Table of Contents

• Run the Hello World example using query language
  • 🗝️ Key Concepts
  • 🏃 Run the Flows
  • Results
  • QueryLanguage

First of all, read up on Jina 101 so you have a clear understanding of how Jina works. We're going to refer to those concepts a lot. We assume you already have some knowledge of Python and machine learning.

This is an extension of the Hello World example, but here we use QueryLanguage to create an index per category and then, with that we can filter by category

In your terminal:

pip install -r requirements.txt

Now that we've got the code to load our data, we're going to dive into writing our app and running our Flows!

To run the index you type:

python app.py index

First up we need to build up an index and then we'll search through this index when we use the query Flow later. So what is happening here?

There are 10 categories in the fashion-mnist-data, but to simplify this example we will only use 3 categories:

0	        T-shirt/top
1	        Trouser
2	        Pullover

We will create a separate index per category, so each category will have its own yml file and then we index the data updating updating the label

d.tags.update({'label': get_mapped_label(label_int)})

This is where you could tweak the code if you would like to see only one category. For example if you would like to see only pullovers, you could do something like this

Then we have ready all the indexes!

Now for the query time, run:

python app.py query

So if we run this as it is, we will see our results filtered by pullovers like this:

So the real magic here, if you compare it with the original HelloWorld, is that we are using QueryLanguage. This is happening in the index and query flow of each category, for example, for the indexer-pullover.yml we have

IndexRequest:
      - !FilterQL
        with:
          lookups: {'tags__label': 'pullover'}

The key is the !FilterQL, here we are filtering with some specific label, like pullover in this example.