export CUDA_VISIBLE_DEVICES=3 python -u classification.py \ --task_name "classification" \ --dataset_name "PEMS-SF" \ --data "UEA" \ --root_path "./datasets/PEMS-SF" \ --forward_layers 2 \ --conv1d True \ --out_channels 32 \ --stride 4 \ --batch_size 32 \ --learning_rate 0.0001 \ --num_workers 0 \ --lradj "type1" \ --patience 10 \