Building Local SLM Inference & Quantized Serving in Pure Python: GGUF Parsing, Q8/Q4 Kernels, LoRA Fusion & Zero-GPU Edge Deployment

In modern enterprise artificial intelligence, relying exclusively on centralized, proprietary cloud APIs (such as OpenAI, Anthropic, or Google) has introduced catastrophic architectural and financial