Whisper

Notation Conventions

• Parentheses () indicate literal parentheses.
• Braces {} are used to bind combinations of options.
• The bracket [] indicates an optional clause.
• An ellipsis following a comma in brackets [,...] means that the preceding item can be repeated as a comma-separated list
• The vertical bar | represents the logic OR.
• VALUE represents a regular value.

• literal: a fixed or unchangeable value, also known as a Constant.

  Each literal has a special data type such as column, in the table.

PREDICT Syntax

Use the "PREDICT" statement to apply AI models to perform prediction, classification, recommendation, and more. The "PREDICT" statement can preprocess the dataset defined by the query_expr that comes after the "AS" clause.

query_statement:
    query_expr

PREDICT USING (model_name_expression)
OPTIONS (
    expression [ , ...]
    )
AS
(query_expr)

OPTIONS Clause

OPTIONS (
    (audio_col=column_name),
    [batch_size=VALUE],
    (language=expression),
    (task={'transcribe'|'translate'}),
    [result_col=column_name]
    )

The "OPTIONS" clause allows you to change the value of a parameter. The definition of each parameter is as follows.

• "audio_col": the name of the column containing the audio path to be used for prediction (str, default: 'audio_path')
• "batch_size": the size of the dataset bundle utilized in a single cycle of prediction (int, optional, default: 16)
• "language": specifies the language of the audio file. If selected as ‘auto’, the model will recognize the language from the available pool of 99 languages (str, default: 'auto')
• "task": type of work to do (str, 'transcribe'|'translate', default: 'transcribe')

  "transcribe": the task of recognizing speech "translate": the task of translating recognized speech back into English. This process is different from the ordinary translating task in that it translates Korean speech directly into English text, skipping the middle task of translating to Korean text

• "result_col": the column that contains the predicted results (str, optional, default: 'predict_result')

PREDICT Example

An example "PREDICT" query can be found in Using a Speech Recognition Model.

%%thanosql
PREDICT USING tutorial_whisper_small
OPTIONS (
    audio_col='audio_path',
    language='auto',
    task='transcribe',
    result_col='predict_result'
    )
AS
SELECT *
FROM korean_voice

EVALUATE Syntax

Use the "EVALUATE" statement to evaluate the AI model. The "EVALUATE" expression evaluates a model using the dataset defined by the query_expr that comes after the "AS" clause.

query_statement:
    query_expr

EVALUATE USING (model_name_expression)
OPTIONS (
    expression [ , ...]
    )
AS
(query_expr)

OPTIONS Clause

OPTIONS (
    (audio_col=column_name),
    [batch_size=VALUE],
    (language=expression),
    (task={'transcribe'|'translate'}),
    (text_col=column_name)
    )

The "OPTIONS" clause allows you to change the value of a parameter. The definition of each parameter is as follows.

• "audio_col": the name of the column containing the audio path to be used for evaluation (str, default: 'audio_path')
• "batch_size": the size of dataset bundle utilized in a single cycle of evaluation (int, optional, default: 16)
• "language": specifies the language of the audio file. If selected as ‘auto’, the model will recognize the language from the available pool of 99 languages (str, default: 'auto')
• "task": type of work to do (str, 'transcribe'|'translate', default: 'transcribe')

  "transcribe": the task of recognizing speech "translate": the task of translating recognized speech back into English. This process is different from the ordinary translating task in that it translates Korean speech directly into English text, skipping the middle task of translating to Korean text

• "text_col": the name of the column containing information about the target (str, default: 'text')

---

Last update: 2023-08-09