-- datasets:
This is Longformer Encoder Decoder [led-base-16384] model for the legal domain, trained for long document abstractive summarization task.