
with
stands.
Machine
learning
bought
to
label
issues
accordingly,
because
the
only
project
that's
currently
there
is
this
tanukistan.
B
Okay,
all
right
so
mia.
I
know
you
spent
a
lot
of
your
time
on
maintainership
responsibilities.
You
know
you're
you're,
the
backend
maintainer
and
database
maintainer
you're
pursuing
both,
which
is
extremely
loudable,
because
we
need
more
of
those.
We
need
more
people
doing
that
at
gitlab
to
keep
things
moving
with
murder
requests
that
require
maintainer
reviews
of
both
types
across
the
entire
company.
C
B
C
I
am
unable
to
say
right
now.
Well,
basically,
the
gist
is
that
I
kind
of
put
the
back
and
maintainership
on
hold,
because
we
have
much
more
of
these
than
we
have
of
the
database
maintainers,
and
I
created
a
issue
to
actually
get
myself
into
the
maintainers
like
to
finally
graduate
from
being
a
training
maintainer.
C
But
it's
still
under
discussion,
mostly
because
of.
C
Let's
say
the
the
list
of
the
issues
I
read
as
a
database,
training
maintainer
wasn't
complete,
so
I
had
to
like
fill
in
the
gaps
and
there
is
a
merge
request
open
and
I
ping
the
database
maintainers
today
to
get
the
discussion
underway.
C
B
Thanks
for
the
updates
on
that,
and
again
thanks
for
volunteering
to
work
on
becoming
a
maintainer
in
two
different
areas,
which
is
great,
I'm
just
curious
about
that.
Eighty
percent
would
not
be
a
problem
right
if
you
were
gonna
be
eighty
percent.
It
would
just
it's
just
really
awareness
so
good.
C
It's
it's.
It
kind
of
varies
because
like
if
we
have
a
rapid
action
going
on,
then
you
get
a
lot
of
these
and
they
need
to
be
like
performed
really
quickly,
but
otherwise
I
just
get
like
two
or
three
merge
requests
per
day
from
both
back
end
and
database.
So
it's
like
six
merge
requests
per
day
which
takes
takes,
doesn't
take
10
minutes,
but
it's
like
more
than
one
more
than
one
hour
to
review
all
of
them
and
to
perform
the
review
on
an
acceptable
level.
C
B
Make
sense
great
doug
remind
me
you're
your
maintainer,
with
on
which
portions
with
which
focuses
for
focus.
D
I
am
a
reviewer
of
the
front
end
on
gitlab
and
a
reviewer
for
database
on
git
lab.
I
was
a
trainee
maintainer
for
the
database,
but
I
opted
out
of
that,
as
is
too
much
to
to
try
to
focus
on.
B
Well
same
of
the
things
to
me
how,
thanks
to
you
for
being
a
maintainer
of
various
things?
It's
I
don't
know
if
it's
often
a
thankless
job,
but
maybe
sometimes
it
is
so.
Thank
you.
C
D
I
just
want
to
say
I'm
here
to
help
out
wherever
I
can
wherever
I
can.
It
may
take
me
a
little
while
to
figure
out
where
that
is
if
at
all,
but
I'm
here
and
I
know
I
I
scheduled
some
time
to
chat
mihael-
is
that
how
you
say
your
name.
D
I've
been
called
dough,
so
if
it
all
works
next
week
and
maybe
alexander
I'll
try
to
catch
up
with
you
as
well.