Unknown Vision Tower

[Blueyyds]

I do the inference, but i met an issue.
Here is my inference script.
`
export CUDA_VISIBLE_DEVICES=5

python -m llava.eval.model_vqa_lora
--model-base liuhaotian/llava-v1.5-13b
--model-path listen2you002/ChartLlama-13b
--question-file playground/data/chartqa/test_aug.jsonl
--image-folder playground/data/chartqa/png
--answers-file playground/eval/chartqa/chartllama/ans.jsonl
--conv-mode v1
--num-chunks 1
`
ValueError: Unknown vision tower: /mnt/share_1227775/yandali/multimodal/models/ft_local/clip-vit-large-patch14-336/

[coffee3699]

You may rename the "mm_vision_tower" in config.json to openai/clip-vit-large-patch14-336 or your local downloaded path. It works for me! @Blueyyds