Hugging Face's logo

Datasets:
optimum-benchmark
/
cpu

Dataset card Viewer Files Community
IlyasMoutawwakil HF staff commited on
Commit
5ff8cc1
•
1 Parent(s): 7bae103

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 +87 -87
cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark_report.json CHANGED
@@ -2,7 +2,7 @@
2
  "load": {
3
  "memory": {
4
  "unit": "MB",
5
- "max_ram": 1110.188032,
6
  "max_global_vram": null,
7
  "max_process_vram": null,
8
  "max_reserved": null,
@@ -11,15 +11,15 @@
11
  "latency": {
12
  "unit": "s",
13
  "values": [
14
- 1.748812111999996
15
  ],
16
  "count": 1,
17
- "total": 1.748812111999996,
18
- "mean": 1.748812111999996,
19
- "p50": 1.748812111999996,
20
- "p90": 1.748812111999996,
21
- "p95": 1.748812111999996,
22
- "p99": 1.748812111999996,
23
  "stdev": 0,
24
  "stdev_": 0
25
  },
@@ -30,7 +30,7 @@
30
  "prefill": {
31
  "memory": {
32
  "unit": "MB",
33
- "max_ram": 965.091328,
34
  "max_global_vram": null,
35
  "max_process_vram": null,
36
  "max_reserved": null,
@@ -39,51 +39,51 @@
39
  "latency": {
40
  "unit": "s",
41
  "values": [
42
- 0.045159581000007165,
43
- 0.04782010300004913,
44
- 0.048109933000034744,
45
- 0.045600812999964546,
46
- 0.04557635800000526,
47
- 0.04686897900000986,
48
- 0.045508582000024944,
49
- 0.04807066099999702,
50
- 0.04799457799998663,
51
- 0.04759849999999233,
52
- 0.0461754049999854,
53
- 0.04738526099998808,
54
- 0.04742278100002295,
55
- 0.04446566699999721
56
  ],
57
  "count": 14,
58
- "total": 0.6537572020000653,
59
- "mean": 0.04669694300000466,
60
- "p50": 0.04712711999999897,
61
- "p90": 0.0480478360999939,
62
- "p95": 0.04808440620001022,
63
- "p99": 0.04810482764002984,
64
- "stdev": 0.0011998660854343115,
65
- "stdev_": 2.569474591589843
66
  },
67
  "throughput": {
68
  "unit": "tokens/s",
69
- "value": 42.82935608868016
70
  },
71
  "energy": {
72
  "unit": "kWh",
73
- "cpu": 1.6855388877314497e-06,
74
- "ram": 7.047918965611992e-08,
75
  "gpu": 0.0,
76
- "total": 1.7560180773875695e-06
77
  },
78
  "efficiency": {
79
  "unit": "tokens/kWh",
80
- "value": 1138940.439027486
81
  }
82
  },
83
  "decode": {
84
  "memory": {
85
  "unit": "MB",
86
- "max_ram": 965.091328,
87
  "max_global_vram": null,
88
  "max_process_vram": null,
89
  "max_reserved": null,
@@ -92,45 +92,45 @@
92
  "latency": {
93
  "unit": "s",
94
  "values": [
95
- 0.029038425999999617,
96
- 0.02787078799997289,
97
- 0.028081450999991375,
98
- 0.027976996000006693,
99
- 0.02838096900001119,
100
- 0.02798565199998393,
101
- 0.028071210999996765,
102
- 0.02770303500000182,
103
- 0.029908277999993516,
104
- 0.028626497000004747,
105
- 0.027926182000044264,
106
- 0.0290267440000207,
107
- 0.026313824999988356,
108
- 0.027343624999957683
109
  ],
110
  "count": 14,
111
- "total": 0.39425367899997354,
112
- "mean": 0.02816097707142668,
113
- "p50": 0.028028431499990347,
114
- "p90": 0.029034921400005943,
115
- "p95": 0.029342874199997482,
116
- "p99": 0.029795197239994308,
117
- "stdev": 0.0008170368758343155,
118
- "stdev_": 2.901308693096859
119
  },
120
  "throughput": {
121
  "unit": "tokens/s",
122
- "value": 35.51013153640334
123
  },
124
  "energy": {
125
  "unit": "kWh",
126
- "cpu": 9.79358107453715e-07,
127
- "ram": 4.095441563995275e-08,
128
  "gpu": 0.0,
129
- "total": 1.020312523093668e-06
130
  },
131
  "efficiency": {
132
  "unit": "tokens/kWh",
133
- "value": 980091.8614307713
134
  }
135
  },
136
  "per_token": {
@@ -138,34 +138,34 @@
138
  "latency": {
139
  "unit": "s",
140
  "values": [
141
- 0.028640944000017043,
142
- 0.02746865799997522,
143
- 0.027691313999980594,
144
- 0.027526566000005914,
145
- 0.027957318999995096,
146
- 0.02758130799998071,
147
- 0.027670403999991322,
148
- 0.02723674500003881,
149
- 0.029461855000022297,
150
- 0.028227052000033837,
151
- 0.02752711700003374,
152
- 0.028434921000041413,
153
- 0.025921832999983963,
154
- 0.026954859999989367
155
  ],
156
  "count": 14,
157
- "total": 0.3883008960000893,
158
- "mean": 0.027735778285720665,
159
- "p50": 0.027625855999986015,
160
- "p90": 0.028579137100024354,
161
- "p95": 0.02892826285001888,
162
- "p99": 0.029355136570021614,
163
- "stdev": 0.000798030435648703,
164
- "stdev_": 2.87725993274022
165
  },
166
  "throughput": {
167
  "unit": "tokens/s",
168
- "value": 36.05451376552265
169
  },
170
  "energy": null,
171
  "efficiency": null
 
2
  "load": {
3
  "memory": {
4
  "unit": "MB",
5
+ "max_ram": 1122.46784,
6
  "max_global_vram": null,
7
  "max_process_vram": null,
8
  "max_reserved": null,
 
11
  "latency": {
12
  "unit": "s",
13
  "values": [
14
+ 1.7530449449999992
15
  ],
16
  "count": 1,
17
+ "total": 1.7530449449999992,
18
+ "mean": 1.7530449449999992,
19
+ "p50": 1.7530449449999992,
20
+ "p90": 1.7530449449999992,
21
+ "p95": 1.7530449449999992,
22
+ "p99": 1.7530449449999992,
23
  "stdev": 0,
24
  "stdev_": 0
25
  },
 
30
  "prefill": {
31
  "memory": {
32
  "unit": "MB",
33
+ "max_ram": 976.846848,
34
  "max_global_vram": null,
35
  "max_process_vram": null,
36
  "max_reserved": null,
 
39
  "latency": {
40
  "unit": "s",
41
  "values": [
42
+ 0.04475181399999428,
43
+ 0.04635989000001928,
44
+ 0.04564587999999503,
45
+ 0.04473374999997759,
46
+ 0.044882556999993994,
47
+ 0.044579342000020006,
48
+ 0.04701173600000175,
49
+ 0.04668405500001427,
50
+ 0.04642194699999891,
51
+ 0.04683937499999047,
52
+ 0.04586274399997592,
53
+ 0.048819994999973915,
54
+ 0.047081325000021934,
55
+ 0.04525933000002169
56
  ],
57
  "count": 14,
58
+ "total": 0.644933739999999,
59
+ "mean": 0.04606669571428564,
60
+ "p50": 0.0461113169999976,
61
+ "p90": 0.04706044830001588,
62
+ "p95": 0.04768985950000513,
63
+ "p99": 0.048593967899980155,
64
+ "stdev": 0.0011552411049127027,
65
+ "stdev_": 2.507757691321571
66
  },
67
  "throughput": {
68
  "unit": "tokens/s",
69
+ "value": 43.41531271103919
70
  },
71
  "energy": {
72
  "unit": "kWh",
73
+ "cpu": 1.813345296212166e-06,
74
+ "ram": 7.582284233777658e-08,
75
  "gpu": 0.0,
76
+ "total": 1.8891681385499426e-06
77
  },
78
  "efficiency": {
79
  "unit": "tokens/kWh",
80
+ "value": 1058667.0181380087
81
  }
82
  },
83
  "decode": {
84
  "memory": {
85
  "unit": "MB",
86
+ "max_ram": 976.846848,
87
  "max_global_vram": null,
88
  "max_process_vram": null,
89
  "max_reserved": null,
 
92
  "latency": {
93
  "unit": "s",
94
  "values": [
95
+ 0.028068726000014976,
96
+ 0.029693993999956092,
97
+ 0.02843563900000845,
98
+ 0.027791218000004392,
99
+ 0.02799295400001256,
100
+ 0.027714114999980666,
101
+ 0.031720549999988634,
102
+ 0.028388199999994868,
103
+ 0.028711142000020118,
104
+ 0.028218223999999736,
105
+ 0.02761716399999159,
106
+ 0.03170004200001131,
107
+ 0.028573396000012963,
108
+ 0.027913205999993806
109
  ],
110
  "count": 14,
111
+ "total": 0.40253856999999016,
112
+ "mean": 0.028752754999999297,
113
+ "p50": 0.028303211999997302,
114
+ "p90": 0.031098227599994744,
115
+ "p95": 0.03170721980000337,
116
+ "p99": 0.03171788395999158,
117
+ "stdev": 0.0013086594938723289,
118
+ "stdev_": 4.551422964068524
119
  },
120
  "throughput": {
121
  "unit": "tokens/s",
122
+ "value": 34.77927593373311
123
  },
124
  "energy": {
125
  "unit": "kWh",
126
+ "cpu": 1.1213189260100731e-06,
127
+ "ram": 4.688936083362013e-08,
128
  "gpu": 0.0,
129
+ "total": 1.1682082868436928e-06
130
  },
131
  "efficiency": {
132
  "unit": "tokens/kWh",
133
+ "value": 856011.7328921164
134
  }
135
  },
136
  "per_token": {
 
138
  "latency": {
139
  "unit": "s",
140
  "values": [
141
+ 0.027695480000033967,
142
+ 0.029309737999994923,
143
+ 0.028077191000022594,
144
+ 0.027402214000005642,
145
+ 0.027616974000011396,
146
+ 0.027318316999981107,
147
+ 0.031161537999992106,
148
+ 0.028007200999979887,
149
+ 0.0283186909999813,
150
+ 0.027850898999986384,
151
+ 0.02726062900001125,
152
+ 0.03134149400000297,
153
+ 0.028211983000005603,
154
+ 0.02756093900001133
155
  ],
156
  "count": 14,
157
+ "total": 0.39713328800002046,
158
+ "mean": 0.02836666342857289,
159
+ "p50": 0.027929049999983135,
160
+ "p90": 0.030605997999992952,
161
+ "p95": 0.031224522599995908,
162
+ "p99": 0.03131809972000155,
163
+ "stdev": 0.0012808980300018803,
164
+ "stdev_": 4.515504734024059
165
  },
166
  "throughput": {
167
  "unit": "tokens/s",
168
+ "value": 35.252647972434076
169
  },
170
  "energy": null,
171
  "efficiency": null