系統設計者的話：學英文根本不用背，你講中文，你有在背中文嗎？學英文也是一樣的道理，那是自然而然就會的東西，用背的方式最後肯定都是不行的。 更多...

最好的學習就是有英文的父母，不斷的跟你說話、耐心的教導你，時間長了，你也就慢慢就融入其中了。

但是我們沒有英文父母怎麼辦？看影片學英文就學的起來嗎？學不起來。

我們從小到大，也看過了不少英文電影，至少也看過1、200部也跑不掉，但是為什麼看了這麼多的英文電影，英文還是很爛呢？

主要的原因是我們沒有從基礎性的強化學習。舉個例子，西方的幼兒出生時，他至少聽了各種大人講英文至少聽了2年，才慢慢開始說英文、與大人互動。

而且都是一字一句慢慢的說、慢慢的教、漸進式的、慢慢的、越來越習慣、越來越聽的懂、越來越會說的過程。

那麼我們看的影片、電影過程中，都是大人成熟型的講話速度、方式、語調，我們怎麼可能跟的上？

一旦跟不上，思想中就會產生覺的好難哦、我不行種種念頭，那麼產生了這種潛在式的排斥心理作用，那麼你就更難學的會了。


所以我們必須打好基礎、深厚的基礎，萬丈高樓平地起，唯有打下深厚的基礎，後面你走的路才會越來越順。

程式界裏有句話，程式債欠的越多，最後就只能放棄。學英文也是一樣，英文單字債欠的越多，你就直接上來文法、句型、語調，最後也只能放棄。

基於上述的思維，我打造了這套系統，從最基礎的單字學習起，不斷的強化、加深你對這個單字的聽力，透過各式各樣的影片、各式各樣的人、各種不同的場景。

就只是為了訓練一個單字的聽力，無數的人跟你講同一個單字，講個20遍、50篇、100遍(且不重覆)，我就不信學不起來。


這套系統做出來後，我自己在使用的過程中，確實真的能聽出來他們講的這個單字，而且他們講的速度再快也能聽的出來。

能聽的出來，就知道他講的大概意思；能夠知道他講的大概意思，就能夠用幾個簡單的單字回應；能夠回應，也就能夠對話；當能夠對話，就能夠再修正文法、句型、語調。


我的目地就是希望自己至少能聽的懂人家在說什麼，用英文跟人家對話，大家共同努力、學習。若有其它想法或對這套系統設計的建議，也可以與我聯絡：ichich2013@gmail.com

Serving Models in Production with TensorFlow Serving (TensorFlow Dev Summit 2017)
觀看次數：395
i'm
noah
fido
and
i'm
going
to
share
couldn't
retrain
the
model
right
and
now
those
so
here's
the
agenda
for
today's
say
you're
showing
video
recommendations
latency
and
this
differs
quite
a
bit
pipelines
you
might
run
your
users
don't
want
to
wait
for
the
recommendations
consistently
fast
recommendations
to
your
users
and
then
of
your
requests
arrive
asynchronously
internally
to
build
our
binaries
and
containers
and
tutorials
and
code
labs
you
want
to
adopt
tensorflow
servable
the
components
of
the
libraries
imagine
they're
also
using
the
ployed
of
abstractions
with
AP
eyes
and
the
yellow
boxes
are
plugins
into
those
abstractions
so
let's
imagine
that
this
recommendations
and
you
have
version
one
provide
better
recommendations
to
users
tensorflow
save
model
and
it's
important
almost
all
scenarios
if
you're
serving
something
like
video
recommendations
to
pipeline
that's
maybe
annotating
these
have
a
little
bit
of
unavailability
in
first
and
foremost
is
they're
optimized
you
know
rath
count
accesses
to
your
of
resizing
we
have
these
nice
plug-in
tutorials
but
for
those
of
you
who've
internally
and
externally
is
that
you
out-of-the-box
and
so
this
should
let
intentionally
short
the
binaries
are
flags
and
configuration
and
monitoring
google's
high-performance
open-source
low-level
tensor
oriented
predict
API
so
classification
interfaces
the
api's
are
optimize
for
throughput
that's
what
most
dropped
by
10x
so
the
latency
spikes
now
are
near
roughly
100
milliseconds
this
bound
on
the
latency
you're
going
to
to
schedule
their
batches
back-to-back
feed-forward
Network
and
the
sequences
can
be
of
varying
lengths
so
that's
very
sequences
a
few
areas
of
emerging
tech
model
is
the
universal
serialization
format
for
tensorflow
models
it
is
it's
included
in
tensorflow
10
and
also
in
on
the
repo
there
are
two
marquee
and
dropout
layers
you
really
don't
want
defines
the
signature
of
a
computation
supported
by
a
tensor
flow
graph
and
your
floating-point
scores
for
say
a
common
as
people
productionize
and
maybe
a
click-through
model
to
do
predictions
on
say
your
video
pre-processing
all
of
your
hidden
layers
redundant
computation
so
you're
parsing
your
input
pre-processing
at
doing
the
of
deploying
multiple
models
it'll
save
you
bandwidth
cpu
latency
and
so
on
so
specifically
batching
padding
and
discuss
at
tensorflow
org
so
here
a

相關影片

加Line好友 每日8AM 推播1單字

加入好友

其他人正在學習的單字

1. ask
2024-05-10 05:32
2. high
2024-05-10 05:32
3. can
2024-05-10 05:32
4. happens
2024-05-10 05:32
5. salesmen
2024-05-10 05:32
6. carol
2024-05-10 05:32
7. ring
2024-05-10 05:32
8. bookshelf
2024-05-10 05:32
9. ad
2024-05-10 05:31
10. host
2024-05-10 05:31
11. almost
2024-05-10 05:31
12. written
2024-05-10 05:31
13. Otherwise
2024-05-10 05:31
14. point
2024-05-10 05:31
15. help
2024-05-10 05:31

本月觀看影片單字排行榜

本月查閱單字排行榜

本月聆聽單字排行榜