| export PATH=/venv/main/bin:/usr/local/bin:/usr/bin:/bin | |
| export PYTHONDONTWRITEBYTECODE=1 | |
| export PYTHONUNBUFFERED=1 | |
| export CUDA_VISIBLE_DEVICES=0 | |
| export WORK=/dev/shm/eval | |
| export QUIP_DIR=$WORK/quip-sharp | |
| export HF_DATASETS_TRUST_REMOTE_CODE=1 | |
| export PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True | |
| cd $WORK/lm-evaluation-harness | |
| python -u $WORK/eval_polish_quip.py \ | |
| --model_path $WORK/model \ | |
| --tokenizer speakleash/Bielik-11B-v2.3-Instruct \ | |
| --output_dir $WORK/results_eq_bench \ | |
| --num_fewshot 5 \ | |
| --batch_size 32 \ | |
| --tasks polish_eq_bench polish_eq_bench_first_turn | |