shmsw25/FActScore

Issues

• AssertionError: `topic` in your data (Jessie Mae Brown Beavers) is likely to be not a valid title in the DB.

  #37 opened 10 months ago by clawnotfound
  2

• Strange token ID mapping

  #49 opened 5 months ago by Yingjia-Wan
  0

• Some cases couldn't be found in the default db

  #35 opened a year ago by Loose-Gu
  2

• Maintaining core factscore in a new repository

  #47 opened 7 months ago by armingh2000
  0

• Improved version for speed and openai updates

  #46 opened 7 months ago by wj210
  0

• Batching for speed.

  #45 opened 7 months ago by wj210
  0

• Issue running get_score() on M1 Mac due to Torch build

  #33 opened 8 months ago by jcblsn
  1

• Any way to run without an OpenAI key?

  #34 opened a year ago by mohummedalee
  3

• About the enwiki-20230401

  #44 opened 9 months ago by Toblame
  5

• sqlite3.DatabaseError: file is not a database error after Installation

  #40 opened 9 months ago by GeroVanMi
  0

• InvalidRequestError: The model `text-davinci-003` has been deprecated

  #39 opened 9 months ago by shrusti-ghela
  2

• problem using the factscorer function with LLaMa

  #36 opened 10 months ago by samuel-LP
  1

• Move to ChatCompletion endpoint for call_GPT3

  #38 opened 10 months ago by martiansideofthemoon
  0

• Having issues on using LLAMA-7B model

  #25 opened a year ago by mungg
  2

• How to modify the code to calculate factscore between any two paragraphs/sentences without retrieval steps?

  #32 opened a year ago by laiviet
  1

• The chatgpt-based AtomicFactGenerator of FactScore sometimes abstain from answering

  #30 opened a year ago by HillZhang1999
  4

• Generation details for models under data/labeled

  #31 opened a year ago by caiqizh
  1

• Regarding human evaluation

  #29 opened a year ago by pat-jj
  2

• Typo here

  #27 opened a year ago by CiaoHe
  1

• FileNotFoundError (demons.json) for custom knowledge base

  #28 opened a year ago by MathiasKraus
  2

• Scoring existing generated text

  #26 opened a year ago by sofi444
  1

• Using unlabeled data to generate atomic facts and retrieving evidence

  #22 opened a year ago by rubaha96
  1

• FActScore can be slow to run

  #23 opened a year ago by martiansideofthemoon
  0

• improve processing of Perplexity AI outputs

  #6 opened a year ago by martiansideofthemoon
  2

• Better ranking between models

  #13 opened a year ago by martiansideofthemoon
  1

• Don’t understand how true_score and false_score are represented in logits

  #17 opened a year ago by rubaha96
  1

• Move abstain detection outside the atomic fact generator

  #19 opened a year ago by martiansideofthemoon
  0

• License?

  #16 opened a year ago by brucewlee
  0

• MAX_LENGTH is not defined in retrieval

  #14 opened a year ago by absolute-nil
  2

• estimated cost for API before running API

  #3 opened a year ago by martiansideofthemoon
  1

• Create HF weight diffs for LLAMA model and add it to the README

  #5 opened a year ago by martiansideofthemoon
  0

• Improving logging, show different levels

  #10 opened a year ago by martiansideofthemoon
  0

• Add citation / bibtex to README

  #4 opened a year ago by martiansideofthemoon
  1

• Return respond rate and number of facts together with FActScore

  #7 opened a year ago by shmsw25
  1