-
Notifications
You must be signed in to change notification settings - Fork 2.5k
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
Segmentation fault (core dumped) #210
Comments
Same problem with version 2. I get pass that part but runs into another problem. (idk how to put in code segment) ****\env\lib\site-packages\torch\nn\modules\conv.py:456: UserWarning: Plan failed with a cudnnException: CUDNN_BACKEND_EXECUTION_PLAN_DESCRIPTOR: cudnnFinalize Descriptor Failed cudnn_status: CUDNN_STATUS_NOT_SUPPORTED (Triggered internally at ..\aten\src\ATen\native\cudnn\Conv_v8.cpp:919.) 0%| | 0/1 [00:00<?, ?it/s]Some weights of the model checkpoint at bert-base-uncased were not used when initializing BertForMaskedLM: ['bert.pooler.dense.bias', 'bert.pooler.dense.weight', 'cls.seq_relationship.bias', 'cls.seq_relationship.weight']
|
I had to add a bunch of CUDA shared libraries to my laptop's search path. No idea why this didn't happen automatically like it does for every other project. I'll probably have to debug this for real eventually. I'm using python through pyenv with the venv module for a virtual environment.
|
The following error occurs.
I am using Ubuntu 22.04, and I utilized the provided conda environment.
The text was updated successfully, but these errors were encountered: