Can’t generate my own dataset using load_dataset

I am trying to upload an image dataset to huggingface but im facing problems with load_dataset function.
the code:
for fold in tqdm(range(FOLDS_N), “Uploading Fold”):
print(“fold:”, fold)
train_dataset = load_dataset(“imagefolder”, data_dir=f"{DRIVE_RP_FOLDER}/{FOLDS_N}-fold/fold-{fold}“, split=f"train”)
train_dataset.push_to_hub(f"{HUGG_DATASET_NAME}fold{fold}“, private=HUGG_DATASET_IS_PRIVATE)
train_dataset = load_dataset(“imagefolder”, data_dir=f”{DRIVE_RP_FOLDER}/{FOLDS_N}-fold/fold-{fold}“, split=f"test”)
train_dataset.push_to_hub(f"{HUGG_DATASET_NAME}fold{fold}", private=HUGG_DATASET_IS_PRIVATE)
The execution msg that I obtained was .

also the images has the following format 128x128 pixels
names 1600x1.png 1600x2.png

I watched the path and seems to be ok.

DRIVE_RP_FOLDER=
/home/adrianoggm/Escritorio/TFG/data/WISDM/plots/recurrence_plots/sampling_loto

I tried to install and reinstall datasets but didnt worked.

the code clearer: