I'm on an Intel iMac, and llama can't leverage its GPU. So, it's 1990s slow. It's literally like talking to a machine in the 90s, from the slow response time to

I'm on an Intel iMac, and llama can't leverage its GPU. So, it's 1990s slow. It's literally like talking to a machine in the 90s, from the slow response time to the 1200-2400 baud output. It's easy to give tasks to, hard to have a conversation with. Just paste the task in and let it churn, come back to it later.