M2 (8-core GPU, 16 GB)
3 records across 3 models (quants: Q4_K - Medium).
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M2 (8-core GPU, 16 GB) | llama-3-1-8b-instruct | Q4_K - Medium | llamafile | 12.906848888888891 | 113.95904577777776 | – | 10.946365963333333 | – | – | sourced | – | source |
| M2 (8-core GPU, 16 GB) | llama-3-2-1b-instruct | Q4_K - Medium | llamafile | 35.320437777777784 | 523.1656983888889 | – | 3.046989798555555 | – | – | sourced | – | source |
| M2 (8-core GPU, 16 GB) | qwen-2-5-14b-instruct | Q4_K - Medium | llamafile | 6.960648666666668 | 59.9820798888889 | – | 21.432082513888886 | – | – | sourced | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085