| #!/bin/bash | |
| #SBATCH --job-name=p5_seq2seq_ft | |
| #SBATCH --partition=akya-cuda | |
| #SBATCH --nodes=1 | |
| #SBATCH --gres=gpu:4 | |
| #SBATCH --cpus-per-task=40 | |
| #SBATCH --mem=192G | |
| #SBATCH --time=24:00:00 | |
| #SBATCH --output=/arf/scratch/stakan/hitit-proje/logs/p5/08_seq2seq_ft_%j.log | |
| #SBATCH --error=/arf/scratch/stakan/hitit-proje/logs/p5/08_seq2seq_ft_%j.err | |
| source /arf/scratch/stakan/hitit-proje/hitit_ocr/scripts/pipeline_p5/_common.sh | |
| python3 hitit_ocr/src/seq2seq/train_image.py \ | |
| --config hitit_ocr/configs/seq2seq_hitit.yaml \ | |
| --phase finetune \ | |
| --encoder hitit_ocr/runs/p5/ssl_dinov3/checkpoint.pt \ | |
| --decoder hitit_ocr/runs/p5/seq2seq_text_pretrain/checkpoint.pt \ | |
| --data datasets/processed/seq2seq_pairs.jsonl \ | |
| --output hitit_ocr/runs/p5/seq2seq_image_ft/ | |