File size: 1,559 Bytes
a88e206 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 | #!/usr/bin/env bash
#SBATCH --qos=regular
#SBATCH --job-name=xnli_llamainstruct70
#SBATCH --cpus-per-task=2
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=1
#SBATCH --mem=64GB
#SBATCH --gres=gpu:4
#SBATCH --constraint=a100-sxm4
#SBATCH --output=/scratch/jbengoetxea/phd/XNLIvar/scripts/generative/logs/xnli-llamainstruct70_%a.log
#SBATCH --error=/scratch/jbengoetxea/phd/XNLIvar/scripts/generative/logs/xnli-llamainstruct70_%a.err
#SBATCH --time=01:00:00 #ee-hh:mm:ss
#SBATCH --mail-type=REQUEUE
#SBATCH --mail-user=jaione.bengoetxea@ehu.eus
#SBATCH --array=0-10%2
source /scratch/jbengoetxea/phd/.phd_venv_new/bin/activate
export TRANSFORMERS_CACHE="/scratch/jbengoetxea/.cache"
# Values
DATASET_VALUES=(xnli-eu-biz xnli-eu-gip xnli-eu-naf xnli-eu-nat-biz xnli-eu-nat-gip xnli-eu-nat-naf)
PROMPT_TYPE_VALUES=(contradiction entailment neutral)
TASK_VALUES=(qa-zero qa-few)
# Get job array working
D=${#DATASET_VALUES[@]}
P=${#PROMPT_TYPE_VALUES[@]}
T=${#TASK_VALUES[@]}
TASK_ID=$SLURM_ARRAY_TASK_ID
IDX_D=$((TASK_ID / (P * T)))
IDX_P=$(((TASK_ID / T) % P))
IDX_T=$((TASK_ID % T))
DATASET="${DATASET_VALUES[$IDX_D]}"
PROMPT_TYPE="${PROMPT_TYPE_VALUES[$IDX_P]}"
TASK="${TASK_VALUES[$IDX_T]}"
# Final values and run script
MODEL=llama3instruct70
OUTPUT=/scratch/jbengoetxea/phd/XNLIvar/scripts/generative/results/$DATASET/$MODEL/$TASK
python3 /scratch/jbengoetxea/phd/XNLIvar/scripts/generative/scripts/zero_shot.py \
--dataset "${DATASET}" \
--model $MODEL \
--output_dir $OUTPUT \
--task $TASK \
--prompt_type "${PROMPT_TYPE}"
|