Error parsing message with type 'onnx.ModelProto': Wire format was corrupt

uv run run.py --execution-provider coreml
warning: No `requires-python` value found in the workspace. Defaulting to `>=3.14`.
[platform] Darwin arm64 | python 3.14.6 | accelerator: CoreML (Apple Neural Engine) | providers: ['CoreMLExecutionProvider', 'AzureExecutionProvider', 'CPUExecutionProvider']
qt.qpa.fonts: Populating font family aliases took 149 ms. Replace uses of missing font family "Segoe UI" with one that exists to avoid this cost. 
2026-08-05 00:31:52.532 Python[71264:1646092] error messaging the mach port for IMKCFRunLoopWakeUpReliable
Applied providers: ['CoreMLExecutionProvider', 'CPUExecutionProvider'], with options: {'CPUExecutionProvider': {}, 'CoreMLExecutionProvider': {}}
model ignore: /Users/keith/.insightface/models/buffalo_l/1k3d68.onnx landmark_3d_68
Applied providers: ['CoreMLExecutionProvider', 'CPUExecutionProvider'], with options: {'CPUExecutionProvider': {}, 'CoreMLExecutionProvider': {}}
find model: /Users/keith/.insightface/models/buffalo_l/2d106det.onnx landmark_2d_106 ['None', 3, 192, 192] 0.0 1.0
Applied providers: ['CoreMLExecutionProvider', 'CPUExecutionProvider'], with options: {'CPUExecutionProvider': {}, 'CoreMLExecutionProvider': {}}
find model: /Users/keith/.insightface/models/buffalo_l/det_10g.onnx detection [1, 3, '?', '?'] 127.5 128.0
Applied providers: ['CoreMLExecutionProvider', 'CPUExecutionProvider'], with options: {'CPUExecutionProvider': {}, 'CoreMLExecutionProvider': {}}
duplicated model task type, ignore: /Users/keith/.insightface/models/buffalo_l/det_10g_coreml.onnx detection
Applied providers: ['CoreMLExecutionProvider', 'CPUExecutionProvider'], with options: {'CPUExecutionProvider': {}, 'CoreMLExecutionProvider': {}}
model ignore: /Users/keith/.insightface/models/buffalo_l/genderage.onnx genderage
Applied providers: ['CoreMLExecutionProvider', 'CPUExecutionProvider'], with options: {'CPUExecutionProvider': {}, 'CoreMLExecutionProvider': {}}
find model: /Users/keith/.insightface/models/buffalo_l/w600k_r50.onnx recognition ['None', 3, 112, 112] 127.5 127.5
set det-size: (640, 640)
Traceback (most recent call last):
  File "/Users/keith/Downloads/Deep-Live-Cam/modules/ui.py", line 931, in _on_live
    get_face_swapper()
    ~~~~~~~~~~~~~~~~^^
  File "/Users/keith/Downloads/Deep-Live-Cam/modules/processors/frame/face_swapper.py", line 252, in get_face_swapper
    model_path = optimize_for_coreml(model_path)
  File "/Users/keith/Downloads/Deep-Live-Cam/modules/onnx_optimize.py", line 66, in optimize_for_coreml
    model = onnx.load(model_path)
  File "/Users/keith/Downloads/Deep-Live-Cam/.venv/lib/python3.14/site-packages/onnx/__init__.py", line 231, in load_model
    model = _get_serializer(format, f).deserialize_proto(_load_bytes(f), ModelProto())
  File "/Users/keith/Downloads/Deep-Live-Cam/.venv/lib/python3.14/site-packages/onnx/serialization.py", line 121, in deserialize_proto
    decoded = typing.cast("int | None", proto.ParseFromString(serialized))
                                        ~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^
google.protobuf.message.DecodeError: Error parsing message with type 'onnx.ModelProto': Wire format was corrupt