This is a sentiment analysis method based on K-BERT and residual cyclic units
This is a sentiment analysis method based on K-BERT and residual cyclic units. For the K-BERT training model, please refer to K-BERT ,For SRU, please refer to SRU..
Because the code needs to be organized and privacy protected, and there are other researches in the sequence, it will be disclosed at an appropriate time.
Before running the code, make sure that the CUDA tookit, CUDNN, nvcc-V, and graphics card drivers used are the same. You can use sru_environment_test.py to test whether the SRU environment is normal;