Hugging Face's logo

Datasets:
optimum-benchmark
/
cpu

Dataset card Viewer Files Community
IlyasMoutawwakil HF staff commited on
Commit
f367b13
·
verified ·
1 Parent(s): 1cd0fb9

Upload cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark_report.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark_report.json +82 -79
cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark_report.json CHANGED
@@ -2,7 +2,7 @@
2
  "prefill": {
3
  "memory": {
4
  "unit": "MB",
5
- "max_ram": 961.970176,
6
  "max_global_vram": null,
7
  "max_process_vram": null,
8
  "max_reserved": null,
@@ -10,51 +10,52 @@
10
  },
11
  "latency": {
12
  "unit": "s",
13
- "count": 14,
14
- "total": 0.6281763009999395,
15
- "mean": 0.04486973578570996,
16
- "stdev": 0.0007107499136287721,
17
- "p50": 0.04480706999999029,
18
- "p90": 0.045821137499999054,
19
- "p95": 0.04597427304999542,
20
- "p99": 0.046066543409987164,
21
  "values": [
22
- 0.0460896109999851,
23
- 0.044436222000001635,
24
- 0.04530898299998398,
25
- 0.0452456639999923,
26
- 0.044727555999998,
27
- 0.04488658399998258,
28
- 0.04439754000000562,
29
- 0.04462246099998879,
30
- 0.044084506000018564,
31
- 0.044999555000003966,
32
- 0.045912168000000975,
33
- 0.04452786400000264,
34
- 0.04560873299999457,
35
- 0.043328853999980765
 
36
  ]
37
  },
38
  "throughput": {
39
  "unit": "tokens/s",
40
- "value": 44.57347396810294
41
  },
42
  "energy": {
43
  "unit": "kWh",
44
- "cpu": 1.7387622796394976e-06,
45
- "ram": 7.266438844681034e-08,
46
  "gpu": 0.0,
47
- "total": 1.811426668086308e-06
48
  },
49
  "efficiency": {
50
  "unit": "tokens/kWh",
51
- "value": 1104102.106497588
52
  }
53
  },
54
  "decode": {
55
  "memory": {
56
  "unit": "MB",
57
- "max_ram": 961.970176,
58
  "max_global_vram": null,
59
  "max_process_vram": null,
60
  "max_reserved": null,
@@ -62,79 +63,81 @@
62
  },
63
  "latency": {
64
  "unit": "s",
65
- "count": 14,
66
- "total": 0.38396165699995777,
67
- "mean": 0.027425832642854125,
68
- "stdev": 0.0007352998775288754,
69
- "p50": 0.02731566999999302,
70
- "p90": 0.028309482199998116,
71
- "p95": 0.028747917999997697,
72
- "p99": 0.0290540420000022,
73
  "values": [
74
- 0.02706296100001282,
75
- 0.027073289999975714,
76
- 0.02773456500000293,
77
- 0.027767237000006162,
78
- 0.027402034999994385,
79
- 0.027203744000019014,
80
- 0.029130573000003324,
81
- 0.026431651999985206,
82
- 0.027199194999980136,
83
- 0.027229304999991655,
84
- 0.02755747599999836,
85
- 0.027540241999986392,
86
- 0.028541872999994666,
87
- 0.026087509000007003
 
88
  ]
89
  },
90
  "throughput": {
91
  "unit": "tokens/s",
92
- "value": 36.461974118424905
93
  },
94
  "energy": {
95
  "unit": "kWh",
96
- "cpu": 9.89037754071124e-07,
97
- "ram": 4.133575090099837e-08,
98
  "gpu": 0.0,
99
- "total": 1.0303735049721216e-06
100
  },
101
  "efficiency": {
102
  "unit": "tokens/kWh",
103
- "value": 970521.8497704447
104
  }
105
  },
106
  "per_token": {
107
  "memory": null,
108
  "latency": {
109
  "unit": "s",
110
- "count": 14,
111
- "total": 0.3789960939999162,
112
- "mean": 0.027071149571422586,
113
- "stdev": 0.000703221778495897,
114
- "p50": 0.026973488999985307,
115
- "p90": 0.027971441800002595,
116
- "p95": 0.02835305064999858,
117
- "p99": 0.0285585229300014,
118
  "values": [
119
- 0.026718358000010767,
120
- 0.026739355999978898,
121
- 0.027364022999989857,
122
- 0.027403718000016397,
123
- 0.027070224999988568,
124
- 0.026845895999997538,
125
- 0.028609891000002108,
126
- 0.026091646999987006,
127
- 0.026876752999982045,
128
- 0.026860505999991346,
129
- 0.02723100599999384,
130
- 0.027196229999987054,
131
- 0.02821475199999668,
132
- 0.02577373299999408
 
133
  ]
134
  },
135
  "throughput": {
136
  "unit": "tokens/s",
137
- "value": 36.9396946872046
138
  },
139
  "energy": null,
140
  "efficiency": null
 
2
  "prefill": {
3
  "memory": {
4
  "unit": "MB",
5
+ "max_ram": 951.169024,
6
  "max_global_vram": null,
7
  "max_process_vram": null,
8
  "max_reserved": null,
 
10
  },
11
  "latency": {
12
  "unit": "s",
13
+ "count": 15,
14
+ "total": 0.651997793999783,
15
+ "mean": 0.04346651959998553,
16
+ "stdev": 0.0018563076160604478,
17
+ "p50": 0.04241198599999052,
18
+ "p90": 0.04671048879998807,
19
+ "p95": 0.04679353020000008,
20
+ "p99": 0.04680695563999279,
21
  "values": [
22
+ 0.04576482500004886,
23
+ 0.042967419999968115,
24
+ 0.042281572000035794,
25
+ 0.04210176699996282,
26
+ 0.04235397700000476,
27
+ 0.042249752999964585,
28
+ 0.042684934999954294,
29
+ 0.04210051599994813,
30
+ 0.04183603200004882,
31
+ 0.04681031199999097,
32
+ 0.046596714999964206,
33
+ 0.04678633800000398,
34
+ 0.04224789999994982,
35
+ 0.042803745999947296,
36
+ 0.04241198599999052
37
  ]
38
  },
39
  "throughput": {
40
  "unit": "tokens/s",
41
+ "value": 46.01242561874371
42
  },
43
  "energy": {
44
  "unit": "kWh",
45
+ "cpu": 1.6525432467460635e-06,
46
+ "ram": 6.906195599910347e-08,
47
  "gpu": 0.0,
48
+ "total": 1.7216052027451668e-06
49
  },
50
  "efficiency": {
51
  "unit": "tokens/kWh",
52
+ "value": 1161706.5264503858
53
  }
54
  },
55
  "decode": {
56
  "memory": {
57
  "unit": "MB",
58
+ "max_ram": 951.169024,
59
  "max_global_vram": null,
60
  "max_process_vram": null,
61
  "max_reserved": null,
 
63
  },
64
  "latency": {
65
  "unit": "s",
66
+ "count": 15,
67
+ "total": 0.40866484299999684,
68
+ "mean": 0.027244322866666455,
69
+ "stdev": 0.0016764205992976003,
70
+ "p50": 0.026517742999999427,
71
+ "p90": 0.029772413199987115,
72
+ "p95": 0.030090030100001285,
73
+ "p99": 0.030550253220022795,
74
  "values": [
75
+ 0.02569557999999006,
76
+ 0.026732293000009122,
77
+ 0.02600638999996363,
78
+ 0.026550244000020484,
79
+ 0.026506561999951828,
80
+ 0.02580723900001658,
81
+ 0.026344940000001316,
82
+ 0.025876207000010254,
83
+ 0.028901275999999143,
84
+ 0.029843481999989763,
85
+ 0.02966580999998314,
86
+ 0.030665309000028174,
87
+ 0.025334357000019736,
88
+ 0.026517742999999427,
89
+ 0.02821741100001418
90
  ]
91
  },
92
  "throughput": {
93
  "unit": "tokens/s",
94
+ "value": 36.70489462681799
95
  },
96
  "energy": {
97
  "unit": "kWh",
98
+ "cpu": 9.659159073123232e-07,
99
+ "ram": 4.0366258564091585e-08,
100
  "gpu": 0.0,
101
+ "total": 1.0062821658764144e-06
102
  },
103
  "efficiency": {
104
  "unit": "tokens/kWh",
105
+ "value": 993757.0533499985
106
  }
107
  },
108
  "per_token": {
109
  "memory": null,
110
  "latency": {
111
  "unit": "s",
112
+ "count": 15,
113
+ "total": 0.4029724250000868,
114
+ "mean": 0.02686482833333912,
115
+ "stdev": 0.0016246870931077404,
116
+ "p50": 0.026195022000024437,
117
+ "p90": 0.02925115600000936,
118
+ "p95": 0.02962354190001406,
119
+ "p99": 0.0301667827799929,
120
  "values": [
121
+ 0.025339305999978023,
122
+ 0.026379845999997542,
123
+ 0.02566390199996249,
124
+ 0.026195022000024437,
125
+ 0.026159093999979177,
126
+ 0.025473046000001887,
127
+ 0.0260016810000252,
128
+ 0.02552204700003813,
129
+ 0.028392617999998038,
130
+ 0.029332520000025397,
131
+ 0.029129109999985303,
132
+ 0.03030259299998761,
133
+ 0.02502849700005072,
134
+ 0.02620063200004097,
135
+ 0.027852510999991864
136
  ]
137
  },
138
  "throughput": {
139
  "unit": "tokens/s",
140
+ "value": 37.22339065755372
141
  },
142
  "energy": null,
143
  "efficiency": null