Hi, First of all Thank you for sharing nice works.
It seems really good in NER tasks.
I wanna check with my own dataset, So I customized my data and have trained.
In this steps, I have some questions as below.
How can I add tensorboard log with metrics(ex. mEA, mEF, mER, mEP).
I added the function self.writer.add_scalar("mEP/steps", val_result_dict['overall']['mEP']) after Step Validation line, because I wanna check the metric results in every steps.
But It returns like below, mEA steps is not same with other metric(gl_loss). I don't know the way to get right steps.
Is there any function to visualize the relations of each entities.(like Key-Values)?
I wanna check that model predicted right Key-Value mapping by graph embedding(GCN).. Do you have any functions about it?
If not so, could you please give some tips to implement function?
Is there any visualization function to inference the result of models on each of images+bbox+predicted labels like below?
Hi, First of all Thank you for sharing nice works. It seems really good in NER tasks.
I wanna check with my own dataset, So I customized my data and have trained. In this steps, I have some questions as below.
How can I add tensorboard log with metrics(ex. mEA, mEF, mER, mEP).
self.writer.add_scalar("mEP/steps", val_result_dict['overall']['mEP'])
after Step Validation line, because I wanna check the metric results in every steps.Is there any function to visualize the relations of each entities.(like Key-Values)?
Is there any visualization function to inference the result of models on each of images+bbox+predicted labels like below?![image](https://user-images.githubusercontent.com/30281608/174536925-e8cf52e5-b37a-4e18-9d87-86bc0fc2a831.png)
I'll wait for your reply. Thank you 😊