torch._dynamo.exc.Unsupported when finetuning QWEN2-VL

[definitelynotadoge]

I am following this notebook to finetune QWEN2-VL on my custom dataset and getting exception on the first training step:
torch._dynamo.exc.Unsupported: Graph break due to unsupported builtin builtins.__import__. This function is either a Python builtin (e.g. _warnings.warn) or a third-party C/C++ Python extension (perhaps created with pybind). If it is a Python builtin, please file an issue on GitHub so the PyTorch team can add support for it and see the next case for a workaround. If it is a third-party C/C++ Python extension, please either wrap it into a PyTorch-understood custom operator (see https://pytorch.org/tutorials/advanced/custom_ops_landing_page.html for more details) or, if it is traceable, use torch.compiler.allow_in_graph.
traceback.txt

Python version: 3.12
CUDA version: 12.4

packages:

bitsandbytes==0.43.3
datasets==2.16.0
flash-attn==2.7.3
numpy==2.2.2
peft==0.14.0
torch==2.5.1
torchvision==0.20.1
transformers==4.46.1
trl==0.11.4
unsloth_zoo==2025.1.5
unsloth @ git+https://github.com/unslothai/unsloth.git@1ef71e8b9c570d59470ba5a40ed49064db295f43

[shimmyshimmer]

Is this local? We have installation instructions, it seems something went wrong with installation: https://docs.unsloth.ai/get-started/installing-+-updating

[jaffe-fly]

same error