Description
Describe the bug
Finetune qwen3 in modelscope notebook.
Processing 2 items: 100%|██████████████████████████████████████████████████████████████████████████████████████████| 2.00/2.00 [00:26<00:00, 13.4s/it]
[INFO:swift] Loading the model using model_dir: /mnt/workspace/.cache/modelscope/models/Qwen/Qwen3-1___7B
[INFO:swift] attn_impl: flash_attn
[INFO:swift] model_kwargs: {'device_map': 'cuda:0'}
Traceback (most recent call last):
File "/usr/local/lib/python3.11/site-packages/swift/cli/sft.py", line 7, in
sft_main()
File "/usr/local/lib/python3.11/site-packages/swift/llm/train/sft.py", line 269, in sft_main
return SwiftSft(args).main()
^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/swift/llm/train/sft.py", line 29, in init
self._prepare_model_tokenizer()
File "/usr/local/lib/python3.11/site-packages/swift/llm/train/sft.py", line 45, in _prepare_model_tokenizer
self.model, self.processor = args.get_model_processor()
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/swift/llm/argument/base_args/base_args.py", line 301, in get_model_processor
return get_model_tokenizer(**kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/swift/llm/model/register.py", line 586, in get_model_tokenizer
model, processor = get_function(model_dir, model_info, model_kwargs, load_model, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/swift/llm/model/register.py", line 286, in get_model_tokenizer_with_flash_attn
return get_model_tokenizer_from_local(model_dir, model_info, model_kwargs, load_model, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/swift/llm/model/register.py", line 240, in get_model_tokenizer_from_local
model = automodel_class.from_pretrained(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/transformers/models/auto/auto_factory.py", line 571, in from_pretrained
return model_class.from_pretrained(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/swift/llm/model/patcher.py", line 280, in _new_from_pretrained
model = from_pretrained(cls, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/transformers/modeling_utils.py", line 309, in _wrapper
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/transformers/modeling_utils.py", line 4508, in from_pretrained
model = cls(config, *model_args, **model_kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/transformers/models/qwen3/modeling_qwen3.py", line 660, in init
self.model = Qwen3Model(config)
^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/transformers/models/qwen3/modeling_qwen3.py", line 389, in init
self.post_init()
File "/usr/local/lib/python3.11/site-packages/transformers/modeling_utils.py", line 1969, in post_init
if v not in ALL_PARALLEL_STYLES:
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: argument of type 'NoneType' is not iterable
Your hardware and system info
modelscope notebook