public class BertQAModelLoader
extends ai.djl.repository.zoo.BaseModelLoader
The model was trained on Gluon and loaded in DJL in MXNet Symbol Block.
See the BERT paper or the jupyter demo for more information about BERT.
MxSymbolBlock| Constructor and Description |
|---|
BertQAModelLoader(ai.djl.repository.Repository repository)
Creates the Model loader from the given repository.
|
| Modifier and Type | Method and Description |
|---|---|
ai.djl.repository.zoo.ZooModel<ai.djl.modality.nlp.qa.QAInput,java.lang.String> |
loadModel(java.util.Map<java.lang.String,java.lang.String> filters,
ai.djl.Device device,
ai.djl.util.Progress progress)
Loads the model with the given search filters.
|
public BertQAModelLoader(ai.djl.repository.Repository repository)
repository - the repository to load the model frompublic ai.djl.repository.zoo.ZooModel<ai.djl.modality.nlp.qa.QAInput,java.lang.String> loadModel(java.util.Map<java.lang.String,java.lang.String> filters,
ai.djl.Device device,
ai.djl.util.Progress progress)
throws java.io.IOException,
ai.djl.repository.zoo.ModelNotFoundException,
ai.djl.MalformedModelException
filters - the search filters to match against the loaded modeldevice - the device the loaded model should useprogress - the progress tracker to update while loading the modeljava.io.IOException - for various exceptions loading data from the repositoryai.djl.repository.zoo.ModelNotFoundException - if no model with the specified criteria is foundai.djl.MalformedModelException - if the model data is malformed