评测InternVL3-8B报错, AttributeError: 'InternVLModel' object has no attribute 'chat'

报错命令如下

Traceback (most recent call last):
  File "/nas/user/projects/VLMEvalKit/run.py", line 376, in main
    model = infer_data_job(
            ^^^^^^^^^^^^^^^
  File "/nas/user/projects/VLMEvalKit/vlmeval/inference.py", line 203, in infer_data_job
    model = infer_data(
            ^^^^^^^^^^^
  File "/nas/user/projects/VLMEvalKit/vlmeval/inference.py", line 166, in infer_data
    response = model.generate(message=struct, dataset=dataset_name)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/nas/user/projects/VLMEvalKit/vlmeval/vlm/base.py", line 116, in generate
    return self.generate_inner(message, dataset)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/nas/user/projects/VLMEvalKit/vlmeval/vlm/internvl/internvl_chat.py", line 467, in generate_inner
    return self.generate_v2(message, dataset)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/.local/lib/python3.11/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/nas/user/projects/VLMEvalKit/vlmeval/vlm/internvl/internvl_chat.py", line 422, in generate_v2
    response = self.model.chat(
               ^^^^^^^^^^^^^^^
  File "/home/user/.local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1931, in __getattr__
    raise AttributeError(
AttributeError: 'InternVLModel' object has no attribute 'chat'

执行的命令为

python run.py --data ChartQA_TEST --model InternVL3-8B --work-dir results/ --judge gpt-4o --mode all  --reuse

Start with vlmeval/vlm/internvl/internvl_chat.py, especially generate_inner and generate_v2, then inspect how the InternVL3-8B model is loaded before the failing self.model.chat call. Re-run the provided ChartQA_TEST command after tracing the model API. Done means the command no longer raises the reported AttributeError.