GS_MODEL_WAREHOUSE

GS_MODEL_WAREHOUSE stores AI engine training models, including the models and detailed description of the training process.

Table 1 GS_MODEL_WAREHOUSE columns

Name	Data Type	Description
oid	oid	Hidden column
modelname	name	Unique constraint
modelowner	oid	OID of a model owner
createtime	timestamp	Time when a model is created
processedtuples	int	Number of tuples involved in training
discardedtuples	int	Number of unqualified tuples not involved in training
preprocesstime	real	Data preprocessing time
exectime	real	Training duration
iterations	int	Iteration round
outputtype	oid	OID of the output data type
modeltype	text	AI operator type
query	text	Query statement executed to create a model
modeldata	bytea	Stored binary model information
weight	real[]	Currently, this column applies only to GD operator models.
hyperparametersnames	text[]	Involved hyperparameter name
hyperparametersvalues	text[]	Hyperparameter value
hyperparametersoids	oid[]	OID of the data type corresponding to a hyperparameter
coefnames	text[]	Model parameter
coefvalues	text[]	Value of a model parameter
coefoids	oid[]	OID of the data type corresponding to a model parameter
trainingscoresname	text[]	Method used to measure model performance
trainingscoresvalue	real[]	Value used to measure model performance
modeldescribe	text[]	Model description