search

allenai / commonsense-kg-completion

MIT License
105 stars 16 forks source link

The scripts kills itself during the evaluation process. #17

Closed RomanShen closed 2 years ago

RomanShen commented 2 years ago

Hi, it's a great job. I'm trying to run the evaluation using the following command:

CUDA_VISIBLE_DEVICES={GPU_ID} python src/run_kbc_subgraph.py --dataset atomic --sim_relations --use_bias --load_model {PATH_TO_PRETRAINED_MODEL} --eval_only --write_results

I create a new vitural environment with conda using the requirements file in the repo. I have downloaded ATOMIC model here and BERT embeddings at this link. However, the script kills itself when finishing the loop without any error or warning. I am running the script on 128GB RAM UBUNTU 20.04 with NVIDIA TITAN RTX. So it may not be related to the resource limitation.

Could you please help me find out why it kills itself? Thank you very much!