feat: ✨ add M1 Max MacBook Pro

[Supporterino]

                    'c.          supporterino@Freya-II
                 ,xNMM.          ---------------------
               .OMMMMo           OS: macOS 26.0.1 25A362 arm64
               OMMM0,            Host: MacBookPro18,4
     .;loddo:' loolloddol;.      Kernel: 25.0.0
   cKMMMMMMMMMMNWMMMMMMMMMM0:    Uptime: 6 days, 4 hours, 22 mins
 .KMMMMMMMMMMMMMMMMMMMMMMMWd.    Packages: 209 (brew)
 XMMMMMMMMMMMMMMMMMMMMMMMX.      Shell: zsh 5.9
;MMMMMMMMMMMMMMMMMMMMMMMM:       Resolution: 1800x1169
:MMMMMMMMMMMMMMMMMMMMMMMM:       DE: Aqua
.MMMMMMMMMMMMMMMMMMMMMMMMX.      WM: Quartz Compositor
 kMMMMMMMMMMMMMMMMMMMMMMMMWd.    WM Theme: Blue (Dark)
 .XMMMMMMMMMMMMMMMMMMMMMMMMMMk   Terminal: iTerm2
  .XMMMMMMMMMMMMMMMMMMMMMMMMK.   Terminal Font: MesloLGS-NF-Regular 13
    kMMMMMMMMMMMMMMMMMMMMMMd     CPU: Apple M1 Max
     ;KMMMMMMMWXXWMMMMMMMk.      GPU: Apple M1 Max
       .cooc,.    .,coo:.        Memory: 4747MiB / 32768MiB

llama3.2:3b

running benchmark 3 times using model: llama3.2:3b

Run	Eval Rate (Tokens/Second)
1	80.32 tokens/s
2	80.11 tokens/s
3	80.17 tokens/s
Average Eval Rate	80.20 tokens/second

deepseek-r1

Running benchmark 3 times using model: deepseek-r1

Run	Eval Rate (Tokens/Second)
1	35.99 tokens/s
2	33.12 tokens/s
3	32.74 tokens/s
Average Eval Rate	33.95 tokens/second

[geerlingguy]

@Supporterino which version of deepseek-r1 was run? I presume it was maybe 8b and not 14b, since that would be the default according to https://ollama.com/library/deepseek-r1

You can run a specific version specifying like deepseek-r1:14b