Tiiiger/bert_score

Issues

• No module named 'transformers.models.beit.configuration_beit'

  #137 opened 2 years ago
  2

• The computation method when dealing with multiple references

  #136 opened 2 years ago
  1

• Error "int too big to convert" when I use the offline model downloaded from hugging face

  #135 opened 2 years ago
  7

• IDF weighting doesn't work

  #134 opened 2 years ago
  2

• Score less than 0

  #133 opened 2 years ago
  1

• Wrong default bert score.

  #132 opened 3 years ago
  1

• lang=en-sci error

  #131 opened 3 years ago
  2

• local tokenizer and checkpoint

  #130 opened 3 years ago
  2

• weights of the model checkpoint at roberta-large

  #129 opened 3 years ago
  2

• Can't load config for 'microsoft/deberta-v2-xlarge-mnli'

  #128 opened 3 years ago
  1

• text similarity between word “happy” and "sad"

  #126 opened 3 years ago
  2

• where should the downloaded pre-trained weight file be placed

  #125 opened 3 years ago
  1

• Running bert score examples

  #124 opened 3 years ago
  4

• Pairwise score

  #123 opened 3 years ago
  3

• can bertscore be -ve?

  #122 opened 3 years ago
  1

• Reproduce COCO captioning results

  #121 opened 3 years ago
  1

• BertScorer breaks with latest current version of matplotlib

  #120 opened 3 years ago
  1

• Feature request

  #119 opened 3 years ago
  2

• I got different results when the value of rescale_with_baseline is different(True/False)

  #118 opened 3 years ago
  0

• Error when I use it on a pair of long text

  #117 opened 3 years ago
  2

• Request for the data in Image Captioning experiment, again

  #116 opened 3 years ago
  2

• Supporting new Languages

  #115 opened 3 years ago
  0

• AttributeError: 'RobertaTokenizerFast' object has no attribute 'max_len'

  #114 opened 3 years ago
  2

• Support for ByT5

  #113 opened 3 years ago
  1

• How to use offline pre-trained model like bert-base-chinese password

  #112 opened 3 years ago
  2

• AssertionError: Different number of candidates and references

  #111 opened 3 years ago
  13

• Script to reporoduce segemet-level WMT17

  #110 opened 3 years ago
  2

• Random sequences giving a higher F1 than real sentences

  #109 opened 3 years ago
  5

• turkish bert support

  #108 opened 3 years ago
  1

• Empty candidate sentence

  #107 opened 3 years ago
  1

• Slow tokenizer is used by default

  #105 opened 3 years ago
  6

• Find relevant word embeddings given a reference word

  #104 opened 3 years ago
  1

• Adaptation of `score()` to run on my custom bert version.

  #103 opened 3 years ago
  2

• Calculate P but the result of the paper is different

  #102 opened 3 years ago
  4

• Is there a way to return the best sentence? Not just the score.

  #101 opened 3 years ago
  1

• Issues using scibert-scivocab-uncased

  #100 opened 3 years ago
  2

• Question: tf_idf feature?

  #99 opened 4 years ago
  1

• Question: how to get the best reference

  #98 opened 4 years ago
  3

• UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 0: ordinal not in range(128)

  #97 opened 4 years ago
  2

• Is text truncation to 512 tokens handled automatically for both candidate and reference texts?

  #96 opened 4 years ago
  3

• Question: Using Vokenization with BERTScore

  #95 opened 4 years ago
  3

• 'T5Stack' object has no attribute 'layer

  #94 opened 4 years ago
  2

• mT5

  #93 opened 4 years ago
  3

• Cache Model if Possible

  #92 opened 4 years ago
  2

• the running is always terminated

  #91 opened 4 years ago
  3

• Error when scoring big text

  #90 opened 4 years ago
  4

• Ways to specify the model download path?

  #89 opened 4 years ago
  1

• Error while running tune_layers.py

  #88 opened 4 years ago
  4

• `score` function failed with the new `transformers` v4.0.0

  #85 opened 4 years ago
  4

• Getting key error for score.py in bert_score

  #82 opened 4 years ago
  1