Submissions from twitter.com/awnihannun

		I've been thinking a bit about continual learning (twitter.com/awnihannun)
		2 points by tosh 3 months ago \| past
		Three thoughts on the future of writing software (twitter.com/awnihannun)
		2 points by tosh 3 months ago \| past
		GLM-5 with mlx-lm on single 512GB M3 Ultra in Q4 (twitter.com/awnihannun)
		1 point by tosh 4 months ago \| past
		voxmlx: MLX implementation of Mistral's Voxtral mini realtime speech recognition (twitter.com/awnihannun)
		2 points by tosh 4 months ago \| past
		Kimi K2.5 1T runs on 2 M3 Ultras with mlx-lm in it's native precision (twitter.com/awnihannun)
		5 points by jeudesprits 4 months ago \| past
		Latest MLX Release Includes Jaccl RDMA Back End over TB5 (twitter.com/awnihannun)
		2 points by geerlingguy 5 months ago \| past
		Kimi K2 1T model runs on 2 512GB M3 Ultras (twitter.com/awnihannun)
		234 points by jeudesprits 6 months ago \| past \| 121 comments
		Transformers are almost adversarially designed for computer memory hierarchy (twitter.com/awnihannun)
		4 points by tosh 10 months ago \| past
		DeepSeek R1 671B running on 2 M2 Ultras faster than reading speed (twitter.com/awnihannun)
		96 points by thyrox on Jan 29, 2025 \| past \| 29 comments
		Apple MLX was open sourced one year ago (twitter.com/awnihannun)
		3 points by amrrs on Dec 5, 2024 \| past
		MLX 0.11: faster generation across model sizes and machines (twitter.com/awnihannun)
		3 points by tosh on April 20, 2024 \| past
		With the latest MLX, 4-bit Llama 3 8B runs nicely on an 8GB M2 mini (twitter.com/awnihannun)
		2 points by mariuz on April 19, 2024 \| past
		100 tokens/s, 4-bit Mistral 7B in MLX on M2 Ultra (faster than llama.cpp) (twitter.com/awnihannun)
		3 points by tosh on April 4, 2024 \| past
		Apple is hiring GPU kernel engineers for the MLX project (twitter.com/awnihannun)
		4 points by behnamoh on Jan 24, 2024 \| past
		Mistral 7B 4-bit quantization runs no problem on an 8GB M2 (twitter.com/awnihannun)
		20 points by tosh on Dec 22, 2023 \| past