SentenceTransformerModel.save_as_onnx

opensearch_py_ml.ml_models.SentenceTransformerModel.save_as_onnx(self, model_id='sentence-transformers/msmarco-distilbert-base-tas-b', model_name: str | None = None, save_json_folder_path: str | None = None, model_output_path: str | None = None, zip_file_name: str | None = None, add_apache_license: bool = False) → str

Download sentence transformer model directly from huggingface, convert model to onnx format, zip the model file and its tokenizer.json file to prepare to upload to the Open Search cluster

Parameters:

• model_id (string) – sentence transformer model id to download model from sentence transformers. default model_id = “sentence-transformers/msmarco-distilbert-base-tas-b”

• model_name (string) – Optional, model name to name the model file, e.g, “sample_model.pt”. If None, default takes the model_id and add the extension with “.pt”

• save_json_folder_path (string) – Optional, path to save model json file, e.g, “home/save_pre_trained_model_json/”). If None, default as default_folder_path from the constructor

• model_output_path (string) – Optional, path to save traced model zip file. If None, default as default_folder_path from the constructor

• zip_file_name (string) – Optional, file name for zip file. e.g, “sample_model.zip”. If None, default takes the model_id and add the extension with “.zip”

• add_apache_license (string) – Optional, whether to add a Apache-2.0 license file to model zip file

Returns:

model zip file path. The file path where the zip file is being saved

Return type:

string