Update with corrected usage guide and working model

Add CPU-optimized quantized SLM model

initial commit