AttributeError: 'LlavaQwenConfig' object has no attribute 'mm_newline_position'

[linxid]

Before you open an issue, please check if a similar issue already exists or has been closed before.

When you open an issue, please be sure to include the following

• A descriptive title: [xxx] XXXX
• A detailed description

Thank you for your contributions!
when I run llava-one-vision, have a error.
My sh file is :

CUDA_VISIBLE_DEVICE=1 python3 -m accelerate.commands.launch --num_processes 1 --main_process_port 12345 -m lmms_eval \
    --model llava_onevision \
    --model_args pretrained=/home/yuxi.dwh/notebook/pre-trained/lmms-lab/llava-onevision-qwen2-0.5b-ov,conv_template=qwen_1_5,model_name=llava_qwen \
    --tasks videochatgpt \
    --batch_size 1 \
    --log_samples \
    --log_samples_suffix llava_onevision \
    --output_path ./logs/

and error is :

    if self.config.mm_newline_position == "grid":
  File "/home/yuxi.dwh/.local/lib/python3.10/site-packages/transformers/configuration_utils.py", line 263, in __getattribute__
    return super().__getattribute__(key)
AttributeError: 'LlavaQwenConfig' object has no attribute 'mm_newline_position'