Open Jingut opened 2 months ago
I tried to run train_caption.py, but it says COCO_val2014_000000184613.jpg image could not be found in the eval() part of coco_karpathy_dataset.py.
Is there any solution?
have you downloaded COCO dataset? The json file created by the script just provided a splitting strategy but no actual data. You have to download data by yourself.
I tried to run train_caption.py, but it says COCO_val2014_000000184613.jpg image could not be found in the eval() part of coco_karpathy_dataset.py.
Is there any solution?