Method has_fp4_kv_cache

tensorrt_llm/quantization/mode.py:170–171

(self)

        return self._any(self.FP8_KV_CACHE)

    def has_fp4_kv_cache(self):
        return self._any(self.NVFP4_KV_CACHE)

    def has_kv_cache_quant(self):
        return (self.has_int8_kv_cache() or self.has_fp8_kv_cache()