Test NVIDIA v5.1 implementations

[anandhu-eng]

We have to test the working of NVIDIA implementation being integrated to the automation for the following models by determining the order of their priority(❕ is marked for models which does not have a default config file for NVIDIA implementation):

• Bert❕
• DeepSeek-r1
• DLRM-v2❕
• Llama2-70b
• Llama3.1-405b
• Llama3.1-8b
• Mixtral-8x7b
• Resnet-50❕
• Retinanet❕
• RGAT
• SDXL
• Whisper