AkariAsai/learning_to_retrieve_reasoning_paths

Issues

• What is the problem？

  #29 opened a year ago by houdawang
  0

• Why are some document titles missing?

  #28 opened a year ago by houdawang
  0

• sqlite3.OperationalError: unable to open database file

  #27 opened 3 years ago by Suki-m730026038
  1

• Why are some document titles missing?

  #26 opened 3 years ago by mukhal
  2

• What do output_masks do?

  #25 opened 3 years ago by NoviScl
  2

• What the TF-IDF retriever data output mean

  #19 opened 3 years ago by ajesujoba
  3

• How to train and evaluate the models in HotpotQA distractor setting?

  #24 opened 4 years ago by solomon-han
  2

• A problem about total tranining steps of reader

  #20 opened 4 years ago by SparkJiao
  2

• How to evaluate the supporting facts in the HotPotQA experiment?

  #21 opened 4 years ago by ditingdapeng
  5

• How many of the first TF-IDF processing needs to be retained?

  #22 opened 4 years ago by ditingdapeng
  5

• The error when training the graph_retriever in the HotpotQA

  #15 opened 4 years ago by Frankszc
  5

• The hyperparameters for training the bert-base reader ?

  #23 opened 4 years ago by SparkJiao
  1

• hotpot model zip file corrupted?

  #7 opened 5 years ago by liuzzi
  7

• Fine-tuning on own documents?

  #18 opened 4 years ago by tmontana
  2

• Using switch_logits of final doc_span only

  #5 opened 4 years ago by talkhaldi
  4

• How to evaluate the pretrained graph retriever model?

  #14 opened 4 years ago by jellying
  5

• json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0)

  #17 opened 4 years ago by ukab23
  1

• Training data construction for reader verifier

  #16 opened 4 years ago by dsl-light
  3

• Inconsistent 'answers' types in the nq_reader_train data

  #12 opened 5 years ago by mjeensung
  1

• demo.py arg error about NQ

  #11 opened 5 years ago by yuchenlin
  4

• `database is locked` while evaluation

  #13 opened 5 years ago by yuchenlin
  1

• Some details regarding generating NQ trainset for the reader model

  #10 opened 5 years ago by mjeensung
  6

• Wrong links to the training data of the reader models

  #9 opened 5 years ago by soheeyang
  9

• negative documents construction for graph retriever of hotpotQA fullwiki

  #8 opened 5 years ago by dsl-light
  2

• question about wikipedia data

  #4 opened 5 years ago by jcyk
  2

• Small typo in the paper

  #6 opened 5 years ago by wenhuchen
  1

• Evaluation input for retriever

  #2 opened 5 years ago by tbright17
  2

• Preprocessing of HotpotQA

  #3 opened 5 years ago by talkhaldi
  3

• Minor fix in demo.py

  #1 opened 5 years ago by kumar-mallikarjuna
  2