[FEA] Support Roberta model

[esnvidia]

As title suggests, support the HF Roberta model with the pretrained weights here: https://huggingface.co/roberta-large

[jdemouth-nvidia]

Hi @esnvidia ,

Thanks for the suggestion. We will add it to the list of feature requests. Not sure if/when we will have time to implement it but we will keep it in mind.

Thanks,
Julien

[AdamzNV]

As more and more new models enter the market, we have prepared comprehensive instructions for TRT-LLM developers on adapting to new models of interest. We encourage our community developers to expand the range of supported models, fostering an open ecosystem with rapid iterations.

Please try following these instructions and let us know if you encounter any issues during the adaptation process. We greatly appreciate your dedication.

[nv-guomingz]

This model had been supported in Tensorrt-llm https://github.com/NVIDIA/TensorRT-LLM/blob/main/examples/bert/build.py#L101

please feel free to reopen this ticket if needed.