[ab] server http://172.17.0.1:8100 model /models/Qwen3.8-27B-exl3-4.0bpw
[transformers] Token indices sequence length is longer than the specified maximum sequence length for this model (317092 > 262144). Running this sequence through the model will result in indexing errors
  [0] code       prompt=8244 out=512 proposed=865 accepted=343 rate=0.3965 tok/s= 63.11 length
  [1] code       prompt=8243 out=512 proposed=895 accepted=333 rate=0.3721 tok/s= 61.67 length
  [2] code       prompt=8243 out=512 proposed=905 accepted=331 rate=0.3657 tok/s= 61.12 length
  [3] code       prompt=8244 out=512 proposed=805 accepted=351 rate=0.4360 tok/s= 66.31 length
  [4] wiki       prompt=8244 out=512 proposed=915 accepted=330 rate=0.3607 tok/s= 60.56 length
  [5] wiki       prompt=8244 out=512 proposed=865 accepted=338 rate=0.3908 tok/s= 62.86 length
  [6] wiki       prompt=8244 out=512 proposed=795 accepted=356 rate=0.4478 tok/s= 66.73 length
  [7] wiki       prompt=8244 out=512 proposed=920 accepted=328 rate=0.3565 tok/s= 60.18 length
  [8] math       prompt=8244 out=512 proposed=830 accepted=348 rate=0.4193 tok/s= 64.68 length
  [9] math       prompt=8244 out=512 proposed=715 accepted=370 rate=0.5175 tok/s= 71.52 length
  [10] math       prompt=8244 out=512 proposed=970 accepted=317 rate=0.3268 tok/s= 58.00 length
  [11] math       prompt=8244 out=352 proposed=505 accepted=251 rate=0.4970 tok/s= 61.04 stop
  [12] longctx    prompt=8244 out=512 proposed=930 accepted=326 rate=0.3505 tok/s= 59.76 length
  [13] longctx    prompt=8244 out=512 proposed=895 accepted=334 rate=0.3732 tok/s= 61.40 length
  [14] longctx    prompt=8244 out=512 proposed=895 accepted=336 rate=0.3754 tok/s= 61.47 length
  [15] longctx    prompt=8244 out=512 proposed=885 accepted=335 rate=0.3785 tok/s= 61.94 length
  [16] sci        prompt=8244 out=129 proposed=200 accepted=90 rate=0.4500 tok/s= 34.14 stop
  [17] sci        prompt=8244 out=512 proposed=835 accepted=346 rate=0.4144 tok/s= 64.35 length
  [18] sci        prompt=8244 out=158 proposed=260 accepted=109 rate=0.4192 tok/s= 37.77 stop
  [19] sci        prompt=8245 out=258 proposed=405 accepted=180 rate=0.4444 tok/s= 50.29 stop
[ab] pooled accept rate 0.3958 over 20 prompts -> /cache/ab/OFFg.json

## pins

## pins (from the server log)
