启用swanlab报错

[SunflowerColor]

Reminder

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

System Info

训练时勾选swanlab，发生报错，请问可能是什么问题？
系统环境如下：
- llamafactory` version: 0.9.2.dev0

• Platform: Windows-11-10.0.26100-SP0
• Python version: 3.12.8
• PyTorch version: 2.6.0+cu126 (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 GeForce RTX 4070 Laptop GPU
• GPU number: 1
• GPU memory: 8.00GB
• Bitsandbytes version: 0.45.3
• vLLM version: 0.7.3`

Reproduction

O^O/ \_/ \    Batch size per device = 2 | Gradient Accumulation steps = 8
\        /    Total batch size = 16 | Total steps = 60
 "-____-"     Number of trainable parameters = 9,232,384
swanlab: swanlab version 0.4.11 is available!  Upgrade: `pip install -U swanlab`
swanlab: Tracking run with swanlab version 0.4.10
swanlab: Run data will be saved locally in D:\Code\myllamafactory\swanlog\run-20250305_225609-a3b1799d
swanlab: 👋 Hi SunflowerColor, welcome to swanlab!
swanlab: Syncing run tiger-4 to the cloud
swanlab: 🌟 Run `swanlab watch D:\Code\myllamafactory\swanlog` to view SwanLab Experiment Dashboard locally
swanlab: 🏠 View project at https://swanlab.cn/@SunflowerColor/SafeKnowlege
swanlab: 🚀 View run at https://swanlab.cn/@SunflowerColor/SafeKnowlege/runs/zdjl66d3kmptlxo5i72gn
swanlab: Error happened while training
swanlab: 🌟 Run `swanlab watch D:\Code\myllamafactory\swanlog` to view SwanLab Experiment Dashboard locally
swanlab: 🏠 View project at https://swanlab.cn/@SunflowerColor/SafeKnowlege
swanlab: 🚀 View run at https://swanlab.cn/@SunflowerColor/SafeKnowlege/runs/zdjl66d3kmptlxo5i72gn
  File "<frozen runpy>", line 198, in _run_module_as_main
  File "<frozen runpy>", line 88, in _run_code
  File "C:\Users\40801\AppData\Local\Programs\Python\Python312\Scripts\llamafactory-cli.exe\__main__.py", line 7, in <module>
    sys.exit(main())
             ^^^^^^
  File "D:\Code\fromgithub\LLaMA-Factory\src\llamafactory\cli.py", line 112, in main
    run_exp()
  File "D:\Code\fromgithub\LLaMA-Factory\src\llamafactory\train\tuner.py", line 93, in run_exp
    _training_function(config={"args": args, "callbacks": callbacks})
  File "D:\Code\fromgithub\LLaMA-Factory\src\llamafactory\train\tuner.py", line 67, in _training_function
    run_sft(model_args, data_args, training_args, finetuning_args, generating_args, callbacks)
  File "D:\Code\fromgithub\LLaMA-Factory\src\llamafactory\train\sft\workflow.py", line 102, in run_sft
    train_result = trainer.train(resume_from_checkpoint=training_args.resume_from_checkpoint)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\40801\AppData\Local\Programs\Python\Python312\Lib\site-packages\transformers\trainer.py", line 2241, in train
    return inner_training_loop(
           ^^^^^^^^^^^^^^^^^^^^
  File "<string>", line 213, in _fast_inner_training_loop
  File "C:\Users\40801\AppData\Local\Programs\Python\Python312\Lib\site-packages\transformers\trainer_callback.py", line 507, in on_train_begin
    return self.call_event("on_train_begin", args, state, control)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\40801\AppData\Local\Programs\Python\Python312\Lib\site-packages\transformers\trainer_callback.py", line 557, in call_event
    result = getattr(callback, event)(
             ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\40801\AppData\Local\Programs\Python\Python312\Lib\site-packages\swanlab\integration\transformers.py", line 215, in on_train_begin
    self.setup(args, state, model, **kwargs)
  File "D:\Code\fromgithub\LLaMA-Factory\src\llamafactory\train\trainer_utils.py", line 603, in setup
    swanlab_public_config = self._experiment.get_run().public.json()
                            ^^^^^^^^^^^^^^^^
'SwanLabCallbackExtension' object has no attribute '_experiment'

Others

No response

[hiyouga]

@Zeyi-Lin Do you have any thoughts?

[Zeyi-Lin]

Due to some changes in SwanLabCallback in version 0.4.10, you can first downgrade to version 0.4.9:

pip install swanlab==0.4.9

I will also submit a compatible PR to LLaMA Factory soon.