jamescalam/youtube-transcriptions · Datasets at Hugging Face

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
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t263.44	So QA index already exists.	263.44	268.56
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t267.2	So it's not going to create a new index.	267.2	271.68
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t269.12	And instead, it's just going to connect that index here.	269.12	272.24
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t271.68	Right.	271.68	278.08
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t272.24	So we just connected or we created our index, our vector database index.	272.24	284.88
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t279.36	And now what I want to do is I'm going to switch back to our data.	279.36	286.88
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t285.6	And I'm going to run through that.	285.6	293.76
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t286.88	So I'm going to load the data set and the squad data set from Hugging Face.	286.88	299.84
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t294.96000000000004	Now I'm going to use a validation split because the model has been trained on the training data.	294.96	304.16
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t299.84	But squad, I want to make it a little bit hard.	299.84	307.92
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t304.15999999999997	So we're going to use a validation split that hasn't seen before.	304.16	313.76
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t307.91999999999996	I'm removing any unique or duplicate context in there.	307.92	317.04
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t313.76	So zoom out a little bit here.	313.76	320.08
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t318.4	Squad depth, we're using this filter.	318.4	323.92
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t320.08	So this is all Hugging Face data sets syntax here.	320.08	326.88
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t325.52	And then we're encoding it.	325.52	330.24
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t326.88	So this model.encode, so this is our sentence transformer.	326.88	335.68
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t330.24	We're encoding it to create a load of sentence vectors for our context.	330.24	339.36
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t335.68	And we're converting these to lists because we are going to be pushing these	335.68	342.48
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t339.36	through an API request to Pinecone.	339.36	346.64
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t342.48	Again, we need a list, not a NumPy array or otherwise you're going to get an error.	342.48	348.42
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t347.92	OK.	347.92	352.24
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t349.12	Then back to the Pinecone side of things.	349.12	358.32
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t352.24	We want to create a list of, it's basically a list of tuples.	352.24	364.16
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t358.32	And those tuples include the ID of each context.	358.32	366
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t364.16	So there's a unique ID for each context.	364.16	370
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t366.48	We want the vector or the encoding, the context vector.	366.48	373.04
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t371.04	And then we also have this dictionary here.	371.04	374.48
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t373.04	Now this is metadata.	373.04	380.08
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t374.48	So metadata in Pinecone is like any other information about your vectors that you want to include.	374.48	386.88
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t380.08	And this is really good if you want to use metadata filtering, which is super powerful in Pinecone.	380.08	392.08
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t387.91999999999996	And I definitely want to leave the option open later on.	387.92	393.68
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t392.08	I'm not sure if we'll use it or not.	392.08	397.44
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t394.4	We'll probably put something in there just so we can play around with it.	394.4	405.2
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t398.96	Now that creates the format that we need to upset everything, which means just like push or upload	398.96	406.4
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t405.2	everything to Pinecone.	405.2	409.76
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t406.4	So then I do that in chunks of 50 at a time.	406.4	416
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t409.76	Just makes things a little bit easier on the API aggressor rather than sending everything at once.	409.76	416.5
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t416.0	OK.	416	420.16
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t417.44	So that's like how we create the index.	417.44	428.96
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t421.2	So now what we're going to do is actually integrate that a little bit in our app.	421.2	432.32
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t429.76	So let's switch back to our app here.	429.76	434.96
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t432.32	Let's move this over here.	432.32	436.64
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t434.96	Let's view it.	434.96	439.36
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t437.84	So first, let's just remove this.	437.84	440.16
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t439.35999999999996	We don't need that.	439.36	443.84
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t440.71999999999997	OK, save will automatically reload.	440.72	451.92
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t445.52	So first thing we want to do here is let's initialize the Pinecone connection.	445.52	456
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t451.91999999999996	So I'm going to just take.	451.92	459.6
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t457.35999999999996	Let's just take this part of the code.	457.36	463.52
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t460.71999999999997	Just copy it and then we'll remove what we don't need in a minute.	460.72	469.52
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t463.52	And so we don't need we do need sentence transformers in a minute.	463.52	470.64
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t469.52	We don't need data sets.	469.52	472.64
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t471.76	We do need Pinecone.	471.76	477.76
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t473.44	So actually here we're initializing our retriever model.	473.44	479.44
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t477.76	It's the same as what we did before.	477.76	480.96
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t479.44	So we do want to keep that in there.	479.44	482.4
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t480.96	Make that bigger.	480.96	486.48
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t483.91999999999996	API key again, just store this somewhere else.	483.92	494
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t486.48	Or if you are using Streamlit Cloud, they have like a secrets management system.	486.48	496.56
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t494.0	And it's something we'll look at in the future for sure.	494	498.48
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t496.56	But for now, I'm just putting it in here.	496.56	505.04
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t500.24	So we have our API key environment and we're just doing the same thing we did before.	500.24	507.28
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t505.04	But actually, we don't want to create an index.	505.04	510.4
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t507.28000000000003	We're assuming we've already created an index if we're in our app.	507.28	511.68
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t510.40000000000003	So we're just going to connect to it.	510.4	516.4
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t512.48	OK, so with that, we've got our API key.	512.48	521.92
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t516.4	We've kind of set up the like the back end part of our app.	516.4	525.04
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t521.92	Like the smart part that's going to handle the open the main Q&A.	521.92	527.6
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t526.0	But it's going to be a little bit slow.	526	532.88
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t527.6	And we will have a look at how to solve that pretty soon.	527.6	536.32
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t532.88	But for now, what we're going to do is actually just implement this.	532.88	541.68
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t536.88	And we're going to actually query and see what we return.	536.88	544
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t542.56	So I'm going to save this.	542.56	548.4
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t544.0	We won't see anything change in our app now other than the fact that it takes longer to load	544	553.76
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t548.4	because it's downloading the driver model.	548.4	558.16
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t553.76	That's the main part of the slowness here and then obviously connecting to Pinecone.	553.76	560.72
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t558.16	Also takes a second as well.	558.16	565.6
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t561.68	So for now, we're going to deal with how slow it is.	561.68	569.84
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t567.36	But we will fix that pretty soon.	567.36	577.28
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t569.84	And now I actually want to do is I would say, OK, if the query is not empty,	569.84	580.32
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t577.2800000000001	because by default it is empty, that's why we add that in there.	577.28	581.92
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t580.32	So I'm going to actually remove this.	580.32	583.52
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t582.8000000000001	Enter.	582.8	589.76
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t583.52	If it is not empty, so if query is not equal to nothing,	583.52	596.88
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t590.72	we're going to query Pinecone for whatever is in that query.	590.72	599.6
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t596.88	So the first thing we need to do is create our context vector.	596.88	603.52
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t599.6	So I'm going to write XQ, just shorthand for context vector.	599.6	609.76
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t605.6	It's pretty standard, especially if you use FICE before they tend to use this.	605.6	613.52
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t610.32	And I say I said context vector, I meant query vector.	610.32	617.68
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t615.12	So we're going to do model and code.	615.12	621.84
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t618.72	And we need to put this in square brackets and we have query.	618.72	624.4
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t622.4	OK, and I'm going to convert that to a list.	622.4	628.64
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t624.4	OK, so this is going to create our query vector.	624.4	629.44
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t628.64	Let's write it down.	628.64	632.4
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t630.24	Create query vector.	630.24	639.28
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t634.3199999999999	And then the next thing we want to do is query Pinecone with this vector.	634.32	646.4
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t640.16	So to do that, we want to write first list.	640.16	649.76
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t646.4	Get relevant context.	646.4	652.88
Streamlit for ML #2 - ML Models and APIs	2022-01-26 16:30:36 UTC	https://youtu.be/U0EoaFFGyTg	U0EoaFFGyTg	UCv83tO5cePwHMt1952IVVHw	U0EoaFFGyTg-t650.96	And we're going to store these in XC.	650.96	659.36

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t263.44

So QA index already exists.

263.44

268.56

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t267.2

So it's not going to create a new index.

267.2

271.68

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t269.12

And instead, it's just going to connect that index here.

269.12

272.24

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t271.68

Right.

271.68

278.08

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t272.24

So we just connected or we created our index, our vector database index.

272.24

284.88

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t279.36

And now what I want to do is I'm going to switch back to our data.

279.36

286.88

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t285.6

And I'm going to run through that.

285.6

293.76

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t286.88

So I'm going to load the data set and the squad data set from Hugging Face.

286.88

299.84

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t294.96000000000004

Now I'm going to use a validation split because the model has been trained on the training data.

294.96

304.16

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t299.84

But squad, I want to make it a little bit hard.

299.84

307.92

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t304.15999999999997

So we're going to use a validation split that hasn't seen before.

304.16

313.76

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t307.91999999999996

I'm removing any unique or duplicate context in there.

307.92

317.04

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t313.76

So zoom out a little bit here.

313.76

320.08

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t318.4

Squad depth, we're using this filter.

318.4

323.92

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t320.08

So this is all Hugging Face data sets syntax here.

320.08

326.88

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t325.52

And then we're encoding it.

325.52

330.24

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t326.88

So this model.encode, so this is our sentence transformer.

326.88

335.68

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t330.24

We're encoding it to create a load of sentence vectors for our context.

330.24

339.36

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t335.68

And we're converting these to lists because we are going to be pushing these

335.68

342.48

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t339.36

through an API request to Pinecone.

339.36

346.64

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t342.48

Again, we need a list, not a NumPy array or otherwise you're going to get an error.

342.48

348.42

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t347.92

OK.

347.92

352.24

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t349.12

Then back to the Pinecone side of things.

349.12

358.32

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t352.24

We want to create a list of, it's basically a list of tuples.

352.24

364.16

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t358.32

And those tuples include the ID of each context.

358.32

366

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t364.16

So there's a unique ID for each context.

364.16

370

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t366.48

We want the vector or the encoding, the context vector.

366.48

373.04

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t371.04

And then we also have this dictionary here.

371.04

374.48

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t373.04

Now this is metadata.

373.04

380.08

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t374.48

So metadata in Pinecone is like any other information about your vectors that you want to include.

374.48

386.88

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t380.08

And this is really good if you want to use metadata filtering, which is super powerful in Pinecone.

380.08

392.08

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t387.91999999999996

And I definitely want to leave the option open later on.

387.92

393.68

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t392.08

I'm not sure if we'll use it or not.

392.08

397.44

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t394.4

We'll probably put something in there just so we can play around with it.

394.4

405.2

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t398.96

Now that creates the format that we need to upset everything, which means just like push or upload

398.96

406.4

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t405.2

everything to Pinecone.

405.2

409.76

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t406.4

So then I do that in chunks of 50 at a time.

406.4

416

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t409.76

Just makes things a little bit easier on the API aggressor rather than sending everything at once.

409.76

416.5

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t416.0

OK.

416

420.16

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t417.44

So that's like how we create the index.

417.44

428.96

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t421.2

So now what we're going to do is actually integrate that a little bit in our app.

421.2

432.32

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t429.76

So let's switch back to our app here.

429.76

434.96

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t432.32

Let's move this over here.

432.32

436.64

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t434.96

Let's view it.

434.96

439.36

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t437.84

So first, let's just remove this.

437.84

440.16

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t439.35999999999996

We don't need that.

439.36

443.84

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t440.71999999999997

OK, save will automatically reload.

440.72

451.92

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t445.52

So first thing we want to do here is let's initialize the Pinecone connection.

445.52

456

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t451.91999999999996

So I'm going to just take.

451.92

459.6

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t457.35999999999996

Let's just take this part of the code.

457.36

463.52

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t460.71999999999997

Just copy it and then we'll remove what we don't need in a minute.

460.72

469.52

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t463.52

And so we don't need we do need sentence transformers in a minute.

463.52

470.64

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t469.52

We don't need data sets.

469.52

472.64

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t471.76

We do need Pinecone.

471.76

477.76

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t473.44

So actually here we're initializing our retriever model.

473.44

479.44

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t477.76

It's the same as what we did before.

477.76

480.96

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t479.44

So we do want to keep that in there.

479.44

482.4

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t480.96

Make that bigger.

480.96

486.48

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t483.91999999999996

API key again, just store this somewhere else.

483.92

494

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t486.48

Or if you are using Streamlit Cloud, they have like a secrets management system.

486.48

496.56

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t494.0

And it's something we'll look at in the future for sure.

494

498.48

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t496.56

But for now, I'm just putting it in here.

496.56

505.04

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t500.24

So we have our API key environment and we're just doing the same thing we did before.

500.24

507.28

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t505.04

But actually, we don't want to create an index.

505.04

510.4

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t507.28000000000003

We're assuming we've already created an index if we're in our app.

507.28

511.68

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t510.40000000000003

So we're just going to connect to it.

510.4

516.4

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t512.48

OK, so with that, we've got our API key.

512.48

521.92

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t516.4

We've kind of set up the like the back end part of our app.

516.4

525.04

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t521.92

Like the smart part that's going to handle the open the main Q&A.

521.92

527.6

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t526.0

But it's going to be a little bit slow.

526

532.88

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t527.6

And we will have a look at how to solve that pretty soon.

527.6

536.32

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t532.88

But for now, what we're going to do is actually just implement this.

532.88

541.68

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t536.88

And we're going to actually query and see what we return.

536.88

544

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t542.56

So I'm going to save this.

542.56

548.4

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t544.0

We won't see anything change in our app now other than the fact that it takes longer to load

544

553.76

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t548.4

because it's downloading the driver model.

548.4

558.16

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t553.76

That's the main part of the slowness here and then obviously connecting to Pinecone.

553.76

560.72

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t558.16

Also takes a second as well.

558.16

565.6

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t561.68

So for now, we're going to deal with how slow it is.

561.68

569.84

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t567.36

But we will fix that pretty soon.

567.36

577.28

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t569.84

And now I actually want to do is I would say, OK, if the query is not empty,

569.84

580.32

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t577.2800000000001

because by default it is empty, that's why we add that in there.

577.28

581.92

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t580.32

So I'm going to actually remove this.

580.32

583.52

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t582.8000000000001

Enter.

582.8

589.76

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t583.52

If it is not empty, so if query is not equal to nothing,

583.52

596.88

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t590.72

we're going to query Pinecone for whatever is in that query.

590.72

599.6

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t596.88

So the first thing we need to do is create our context vector.

596.88

603.52

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t599.6

So I'm going to write XQ, just shorthand for context vector.

599.6

609.76

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t605.6

It's pretty standard, especially if you use FICE before they tend to use this.

605.6

613.52

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t610.32

And I say I said context vector, I meant query vector.

610.32

617.68

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t615.12

So we're going to do model and code.

615.12

621.84

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t618.72

And we need to put this in square brackets and we have query.

618.72

624.4

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t622.4

OK, and I'm going to convert that to a list.

622.4

628.64

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t624.4

OK, so this is going to create our query vector.

624.4

629.44

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t628.64

Let's write it down.

628.64

632.4

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t630.24

Create query vector.

630.24

639.28

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t634.3199999999999

And then the next thing we want to do is query Pinecone with this vector.

634.32

646.4

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t640.16

So to do that, we want to write first list.

640.16

649.76

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t646.4

Get relevant context.

646.4

652.88

Streamlit for ML #2 - ML Models and APIs

2022-01-26 16:30:36 UTC

https://youtu.be/U0EoaFFGyTg

U0EoaFFGyTg

UCv83tO5cePwHMt1952IVVHw

U0EoaFFGyTg-t650.96

And we're going to store these in XC.

650.96

659.36