Here is another test with Llama 4 Scout - 6 bit quantized version on M4 Max. Got ~30t/s which is faster than most of the hosted models.

Tested Llama 4 Scout 4-bit quantized version on MLX and it is running at ~40t/s completely offline. It consumes around ~61GB of memory. It seems 64GB Macs won't be enough to run this without spilling into swap.

@jay.bsky.team there will be general elections in Turkey tomorrow. Would you mind handing some invites out in case Twitter gets blocked in Turkey? It can keep the communication open during critical times.

Had a boat tour today. Great weather, great views.

What a nice place to vote today! 😊

Having a boat trip for the closing party (funeral?) of the Book Depository.

Goodbye Book Depository! You will be missed

Happy to announce that we win the global hackathon in Amazon about LLMs! 🎉Good experience to mash up different tools to create a project in a short time.

Hello world!