Paligemma2-3b-pt-448 could train but failed to chat

[hanggun]

Reminder

• I have read the above rules and searched the existing issues.

System Info

• llamafactory version: 0.9.2.dev0
• Platform: Linux-5.15.0-86-generic-x86_64-with-glibc2.35
• Python version: 3.12.3
• PyTorch version: 2.6.0+cu124 (GPU)
• Transformers version: 4.49.0
• Datasets version: 3.2.0
• Accelerate version: 1.2.1
• PEFT version: 0.12.0
• TRL version: 0.9.6
• GPU type: NVIDIA A100-PCIE-40GB
• GPU number: 8
• GPU memory: 39.39GB

Reproduction

Please help me, I used this setting but the paligemma2 can not chat

the error says:

Traceback (most recent call last):
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/gradio/queueing.py", line 625, in process_events
    response = await route_utils.call_process_api(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/gradio/route_utils.py", line 322, in call_process_api
    output = await app.get_blocks().process_api(
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/gradio/blocks.py", line 2042, in process_api
    result = await self.call_function(
             ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/gradio/blocks.py", line 1601, in call_function
    prediction = await utils.async_iteration(iterator)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/gradio/utils.py", line 728, in async_iteration
    return await anext(iterator)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/gradio/utils.py", line 722, in __anext__
    return await anyio.to_thread.run_sync(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/anyio/to_thread.py", line 56, in run_sync
    return await get_async_backend().run_sync_in_worker_thread(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/anyio/_backends/_asyncio.py", line 2461, in run_sync_in_worker_thread
    return await future
           ^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/anyio/_backends/_asyncio.py", line 962, in run
    result = context.run(func, *args)
             ^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/gradio/utils.py", line 705, in run_sync_iterator_async
    return next(iterator)
           ^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/gradio/utils.py", line 866, in gen_wrapper
    response = next(iterator)
               ^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/src/llamafactory/webui/chatter.py", line 188, in stream
    for new_text in self.stream_chat(
  File "/home/ps/data/LLaMA-Factory/src/llamafactory/chat/chat_model.py", line 112, in stream_chat
    yield task.result()
          ^^^^^^^^^^^^^
  File "/home/ps/anaconda3/envs/py3.12.3/lib/python3.12/concurrent/futures/_base.py", line 456, in result
    return self.__get_result()
           ^^^^^^^^^^^^^^^^^^^
  File "/home/ps/anaconda3/envs/py3.12.3/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
    raise self._exception
  File "/home/ps/data/LLaMA-Factory/src/llamafactory/chat/chat_model.py", line 129, in astream_chat
    async for new_token in self.engine.stream_chat(
  File "/home/ps/data/LLaMA-Factory/src/llamafactory/chat/hf_engine.py", line 390, in stream_chat
    stream = self._stream_chat(*input_args)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/venv/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/src/llamafactory/chat/hf_engine.py", line 275, in _stream_chat
    gen_kwargs, _ = HuggingfaceEngine._process_args(
                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ps/data/LLaMA-Factory/src/llamafactory/chat/hf_engine.py", line 184, in _process_args
    value = torch.stack([torch.stack(v) for v in value])
                         ^^^^^^^^^^^^^^
TypeError: expected Tensor as element 0 in argument 0, but got int

Others

No response

[hanggun]

我发现输入不是tensor

pixel_values
tensor([[[[1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.],
          ...,
          [1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.]],

         [[1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.],
          ...,
          [1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.]],

         [[1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.],
          ...,
          [1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.],
          [1., 1., 1.,  ..., 1., 1., 1.]]]])
token_type_ids
[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]]

然后hf_engine的第186行是这样写的

if isinstance(value, list) and all(isinstance(v, torch.Tensor) for v in value):  # for pixtral inputs
                value = torch.stack(value)  # assume they have same sizes
elif isinstance(value, list) and all(isinstance(v, list) for v in value):  # for minicpmv inputs
                value = torch.stack([torch.stack(v) for v in value])
elif not isinstance(value, torch.Tensor):
    value = torch.tensor(value)

所以我改成了如下

if isinstance(value, list) and all(isinstance(v, torch.Tensor) for v in value):  # for pixtral inputs
                value = torch.stack(value)  # assume they have same sizes
elif not isinstance(value, torch.Tensor):
    value = torch.tensor(value)
elif isinstance(value, list) and all(isinstance(v, list) for v in value):  # for minicpmv inputs
    value = torch.stack([torch.stack(v) for v in value])

改成这样之后可以运行了
但是结果如下

请问我的做法是否正确？