From 81cc9527e786c8667201d6435eba42804caca8f7 Mon Sep 17 00:00:00 2001 From: tooyosi Date: Thu, 19 Dec 2024 15:50:42 +0000 Subject: [PATCH] remove gpu all nvidia error command --- bajor/batch/train_finetuning.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/bajor/batch/train_finetuning.py b/bajor/batch/train_finetuning.py index 60bebf2..1b52041 100644 --- a/bajor/batch/train_finetuning.py +++ b/bajor/batch/train_finetuning.py @@ -234,7 +234,7 @@ def create_job_tasks(job_id, task_id=1, run_opts=''): '/bin/bash -c "' 'set -ex; ' f'{wait_for_preparation_task_completion}; ' - 'nvidia-smi; || echo \\"nvidia-smi command failed\\"; ' + 'nvidia-smi; ' 'python -c \\"import torch; print(torch.cuda.is_available()); print(torch.cuda.device_count())\\"; ' f'{setup_pytorch_kernel_cache_env_var}; ' f'{setup_hugging_face_cache_env_var}; '