I've found threads online that suggest that running gpt-oss-20b on ollama is slow for some reason. I'm running the 20b model via LM Studio on a 2021 M1 and I'm

I've found threads online that suggest that running gpt-oss-20b on ollama is slow for some reason. I'm running the 20b model via LM Studio on a 2021 M1 and I'm consistently getting around 50-60 T/s.