ModelPath

The path to the exported model file to use. This parameter applies when Type is set to Transformer.

You can generate an appropriate model file to use by using the export_transformers_model.py script, which is provided in your Answer Server installation.

The script converts a third-party model into a format that Answer Server can use to generate the embeddings. For more information about how to use the script to create the model files, see Configure Candidate Retrieval for Passage Extractor.

Type: String
Default:  
Required: Yes, when Type is set to Transformer
Configuration Section: MyEmbeddings
Example: ModelPath=C:\Models\model.pt
See Also: TokenizerPath