Support Automotive 3d Object Detection Benchmark Automation

[arjunsuresh]

• Add MLC scripts for validation dataset download
• Add MLC scripts for calibration dataset download
• Add MLC scripts for model download
• Add MLC scripts for benchmark run
• Add MLC scripts for accuracy verification
• Add MLC scripts for compliance runs
• Document the benchmark details in inference docs page
• Document the benchmark run commands in inference docs
• Update the README in the Inference GitHub repo with the relevant MLC commands
• Update the GitHub action for calibration dataset download (calibration dataset download is not checked in the GitHub action for reference implementation)
• Upload valid submission log files to the unofficial results repository

[anandhu-eng]

README in inference repo has been updated in this PR

[anandhu-eng]

Hi @arjunsuresh , I think this issue can be closed. The GitHub runner storage is insufficient for downloading the waymo calibration dataset. We could open a separate issue for adding the calibration dataset downloads for large datasets(to be implemented when we get run-pod instance).

Edit: Created a separate issue for larger calibration datasets here