M1 (8-core GPU, 16 GB)
2 records across 2 models (quants: Q4_K - Medium).
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M1 (8-core GPU, 16 GB) | llama-3-2-1b-instruct | Q4_K - Medium | llamafile | 40.18323162962963 | 585.7259774074074 | – | 2.24642097837037 | – | – | sourced | – | source |
| M1 (8-core GPU, 16 GB) | qwen-2-5-14b-instruct | Q4_K - Medium | llamafile | 5.429473666666667 | 53.26808444444444 | – | 24.55238074088889 | – | – | sourced | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085