[Bug] ROCM/AMD not working with latest Studio builds.

Note: Please do not remove the questions. Answer beside them. 1. Did you update? `pip install --upgrade unsloth unsloth_zoo` yes 2. `Colab` or `Kaggle` or local / cloud local 3. Number GPUs used, use `nvidia-smi` 1 gpu, halo strip 395+ 4. Which notebook? Please link! not using a notebook 5. Which Unsloth version, TRL version, transformers version, PyTorch version? V0.1.464-beta 2026.6.7. 6. Which trainer? `SFTTrainer`, `GRPOTrainer` etc not training, just loading or model or doing anything. Llama.cpp that is bundled with the latest release wont even run, just immediately gets a sigfault.