Add Qwen3.5-4B W8A8 (compressed-tensors int8) for vLLM 635f962 verified NotaMG commited on 24 days ago