Getting an ValueError: Expected a xpu device, but got: cuda:0 how to fix

[SeanModo]

When I'm using Adetailer, I'm getting this error. Is there something else I need to install?

ADetailer                                                                                        │

│ ┏━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓ │
│ ┃ ┃ Value ┃ │
│ ┡━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩ │
│ │ version │ 24.5.1 │ │
│ │ ad_model │ face_yolov8n.pt │ │
│ │ ad_prompt │ │ │
│ │ ad_negative_prompt │ │ │
│ │ ad_controlnet_model │ None │ │
│ │ is_api │ False │ │
│ └─────────────────────┴─────────────────┘ │
│ ╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮ │
│ │ /home/seanmodo/automatic/extensions/adetailer/aaaaaa/traceback.py:147 in wrapper │ │
│ │ │ │
│ │ 146 │ │ try: │ │
│ │ ❱ 147 │ │ │ return func(*args, **kwargs) │ │
│ │ 148 │ │ except Exception as e: │ │
│ │ │ │
│ │ /home/seanmodo/automatic/extensions/adetailer/scripts/!adetailer.py:821 in postprocess_image │ │
│ │ │ │
│ │ 820 │ │ │ │ │ continue │ │
│ │ ❱ 821 │ │ │ │ is_processed |= self._postprocess_image_inner(p, pp, args, n=n) │ │
│ │ 822 │ │
│ │ │ │
│ │ /home/seanmodo/automatic/extensions/adetailer/scripts/!adetailer.py:742 in │ │
│ │ _postprocess_image_inner │ │
│ │ │ │
│ │ 741 │ │ with change_torch_load(): │ │
│ │ ❱ 742 │ │ │ pred = predictor(ad_model, pp.image, args.ad_confidence, **kwargs) │ │
│ │ 743 │ │
│ │ │ │
│ │ /home/seanmodo/automatic/extensions/adetailer/adetailer/ultralytics.py:29 in ultralytics_predict │ │
│ │ │ │
│ │ 28 │ apply_classes(model, model_path, classes) │ │
│ │ ❱ 29 │ pred = model(image, conf=confidence, device=device) │ │
│ │ 30 │ │
│ │ │ │
│ │ /home/seanmodo/automatic/venv/lib/python3.10/site-packages/ultralytics/engine/model.py:179 in │ │
│ │ call │ │
│ │ │ │
│ │ 178 │ │ """ │ │
│ │ ❱ 179 │ │ return self.predict(source, stream, **kwargs) │ │
│ │ 180 │ │
│ │ │ │
│ │ /home/seanmodo/automatic/venv/lib/python3.10/site-packages/ultralytics/engine/model.py:557 in │ │
│ │ predict │ │
│ │ │ │
│ │ 556 │ │ │ self.predictor.set_prompts(prompts) │ │
│ │ ❱ 557 │ │ return self.predictor.predict_cli(source=source) if is_cli else self.predictor(s │ │
│ │ 558 │ │
│ │ │ │
│ │ /home/seanmodo/automatic/venv/lib/python3.10/site-packages/ultralytics/engine/predictor.py:173 │ │
│ │ in call │ │
│ │ │ │
│ │ 172 │ │ else: │ │
│ │ ❱ 173 │ │ │ return list(self.stream_inference(source, model, *args, **kwargs)) # merge │ │
│ │ 174 │ │
│ │ │ │
│ │ /home/seanmodo/automatic/venv/lib/python3.10/site-packages/torch/utils/_contextlib.py:35 in │ │
│ │ generator_context │ │
│ │ │ │
│ │ 34 │ │ │ with ctx_factory(): │ │
│ │ ❱ 35 │ │ │ │ response = gen.send(None) │ │
│ │ 36 │ │
│ │ │ │
│ │ /home/seanmodo/automatic/venv/lib/python3.10/site-packages/ultralytics/engine/predictor.py:254 │ │
│ │ in stream_inference │ │
│ │ │ │
│ │ 253 │ │ │ │ # Preprocess │ │
│ │ ❱ 254 │ │ │ │ with profilers[0]: │ │
│ │ 255 │ │ │ │ │ im = self.preprocess(im0s) │ │
│ │ │ │
│ │ /home/seanmodo/automatic/venv/lib/python3.10/site-packages/ultralytics/utils/ops.py:46 in │ │
│ │ enter │ │
│ │ │ │
│ │ 45 │ │ """Start timing.""" │ │
│ │ ❱ 46 │ │ self.start = self.time() │ │
│ │ 47 │ │ return self │ │
│ │ │ │
│ │ /home/seanmodo/automatic/venv/lib/python3.10/site-packages/ultralytics/utils/ops.py:61 in time │ │
│ │ │ │
│ │ 60 │ │ if self.cuda: │ │
│ │ ❱ 61 │ │ │ torch.cuda.synchronize(self.device) │ │
│ │ 62 │ │ return time.time() │ │
│ │ │ │
│ │ /home/seanmodo/automatic/venv/lib/python3.10/site-packages/torch/xpu/init.py:384 in │ │
│ │ synchronize │ │
│ │ │ │
│ │ 383 │ _lazy_init() │ │
│ │ ❱ 384 │ device = _get_device_index(device, optional=True) │ │
│ │ 385 │ return torch._C._xpu_synchronize(device) │ │
│ │ │ │
│ │ /home/seanmodo/automatic/venv/lib/python3.10/site-packages/torch/xpu/_utils.py:35 in │ │
│ │ _get_device_index │ │
│ │ │ │
│ │ 34 │ │ elif device.type != "xpu": │ │
│ │ ❱ 35 │ │ │ raise ValueError(f"Expected a xpu device, but got: {device}") │ │
│ │ 36 │ if not torch.jit.is_scripting(): │ │
│ ╰──────────────────────────────────────────────────────────────────────────────────────────────────╯ │
│ ValueError: Expected a xpu device, but got: cuda:0

[dosubot[bot]]

Hey @SeanModo! I'm here to help you with any bugs, questions, or contributions you have regarding ADetailer. Let's work together to solve this issue.

The error you're encountering, "ValueError: Expected a xpu device, but got: cuda:0," suggests that the code is trying to use an XPU device, but it's currently set to use a CUDA device. This mismatch is likely causing the issue.

To resolve this, you can try the following steps:

1. Check Device Configuration: Ensure that the device configuration in your setup is correctly set to use the intended device type. If you want to use a CUDA device, make sure that the configuration or environment is not set to expect an XPU device.

2. Modify Device Selection Logic: In the adetailer codebase, the device is determined by the get_ultralytics_device method. You might need to adjust this logic to ensure it selects the correct device type for your setup. For example, if you are on a system that supports CUDA, ensure that the conditions in this method allow for CUDA usage.

3. Environment Variables: Check if there are any environment variables or command-line options that might be forcing the use of an XPU device. Adjust these settings to align with your hardware capabilities.

4. Dependencies and Libraries: Ensure that all necessary libraries and dependencies for CUDA are correctly installed and configured. This includes having the correct version of PyTorch that supports CUDA.

If these steps do not resolve the issue, you may need to further investigate the specific configuration and setup of your environment to ensure compatibility with the desired device type.

To continue talking to Dosu, mention @dosu.