File size: 1,826 Bytes
f741e18
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
09c008d
 
 
 
 
f741e18
 
 
 
 
 
 
 
 
 
 
 
 
09c008d
 
 
 
 
f741e18
 
 
 
 
 
 
 
 
 
 
 
 
09c008d
 
 
 
 
f741e18
 
 
 
 
 
 
09c008d
 
f741e18
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
{
 "command_a": {
  "n": 331,
  "verified": [
   0.8096676737160121,
   0.7639169970960494,
   0.8483127855832832
  ],
  "single": [
   0.8277945619335347,
   0.7834184000616147,
   0.8646492235550447
  ],
  "usd_per_1k": 4.244501510574018,
  "tokens_per_query": [
   666.404833836858,
   257.8489425981873
  ],
  "latency_s_median": 9.74,
  "latency_s_p90": 42.22
 },
 "sft_ho": {
  "n": 2648,
  "verified": [
   0.09441087613293052,
   0.08385386452215371,
   0.10614300406909247
  ],
  "single": [
   0.145392749244713,
   0.13248019213082912,
   0.15933271253134978
  ],
  "gen_s_laptop": 465.57,
  "gen_s_l4": 698.1727373245488,
  "l4_time": "laptop time x B's L4/laptop ratio",
  "trajectories_per_s": 3.7927576635938807,
  "usd_per_1k": 0.05419685985441488
 },
 "B": {
  "n": 2648,
  "verified": [
   0.1268882175226586,
   0.11474862442249444,
   0.1401088301102348
  ],
  "single": [
   0.1782477341389728,
   0.16413956667413937,
   0.19328811678930724
  ],
  "gen_s_laptop": 523.65,
  "gen_s_l4": 785.27,
  "l4_time": "measured on L4",
  "trajectories_per_s": 3.3720885809976187,
  "usd_per_1k": 0.06095793470963409
 },
 "C": {
  "n": 2648,
  "verified": [
   0.08950151057401813,
   0.07921473537422782,
   0.10097762603981741
  ],
  "single": [
   0.1404833836858006,
   0.12776814975928472,
   0.15424024772569325
  ],
  "gen_s_laptop": 443.21,
  "gen_s_l4": 664.6414908813138,
  "l4_time": "laptop time x B's L4/laptop ratio",
  "trajectories_per_s": 3.984102762661951,
  "usd_per_1k": 0.05159393916290829
 },
 "assumptions": {
  "command_a_usd_per_1M": [
   2.5,
   10.0
  ],
  "gpu_usd_hr": 0.74,
  "l4_over_laptop_ratio": 1.4996085171393105,
  "throughput": "vLLM batched, 331 items x 8 samples, T=0.7; B timed on the L4 (785 s) and on an RTX 5070 Ti laptop; per-query latency not measured for Tiny Aya"
 }
}