Use --use_cache but the responses are not saved

[XXXCARREY]

Thank you for your excellent contribution in lmms evaluation, I met a problem when saving the response when evaluate lmms on my curated dataset. The code was executed correctly except for there was no cache files in the folder I initiated in command:
I run:
python3 -m accelerate.commands.launch --num_processes=1 -m lmms_eval --model qwen2_vl --model_args pretrained="path/to//Qwen2-VL-7B-Instruct" --tasks dgqa --batch_size 1 --log_samples --log_samples_suffix Qwen2-VL-7B-Instruct_dgqa --output_path ./logs/ --limit 10 --use_cache ./cache/qwen_dgqa_
dgqa is my new task. While all other procedues(response and evaluation) could be executed correctly, there was no cache file in ./cache/qwen_dgqa_ folder. Please let me know how I can save this problem, thank you very much!