Minimum VRAM?

[yuesam]

What's the min vram for this? Thanks!

[jryebread]

+1 Same Question, Tried to run on an A10G with 24GB VRAM and got onnxruntime.capi.onnxruntime_pybind11_state.InvalidProtobuf: [ONNXRuntimeError] : 7 : INVALID_PROTOBUF : Load model from ./pretrained_models/face_analysis/models/1k3d68.onnx failed:Protobuf parsing failed. Think the model too big to load?

[nitinmukesh]

I am testing on RTX4060 8 GB and it is not able to use full VRAM. The only problem is it does not show how much is remaining, just going on endlessely for 7 hours

[AricGamma]

it cost about 9G VRAM at runtime.

[AricGamma]

+1 Same Question, Tried to run on an A10G with 24GB VRAM and got onnxruntime.capi.onnxruntime_pybind11_state.InvalidProtobuf: [ONNXRuntimeError] : 7 : INVALID_PROTOBUF : Load model from ./pretrained_models/face_analysis/models/1k3d68.onnx failed:Protobuf parsing failed. Think the model too big to load?

That may be another problem. please open a new issue.

[AricGamma]

+1 Same Question, Tried to run on an A10G with 24GB VRAM and got onnxruntime.capi.onnxruntime_pybind11_state.InvalidProtobuf: [ONNXRuntimeError] : 7 : INVALID_PROTOBUF : Load model from ./pretrained_models/face_analysis/models/1k3d68.onnx failed:Protobuf parsing failed. Think the model too big to load?

Maybe the model is incomplete. Refer to #42