sentence_bert_config.json

{
  "transformer_task": "feature-extraction",
  "modality_config": {
    "image": {
      "method": "forward",
      "method_output_name": "embedding"
    }
  },
  "module_output_name": "sentence_embedding"
}