model_config.json

{
  "architecture": "WECHSEL-XLM-R-Dense",
  "backbone": "xlm-roberta-base",
  "vocab_size": 61028,
  "d_model": 768,
  "max_length": 128,
  "freeze_layers": 8
}