[NvTesnorRtRtx] Add mixed precision nvModelOpt recipes for Llama-3.2-1B-Instruct, DeepSeek-R1-Distill-Qwen-1.5B, Qwen-Qwen2.5-1.5B-Instruct

[ynankani]

• Add mixed precision nvModelOpt recipes for Llama-3.2-1B-Instruct
• Add mixed precision nvModelOpt recipes for DeepSeek-R1-Distill-Qwen-1.5B
• Add mixed precision nvModelOpt recipes for Qwen-Qwen2.5-1.5B-Instruct

Observed improvement in mmlu and perplexity score for the above model with mixed (Int4+Int8) precision quantization compared to standard int4 quantization.

<style> </style>

MMLU
Model	FP16-MB	Mixed AWQ -MO	Mixed RTN-MO	Pure INT4 AWQ-MO	Pure INT4 RTN-MO
DeepSeek R1 Distill Qwen 1.5B	36.60%	34.80%	33.90%	33.10%	32.40%
Llama 3.2 1B Instruct	47.30%	44.40%	44.70%	43.20%	39.90%
Qwen 2.5 1.5B Instruct	60.00%	57.50%	57.50%	56.60%	56.70%
Perplexity (isl=1024, stride=512)
Model	FP16-MB	Mixed AWQ -MO	Mixed RTN-MO	Pure INT4 AWQ-MO	Pure INT4 RTN-MO
DeepSeek R1 Distill Qwen 1.5B	39.447	41.699	44.332	44.213	46.304
Llama 3.2 1B Instruct	12.631	13.852	14.176	14.549	16.9
Qwen 2.5 1.5B Instruct	9.216	10.084	10.338	10.495	10.933

[ynankani]

Please review, merge.

CC @devang-ml

[shaahji]

Please fix the failing pre-commit checks. Also, update the branch to the latest.