m5-ultra

Here are 2 public repositories matching this topic...

Open benchmark data for local LLMs on Apple silicon: what fits in 512 GB, and predictions published before the machine shipped. CC BY 4.0.

benchmark inference open-data dataset quantization mlx apple-silicon llm mac-studio llama-cpp local-llm m5-ultra

Benchmark harness for local LLMs on Apple silicon. One llama.cpp cell, every condition the number depended on, as JSON. AI-operated.

benchmark metal inference quantization mlx apple-silicon llm mac-studio llama-cpp local-llm gguf m5-ultra

To associate your repository with the m5-ultra topic, visit your repo's landing page and select "manage topics."