trained_model_stsbenchmark_bert-base-uncased/sentence_bert_config.json

{
  "max_seq_length": 512,
  "do_lower_case": false
}