This tutorial uses the Lightwood handler to create an embedding model.

​
Tutorial

The following example shows how to create an embedding model using the Lightwood engine.

Start by creating an engine from the Lightwood handler.

CREATE ML_ENGINE lightwood
FROM lightwood;

Verify that the engine has been created successfully.

SHOW ML_ENGINES
WHERE name='lightwood';

Connect our sample database to use it for training the model:

CREATE DATABASE example_db
WITH ENGINE = "postgres",
PARAMETERS = {
"user": "demo_user",
"password": "demo_password",
"host": "samples.mindsdb.com",
"port": "5432",
"database": "demo",
"schema": "demo_data"
};

Create a model using this engine:

CREATE MODEL home_rentals_model_embeddings
FROM example_db
  (SELECT * FROM demo_data.home_rentals)
PREDICT rental_price
USING
    problem_definition.embedding_only = True;

Check the status of the model.

DESCRIBE home_rentals_model_embeddings;

Now you can use the model to predict home rental prices for specific criteria and get predictions in the form of embeddings.

SELECT rental_price, rental_price_explain
FROM home_rentals_model_embeddings
WHERE sqft = 823
AND location='good'
AND neighborhood='downtown'
AND days_on_market=10
USING
    return_embedding = True;

Here is the output:

+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| rental_price                                                                                                                                                                                                                 | rental_price_explain                                                                                                                                                                                                                                                                                                                                                             |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| [1,6.712956428527832,1.247057318687439,0,0,1,0,0,0,0,0,1,0,0,0,1,2.3025851249694824,0.5629426836967468,0,1,0,0,0,0,0,0,0.7540000081062317,0.3333333432674408,0.35483869910240173,0.9583333134651184,0.7833333611488342,0.25] | {"predicted_value": [1.0, 6.712956428527832, 1.247057318687439, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0, 2.3025851249694824, 0.5629426836967468, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.7540000081062317, 0.3333333432674408, 0.35483869910240173, 0.9583333134651184, 0.7833333611488342, 0.25], "confidence": null, "anomaly": null, "truth": null} |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+