unable to achive 91.82 in ner

[omerarshad]

Thanks for the code, but i am unable to achieve F1 91.82 in ner by using the same config given in repo. Any help?

[TanyaZhao]

@omerarshad
Hi friend~I also can not achieve the result 91.82.

And I quite doubt that the hyperparameters used in this project are quite different from that used in Ma's paper (End-to-end Sequence Labeling via Bi-directional LSTM-CNNs-CRF). The dimensions of the hyperparameters are much larger.

So, I doubt that is this comparable to Ma's model ?
btw, could you share with me what F1 score you achieve ?

Thx~

[omerarshad]

I was able to achieve 91.05 after lot of hyper parameter tuning using Glove embeddings.

[TanyaZhao]

cool !
Could you tell me some key parameters ?
I was suffering from tuning parameters for a long time, and haven't got an idea result...

[omerarshad]

what's your best score achieved?

[26hzhang]

Sorry guys, I already changed the error information in README.md, since I was wrongly use the CoNLLeval.py function.

[omerarshad]

What did you changed? Do we need to pull your code again?