naive approach model run with unknown token

main.py

@@ -444,19 +444,20 @@ def custom_collate_fn(batch, max_len_specs, vocabulary, folder, specs_folder, de
 def setup_parser():
     out = argparse.ArgumentParser()
     # # cluster version
-    folder_given = "naive_approach"
-    out.add_argument('--get_vocabulary', default=f'/scratch/grzonkow/vocabulary.pkl', type=str, help="Path to load vocabulary")
-    out.add_argument('--save_vocabulary', default=f'/scratch/grzonkow/{folder_given}/vocabulary.pkl',
+    folder_given = "Model_naive_approach"
+    out.add_argument('--get_vocabulary', default=f'/scratch/grzonkow/{folder_given}/vocabulary_{folder_given}.pkl', type=str, help="Path to load vocabulary")
+    out.add_argument('--save_vocabulary', default=f'/scratch/grzonkow/{folder_given}/vocabulary_new_{folder_given}.pkl',
                      type=str, help="Path to safe vocabulary")
 
     # if built_Dataset, dataset does not matter
     out.add_argument('--built_dataset', default=f'/scratch/grzonkow/output/', type=str, help="Path to dataset folder")
+    out.add_argument('--built_dataset_charts', default=f'/scratch/grzonkow/{folder_given}/output/', type=str, help="Path to dataset folder")
     out.add_argument('--dataset', default=None, type=str, help="Path to dataset")
     out.add_argument('--save_dataset', default=f'/scratch/grzonkow/{folder_given}/train_dataset_{folder_given}.pkl',
                      type=str, help="Path to safe dataset")
     out.add_argument('--old_version', default=False, type=str,
                      help="Indicates if old version without start end tokens need to be loaded")
-    out.add_argument('--old_version_voc', default=True, type=str,
+    out.add_argument('--old_version_voc', default=False, type=str,
                      help="Indicates if old version without start end tokens need to be loaded")
 
     out.add_argument('--folder', default=r'/work/MLShare/StepMania/data/cleaned/allowed_meter_difference_of_2/',
@@ -693,7 +694,7 @@ if __name__ == '__main__':
             f'{args.built_dataset}/all_waves/',
             f'{args.built_dataset}/all_sample_rates/',
             f'{args.built_dataset}/all_difficulties/',
-            f'{args.built_dataset}/all_charts/',
+            f'{args.built_dataset_charts}/all_charts/',
             f"{args.built_dataset}/mel_specs_{win_length_list_ms[0]}/",
             f"{args.built_dataset}/mel_specs_{win_length_list_ms[1]}/",
             f"{args.built_dataset}/mel_specs_{win_length_list_ms[2]}/",

model_run.sbatch

 #!/bin/bash
 #SBATCH -t 7-00:00:00                    # time limit set to 1 week, 1 day 1-00:00:00
-#SBATCH -J Naive_model                  # the job name
+#SBATCH -J Model_naive_approach                  # the job name
 #SBATCH --mail-type=END,FAIL,TIME_LIMIT  # send notification emails
 #SBATCH -n 5                             # use 5 tasks
 #SBATCH --cpus-per-task=1                # use 1 thread per taks
@@ -9,8 +9,8 @@
 #SBATCH --partition=informatik-mind
 #SBATCH --gpus=1                         # request 1 GPU
 #SBATCH --gpu_cmode=shared               # Set the GPU into shared mode, so that multiple processes can run on it
-#SBATCH --output=/scratch/grzonkow/naive_approach/model.txt         # capture output
-#SBATCH --error=/scratch/grzonkow/naive_approach/err.txt          # and error streams
+#SBATCH --output=/scratch/grzonkow/Model_naive_approach/model.txt         # capture output
+#SBATCH --error=/scratch/grzonkow/Model_naive_approach/err.txt          # and error streams
 
 
 
@@ -22,7 +22,8 @@ unset LD_LIBRARY_PATH
 #pip install -r requirements.txt
 #pip install simfile
 #python -u main.py --get_vocabulary /scratch/grzonkow/vocabulary.pkl --dataset /scratch/grzonkow/train_dataset_latest.pkl --processes $SLURM_NTASKS --threads $SLURM_CPUS_PER_TASK "$@"
-python -u main.py --get_vocabulary /scratch/grzonkow/vocabulary.pkl --built_dataset /scratch/grzonkow/output  --processes $SLURM_NTASKS --threads $SLURM_CPUS_PER_TASK "$@"
+#python -u main.py --get_vocabulary /scratch/grzonkow/vocabulary.pkl --built_dataset /scratch/grzonkow/output  --processes $SLURM_NTASKS --threads $SLURM_CPUS_PER_TASK "$@"
+python -u main.py --processes $SLURM_NTASKS --threads $SLURM_CPUS_PER_TASK "$@"
 #python -u main.py --get_vocabulary /scratch/grzonkow/vocabulary.pkl --processes $SLURM_NTASKS --threads $SLURM_CPUS_PER_TASK "$@"
 #python -u main.py --processes $SLURM_NTASKS --threads $SLURM_CPUS_PER_TASK "$@"
 conda deactivate