| 12345678910 |
- c1-prose decode 33.86 tok/s (min 33.49 max 34.04 sd 0.20) ttft 0.273s toks [512, 512, 512, 512, 512] distinct 1 err 0 2026-09-24T06:32:33.259239Z..2026-09-24T06:33:50.275805Z
- c1-code decode 61.67 tok/s (min 60.44 max 62.28 sd 0.60) ttft 0.284s toks [512, 512, 512, 512, 512] distinct 3 err 0 2026-09-24T06:33:50.276168Z..2026-09-24T06:34:33.215461Z
- c1-chat_sampled decode 36.73 tok/s (min 35.73 max 37.73 sd 0.67) ttft 0.262s toks [512, 512, 512, 512, 512] distinct 5 err 0 2026-09-24T06:34:33.215784Z..2026-09-24T06:35:43.997466Z
- c1-prose2 decode 32.25 tok/s (min 31.56 max 32.53 sd 0.32) ttft 0.250s toks [512, 512, 512, 512, 512] distinct 4 err 0 2026-09-24T06:35:43.997651Z..2026-09-24T06:37:04.733585Z
- c4 agg 76.80 tok/s (min 75.33 max 77.26) per-stream 24.77 2026-09-24T06:37:04.733837Z..2026-09-24T06:39:18.373803Z
- c1-prose accept_len 1.93 step 58.4 ms (n=33 intervals, bs=1)
- c1-code accept_len 4.12 step 68.2 ms (n=16 intervals, bs=1.0)
- c1-chat_sampled accept_len 2.23 step 60.3 ms (n=29 intervals, bs=1)
- c1-prose2 accept_len 1.73 step 57.5 ms (n=35 intervals, bs=1)
- c4 accept_len 2.49 step 78.1 ms (n=32 intervals, bs=2.0)
|
