-
Notifications
You must be signed in to change notification settings - Fork 26
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Evaluation and Prediction #82
Comments
Hi @mayinghan, I can't repro this issue and might need more information for investigating. For example, full log, config file, data size and environment setting will be helpful |
@stevezheng23 thanks for the reply. In the dev set, there are 30426 entities (3209 sentences). In the test set, there are around 442180 entities (45053 sentences).
My tensorflow is using 1.13.0 , python version is 3.7.3 |
This is the log for prediction. The one for evaluation is very similar to this one
|
the log looks normal to me, you might want to check the GPU usage to confirm whether the evaluation job is still running, since the evaluation method doesn't print any log while generating the result, and you can also use a smaller evaluation set (10 - 20 examples) to confirm this |
@stevezheng23 thanks. Speaking of the cpu usage, I am currently using xlnet large as the pretrain model. My GPU has 24 GiB memory. However, no matter how I decrease batch size and max_seq_length, the model always eats up like 23 GiB. Is that normal? |
It should not take that much GPU memory for small batch_size and short max_seq_length, but it's possible that most GPU memory is occupied even though not fully utilized. You can try re-config |
Hi,
I was trying to run the NER task on a customized dataset. The training process was successful. However, when it went to evaluation and prediction step, the program stuck at
INFO:tensorflow:Done running local_init_op.
and not moving forward. Is there any potential fix on this problem?Here is the log
The text was updated successfully, but these errors were encountered: