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
|
|---|---|---|---|---|---|---|---|---|
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1162.32
|
So it's not really a problem.
| 1,162.32 | 1,169.12 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1164.08
|
But you can see that we have added answers and into there now which is what we wanted to do.
| 1,164.08 | 1,177.2 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1170.72
|
And also importantly is if I let me copy this bring down here.
| 1,170.72 | 1,182.88 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1178.48
|
We'll notice that we do still have all of our data sets.
| 1,178.48 | 1,186.56 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1182.8799999999999
|
So if I go here.
| 1,182.88 | 1,189.2 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1187.6799999999998
|
I don't really need to remove that's fine.
| 1,187.68 | 1,191.44 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1190.0
|
I'll just break straight away.
| 1,190 | 1,192.4 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1191.44
|
That's fine.
| 1,191.44 | 1,198.24 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1196.64
|
So sample sorry here.
| 1,196.64 | 1,200.32 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1199.28
|
So you see the whole thing.
| 1,199.28 | 1,204.24 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1201.52
|
And we see that we still have the ID.
| 1,201.52 | 1,205.44 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1204.24
|
We have the text.
| 1,204.24 | 1,206.56 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1205.44
|
We have the context.
| 1,205.44 | 1,207.92 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1206.56
|
We have everything in there.
| 1,206.56 | 1,213.44 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1207.92
|
Now I'm just going to show you you know why this breaks.
| 1,207.92 | 1,216.64 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1213.44
|
Or what happens if I remove these.
| 1,213.44 | 1,222.8 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1216.64
|
Okay so let me rerun that and this as well.
| 1,216.64 | 1,226.24 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1222.8000000000002
|
So yeah so this should look the same.
| 1,222.8 | 1,228.64 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1227.0400000000002
|
Do we have yet that's fine.
| 1,227.04 | 1,230.4 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1229.1200000000001
|
But then if I run this.
| 1,229.12 | 1,238.32 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1231.2
|
So before this had all the all the features but now we only have the single feature that we specified in this formula.
| 1,231.2 | 1,239.36 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1238.3200000000002
|
So the answers.
| 1,238.32 | 1,243.76 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1239.3600000000001
|
So that's why you need to when shuffle is set to true.
| 1,239.36 | 1,247.6 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1243.76
|
That's why you need to add every single feature in there.
| 1,243.76 | 1,250.96 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1247.6
|
Otherwise it's just going to remove them when you perform the map operation.
| 1,247.6 | 1,257.52 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1251.52
|
But that's only the case when shuffle is actually set to true.
| 1,251.52 | 1,260.24 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1257.52
|
Shuffle? Why am I saying shuffle? Streaming is set to true.
| 1,257.52 | 1,262.8 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1260.96
|
So let me bring this down here.
| 1,260.96 | 1,268.56 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1264.16
|
And let me also copy our initial loading code.
| 1,264.16 | 1,269.84 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1268.56
|
So here.
| 1,268.56 | 1,272.88 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1269.84
|
Because we're going to need to reload our data set now anyway.
| 1,269.84 | 1,275.36 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1272.8799999999999
|
Because we just removed all the features from it.
| 1,272.88 | 1,285.44 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1279.4399999999998
|
Okay and what I'm going to do now is just set streaming to false.
| 1,279.44 | 1,291.6 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1285.4399999999998
|
And I'm going to run this same code where we still don't have our ids or anything like that in there.
| 1,285.44 | 1,293.84 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1292.3999999999999
|
And we'll see what happens.
| 1,292.4 | 1,296.48 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1293.84
|
As well we'll also notice we'll get a loading bar here.
| 1,293.84 | 1,298.96 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1296.48
|
And it's going to take a little bit of time to process this.
| 1,296.48 | 1,302 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1298.96
|
Although actually with this it's probably going to be super fast.
| 1,298.96 | 1,304.16 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1302.0
|
So probably ignore that.
| 1,302 | 1,308.64 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1304.8
|
But it will you see okay it's taking a little bit of time.
| 1,304.8 | 1,310.4 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1308.64
|
So now it's going through a whole data set.
| 1,308.64 | 1,312.4 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1310.4
|
We haven't called the date set.
| 1,310.4 | 1,315.52 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1313.1200000000001
|
But we have used this map function.
| 1,313.12 | 1,320.8 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1315.52
|
When streaming is set to false the data set isn't lazily loaded.
| 1,315.52 | 1,324.8 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1320.8
|
And so the operation is going to be a bit more complicated.
| 1,320.8 | 1,329.2 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1324.8
|
So the operation the map operation is performed as soon as you call it.
| 1,324.8 | 1,332.72 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1329.84
|
So it's a slightly different behavior.
| 1,329.84 | 1,335.76 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1332.72
|
And the other behavior which is different is the fact that
| 1,332.72 | 1,338.88 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1335.76
|
we only need to specify the answers feature here.
| 1,335.76 | 1,341.68 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1338.8799999999999
|
So we only when we have streaming set to false.
| 1,338.88 | 1,346.24 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1342.24
|
We don't need to include every feature within the map operation.
| 1,342.24 | 1,350.88 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1346.24
|
We only need to include the feature that we are modifying or creating.
| 1,346.24 | 1,355.36 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1350.88
|
Which you know it's weird I don't know why there's a behavior difference
| 1,350.88 | 1,357.36 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1355.3600000000001
|
when streaming is true or false.
| 1,355.36 | 1,358.96 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1357.7600000000002
|
But it is there.
| 1,357.76 | 1,362.32 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1358.96
|
So if I now take this again.
| 1,358.96 | 1,366.48 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1363.3600000000001
|
Come down here and run that.
| 1,363.36 | 1,369.92 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1366.48
|
We see now that we have all of our features again.
| 1,366.48 | 1,374.16 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1370.96
|
Right so before when streaming was true.
| 1,370.96 | 1,378.48 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1374.5600000000002
|
If I run this code it would have only included our answers.
| 1,374.56 | 1,383.2 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1378.48
|
The id, title, context, question they all would have been removed.
| 1,378.48 | 1,388.8 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1383.84
|
But now with streaming equal to false they're still there.
| 1,383.84 | 1,392 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1389.92
|
So weird a weird.
| 1,389.92 | 1,396.64 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1393.92
|
So it's a weird feature or a weird behavior.
| 1,393.92 | 1,401.04 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1396.64
|
But it's how it is and we obviously just need to deal with it.
| 1,396.64 | 1,406.96 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1402.0
|
Now the next thing I want to show you is how we can also add batching
| 1,402 | 1,409.52 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1406.96
|
to our mapping process.
| 1,406.96 | 1,417.76 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1409.52
|
So typically with well pretty much every or any as far as I can think of any
| 1,409.52 | 1,423.12 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1417.76
|
NLP tasks we're going to want to tokenize our text.
| 1,417.76 | 1,426.72 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1424.0
|
So we're going to go ahead and do that for Q&A.
| 1,424 | 1,433.28 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1427.44
|
So we would import transformers or from transformers import a BERT tokenizer.
| 1,427.44 | 1,438 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1433.28
|
Let's say and I would initialize that.
| 1,433.28 | 1,440.08 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1438.0
|
So this is you know what we typically do.
| 1,438 | 1,445.6 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1440.08
|
We do tokenizer equals BERT tokenizer from pre-trained.
| 1,440.08 | 1,451.44 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1448.0
|
And let's say BERT base un-gased.
| 1,448 | 1,458.4 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1456.3999999999999
|
Okay I'll initialize that.
| 1,456.4 | 1,464.4 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1458.4
|
And then what I want to do is I'm going to tokenize my context or question and context.
| 1,458.4 | 1,471.52 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1465.2800000000002
|
In the format that SQUAD would usually expect when you're doing Q&A or building a Q&A model.
| 1,465.28 | 1,475.52 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1472.3200000000002
|
And I want to do that using the map function.
| 1,472.32 | 1,480.8 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1475.52
|
So you can do this in both streaming and non-streaming by the way.
| 1,475.52 | 1,484.48 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1481.6000000000001
|
So we just write date set.
| 1,481.6 | 1,488.48 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1484.48
|
It was train so same as before.
| 1,484.48 | 1,492.48 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1488.48
|
Date set it was train or date set train dot map.
| 1,488.48 | 1,499.04 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1493.04
|
We are using a lambda function so lambda x.
| 1,493.04 | 1,503.04 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1500.48
|
And in here we just want to say tokenizer.
| 1,500.48 | 1,510.24 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1504.24
|
So I'm not doing the usually when you write this you would include a dictionary here.
| 1,504.24 | 1,516 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1510.24
|
But the tokenizer the output from the tokenizer is already in dictionary format.
| 1,510.24 | 1,518.96 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1516.0
|
So we don't need to I don't need to do it in this case.
| 1,516 | 1,523.04 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1519.52
|
But basically what we have here is is still a dictionary.
| 1,519.52 | 1,530.96 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1524.0
|
And what I want to do is so with Q&A in your tokenizer you pass two text inputs.
| 1,524 | 1,536.16 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1530.96
|
You pass your question and you would also pass your question.
| 1,530.96 | 1,540.72 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1536.16
|
And you would also then pass your context.
| 1,536.16 | 1,546.72 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1542.72
|
And as usual we would we set our max length.
| 1,542.72 | 1,550.24 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1547.76
|
So usually 512.
| 1,547.76 | 1,554.16 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1551.2
|
I would set padding equal to the max length.
| 1,551.2 | 1,559.04 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1556.3200000000002
|
And also do truncation as well.
| 1,556.32 | 1,563.12 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1559.76
|
Okay so very typical tokenization process.
| 1,559.76 | 1,565.92 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1563.12
|
Nothing there's nothing different going on here.
| 1,563.12 | 1,572.88 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1565.9199999999998
|
This is what we normally do when we tokenize our text going into a transform model.
| 1,565.92 | 1,577.04 |
Build NLP Pipelines with HuggingFace Datasets
|
2021-09-23 13:30:07 UTC
|
https://youtu.be/r-zQQ16wTCA
|
r-zQQ16wTCA
|
UCv83tO5cePwHMt1952IVVHw
|
r-zQQ16wTCA-t1573.6799999999998
|
And then we want to say okay batched equals true.
| 1,573.68 | 1,581.12 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.