Edouard Belval
Edouard Belval
@sirfz This is quite unsettling because they fixed it less than 3 weeks ago, I'll check if I can reproduce the issue tonight and reopen an issue on their repo...
@sirfz Yes but I opened [this issue](https://github.com/tesseract-ocr/tesseract/issues/1045) and the next day they changed it to add `-DUSING_STD_NAMESPACE` as a fix. The commit I pointed to should work.
@sirfz Ok, I'll report back with results
@sirfz @zqy2084 Can confirm that it is broken again... I'll report back when I find a decent fix.
I am having a similar issue converting a TorchScript model with TorchTensorRT. I see that the consensus is that you can refactor your code as a workaround to the issue,...
Unfortunately the code is not open source, I will try and see if I can get a small repro case.
Still working on the repro, but I just build torch_tensorrt from source (using master) to see if it helped and it it did partially solve the issue above. Unfortunately, I...
` torch_tensorrt.compile(model, inputs=[torch_tensorrt.Input((1, 3, 1440, 1440))], torch_executed_ops=["prim::ListConstruct"])` returns a very similar stack trace as before: ``` RuntimeError: [Error thrown at core/partitioning/shape_analysis.cpp:68] Expected ivalues_maps.count(input) to be true but got falseCould not...
I am not sure that I understand your question. If you are referring to the repro package I sent you, then it could be the `torch.ops.load_library` in `repro.py` that does...
Do you have a PDF I could use to reproduce the error on my side?