# Scientific Paper Processing

A long-sequence processing model based on LSG attention mechanism, fine-tuned specifically for scientific paper summarization tasks

A long-sequence processing model based on the BART architecture, optimized for scientific paper summarization tasks, supporting long-text input up to 16,384 tokens

BigBird is a Transformer model based on sparse attention, capable of handling longer sequences, suitable for tasks like long document summarization.

BigBirdPegasus is a Transformer model based on sparse attention, capable of handling longer sequences, especially suitable for long document summarization tasks.