Oliver Borchers

@vichaos it works nicely! Can I give you a coffee?

I do have the same issue, installing from pypi.

@jianqianzhou: If you have the model saved somewhere, you can add the `from_tf=True` argument to the pytorch version and the transformers library will automatically convert the tf model to pytorsch.

Same for me. For the huggingface predictor it actually works, but it doesn't use the image I built but rather the default one...

Update: After figuring out how to work with the repository for sagemaker images I was able to get my problems fixed (which have been solely regarding the `HuggingfaceModel` not being...

> I hope they fix it soon 😄 Me looking at my inbox and laughing frenetically: No.

@deliahu: Is there any update on this one?

@ZHAOTING: Am I the only one where the provided script isn't working anymore? I tried this on the raw checkpoints, as well as on the merged checkpoints from the `mp_merge`...

@pommedeterresautee: I caution that this is probably more difficult as expected due to (currently trying to figure that out): https://github.com/onnx/onnx-tensorrt/issues/818 and maybe even also https://forums.developer.nvidia.com/t/how-to-convert-a-large-pytorch-model-to-trt-model/182782/4