please make me to understand
Closed this issue · 2 comments
I'm facing the same problem, is there any ideas?
@IvyGao58
What has made me come out from the issue
facing problems in training, the code becomes silent after a few minutes for a long long time
is correcting file paths and running the setup_training.sh
properly without failure. Failure in running setup_training.sh
occurs due to incorrect file paths and failure in accessing Ontonotes-5.0.
If your code stands silently then it is not running, check your json files it may be empty.
about
* It does not use GPUs by default. Instead, it looks for the `GPU` environment variable, which the code treats as shorthand for `CUDA_VISIBLE_DEVICES`.
* The training runs indefinitely and needs to be terminated manually. The model generally converges at about 400k steps.
I don't know; But it identifies GPU automatically.
The issue may have caused because of requirements.txt
installs tensorflow-gpu>=1.13.1 (pip version) and your cudda(cudnn) may be incompatible with it.
Remove TensorFlow from requirements.txt
and install it vial conda, it may resolve the issue