2.0.6怎么使用cuda？ #2822

doublex640 · 2025-06-28T13:19:09Z

doublex640
Jun 28, 2025

windows环境下，我记得老版本在magic-pdf.json里可以设置"device-mode": "cuda"，现在2.0.6版本的json文件里没有这个项目了。
请问如何设置使用cuda？
我第一次运行mineru -p <input_path> -o <output_path> --source local，看到应该是cpu运行。
在安装pytorch之后，运行就报错了。print(torch.cuda.is_available())显示True。版本为2.7.1+cu128。

Answered by myhloli

Jun 29, 2025

用nvidia-smi命令看下驱动版本，可能是驱动太老了

View full answer

doublex640 · 2025-06-28T13:27:01Z

doublex640
Jun 28, 2025
Author

2025-06-28 21:26:23.556 | WARNING | mineru.backend.vlm.predictor::35 - sglang is not installed. If you are not using sglang, you can ignore this warning.
2025-06-28 21:26:36.606 | INFO | mineru.backend.pipeline.pipeline_analyze:doc_analyze:124 - Batch 1/3: 100 pages/275 pages
2025-06-28 21:26:36.606 | INFO | mineru.backend.pipeline.pipeline_analyze:batch_image_analyze:187 - gpu_memory: 16 GB, batch_ratio: 16
2025-06-28 21:26:36.607 | INFO | mineru.backend.pipeline.model_init:init:137 - DocAnalysis init, this may take some times......
2025-06-28 21:26:39.137 | INFO | mineru.backend.pipeline.model_init:init:182 - DocAnalysis init done!
2025-06-28 21:26:39.137 | INFO | mineru.backend.pipeline.pipeline_analyze:custom_model_init:64 - model init cost: 2.530822515487671
Layout Predict: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 100/100 [00:03<00:00, 31.85it/s]
MFD Predict: 3%|████▌ | 3/100 [00:00<00:15, 6.15it/s]
2025-06-28 21:26:42.767 | ERROR | mineru.cli.client:parse_doc:189 - Could not run 'torchvision::nms' with arguments from the 'CUDA' backend. This could be because the operator doesn't exist for this
backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible
resolutions. 'torchvision::nms' is only available for these backends: [CPU, Meta, QuantizedCPU, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTens
or, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradXLA, AutogradMPS, AutogradXPU, AutogradHPU, AutogradLazy, AutogradMTIA, AutogradMeta, Tracer, AutocastCPU, AutocastMTIA, AutocastX
PU, AutocastMPS, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].

CPU: registered at C:\actions-runner_work\vision\vision\pytorch\vision\torchvision\csrc\ops\cpu\nms_kernel.cpp:112 [kernel]
Meta: registered at /dev/null:214 [kernel]
QuantizedCPU: registered at C:\actions-runner_work\vision\vision\pytorch\vision\torchvision\csrc\ops\quantized\cpu\qnms_kernel.cpp:124 [kernel]
BackendSelect: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
Python: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\PythonFallbackKernel.cpp:194 [backend fallback]
FuncTorchDynamicLayerBackMode: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\DynamicLayer.cpp:479 [backend fallback]
Functionalize: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\FunctionalizeFallbackKernel.cpp:349 [backend fallback]
Named: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
Conjugate: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
Negative: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
ZeroTensor: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
ADInplaceOrView: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:100 [backend fallback]
AutogradOther: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:63 [backend fallback]
AutogradCPU: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:67 [backend fallback]
AutogradCUDA: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:75 [backend fallback]
AutogradXLA: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:83 [backend fallback]
AutogradMPS: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:91 [backend fallback]
AutogradXPU: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:71 [backend fallback]
AutogradHPU: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:104 [backend fallback]
AutogradLazy: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:87 [backend fallback]
AutogradMTIA: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:79 [backend fallback]
AutogradMeta: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:95 [backend fallback]
Tracer: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\torch\csrc\autograd\TraceTypeManual.cpp:294 [backend fallback]
AutocastCPU: registered at C:\actions-runner_work\vision\vision\pytorch\vision\torchvision\csrc\ops\autocast\nms_kernel.cpp:34 [kernel]
AutocastMTIA: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\autocast_mode.cpp:466 [backend fallback]
AutocastXPU: registered at C:\actions-runner_work\vision\vision\pytorch\vision\torchvision\csrc\ops\autocast\nms_kernel.cpp:41 [kernel]
AutocastMPS: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\autocast_mode.cpp:209 [backend fallback]
AutocastCUDA: registered at C:\actions-runner_work\vision\vision\pytorch\vision\torchvision\csrc\ops\autocast\nms_kernel.cpp:27 [kernel]
FuncTorchBatched: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:731 [backend fallback]
BatchedNestedTensor: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:758 [backend fallback]
FuncTorchVmapMode: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\VmapModeRegistrations.cpp:27 [backend fallback]
Batched: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
VmapMode: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
FuncTorchGradWrapper: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\TensorWrapper.cpp:208 [backend fallback]
PythonTLSSnapshot: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\PythonFallbackKernel.cpp:202 [backend fallback]
FuncTorchDynamicLayerFrontMode: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\DynamicLayer.cpp:475 [backend fallback]
PreDispatch: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\PythonFallbackKernel.cpp:206 [backend fallback]
PythonDispatcher: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\PythonFallbackKernel.cpp:198 [backend fallback]

Traceback (most recent call last):

File "", line 198, in _run_module_as_main
File "", line 88, in _run_code

File "D:\anaconda_envs\envs\MinerU\Scripts\mineru.exe_main_.py", line 10, in
sys.exit(client.main())
│ │ │ └
│ │ └ <module 'mineru.cli.client' from 'D:\anaconda_envs\envs\MinerU\Lib\site-packages\mineru\cli\client.py'>
│ └
└ <module 'sys' (built-in)>

File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\click\core.py", line 1442, in call
return self.main(*args, **kwargs)
│ │ │ └ {}
│ │ └ ()
│ └ <function Command.main at 0x000001B073A16F20>
└
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\click\core.py", line 1363, in main
rv = self.invoke(ctx)
│ │ └ <click.core.Context object at 0x000001B02C5486E0>
│ └ <function Command.invoke at 0x000001B073A16C00>
└
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\click\core.py", line 1226, in invoke
return ctx.invoke(self.callback, **ctx.params)
│ │ │ │ │ └ {'input_path': './input/经济解释卷1 科学说需求.pdf', 'output_dir': 'output', 'model_source': 'local', 'method': 'auto', 'backend': 'pi...
│ │ │ │ └ <click.core.Context object at 0x000001B02C5486E0>
│ │ │ └ <function main at 0x000001B06850FCE0>
│ │ └
│ └ <function Context.invoke at 0x000001B073A15E40>
└ <click.core.Context object at 0x000001B02C5486E0>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\click\core.py", line 794, in invoke
return callback(*args, **kwargs)
│ │ └ {'input_path': './input/经济解释卷1 科学说需求.pdf', 'output_dir': 'output', 'model_source': 'local', 'method': 'auto', 'backend': 'pi...
│ └ ()
└ <function main at 0x000001B06850FCE0>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\mineru\cli\client.py", line 198, in main
parse_doc([Path(input_path)])
│ │ └ './input/经济解释卷1 科学说需求.pdf'
│ └ <class 'pathlib._local.Path'>
└ <function main..parse_doc at 0x000001B06850F420>

File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\mineru\cli\client.py", line 175, in parse_doc
do_parse(
└ <function do_parse at 0x000001B0684C2840>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\mineru\cli\common.py", line 108, in do_parse
infer_results, all_image_lists, all_pdf_docs, lang_list, ocr_enabled_list = pipeline_doc_analyze(pdf_bytes_list, p_lang_list, parse_method=parse_method, formula_enable=p_formula_enable,table_enable=p_table_enable)
│ │ │ │ │
└ True
│ │ │ │ └ True
│ │ │ └ 'auto'
│ │ └ ['ch']
│ └ [b'%PDF-1.7\r\n%\xa1\xb3\xc5\xd7\r\n1 0 obj\r\n<</Pages 2 0 R /Type/Catalog>>\r\nendobj\r\n2 0 obj\r\n<</Count 275/Kids[ 4 0 ...
└ <function doc_analyze at 0x000001B0729CBBA0>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\mineru\backend\pipeline\pipeline_analyze.py", line 128, in doc_analyze
batch_results = batch_image_analyze(batch_image, formula_enable, table_enable)
│ │ │ └ True
│ │ └ True
│ └ [(<PIL.Image.Image image mode=RGB size=984x1600 at 0x1B072873390>, True, 'ch'), (<PIL.Image.Image image mode=RGB size=990x161...
└ <function batch_image_analyze at 0x000001B0729CBC40>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\mineru\backend\pipeline\pipeline_analyze.py", line 194, in batch_image_analyze
results = batch_model(images_with_extra_info)
│ └ [(<PIL.Image.Image image mode=RGB size=984x1600 at 0x1B072873390>, True, 'ch'), (<PIL.Image.Image image mode=RGB size=990x161...
└ <mineru.backend.pipeline.batch_analyze.BatchAnalyze object at 0x000001B0726A3CB0>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\mineru\backend\pipeline\batch_analyze.py", line 52, in call
images_mfd_res = self.model.mfd_model.batch_predict(
│ │ │ └ <function YOLOv8MFDModel.batch_predict at 0x000001B071DD8CC0>
│ │ └ <mineru.model.mfd.yolo_v8.YOLOv8MFDModel object at 0x000001B07289DA90>
│ └ <mineru.backend.pipeline.model_init.MineruPipelineModel object at 0x000001B0726A3B60>
└ <mineru.backend.pipeline.batch_analyze.BatchAnalyze object at 0x000001B0726A3CB0>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\mineru\model\mfd\yolo_v8.py", line 22, in batch_predict
for image_res in self.mfd_model.predict(
│ │ │ └ <function Model.predict at 0x000001B071F7CA40>
│ │ └ YOLO(
│ │ (model): DetectionModel(
│ │ (model): Sequential(
│ │ (0): Conv(
│ │ (conv): Conv2d(3, 64, kernel_size=(3, 3), ...
│ └ <mineru.model.mfd.yolo_v8.YOLOv8MFDModel object at 0x000001B07289DA90>
└ ultralytics.engine.results.Results object with attributes:

      boxes: ultralytics.engine.results.Boxes object
      keypoints: None
      ma...

File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\ultralytics\engine\model.py", line 555, in predict
return self.predictor.predict_cli(source=source) if is_cli else self.predictor(source=source, stream=stream)
│ │ │ │ │ │ │ │ └ False
│ │ │ │ │ │ │ └ [<PIL.Image.Image image mode=RGB size=987x1609 at 0x1B072961F20>]
│ │ │ │ │ │ └ <ultralytics.models.yolo.detect.predict.DetectionPredictor object at 0x000001B0963FA270>
│ │ │ │ │ └ YOLO(
│ │ │ │ │ (model): DetectionModel(
│ │ │ │ │ (model): Sequential(
│ │ │ │ │ (0): Conv(
│ │ │ │ │ (conv): Conv2d(3, 64, kernel_size=(3, 3), ...
│ │ │ │ └ False
│ │ │ └ [<PIL.Image.Image image mode=RGB size=987x1609 at 0x1B072961F20>]
│ │ └ <function BasePredictor.predict_cli at 0x000001B071F7E700>
│ └ <ultralytics.models.yolo.detect.predict.DetectionPredictor object at 0x000001B0963FA270>
└ YOLO(
(model): DetectionModel(
(model): Sequential(
(0): Conv(
(conv): Conv2d(3, 64, kernel_size=(3, 3), ...
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\ultralytics\engine\predictor.py", line 227, in call
return list(self.stream_inference(source, model, *args, **kwargs)) # merge list of Result into one
│ │ │ │ │ └ {}
│ │ │ │ └ ()
│ │ │ └ None
│ │ └ [<PIL.Image.Image image mode=RGB size=987x1609 at 0x1B072961F20>]
│ └ <function BasePredictor.stream_inference at 0x000001B071F7E980>
└ <ultralytics.models.yolo.detect.predict.DetectionPredictor object at 0x000001B0963FA270>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\torch\utils_contextlib.py", line 36, in generator_context
response = gen.send(None)
│ └ <method 'send' of 'generator' objects>
└ <generator object BasePredictor.stream_inference at 0x000001B09A9F9DF0>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\ultralytics\engine\predictor.py", line 337, in stream_inference
self.results = self.postprocess(preds, im, im0s)
│ │ │ │ │ │ └ [array([[[255, 255, 255],
│ │ │ │ │ │ [255, 255, 255],
│ │ │ │ │ │ [255, 255, 255],
│ │ │ │ │ │ ...,
│ │ │ │ │ │ [255, 255, 255],
│ │ │ │ │ │ [25...
│ │ │ │ │ └ tensor([[[[0.4471, 0.4471, 0.4471, ..., 0.4471, 0.4471, 0.4471],
│ │ │ │ │ [0.4471, 0.4471, 0.4471, ..., 0.4471, 0.4471, 0....
│ │ │ │ └ [tensor([[[ 2.4878e-01, -1.5131e-01, -4.9829e-01, ..., 6.7578e+02, 7.3308e+02, 7.9168e+02],
│ │ │ │ [ 1.6663e-01, 8.082...
│ │ │ └ <function DetectionPredictor.postprocess at 0x000001B071FA7420>
│ │ └ <ultralytics.models.yolo.detect.predict.DetectionPredictor object at 0x000001B0963FA270>
│ └ [ultralytics.engine.results.Results object with attributes:
│
│ boxes: ultralytics.engine.results.Boxes object
│ keypoints: None
│ m...
└ <ultralytics.models.yolo.detect.predict.DetectionPredictor object at 0x000001B0963FA270>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\ultralytics\models\yolo\detect\predict.py", line 56, in postprocess
preds = ops.non_max_suppression(
│ └ <function non_max_suppression at 0x000001B071EF1EE0>
└ <module 'ultralytics.utils.ops' from 'D:\anaconda_envs\envs\MinerU\Lib\site-packages\ultralytics\utils\ops.py'>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\ultralytics\utils\ops.py", line 330, in non_max_suppression
i = torchvision.ops.nms(boxes, scores, iou_thres) # NMS
│ │ │ │ │ └ 0.45
│ │ │ │ └ tensor([0.2888, 0.3000, 0.2601, 0.2694], device='cuda:0')
│ │ │ └ tensor([[ 382.7933, 1373.9725, 397.9456, 1392.4742],
│ │ │ [ 382.8364, 1374.0940, 398.1491, 1392.5159],
│ │ │ [ 382.699...
│ │ └ <function nms at 0x000001B0381637E0>
│ └ <module 'torchvision.ops' from 'D:\anaconda_envs\envs\MinerU\Lib\site-packages\torchvision\ops\init.py'>
└ <module 'torchvision' from 'D:\anaconda_envs\envs\MinerU\Lib\site-packages\torchvision\init.py'>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\torchvision\ops\boxes.py", line 41, in nms
return torch.ops.torchvision.nms(boxes, scores, iou_threshold)
│ │ │ │ │ │ └ 0.45
│ │ │ │ │ └ tensor([0.2888, 0.3000, 0.2601, 0.2694], device='cuda:0')
│ │ │ │ └ tensor([[ 382.7933, 1373.9725, 397.9456, 1392.4742],
│ │ │ │ [ 382.8364, 1374.0940, 398.1491, 1392.5159],
│ │ │ │ [ 382.699...
│ │ │ └ <OpOverloadPacket(op='torchvision.nms')>
│ │ └ <module 'torch.ops.torchvision' from 'torch.ops'>
│ └ <module 'torch.ops' from '_ops.py'>
└ <module 'torch' from 'D:\anaconda_envs\envs\MinerU\Lib\site-packages\torch\init.py'>
File "D:\anaconda_envs\envs\MinerU\Lib\site-packages\torch_ops.py", line 1158, in call
return self._op(*args, **(kwargs or {}))
│ │ │ └ {}
│ │ └ (tensor([[ 382.7933, 1373.9725, 397.9456, 1392.4742],
│ │ [ 382.8364, 1374.0940, 398.1491, 1392.5159],
│ │ [ 382.69...
│ └ <built-in method nms of PyCapsule object at 0x000001B072D60E50>
└ <OpOverloadPacket(op='torchvision.nms')>

NotImplementedError: Could not run 'torchvision::nms' with arguments from the 'CUDA' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/cust
om build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'torchvision::nms' is only available
for these backends: [CPU, Meta, QuantizedCPU, BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, A
utogradCUDA, AutogradXLA, AutogradMPS, AutogradXPU, AutogradHPU, AutogradLazy, AutogradMTIA, AutogradMeta, Tracer, AutocastCPU, AutocastMTIA, AutocastXPU, AutocastMPS, AutocastCUDA, FuncTorchBatched, BatchedNestedTensor, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PreDispatch, PythonDispatcher].

CPU: registered at C:\actions-runner_work\vision\vision\pytorch\vision\torchvision\csrc\ops\cpu\nms_kernel.cpp:112 [kernel]
Meta: registered at /dev/null:214 [kernel]
QuantizedCPU: registered at C:\actions-runner_work\vision\vision\pytorch\vision\torchvision\csrc\ops\quantized\cpu\qnms_kernel.cpp:124 [kernel]
BackendSelect: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\BackendSelectFallbackKernel.cpp:3 [backend fallback]
Python: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\PythonFallbackKernel.cpp:194 [backend fallback]
FuncTorchDynamicLayerBackMode: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\DynamicLayer.cpp:479 [backend fallback]
Functionalize: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\FunctionalizeFallbackKernel.cpp:349 [backend fallback]
Named: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\NamedRegistrations.cpp:7 [backend fallback]
Conjugate: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\ConjugateFallback.cpp:17 [backend fallback]
Negative: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\native\NegateFallback.cpp:18 [backend fallback]
ZeroTensor: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\ZeroTensorFallback.cpp:86 [backend fallback]
ADInplaceOrView: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:100 [backend fallback]
AutogradOther: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:63 [backend fallback]
AutogradCPU: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:67 [backend fallback]
AutogradCUDA: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:75 [backend fallback]
AutogradXLA: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:83 [backend fallback]
AutogradMPS: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:91 [backend fallback]
AutogradXPU: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:71 [backend fallback]
AutogradHPU: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:104 [backend fallback]
AutogradLazy: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:87 [backend fallback]
AutogradMTIA: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:79 [backend fallback]
AutogradMeta: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\VariableFallbackKernel.cpp:95 [backend fallback]
Tracer: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\torch\csrc\autograd\TraceTypeManual.cpp:294 [backend fallback]
AutocastCPU: registered at C:\actions-runner_work\vision\vision\pytorch\vision\torchvision\csrc\ops\autocast\nms_kernel.cpp:34 [kernel]
AutocastMTIA: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\autocast_mode.cpp:466 [backend fallback]
AutocastXPU: registered at C:\actions-runner_work\vision\vision\pytorch\vision\torchvision\csrc\ops\autocast\nms_kernel.cpp:41 [kernel]
AutocastMPS: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\autocast_mode.cpp:209 [backend fallback]
AutocastCUDA: registered at C:\actions-runner_work\vision\vision\pytorch\vision\torchvision\csrc\ops\autocast\nms_kernel.cpp:27 [kernel]
FuncTorchBatched: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:731 [backend fallback]
BatchedNestedTensor: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\LegacyBatchingRegistrations.cpp:758 [backend fallback]
FuncTorchVmapMode: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\VmapModeRegistrations.cpp:27 [backend fallback]
Batched: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\LegacyBatchingRegistrations.cpp:1075 [backend fallback]
VmapMode: fallthrough registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\VmapModeRegistrations.cpp:33 [backend fallback]
FuncTorchGradWrapper: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\TensorWrapper.cpp:208 [backend fallback]
PythonTLSSnapshot: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\PythonFallbackKernel.cpp:202 [backend fallback]
FuncTorchDynamicLayerFrontMode: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\functorch\DynamicLayer.cpp:475 [backend fallback]
PreDispatch: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\PythonFallbackKernel.cpp:206 [backend fallback]
PythonDispatcher: registered at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\core\PythonFallbackKernel.cpp:198 [backend fallback]

-> Cannot close object, library is destroyed. This may cause a memory leak!

0 replies

myhloli · 2025-06-29T01:25:11Z

myhloli
Jun 29, 2025
Maintainer

可能torchvision和torch不匹配吧，你装torch的时候带上torchvision

1 reply

doublex640 Jun 29, 2025
Author

可能torchvision和torch不匹配吧，你装torch的时候带上torchvision

啊，我是直接用官网给的命令安装的。
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu128
这里面不是带了torchvision了嘛。具体需要怎么操作？需要怎么样才能让torchvision和torch匹配？

doublex640 · 2025-06-29T05:34:00Z

doublex640
Jun 29, 2025
Author

解决了。mineru 2.0.6 暂时不支持pytorch的cuda12.8版本（我安装的版本）。把pytorch降级到对应cuda12.6的版本后就没有报错了。

6 replies

myhloli Jun 29, 2025
Maintainer

12.8支持的，装之前需要把驱动版本升级到最新，如果不更新驱动是可能不兼容12.8

doublex640 Jun 29, 2025
Author

12.8支持的，装之前需要把驱动版本升级到最新，如果不更新驱动是可能不兼容12.8

我电脑里驱动安装的就是cuda12.8

doublex640 Jun 29, 2025
Author

12.8支持的，装之前需要把驱动版本升级到最新，如果不更新驱动是可能不兼容12.8

起码在我的电脑上：windows10专业版19045.5965，cuda安装的是12.8.0_571.96。
pytorch不能选cuda12.8，会报错，上面就是错误信息。
pytorch降级到cuda12.6，就正常了。
只适用于我的电脑。
可能在其他客户端mineru确实是支持12.8的吧。

myhloli Jun 29, 2025
Maintainer

用nvidia-smi命令看下驱动版本，可能是驱动太老了

Answer selected by doublex640

doublex640 Jun 29, 2025
Author

用nvidia-smi命令看下驱动版本，可能是驱动太老了

好的。这是电脑目前安装的驱动版本。

doublex640 Jun 29, 2025
Author

用nvidia-smi命令看下驱动版本，可能是驱动太老了

把显卡驱动从571.96升级到576.80，再安装pytorch cuda12.8没有报错。特此反馈！

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

2.0.6怎么使用cuda？ #2822

Uh oh!

{{title}}

Uh oh!

Replies: 3 comments 7 replies

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Select a reply

Uh oh!

2.0.6怎么使用cuda？ #2822

Uh oh!

doublex640 Jun 28, 2025

Replies: 3 comments · 7 replies

Uh oh!

doublex640 Jun 28, 2025 Author

Uh oh!

myhloli Jun 29, 2025 Maintainer

Uh oh!

doublex640 Jun 29, 2025 Author

Uh oh!

doublex640 Jun 29, 2025 Author

Uh oh!

myhloli Jun 29, 2025 Maintainer

Uh oh!

doublex640 Jun 29, 2025 Author

Uh oh!

doublex640 Jun 29, 2025 Author

Uh oh!

myhloli Jun 29, 2025 Maintainer

Uh oh!

doublex640 Jun 29, 2025 Author

Uh oh!

doublex640 Jun 29, 2025 Author

doublex640
Jun 28, 2025

Replies: 3 comments 7 replies

doublex640
Jun 28, 2025
Author

myhloli
Jun 29, 2025
Maintainer

doublex640 Jun 29, 2025
Author

doublex640
Jun 29, 2025
Author

myhloli Jun 29, 2025
Maintainer

doublex640 Jun 29, 2025
Author

doublex640 Jun 29, 2025
Author

myhloli Jun 29, 2025
Maintainer

doublex640 Jun 29, 2025
Author

doublex640 Jun 29, 2025
Author