Add comprehensive quantization guide for DeepSeek-Coder-V2

[isaacmujuni]

• Addresses Issue #79: How to quantize DeepSeek-Coder-V2 for VLLM inference
• Provides detailed quantization methods for vLLM, SGLang, llama.cpp, and AutoGPTQ
• Includes performance comparisons and memory requirements
• Adds troubleshooting section for common issues
• Updates README.md with reference to the quantization guide

This guide helps users efficiently deploy DeepSeek-Coder-V2 models with reduced memory usage while maintaining high code generation quality.