Add Hugging Face AWQ for quantization

[qte77]

• HF AWQ
• Transformer meets AWQ quantization (AutoAWQ and LLM-AWQ) for lighter and faster quantized inference of LLMs