LAVT-RIS/donghwa/scripts/submit_train_refcoco+.sh

#!/bin/bash

#SBATCH --job-name=lavt_cc+      # Submit a job named "example"
#SBATCH [email protected]
#SBATCH --mail-type=BEGIN,END,FAIL
#SBATCH --partition=a100         # a6000 or a100
#SBATCH --gres=gpu:2
#SBATCH --time=7-00:00:00        # d-hh:mm:ss, max time limit
#SBATCH --mem=84000              # cpu memory size
#SBATCH --cpus-per-task=8        # cpu num
#SBATCH --output=log_refcoco+_lavt_one.txt         # std output filename

ml cuda/11.0                # 필요한 쿠다 버전 로드
eval "$(conda shell.bash hook)"  # Initialize Conda Environment
conda activate lavt             # Activate your conda environment



# train
# mkdir ./models

# mkdir ./models/refcoco+
# CUDA_VISIBLE_DEVICES=0,1,2,3 python -m torch.distributed.launch --nproc_per_node 4 --master_port 12345 train.py --model lavt --dataset refcoco+ --model_id refcoco+ --batch-size 8 --lr 0.00005 --wd 1e-2 --swin_type base --pretrained_swin_weights ./pretrained_weights/swin_base_patch4_window12_384_22k.pth --epochs 40 --img_size 480 2>&1 | tee ./models/refcoco+/output

mkdir ./models/refcoco+_lavt_one
srun python -m torch.distributed.launch --nproc_per_node 2 --master_port 11115 train.py --model lavt_one --dataset refcoco+ --model_id refcoco+_lavt_one --batch-size 14 --lr 0.00005 --wd 1e-2 --swin_type base --pretrained_swin_weights ./pretrained_weights/swin_base_patch4_window12_384_22k.pth --epochs 50 --img_size 480 2>&1 | tee ./models/refcoco+_lavt_one/output