BrikerMan/Kashgari

Issues

• model.save：TypeError: Object of type module is not JSON serializable

  #461 opened 3 years ago by asd8095075
  8

• [Question] NER任务，使用load_model.predict()方法进行标签预测时，若句子长度超过128会被截断，应该如何处理。

  #499 opened 7 months ago by Doraemon1203-yue
  0

• Columns and DataType Not Explicitly Set on line 77 of classifications.py

  #498 opened a year ago by CodeSmileBot
  0

• [Question] 使用keras导入RAdam时报错

  #489 opened 2 years ago by RoseWang1
  1

• [Question] https://eliyar.biz/ can't open

  #490 opened 2 years ago by xiaoxiaofengzi
  2

• ner: cnn+lstm and bigru mod ，The code is the same

  #491 opened 2 years ago by weil0258
  1

• [BUG] BiLSTM_Model.load_model('saved_ner_model')源代码加载模型报错

  #493 opened 2 years ago by AnitaSherry
  1

• [BUG] 自定义模型，多个特征输入使用多个embed，模型fit报错，还需要重定义哪些方法来支持？

  #495 opened 2 years ago by Zikangli
  2

• 自定义模型 训练 保存 再倒入使用报错

  #463 opened 3 years ago by xrzlizheng
  5

• [Question]

  #494 opened 2 years ago by OmarMohammed88
  1

• Got 'Blas xGEMMBatched launch failed' using BERT + BiLSTM

  #497 opened a year ago by yanwii
  0

• [Question] 请问如何进行多任务训练和预测呢？

  #474 opened 2 years ago by GinkgoX
  1

• 报错: File "\Kashgari\kashgari\processors\sequence_processor.py", line 129, in transform seq = [self.token_bos] + seq + [self.token_eos] TypeError: can only concatenate list (not "str") to list[BUG]

  #466 opened 2 years ago by ouyangmingyuan
  2

• 1x版本利用的keras比tf.keras运行速度倍增

  #473 opened 2 years ago by xrzlizheng
  1

• Why can't I get reproducible results in even though I set the random seeds?

  #454 opened 2 years ago by ahmad-alismail
  5

• Seq2seq model tend to repeat

  #472 opened 2 years ago by TechSang
  7

• [Question] 模型保存出错

  #455 opened 2 years ago by ppvv6688
  6

• [Question] 如何输出predict的中间层value

  #457 opened 2 years ago by james-robinhood
  2

• [Question] Errors of Bi_LSTM classification model when using TF 2.4.1 and kashgari 2.0.1

  #459 opened 2 years ago by Harrypeas
  2

• 使用BiLSTM_CRF训练几个epoch后, val_accuracy断崖式下降

  #465 opened 2 years ago by who0sy
  4

• tensorflow.python.framework.errors_impl.InvalidArgumentError: Incompatible shapes: [1,512,768] vs. [128,1132,768]

  #476 opened 2 years ago by zcyzhuangzhou
  1

• 您好！序列标注中，想预测很多句子，一个一个去预测时间好慢，可以实现批量的运算吗？[Question]

  #477 opened 2 years ago by baketbek
  3

• [BUG] :Gradients do not exist for variables ['k_conditional_random_field/transitions:0']

  #478 opened 2 years ago by cccccrj
  1

• [Question] 关于继续学习添加新的实体

  #485 opened 2 years ago by SkyHunterC
  2

• [Question] How to adjust learning rate when I use Bi-LSTM model on classification task?

  #484 opened 2 years ago by honeysuckcle
  2

• [Question] labeling模型进行推理时随次数的增加而进程占用的内存逐渐增多，是否存在内存内存泄漏现象？

  #483 opened 2 years ago by mejackomg
  2

• [Question] 请问这个错误如何解决？

  #481 opened 2 years ago by tianjiajun-com
  2

• [Question] Can i adjust the learning rate of CRF?

  #482 opened 2 years ago by cccccrj
  1

• [Question] 分类采用bert embedding output_feature

  #479 opened 2 years ago by shiwl0329
  4

• 使用albert做嵌入的时候报错

  #486 opened 2 years ago by hwq458362228
  1

• [Question] 文本分类中的CNN开头的模型accuracy不管换数据集还是调参数都只有0.2

  #488 opened 2 years ago by hwq458362228
  1

• 简单调用BiLSTM_CRF模型，使用最基本bert-chinese作为embedding，运行报错layer_crf does not support masking

  #496 opened 2 years ago by Zikangli
  1

• [Question] 如何在GPU上进行训练

  #492 opened 2 years ago by AnitaSherry
  0

• [Question] v1.1.1 kashgari.embeddings.bert_embedding_v2.BERTEmbeddingV2 怎么获取得到的向量

  #467 opened 2 years ago by xqrshine
  1

• 导入BERTEmbedding失败

  #487 opened 2 years ago by Robinaoko
  2

• [val_loss值降不下来，最低也有10以上]

  #462 opened 3 years ago by songsa1
  1

• [Question]1.模型训练val_loss的值在40左右，这是什么问题？ 2.如何修改模型的学习率等相关的参数？ 3.no module named 'keras_radam'

  #464 opened 3 years ago by tianjiajun-com
  1

• ner任务 tf_serving 调用问题

  #480 opened 3 years ago by kehlaaa
  0

• [Question] 如何输出 top k 结果或者概率

  #456 opened 3 years ago by james-robinhood
  2

• 环境配置问题

  #475 opened 3 years ago by mading0817
  1

• Transformer Embedding

  #453 opened 3 years ago by ahmad-alismail
  1

• AttributeError: 'DoubleBLSTMModel' object has no attribute 'processor'

  #452 opened 3 years ago by ahmad-alismail
  2

• [Question] v2版本为什么模型分了3个文件，体积翻倍？

  #449 opened 3 years ago by yaleimeng
  7

• [BUG] 'BareEmbedding' object has no attribute 'analyze_corpus'

  #458 opened 3 years ago by ahmad-alismail
  3

• [BUG] AttributeError: module 'tensorflow.python.keras.utils.generic_utils' has no attribute 'populate_dict_with_module_objects'

  #450 opened 3 years ago by learnpythontheew
  1

• [Question] 为什么使用Bilstm+Crf时，实际有33个类别，但CRF转移矩阵尺寸大小为35*35，导致无法预测

  #468 opened 3 years ago by LiangYuHai
  0

• Add integrated gradients for more interpretability

  #448 opened 3 years ago by AliAshraf14
  1

• [Question] How to add customized tokenizer and dictionary(如何替换分词和加入自定义字典)

  #460 opened 3 years ago by gguan
  1

• [Question] Checkpoints

  #447 opened 3 years ago by OmarMohammed88
  1

• [BUG] BERT embedding

  #451 opened 3 years ago by ahmad-alismail
  1