ASDA/scripts/train1.sh

#!/bin/bash
#SBATCH --job-name=asda-sanity
#SBATCH --partition=a6000
#SBATCH --gres=gpu:1
#SBATCH --time=13-11:30:00
#SBATCH --mem=28000
#SBATCH --cpus-per-task=3
#SBATCH --output=./exp_oiou/refcocop_unc_oiou_bs28.log

ml purge
ml load cuda/11.8
eval "$(conda shell.bash hook)"
conda activate asda

cd /data2/projects/chaeyun/ASDA

export NCCL_P2P_DISABLE=1
export NVIDIA_TF32_OVERRIDE=1
export NCCL_IB_TIMEOUT=100
export NCCL_IB_RETRY_CNT=15

export MASTER_PORT=8820

export CUDA_VISIBLE_DEVICES=0

# python train_oiou.py --dataset refcocog --splitBy umd --ngpu 1 --batch_size 28 --time 17 --savename gref_umd_oiou_bs28

python train_oiou.py --dataset refcoco+ --splitBy unc --ngpu 1 --batch_size 28 --time 17 --savename refcocop_unc_oiou_bs28


# python train.py --dataset refcoco+ --splitBy unc --ngpu 1 --batch_size 28 --time 17 --savename refcocop_bs28_repro 
# python train.py --dataset refcoco+ --splitBy unc --ngpu 1 --batch_size 36 --time 17 --savename refcocop_bs36_repro 


# python train.py --dataset refcoco --splitBy unc --ngpu 1 --batch_size 36 --time 17 --savename refcoco_bs36_repro 
# python train.py --dataset refcocog --splitBy umd --ngpu 1 --batch_size 28 --time 17 --savename gref_umd_bs28_repro
# python train.py --dataset refcocog --splitBy umd --ngpu 1 --batch_size 36 --time 17 --savename gref_umd_bs36_repro


# export CUDA_VISIBLE_DEVICES=0,1; python train.py --dataset refcocog --splitBy umd --ngpu 2 --batch_size 64 --time 17 --savename gref_umd_bs64_repro

# export CUDA_VISIBLE_DEVICES=0,1; python train.py --dataset refcocog --splitBy umd --ngpu 1 --batch_size 64 --time 17 --savename gref_umd_bs64_repro