NerReplace Syntax Elements | Vantage BYOM - NerReplace Syntax Elements - Teradata Vantage

Teradata Vantage™ - Bring Your Own Model User Guide

Deployment
VantageCloud
VantageCore
Edition
VMware
Enterprise
IntelliFlex
Lake
Product
Teradata Vantage
Release Number
20.00
Published
May 2026
ft:locale
en-US
ft:lastEdition
2026-06-18
dita:mapPath
fee1607120608274.ditamap
dita:ditavalPath
ayr1485454803741.ditaval
dita:id
fee1607120608274
Accumulate
Specify one or more column names to add to the output table. Use an asterisk (*) to specify all columns.
Specify column names as strings enclosed in single quotes and separated by commas.
Only column names are supported. Do not specify column IDs as integers (for example, 1, 4) or ranges (for example, [1:5]).
Does not support BLOB or CLOB columns.
ModelOutputTensor
[Optional] Specify the model output tensor to use as a string.
Default: 'logits'
Use this element only when the output tensor of the model is different from 'logits'.
ModelOutputFields
[Optional] Specify additional output fields to include as strings enclosed in single quotes and separated by commas.
Valid values:
  • 'replaced_entities': Adds a column with details (JSON) for each replaced entity, including begin position, end position, score, text, and label. Example:
    {"begin":109,"end":130,"text":"Quantum Analytics
            LLC","entity":"COMPANYNAME","score":0.9395}
  • 'logits': Adds a column with the logits (embeddings) calculated by the model.
EntitiesToReplace
[Optional] Specify the entity labels to replace as strings enclosed in single quotes and separated by commas.
Default: All entity labels.
With the 'NONE' aggregation strategy, BIO prefixes are present (for example, B-EMAIL).
An error occurs if a specified label does not exist for the model.
AggregationStrategy
[Optional] Specify how tokens are grouped into entities.
Valid values:
  • 'NONE': Processes each token independently
  • 'SIMPLE': Groups tokens by word
  • 'FIRST': Uses the first token
  • 'AVERAGE': Uses the average logits
  • 'MAX': Uses the maximum logit
Default: 'SIMPLE'
AVERAGE, FIRST, and MAX require a word-aware tokenizer.
ReplacementText
[Optional] Specify the text used to replace detected entities as a string enclosed in single quotes. If not specified, the function replaces entities with their label from the model.
To use this element, set ReplaceWithEntityName('false').
ReplacedEntityInfoLength
[Optional] Specify the maximum length, in characters, of the replaced_entities column.
Default: 6000
Enter an integer value from 1 to 32000.
Applies only when OutputDetails is set to 'true'.
EnableMemoryCheck
[Optional] Specify whether to verify that sufficient native memory is available for large models.
Valid values: 'true' or 'false'
Default: 'true'
IsDebug
[Optional] When true, debug statements are added to a trace table that must be created beforehand. See Error Logs.

Only available with BYOM version 3.00.00.02 and later.

Default: False
Unicode values are not supported in argument name-value pairs.