Download code/scripts/pipeline/evaluate_e2e.slurm from savastakan/hitit-cuneiform-ocr: direct link, hf CLI and curl.
- Browser
- Download file 649 Bytes
-
https://huggingface.co/savastakan/hitit-cuneiform-ocr/resolve/main/code/scripts/pipeline/evaluate_e2e.slurm
- Command line
-
hf download hf://savastakan/hitit-cuneiform-ocr/code/scripts/pipeline/evaluate_e2e.slurm
-
curl -L -o evaluate_e2e.slurm https://huggingface.co/savastakan/hitit-cuneiform-ocr/resolve/main/code/scripts/pipeline/evaluate_e2e.slurm
649 Bytes
| #SBATCH --job-name=eval_e2e | |
| #SBATCH --partition=akya-cuda | |
| #SBATCH --nodes=1 | |
| #SBATCH --gres=gpu:2 | |
| #SBATCH --cpus-per-task=20 | |
| #SBATCH --mem=96G | |
| #SBATCH --time=12:00:00 | |
| #SBATCH --output=/arf/scratch/stakan/hitit-proje/logs/pipeline/eval_%j.log | |
| cd /arf/scratch/stakan/hitit-proje | |
| source ~/.bashrc 2>/dev/null || true | |
| python3 hitit_ocr/src/evaluate_e2e.py \ | |
| --detection-ckpts hitit_ocr/runs/detection_v1_fold{0..4}/best.pt \ | |
| --classifier-ckpt hitit_ocr/runs/classification_v1/best.pt \ | |
| --translit-ckpt hitit_ocr/runs/transliteration_v1/best.pt \ | |
| --gold-set datasets/eval_gold/manifest.jsonl \ | |
| --output hitit_ocr/runs/eval_e2e_v1/ | |