| 12345678910 |
- c1-prose decode 30.20 tok/s (min 30.20 max 30.30 sd 0.04) ttft 0.295s toks [512, 512, 512, 512, 512] distinct 3 err 0 2026-09-24T03:10:08.813109Z..2026-09-24T03:11:34.837253Z
- c1-code decode 52.67 tok/s (min 52.27 max 52.82 sd 0.20) ttft 0.311s toks [512, 512, 512, 512, 512] distinct 4 err 0 2026-09-24T03:11:34.837672Z..2026-09-24T03:12:24.922433Z
- c1-chat_sampled decode 31.77 tok/s (min 29.70 max 32.89 sd 1.08) ttft 0.285s toks [512, 512, 512, 512, 512] distinct 5 err 0 2026-09-24T03:12:24.922745Z..2026-09-24T03:13:47.283789Z
- c1-prose2 decode 28.10 tok/s (min 27.98 max 28.46 sd 0.17) ttft 0.275s toks [512, 512, 512, 512, 512] distinct 4 err 0 2026-09-24T03:13:47.284098Z..2026-09-24T03:15:19.360703Z
- c4 agg 68.90 tok/s (min 67.26 max 70.39) per-stream 22.82 2026-09-24T03:15:19.360900Z..2026-09-24T03:17:48.375919Z
- c1-prose accept_len 2.00 step 68.4 ms (n=31 intervals, bs=1)
- c1-code accept_len 4.15 step 78.8 ms (n=16 intervals, bs=1.0)
- c1-chat_sampled accept_len 2.21 step 69.8 ms (n=30 intervals, bs=1.0)
- c1-prose2 accept_len 1.85 step 66.0 ms (n=35 intervals, bs=1)
- c4 accept_len 2.31 step 112.1 ms (n=32 intervals, bs=3.0)
|
