I tried using the ALBEF VQA finetuned model along with the the vqa_test_dev.json (provided for downstream tasks) and ran evaluation on it. But when I uploaded the result.json file to the official server it failed (multiple times). Screenshot and error report attached as screenshots. Any feedback?
I tried using the ALBEF VQA finetuned model along with the the vqa_test_dev.json (provided for downstream tasks) and ran evaluation on it. But when I uploaded the result.json file to the official server it failed (multiple times). Screenshot and error report attached as screenshots. Any feedback?