资源分类
》 资源 》算法》BERT-tensorflow

BERT-tensorflow

2020-03-10 | |  51 |   0 |   0

PROGRESS

10%

USAGE

Edit the file path in bert.py

Put bert.py to https://github.com/tensorflow/tensor2tensor/tree/master/tensor2tensor/data_generators

Edit https://github.com/tensorflow/tensor2tensor/blob/master/tensor2tensor/data_generators/all_problems.py

Then run https://github.com/tensorflow/tensor2tensor/blob/master/tensor2tensor/bin/t2t_trainer.py

With parameters:

--generate_data --problem=bert  --data_dir=../../t2t_data --tmp_dir=../../t2t_data/tmp  --model=transformer_encoder --hparams_set=transformer_tiny --output_dir=../../t2t_train/lm  --train_steps=1000  --eval_steps=100

EXPERIMENT RESULT


next sentence predict
accuracy90%


上一篇：BERT-chinese

下一篇：BERT-Keyword-Extractor

用户评价
全部评价

热门资源

  • Keras-ResNeXt

    Keras ResNeXt Implementation of ResNeXt models...

  • seetafaceJNI

    项目介绍 基于中科院seetaface2进行封装的JAVA...

  • spark-corenlp

    This package wraps Stanford CoreNLP annotators ...

  • capsnet-with-caps...

    CapsNet with capsule-wise convolution Project ...

  • inferno-boilerplate

    This is a very basic boilerplate example for pe...