| #SBATCH --job-name=wavlm_ssl_sv | |
| #SBATCH --output=slurm_%j | |
| #SBATCH --nodes=1 | |
| #SBATCH --ntasks=2 | |
| #SBATCH --gres=gpu:2 | |
| #SBATCH --cpus-per-task=10 | |
| #SBATCH --constraint=a100 | |
| #SBATCH --time=20:00:00 | |
| #SBATCH --hint=nomultithread | |
| #SBATCH --account=kdp@a100 | |
| module purge | |
| module load cpuarch/amd | |
| module load pytorch-gpu/py3/1.12.1 | |
| srun python -u trainSpeakerNet.py --config configs/wavlm_mhfa_dlg_lc.yaml --train_list exp/train_list_dino.txt --distributed |