how to compare ckiplab/bert-base-chinese with bert-base-chinese?
bmkor opened this issue · comments
Thanks so much for this excellent model and having it accessible in huggingface.
Would like to know why the ckiplab/bert-base-chinese
seems a bit strange to me when compared to the usual bert-base-chinese
which I think it mainly trained on simplified chinese. For instance, when I masked the word 風
of the phrase 颱風預測。
in the usual bert-base-chinese
it managed to give me back 風
with high probability 0.992; in contrast, in the ckiplab/bert-base-chinese
it didn't give back the masked word 風
in the top 5 but giving the word 的
with highest probability albeit only around 0.3 something which I am wondering.
Is it supposed that we have to fine-tune this MLM first? Or perhaps I interpreted it wrongly (as I'm very new in this field). Mind sharing a bit on your thought? Thanks very much and thanks in advance.
Based on the property of the training data and the task objective, this model performs poorly on short or incomplete sentences.
For longer sentences like 新聞報導颱[MASK]預測。
, this model managed to give me back 風
in high probability.
Thanks for pointing out the weakness of our model. We will keep working to make it better.