title
stringlengths 12
112
| published
stringlengths 19
23
| url
stringlengths 28
28
| video_id
stringlengths 11
11
| channel_id
stringclasses 5
values | id
stringlengths 16
31
| text
stringlengths 0
596
| start
float64 0
37.8k
| end
float64 2.18
37.8k
|
|---|---|---|---|---|---|---|---|---|
How-to do Sentiment Analysis with Flair in Python
|
2020-12-04 14:00:03 UTC
|
https://youtu.be/DFtP1THE8fE
|
DFtP1THE8fE
|
UCv83tO5cePwHMt1952IVVHw
|
DFtP1THE8fE-t799.54
|
So it's quite good even here where we're not saying anything particularly like a negative or positive word.
| 799.54 | 814.26 |
How-to do Sentiment Analysis with Flair in Python
|
2020-12-04 14:00:03 UTC
|
https://youtu.be/DFtP1THE8fE
|
DFtP1THE8fE
|
UCv83tO5cePwHMt1952IVVHw
|
DFtP1THE8fE-t807.06
|
We're just saying that the movie is or the movie delivers on the performance of striking skill and depth.
| 807.06 | 820.58 |
How-to do Sentiment Analysis with Flair in Python
|
2020-12-04 14:00:03 UTC
|
https://youtu.be/DFtP1THE8fE
|
DFtP1THE8fE
|
UCv83tO5cePwHMt1952IVVHw
|
DFtP1THE8fE-t815.3
|
Which must be pretty hard for a machine to understand and actually get right.
| 815.3 | 824.1 |
How-to do Sentiment Analysis with Flair in Python
|
2020-12-04 14:00:03 UTC
|
https://youtu.be/DFtP1THE8fE
|
DFtP1THE8fE
|
UCv83tO5cePwHMt1952IVVHw
|
DFtP1THE8fE-t821.14
|
But looking at all these it's doing really well.
| 821.14 | 828.98 |
How-to do Sentiment Analysis with Flair in Python
|
2020-12-04 14:00:03 UTC
|
https://youtu.be/DFtP1THE8fE
|
DFtP1THE8fE
|
UCv83tO5cePwHMt1952IVVHw
|
DFtP1THE8fE-t824.0999999999999
|
And I think it's really cool that we can actually do this with so little effort.
| 824.1 | 833.06 |
How-to do Sentiment Analysis with Flair in Python
|
2020-12-04 14:00:03 UTC
|
https://youtu.be/DFtP1THE8fE
|
DFtP1THE8fE
|
UCv83tO5cePwHMt1952IVVHw
|
DFtP1THE8fE-t829.8599999999999
|
And we've only actually written a few lines of code in reality.
| 829.86 | 837.86 |
How-to do Sentiment Analysis with Flair in Python
|
2020-12-04 14:00:03 UTC
|
https://youtu.be/DFtP1THE8fE
|
DFtP1THE8fE
|
UCv83tO5cePwHMt1952IVVHw
|
DFtP1THE8fE-t833.06
|
And it's producing really good accurate results which is really impressive to me.
| 833.06 | 840.42 |
How-to do Sentiment Analysis with Flair in Python
|
2020-12-04 14:00:03 UTC
|
https://youtu.be/DFtP1THE8fE
|
DFtP1THE8fE
|
UCv83tO5cePwHMt1952IVVHw
|
DFtP1THE8fE-t837.8599999999999
|
So that's it for this video.
| 837.86 | 863.62 |
How-to do Sentiment Analysis with Flair in Python
|
2020-12-04 14:00:03 UTC
|
https://youtu.be/DFtP1THE8fE
|
DFtP1THE8fE
|
UCv83tO5cePwHMt1952IVVHw
|
DFtP1THE8fE-t840.42
| 840.42 | 863.62 |
|
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t0.0
|
Hi and welcome to the video. Today we're going to go through how we can fine-tune a Q&A Transformer model.
| 0 | 18.4 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t7.6000000000000005
|
So for those of you that don't know, Q&A just means question answering and it's one of the biggest topics in NLP at the moment.
| 7.6 | 24.4 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t18.400000000000002
|
There's a lot of models out there where you ask a question and it will give you an answer.
| 18.4 | 33.2 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t24.4
|
And one of the biggest things that you need to know how to do when you are working with transformers,
| 24.4 | 41.2 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t33.2
|
whether that's Q&A or any of the other transformer based solutions, is how to actually fine-tune those.
| 33.2 | 51.6 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t41.2
|
So that's what we're going to be doing in this video. We're going to go through how we can fine-tune a Q&A Transformer model in Python.
| 41.2 | 60.8 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t51.6
|
So I think it's really interesting and I think you will enjoy it a lot. So let's just go ahead and we can get started.
| 51.6 | 66.8 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t60.800000000000004
|
Okay, so first thing we need to do is actually download our data.
| 60.8 | 74 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t66.8
|
So we're going to be using the SQuAD dataset, which is the Stanford question answering dataset,
| 66.8 | 82.4 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t74.0
|
which is essentially one of the better known Q&A datasets out there that we can use to fine-tune our model.
| 74 | 88.8 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t82.4
|
So let's first create a folder.
| 82.4 | 97.2 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t88.8
|
It's going to use OS and OSMaker.
| 88.8 | 104.8 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t97.2
|
We'll just call it SQuAD. Obviously, call this and organize it as you want. This is what I will be doing.
| 97.2 | 112 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t104.8
|
Now, the URL that we are going to be downloading this from is this.
| 104.8 | 120.4 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t112.0
|
Okay, and there are actually two files here that we're going to be downloading, but both will be coming from the same URL.
| 112 | 126 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t120.4
|
So because we're making a request through URL, we're going to import requests.
| 120.4 | 133.6 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t126.0
|
We can also use the Wget library as well, or if you're on Linux, you can just use Wget directly in the terminal.
| 126 | 142.4 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t133.6
|
It's up to you what we're going to be using requests.
| 133.6 | 148.8 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t142.4
|
Okay, and to request our data, we're going to be doing this.
| 142.4 | 158.8 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t148.8
|
So it's just a get request, use a FString, and we have the URL that we've already defined.
| 148.8 | 172 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t158.8
|
And then the training data that we'll be using is this file here.
| 158.8 | 181.36 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t172.0
|
Okay, requests. Okay, we can see that we've successfully put that data in there.
| 172 | 188.72 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t181.36
|
Okay, so like I said before, there's actually two of these files that we want to extract.
| 181.36 | 197.84 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t188.72
|
So what I'm going to do is just put this into a for loop, which will go through both of them.
| 188.72 | 202.56 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t197.84
|
Just copy and paste this across.
| 197.84 | 210.48 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t202.56
|
Rename this file.
| 202.56 | 215.2 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t210.48000000000002
|
And the other file is the same, but instead of train, we have dev.
| 210.48 | 220.48 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t215.2
|
Okay, so here we're making our request.
| 215.2 | 228 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t220.48
|
And then next thing we want to do after making our request is actually saving this file to our drive.
| 220.48 | 233.2 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t228.0
|
Okay, and we want to put that inside this squad folder here.
| 228 | 238.32 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t233.2
|
So to do that, we use open.
| 233.2 | 243.2 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t238.32
|
And again, we're going to use a string here.
| 238.32 | 248.16 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t243.2
|
I want to put inside the squad folder here.
| 243.2 | 254.48 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t248.16
|
And then here we are just going to put our file name, which is file.
| 248.16 | 257.68 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t254.48
|
Now we're writing this in binary because it's JSON.
| 254.48 | 262.88 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t257.68
|
So we put wb for our flags here.
| 257.68 | 270.08 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t262.88
|
And then within this namespace, we are going to run through the file and download it in chunks.
| 262.88 | 273.92 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t270.08
|
So we do for chunk.
| 270.08 | 279.44 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t273.92
|
And then we iterate through the response.
| 273.92 | 281.36 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t279.44
|
Like this.
| 279.44 | 288.4 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t281.36
|
We'll just use a chunk size of four.
| 281.36 | 292.72 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t288.40000000000003
|
And then we just want to write to the file like that.
| 288.4 | 296 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t292.72
|
So that will download both files.
| 292.72 | 297.92 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t296.0
|
Just add the colon there.
| 296 | 301.36 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t297.92
|
So that will download both files.
| 297.92 | 303.36 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t301.36
|
We should be able to see them here now.
| 301.36 | 305.6 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t303.36
|
So in here we have data.
| 303.36 | 308.32 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t305.6
|
We have essentially a lot of different topics.
| 305.6 | 310.16 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t308.32
|
So the first one is Beyonce.
| 308.32 | 318.08 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t310.16
|
And then in here, we will see, if we just come to here, we get a context.
| 310.16 | 323.84 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t318.08000000000004
|
But alongside this context, we also have QAS, which is question and answers.
| 318.08 | 330.08 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t323.84000000000003
|
And each one of these contains a question and answer pair.
| 323.84 | 334.08 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t330.08
|
So we have this question, when did Beyonce start becoming popular?
| 330.08 | 337.28 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t334.08
|
So this answer is actually within this context.
| 334.08 | 343.04 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t337.28
|
And what we want our model to do is extract the answer from that context by telling us
| 337.28 | 348 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t343.03999999999996
|
the start and end token of the answer within that context.
| 343.04 | 352.08 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t348.0
|
So we go zero and it is in the late 1990s.
| 348 | 355.52 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t352.08
|
And we have answer start 269.
| 352.08 | 362 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t355.52
|
So that means that a character 269, we get I.
| 355.52 | 366.4 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t362.0
|
So if we go through here, we can find it here.
| 362 | 368.56 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t366.4
|
OK, so this is the extract.
| 366.4 | 373.2 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t368.56
|
And that's what we will be aiming for our model to actually extract.
| 368.56 | 378.16 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t373.2
|
But there will be a start point and also the end point as well, which is not included in here.
| 373.2 | 383.04 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t378.15999999999997
|
But we will add that manually quite soon.
| 378.16 | 388.48 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t383.04
|
So that's our data, and then we'll also be testing on the dev data as well,
| 383.04 | 390.96 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t389.52000000000004
|
which is exactly the same.
| 389.52 | 402.16 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t399.28000000000003
|
OK, so we move on to the data prep.
| 399.28 | 405.36 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t403.6
|
So now we have our files here.
| 403.6 | 408.32 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t406.16
|
We're going to want to read them in.
| 406.16 | 410.96 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t408.32000000000005
|
So we're going to use the JSON library for that.
| 408.32 | 417.12 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t410.96
|
And like we saw before, there's quite a complex structure in these JSONs.
| 410.96 | 418.64 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t417.12
|
There's a lot of different layers.
| 417.12 | 423.76 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t418.64
|
So we need to use a few for loops to fill through each of these and extract what we want,
| 418.64 | 430.8 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t423.76
|
which is the context, questions and answers all corresponding to each other.
| 423.76 | 435.52 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t430.79999999999995
|
So in the end, we're going to have a list of strings, which is going to be all of these.
| 430.8 | 438.48 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t435.52
|
And in the case of the answers, we also have the starting position.
| 435.52 | 446.4 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t438.48
|
So it will be a list of dictionaries where one value is a text and one value is the starting position.
| 438.48 | 452.32 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t447.84000000000003
|
So to do that, we're going to define a function called Rebsquad.
| 447.84 | 458.8 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t457.28000000000003
|
We'll define our path here as well.
| 457.28 | 464.88 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t460.72
|
And the first thing we need to do is actually open the JSON file.
| 460.72 | 467.68 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t464.88
|
So we do with open path.
| 464.88 | 472 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t468.4
|
And again, we are using a binary file.
| 468.4 | 474 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t472.0
|
So we're going to have B as a flag.
| 472 | 476.4 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t474.48
|
But instead of writing, we are reading.
| 474.48 | 477.52 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t476.4
|
So use R here.
| 476.4 | 478.48 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t477.52
|
So RB.
| 477.52 | 491.6 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t488.96
|
I'm just going to do JSON load F here.
| 488.96 | 496.8 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t491.6
|
So now we have our dictionary within this squad dict here.
| 491.6 | 503.68 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t497.44
|
So maybe whilst we're just building this function up, it's probably more useful to do it here.
| 497.44 | 505.44 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t503.68
|
So we can see what we're actually doing.
| 503.68 | 507.04 |
How to Build Custom Q&A Transformer Models in Python
|
2021-02-12 13:30:03 UTC
|
https://youtu.be/ZIRmXkHp0-c
|
ZIRmXkHp0-c
|
UCv83tO5cePwHMt1952IVVHw
|
ZIRmXkHp0-c-t505.44
|
So let's copy that across.
| 505.44 | 513.52 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.