caixiong
caixiong
  ubuntu系统版本时20.04,系统语言时英文,并且也是乱码状态,一直还没解决, 感觉还不如用网页版来来得方便啊
Hi, Thanks for your this great work! A little problem confuses me in the paper is the baseline model: WaveNet(MoL). In the paper, WaveNet(MoL) model is cited as "Parallel wavenet:...
就是按照教程在:ChineseGLUE/baselines/models/ernie 目录下执行: bash run_classifier_inews.sh, 然后执行之前改了epoch为100, 因为之前epoch是3也是0.1的accuracy, 但是改成epoch=100也是0.1 ps: 下面的是在test set上的结果, 在dev上是0.81, 请问为什么差别这么大呢? I0721 19:18:55.035024 139930260309824 run_classifier.py:1507] eval_accuracy = 0.1031031 INFO:tensorflow: eval_loss = 9.384537 I0721 19:18:55.035102 139930260309824 run_classifier.py:1507] eval_loss = 9.384537...
Hi, guys! I tested some common Chinese Mandarin texts. The g2pM model gets all error results, and pypinyin get all correct results. Here are the examples I tested. 
 As show in the above picture, it should be 'e_control' instead of 'p_control' (red rectangle) in model/modules.py
点击+选中几个topic后,然后点击save,没有反应,刷新后,在首页还是没有添加上
Hello, how many languages are the training data? Aand what is the amount of data in various languages look like?