bojone/bert_in_keras

Issues

• 请问nl2sql的数据集可以提供一下吗？谢谢

  #20 opened 5 years ago by lkkljk
  5

• 大佬您好，我在调用官方预训练模型，即代码为bert_model = load_trained_model_from_checkpoint(config_path, checkpoint_path, seq_len=None)这一语句时，出现了如下问题，可以请教一下为何吗，谢谢

  #28 opened 4 years ago by 2941894638
  1

• 加载自己训练完成的模型后预测用时很久

  #25 opened 5 years ago by zhoujuanshanghai
  1

• 关于relation_extraction例子中的反例样本

  #23 opened 5 years ago by lan2720
  2

• run nl2sql_baseline.py meet the error!

  #9 opened 5 years ago by FakerYFX
  5

• 使用sentiment.py进行训练完，模型保持之后，加载提示存在自定义layer，此处需要如何处理

  #18 opened 5 years ago by gm19900510
  4

• k1, k2 = np.array(items.keys()).T

  #27 opened 5 years ago by a461925358
  0

• AttributeError: module 'keras.backend' has no attribute 'tf'

  #26 opened 5 years ago by a461925358
  0

• NL2SQL复现X-SQL问题

  #24 opened 5 years ago by ningshixian
  0

• relation_extract 这个模型在原数据集上，始终是0分。

  #22 opened 5 years ago
  0

• sentiment.py中是数据文件从哪来的？

  #13 opened 5 years ago by wan560
  1

• 请教关系抽取中的repair函数

  #21 opened 5 years ago
  0

• mask在bert中的问题

  #19 opened 5 years ago by xucncn
  0

• 希望修复的BUG，关于max_position_embeddings超参数

  #17 opened 5 years ago by sangensong
  0

• NL2SQL

  #14 opened 5 years ago by goodnessSZW
  0

• 关于关系抽取的数据集，请教下

  #16 opened 5 years ago by smartcatdog
  0

• nl2sql

  #15 opened 5 years ago by 489597448
  0

• 关于主题抽取的数据集是哪个

  #12 opened 5 years ago by captain2017
  4

• 验证集loss降不下去怎么办

  #5 opened 5 years ago by hc944822354
  7

• 关于主题抽取中的代码

  #4 opened 5 years ago by acmilannesta
  4

• 您好，请教一下，在nl2sql的例子中，pcsel的shape=(?, ?, ?, ?)，会导致pcsel_loss的计算报如下错误，能否指导一下怎么处理？谢谢！

  #11 opened 5 years ago by axdwss
  2

• sentiment analysis 那个例子中如何构建 prediction 函数？

  #8 opened 5 years ago by acacess
  2

• run subject_extract.py meet the error!

  #10 opened 5 years ago by FakerYFX
  1

• 如何对bert微调后的模型进行加载？

  #7 opened 5 years ago by wangxing9315
  3

• 不好意思 作者 能不能问一下数据的格式是怎么样的

  #3 opened 5 years ago by kedimomo
  5

• 将情感分类改为三分类时遇到expected dense_1 to have 2 dimensions, but got array with shape (32, 1, 3)

  #6 opened 5 years ago by xxxxxxxxy
  0

• where is the dataset

  #2 opened 5 years ago by SeekPoint
  1

• 关于详细介绍

  #1 opened 5 years ago by Cooper111
  5