This section describes how to generate new text-based content using HeatWave GenAI.
Connect to your HeatWave Database System.
For Running Batch Queries, add the natural-language queries to a column in a new or existing table.
To generate text-based content using HeatWave GenAI, perform the following steps:
-
To load the LLM in HeatWave memory, use the
ML_MODEL_LOAD
routine:call sys.ML_MODEL_LOAD("LLM", NULL);
Replace
LLM
with the name of the LLM that you want to use. To view the lists of supported LLMs, see HeatWave In-Database LLMs and OCI Generative AI Service LLMs.For example:
call sys.ML_MODEL_LOAD("mistral-7b-instruct-v1", NULL);
This step is optional. The
ML_GENERATE
routine loads the specified LLM too. But it takes a bit longer to load the LLM and generate the output when you run it for the first time. -
To define your natural-language query, set the
@query
session variable:set @query="QueryInNaturalLanguage";
Replace
QueryInNaturalLanguage
with a natural-language query of your choice. For example:set @query="Write an article on Artificial intelligence in 200 words.";
-
To generate text-based content, pass the query to the LLM using the
ML_GENERATE
routine with thetask
parameter set togeneration
:select sys.ML_GENERATE(@query, JSON_OBJECT("task", "generation", "model_id", "LLM", "language", "Language"));
Replace the following:
LLM
: LLM to use, which must be the same as the one you loaded in the previous step.-
Language
: the two-letterISO 639-1
code for the language you want to use. Default language isen
, which is English. To view the list of supported languages, see Languages.NoteThe
language
parameter is supported in HeatWave9.0.1-u1
and later versions.
For example:
select sys.ML_GENERATE(@query, JSON_OBJECT("task", "generation", "model_id", "mistral-7b-instruct-v1", "language", "en"));
Text-based content that is generated by the LLM in response to your query is printed as output. It looks similar to the text output shown below:
| {"text": " Artificial Intelligence, commonly referred to as AI, is a rapidly growing field that focuses on creating intelligent machines capable of performing tasks that typically require human intelligence. These tasks include things like understanding natural language, recognizing images, and making decisions.\n\nAI technology has come a long way in recent years, thanks to advances in machine learning and deep learning algorithms. These algorithms allow machines to learn from data and improve their performance over time. This has led to the development of more advanced AI systems, such as virtual assistants like Siri and Alexa, which can help users with tasks like setting reminders and answering questions.\n\nAI is also being used in a variety of other industries, including healthcare, finance, and transportation. In healthcare, AI is being used to help doctors diagnose diseases and develop treatment plans. In finance, AI is being used to detect fraud and make investment decisions. In transportation, AI is being used to develop self-driving cars and improve traffic flow.\n\nDespite the many benefits of AI, there are also concerns about its potential impact on society. Some worry that AI could lead to job displacement and a loss of privacy. Others worry that AI could be used for malicious purposes, such as cyber attacks or surveillance.\n"} |
To run multiple generation
queries in
parallel, use the
ML_GENERATE_TABLE
routine. This method is faster than running the
ML_GENERATE
routine multiple times.
The ML_GENERATE_TABLE
routine is
supported in HeatWave 9.0.1-u1
and later
versions.
To run batch queries using
ML_GENERATE_TABLE
, perform the following
steps:
-
To load the LLM in HeatWave memory, use the
ML_MODEL_LOAD
routine:call sys.ML_MODEL_LOAD("LLM", NULL);
Replace
LLM
with the name of the LLM that you want to use. To view the lists of supported LLMs, see HeatWave In-Database LLMs and OCI Generative AI Service LLMs.For example:
call sys.ML_MODEL_LOAD("mistral-7b-instruct-v1", NULL);
This step is optional. The
ML_GENERATE_TABLE
routine loads the specified LLM too. But it takes a bit longer to load the LLM and generate the output when you run it for the first time. -
In the
ML_GENERATE_TABLE
routine, specify the table columns containing the input queries and for storing the generated text-based responses:call sys.ML_GENERATE_TABLE("InputDBName.InputTableName.InputColumn", "OutputDBName.OutputTableName.OutputColumn", JSON_OBJECT("task", "generation", "model_id", "LLM", "language", "Language"));
Replace the following:
InputDBName
: the name of the database that contains the table column where your input queries are stored.InputTableName
: the name of the table that contains the column where your input queries are stored.InputColumn
: the name of the column that contains input queries.OutputDBName
: the name of the database that contains the table where you want to store the generated outputs. This can be the same as the input database.OutputTableName
: the name of the table where you want to create a new column to store the generated outputs. This can be the same as the input table. If the specified table doesn't exist, a new table is created.OutputColumn
: the name for the new column where you want to store the output generated for the input queries.LLM
: LLM to use, which must be the same as the LLM you loaded in the previous step.Language
: the two-letterISO 639-1
code for the language you want to use. Default language isen
, which is English. To view the list of supported languages, see Languages.
For example:
call sys.ML_GENERATE_TABLE("demo_db.input_table.Input", "demo_db.output_table.Output", JSON_OBJECT("task", "generation", "model_id", "mistral-7b-instruct-v1", "language", "en"));
To learn more about the available routine options, see ML_GENERATE_TABLE Syntax.