amazon-science/sccl

Issues

• Help running a custom dataset of short texts via a notebook

  #31 opened a year ago by Viole-Grace
  0

• What is the training dataset used in the paper?

  #30 opened a year ago by rxlian
  2

• csv files doesn't exist

  #16 opened 3 years ago by SouthWindShiB
  9

• The original clustering results for News-T, News-S, News-TS and Tweet

  #28 opened 2 years ago by srzxDragon
  0

• contrastive is inconsistency between papers and code

  #27 opened 2 years ago by pickwu
  0

• Where is the model saved? How do I use it?

  #25 opened 2 years ago by AnitaSherry
  1

• Virtual Or Explit

  #19 opened 2 years ago by Tiantian-ZHU
  1

• BUG?: There is an extra log() in the code when calculating cluster_loss?

  #26 opened 2 years ago by wulaoshi
  1

• performance with virtual augmentation

  #24 opened 2 years ago by kiucho
  0

• Stopping criterion

  #17 opened 2 years ago by urospet
  1

• Has anyone run out the results of the paper through the hyper parameters provided in the paper or in the source code?

  #22 opened 2 years ago by hmllmh
  4

• Did you get the score for searchsnippets in the paper by using the plain distilbert as backbone rather than sentence bert?Can you do early stopping whenever the losses stop decreases or the prediction stops changing much?

  #23 opened 2 years ago by hmllmh
  0

• Is the model effective?

  #4 opened 3 years ago by yanhan19940405
  3

• Can‘t achieve the scores in paper

  #3 opened 2 years ago by PaffxAroma
  9

• No code found for dataset partitioning

  #21 opened 2 years ago by gzpbbd
  1

• Not sure how the augmented pairs generated by the Contextual Augmenter

  #14 opened 2 years ago by zexuanqiu
  1

• Can't achive the same score in paper

  #8 opened 3 years ago by kirikokonoka
  4

• How to run clustering on multiple GPUs?

  #11 opened 2 years ago by abhi1nandy2
  0

• How is the training data generated?

  #9 opened 2 years ago by jx1100370217
  1

• Twitter dataset not found?

  #2 opened 2 years ago by urospet
  5

• Data Augmentation？

  #1 opened 2 years ago by WuDiDaBinGe
  5

• I have bug about AutoConfig.from_pretrained()

  #20 opened 2 years ago by Tiantian-ZHU
  1

• Bug in confusion.add

  #12 opened 3 years ago by MrRace
  5

• I have a bug about class Confusion.add()

  #10 opened 3 years ago by oYoungCo
  10

• Results from HAC-SD

  #7 opened 3 years ago by urospet
  1

• Worse results with Banking Dataset

  #13 opened 3 years ago by rajat-tech-002
  2

• use embedder.encode in get_kmeans_centers, but use sentbert.forward() to get the representation in SCCLBert

  #15 opened 3 years ago by jkkl
  0

• Diff between acc and acc_model,[representation] and [model]

  #6 opened 3 years ago by nzy1998
  0

• Doubt regarding dataloader file

  #5 opened 3 years ago by Sohanpatnaik106
  1