Hugging Face's logo

Datasets:
optimum-benchmark
/
cuda

Dataset card Data Studio Files Community
IlyasMoutawwakil HF staff commited on
Commit
ce86da9
·
verified ·
1 Parent(s): 7359135

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

Browse files
Files changed (1) hide show
  1. cuda_training_transformers_text-generation_openai-community/gpt2/benchmark_report.json +43 -43
cuda_training_transformers_text-generation_openai-community/gpt2/benchmark_report.json CHANGED
@@ -2,33 +2,33 @@
2
  "overall": {
3
  "memory": {
4
  "unit": "MB",
5
- "max_ram": 1123.81952,
6
- "max_global_vram": 3406.299136,
7
  "max_process_vram": 0.0,
8
- "max_reserved": 2759.852032,
9
  "max_allocated": 2523.776
10
  },
11
  "latency": {
12
  "unit": "s",
13
  "count": 5,
14
- "total": 1.1130573043823244,
15
- "mean": 0.22261146087646488,
16
- "stdev": 0.2966186784115287,
17
- "p50": 0.07342182159423828,
18
- "p90": 0.5207038085937501,
19
- "p95": 0.6682699951171873,
20
- "p99": 0.7863229443359374,
21
  "values": [
22
- 0.815836181640625,
23
- 0.0780052490234375,
24
- 0.07342182159423828,
25
- 0.07328665924072265,
26
- 0.07250739288330078
27
  ]
28
  },
29
  "throughput": {
30
  "unit": "samples/s",
31
- "value": 44.92131699162318
32
  },
33
  "energy": null,
34
  "efficiency": null
@@ -36,30 +36,30 @@
36
  "warmup": {
37
  "memory": {
38
  "unit": "MB",
39
- "max_ram": 1123.81952,
40
- "max_global_vram": 3406.299136,
41
  "max_process_vram": 0.0,
42
- "max_reserved": 2759.852032,
43
  "max_allocated": 2523.776
44
  },
45
  "latency": {
46
  "unit": "s",
47
  "count": 2,
48
- "total": 0.8938414306640625,
49
- "mean": 0.44692071533203126,
50
- "stdev": 0.36891546630859373,
51
- "p50": 0.44692071533203126,
52
- "p90": 0.7420530883789063,
53
- "p95": 0.7789446350097656,
54
- "p99": 0.8084578723144531,
55
  "values": [
56
- 0.815836181640625,
57
- 0.0780052490234375
58
  ]
59
  },
60
  "throughput": {
61
  "unit": "samples/s",
62
- "value": 8.950133352015863
63
  },
64
  "energy": null,
65
  "efficiency": null
@@ -67,31 +67,31 @@
67
  "train": {
68
  "memory": {
69
  "unit": "MB",
70
- "max_ram": 1123.81952,
71
- "max_global_vram": 3406.299136,
72
  "max_process_vram": 0.0,
73
- "max_reserved": 2759.852032,
74
  "max_allocated": 2523.776
75
  },
76
  "latency": {
77
  "unit": "s",
78
  "count": 3,
79
- "total": 0.21921587371826173,
80
- "mean": 0.07307195790608724,
81
- "stdev": 0.0004030032788678646,
82
- "p50": 0.07328665924072265,
83
- "p90": 0.07339478912353516,
84
- "p95": 0.07340830535888672,
85
- "p99": 0.07341911834716797,
86
  "values": [
87
- 0.07342182159423828,
88
- 0.07328665924072265,
89
- 0.07250739288330078
90
  ]
91
  },
92
  "throughput": {
93
  "unit": "samples/s",
94
- "value": 82.11084213332911
95
  },
96
  "energy": null,
97
  "efficiency": null
 
2
  "overall": {
3
  "memory": {
4
  "unit": "MB",
5
+ "max_ram": 1716.031488,
6
+ "max_global_vram": 3559.391232,
7
  "max_process_vram": 0.0,
8
+ "max_reserved": 2910.846976,
9
  "max_allocated": 2523.776
10
  },
11
  "latency": {
12
  "unit": "s",
13
  "count": 5,
14
+ "total": 0.856558563232422,
15
+ "mean": 0.1713117126464844,
16
+ "stdev": 0.25528712586502056,
17
+ "p50": 0.04334796905517578,
18
+ "p90": 0.42704709930419926,
19
+ "p95": 0.5544658714294433,
20
+ "p99": 0.6564008891296387,
21
  "values": [
22
+ 0.6818846435546875,
23
+ 0.0447907829284668,
24
+ 0.04330086517333984,
25
+ 0.04334796905517578,
26
+ 0.04323430252075195
27
  ]
28
  },
29
  "throughput": {
30
  "unit": "samples/s",
31
+ "value": 58.37312490498423
32
  },
33
  "energy": null,
34
  "efficiency": null
 
36
  "warmup": {
37
  "memory": {
38
  "unit": "MB",
39
+ "max_ram": 1716.031488,
40
+ "max_global_vram": 3559.391232,
41
  "max_process_vram": 0.0,
42
+ "max_reserved": 2910.846976,
43
  "max_allocated": 2523.776
44
  },
45
  "latency": {
46
  "unit": "s",
47
  "count": 2,
48
+ "total": 0.7266754264831543,
49
+ "mean": 0.36333771324157715,
50
+ "stdev": 0.3185469303131104,
51
+ "p50": 0.36333771324157715,
52
+ "p90": 0.6181752574920654,
53
+ "p95": 0.6500299505233764,
54
+ "p99": 0.6755137049484252,
55
  "values": [
56
+ 0.6818846435546875,
57
+ 0.0447907829284668
58
  ]
59
  },
60
  "throughput": {
61
  "unit": "samples/s",
62
+ "value": 11.009041600205336
63
  },
64
  "energy": null,
65
  "efficiency": null
 
67
  "train": {
68
  "memory": {
69
  "unit": "MB",
70
+ "max_ram": 1716.031488,
71
+ "max_global_vram": 3559.391232,
72
  "max_process_vram": 0.0,
73
+ "max_reserved": 2910.846976,
74
  "max_allocated": 2523.776
75
  },
76
  "latency": {
77
  "unit": "s",
78
  "count": 3,
79
+ "total": 0.12988313674926757,
80
+ "mean": 0.04329437891642252,
81
+ "stdev": 4.663027563417732e-05,
82
+ "p50": 0.04330086517333984,
83
+ "p90": 0.04333854827880859,
84
+ "p95": 0.043343258666992186,
85
+ "p99": 0.04334702697753906,
86
  "values": [
87
+ 0.04330086517333984,
88
+ 0.04334796905517578,
89
+ 0.04323430252075195
90
  ]
91
  },
92
  "throughput": {
93
  "unit": "samples/s",
94
+ "value": 138.586120188551
95
  },
96
  "energy": null,
97
  "efficiency": null