--- language: en license: apache-2.0 tags: - summarization datasets: arxiv-summarization model-index: - name: ArtifactAI/led_base_16384_arxiv_summarization results: - task: type: summarization name: Summarization dataset: name: ccdv/arxiv-summarization type: ccdv/arxiv-summarization config: section split: test metrics: - type: rouge value: 37.3255 name: ROUGE-1 verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiODIyZWNhMTgxNDlkYThhNGFlOGIxYjhhMTU4Y2JjN2I2ZDVkYWVhMmU5ZjQxZmQ3ZGY4ZmY1Y2Y2YzYwZjg5MCIsInZlcnNpb24iOjF9.Q5rZaUa1WvJThE1dOVOWEAOTweDkQPilaP9OCdM1W7ypC-XVTrKC-XjeYvgpET8GSqMROoYP9Z0oJdD1KcWeCw - type: rouge value: 10.8948 name: ROUGE-2 verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiYmJkN2E4YzE3MWY0ZTg4YjFkMGY5MjY2YjhmYzBjZGU3Mjc2NjNhYzkwMDkwOTMwNjdmYWI1ZmY2YmQ3OTA2MiIsInZlcnNpb24iOjF9.u9SrzD-QRXU2mboRwkhgyJcDGPfZoGY5vCoC4ROUc2WLB9IcSypzCAfGsIg488aWJ-iGUmfwbGQqj8Vb50mmCA - type: rouge value: 20.3875 name: ROUGE-L verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNzk5NTM4MjM0MGU0OTdmZWEzMjhkMmUxMTY3YTVmMzUzODllZWEwMWEwNjE5ZWNiYzY0MjM1MTFlZWE3NmNmNiIsInZlcnNpb24iOjF9.tJxNOMKwjJlTVhcjoLdy8phj4cSG3b5YaQd5vzl9RJc-kCLcC7Q_F7LDYlEFa7L2S04b6YAcn1JzPsCNy9avAA - type: rouge value: 33.3014 name: ROUGE-LSUM verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiZDVlZjZhMWNlZmE1YmQ1NDQ4ZmIyMzU5YjgxZmE5ZDEzYWJlNDBiODJjZDBhZWYyMmJhYmE4MWQ3ZGE4ZDUxMCIsInZlcnNpb24iOjF9.NGxXK6cEvyIia_iCjuIeR_JL0fKNONDmnaPKslwf56p7Hletg44oi17jM7LIkZ6ToZb31vvcKjx2DO4-k1V0CQ - type: loss value: 3.182162284851074 name: loss verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiM2U1YjkzMmIyNmEzYjlhNWZkNTNmMjgzNjZlMmY2ZWY1OGIyNzM2YmU1MzdiMDAxZDVmNmE5OGNiYThlNTA4ZiIsInZlcnNpb24iOjF9.CeWkK2aAodOUyj7omgJ0sq66GDTuEBRIuDOxLCkw6h1UshWCY2KT-uCUNcQfKMIvPaEjqIKjvtbWBKkmHipHAw - type: gen_len value: 145.5905 name: gen_len verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiZDMzNTIzZTcwMzczNGEzMmU4YjAyOTZhNDVlNmUyNzVjZGE5MjNhYzQ4MGNlYmQ4MjNlOWY4YzY0NDExNDhiZCIsInZlcnNpb24iOjF9.fX3AuS-fWZfYe5KPDr8FSxuVZYwcUKglSIhKYIVdwTsfXgUVTdDzC6wBiBRpS3ybW0yFSxlKnAbBdJEshOpDBw --- ## Introduction A led-base-16384 model to summarize ArXiv papers. Inputs are the abstracts of papers and full documents, and outputs are the summaries of the papers. [Allenai's Longformer Encoder-Decoder (LED)](https://github.com/allenai/longformer#longformer). As described in [Longformer: The Long-Document Transformer](https://arxiv.org/pdf/2004.05150.pdf) by Iz Beltagy, Matthew E. Peters, Arman Cohan, *led-base-16384* was initialized from [*bart-base*](https://huggingface.co/facebook/bart-base) since both models share the exact same architecture. To be able to process 16K tokens, *bart-base*'s position embedding matrix was simply copied 16 times. ### Rouge 2 | Type | Score | | --- | --- | | `precision` | 0.1839148953011932 | | `recall` | 0.14904707945189774 | | `fmeasure` | 0.1580026685776864 |