JetRunner/BERT-of-Theseus

Issues

• CoLA reproducibility

  #15 opened 3 years ago by mcps5601
  2

• What does “max_length” mean in config.json of successor

  #13 opened 3 years ago by SuMeng123
  0

• Duplicate definition of name (last_hidden_state) for float 16 onnx

  #16 opened 2 years ago by MrRace
  0

• fine-tune a predecessor model

  #14 opened 3 years ago by mcps5601
  1

• are predecessor modules weights frozen or not

  #9 opened 4 years ago by TobiasLee
  4

• how to draw the cover picture?

  #11 opened 4 years ago by MrRace
  1

• Which version of transformers is this repo using?

  #12 opened 4 years ago by failable
  1

• ImportError: cannot import name 'BERT_PRETRAINED_MODEL_ARCHIVE_MAP'

  #7 opened 4 years ago by MedSao
  5

• Comparison against six layer BERT

  #10 opened 4 years ago by david-macleod
  2

• The performance of DistilBERT

  #8 opened 4 years ago by forjiuzhou
  1

• Can I easily changed the number of hidden layers for scc_layer?

  #6 opened 4 years ago by Jaeker0512
  2

• The config.json of scc_net should be modified?

  #5 opened 4 years ago by janyChan
  2

• Why the model of pytorch_model.bin is also 579.3MB?

  #4 opened 4 years ago by janyChan
  3

• [Answered] MRPC Reproducibility

  #1 opened 4 years ago by YaNjIeE
  5

• [Answered] Does Theseus support pruning width of network?

  #2 opened 4 years ago by jiezhangGt
  2