MAGICS-LAB/DNABERT_2

Issues

• Question About DNAbert2 Pre-training Dataset Sources

  #124 opened 11 days ago by yanghu819
  1

• Unable to reproduce covid results

  #103 opened 5 months ago by anihab
  2

• Results obtained in the original paper

  #127 opened a month ago by Zehui127
  0

• Shapley plot

  #126 opened 2 months ago by Anshullllllll
  0

• finetune results seems not very stable

  #125 opened 2 months ago by maris205
  1

• When will the code for pre-training model and training BPE tokenizer be available?

  #74 opened 7 months ago by a-green-hand-jack
  2

• Fine-tune for continuous labels

  #79 opened 9 months ago by buwanim
  4

• Is there a way to turn off the setting to use flash attention/triton library?

  #104 opened 5 months ago by vivektreddy
  3

• finetune error

  #123 opened 2 months ago by maris205
  1

• What is the appropriate accuracy of a pre-trained model?

  #122 opened 2 months ago by yuyadanyadan
  2

• The shape of embedding issue

  #121 opened 2 months ago by WhiteNan
  4

• Specific processes for pre-training

  #118 opened 2 months ago by yuyadanyadan
  3

• readme for GUE datasets?

  #120 opened 2 months ago by maris205
  0

• Logits and Labels are different shapes

  #119 opened 2 months ago by SamuelTWu
  0

• cls_token

  #115 opened 2 months ago by lilian-zh
  2

• Language Modelling Head

  #116 opened 3 months ago by tdsone
  2

• DNA Representation

  #117 opened 3 months ago by ChangShaole
  1

• Suggest better error message than 'assert q.is_cuda and k.is_cuda and v.is_cuda'

  #112 opened 4 months ago by richelbilderbeek
  2

• Suggest to supply example code that does not use Triton

  #114 opened 3 months ago by richelbilderbeek
  1

• How to use fine tuned model for prediction?

  #111 opened 4 months ago by Anshullllllll
  0

• Where are your Species Classification datasets?

  #110 opened 4 months ago by andreyfsch
  0

• I always encounter this error during the fine-tuning evaluation phase

  #78 opened 4 months ago by zlw1747832053
  3

• Is it possible to publish the detailed requirement file?

  #107 opened 4 months ago by nettanetta
  0

• Whether huggingface released model has been further pretrained on GUE benchmark

  #106 opened 5 months ago by Mamingqian
  1

• GUE+ datasets?

  #87 opened 7 months ago by leannmlindsey
  3

• How to specifically implement the task of Enhancer promoter interaction?

  #105 opened 5 months ago by yangzhao1230
  1

• Got negative train loss when do pretrain process

  #97 opened 5 months ago by Jingyao711
  3

• Data distribution in pretraining dataset

  #101 opened 5 months ago by leannmlindsey
  1

• About the pretrain data

  #100 opened 5 months ago by wyhsleep
  1

• Instability in reproducing GUE dataset result

  #102 opened 5 months ago by Mamingqian
  1

• What is the loss of pre-training of the published model?

  #98 opened 6 months ago by Wubeizhongxinghua
  2

• EPI datasets, not getting published results

  #99 opened 6 months ago by leannmlindsey
  3

• About the motif prediction function

  #94 opened 6 months ago by josecar24
  3

• Attention error

  #96 opened 6 months ago by HelloWorldLTY
  1

• Random issues still come up in use

  #95 opened 6 months ago by josecar24
  1

• About random factor in the embedding/tokenization process

  #91 opened 7 months ago by josecar24
  7

• CUDA out of memory

  #90 opened 6 months ago by chenruipu
  8

• GUE labels

  #93 opened 6 months ago by nehmea
  0

• TypeError: __init__() got an unexpected keyword argument 'token'

  #92 opened 6 months ago by Jingyao711
  1

• Getting embedding of a sequence

  #89 opened 7 months ago by CorvusVaine
  2

• Quickstart Does not work and Embedding Dim is not 768

  #75 opened 7 months ago by Leo-T-Zang
  1

• Discuss a question about k-mer

  #73 opened 7 months ago by xueleecs
  1

• Pretraining, Pretraining, Pretraining!!!

  #76 opened 7 months ago by multydoffer
  2

• Special token treatment.

  #81 opened 7 months ago by prwoolley
  1

• Cannot Reproduce DNA-BERT2‘s Result

  #86 opened 7 months ago by KatarinaYuan
  1

• Is it neccessary to train a specific BPE tokenizer on own datasets?

  #88 opened 7 months ago by amssljc
  1

• How do I output the attention from the model?

  #80 opened 9 months ago by jkb-ag
  1

• Unable to Retrieve ' hidden_states ' Despite ' Setting return_dict=True ' and ' output_hidden_states=True '

  #85 opened 8 months ago by biglittleme
  3

• splice site predictions

  #83 opened 8 months ago by amitpande74
  0

• hidden_states = model(inputs)[0] # [1, sequence_length, 768]-- Is the second dimension really the sequence length?

  #72 opened 10 months ago by xueleecs
  1