No, seed and temperature are separate parameters accepted by the inference engine. You can still get deterministic outputs with high temp if you're using the same seed, provided the inference engine itself operates in a deterministic manner, and the hardware is deterministic (in testing, we did observe small non-deterministic variations when running the same prompt on the same stack but a different model of GPU).