SigmaWe/DocAsRef

Issues

• Dump results

  #22 opened 2 years ago by forrestbao
  1

• PageRank-based sentence weighting

  #20 opened 2 years ago by forrestbao
  0

• Summary-only coherence

  #25 opened 2 years ago by forrestbao
  0

• Why is our top-k or top-p worse than SUEPRT?

  #21 opened 2 years ago by forrestbao
  0

• Adding MoverScore

  #11 opened 2 years ago by forrestbao
  5

• Skip `dar_env.py`

  #12 opened 2 years ago by forrestbao
  0

• A new way to work between DocAsRef and Evalbase

  #16 opened 2 years ago by forrestbao
  1

• The right way to use Roberta-large-mnli

  #17 opened 2 years ago by forrestbao
  0

• Pseudoreference using decent enough summarizers

  #18 opened 2 years ago by forrestbao
  0

• `min_length` and `max_length` for summarizers

  #19 opened 2 years ago by forrestbao
  0

• discrepant result due to sentence segmentation

  #8 opened 2 years ago by forrestbao
  2

• redundancy in word tokenization and sentence segmentation

  #7 opened 2 years ago by forrestbao
  2

• Vectorizing MNLI sentence segmentation and similarity computation

  #6 opened 2 years ago by forrestbao
  1

• Vectorizing MNLI inference

  #5 opened 2 years ago by forrestbao
  4

• Signatures of all metrics

  #13 opened 2 years ago by forrestbao
  0

• top-k and top-p

  #15 opened 2 years ago by forrestbao
  0

• versions of `transformers`, `huggingface_hub` and `sentence-transformers`

  #14 opened 2 years ago by forrestbao
  0

• Fair and square model selection

  #10 opened 2 years ago by forrestbao
  0

• Printing results with less rows

  #9 opened 2 years ago by forrestbao
  0

• BERTScore

  #3 opened 2 years ago by forrestbao
  0

• BLEURT

  #2 opened 2 years ago by forrestbao
  9

• Turn off use_aggregator for ROUGE

  #1 opened 2 years ago by forrestbao
  3