Hugging Face's logo

Datasets:
optimum-benchmark
/
rocm

Dataset card Viewer Files Community
IlyasMoutawwakil HF staff commited on
Commit
e81e128
·
verified ·
1 Parent(s): 516bfa0

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
@@ -2,34 +2,34 @@
2
  "overall": {
3
  "memory": {
4
  "unit": "MB",
5
- "max_ram": 1350.877184,
6
  "max_global_vram": 68702.69952,
7
- "max_process_vram": 348397.09696,
8
  "max_reserved": 2894.06976,
9
  "max_allocated": 2506.73664
10
  },
11
  "latency": {
12
  "unit": "s",
13
  "values": [
14
- 0.6326380615234375,
15
- 0.04232443618774414,
16
- 0.04159115982055664,
17
- 0.041658199310302735,
18
- 0.04158892059326172
19
  ],
20
  "count": 5,
21
- "total": 0.7998007774353028,
22
- "mean": 0.15996015548706055,
23
- "p50": 0.041658199310302735,
24
- "p90": 0.3965126113891602,
25
- "p95": 0.5145753364562987,
26
- "p99": 0.6090255165100097,
27
- "stdev": 0.23633911505742286,
28
- "stdev_": 147.74874051465943
29
  },
30
  "throughput": {
31
  "unit": "samples/s",
32
- "value": 62.51556813977289
33
  },
34
  "energy": null,
35
  "efficiency": null
@@ -37,31 +37,31 @@
37
  "warmup": {
38
  "memory": {
39
  "unit": "MB",
40
- "max_ram": 1350.877184,
41
  "max_global_vram": 68702.69952,
42
- "max_process_vram": 348397.09696,
43
  "max_reserved": 2894.06976,
44
  "max_allocated": 2506.73664
45
  },
46
  "latency": {
47
  "unit": "s",
48
  "values": [
49
- 0.6326380615234375,
50
- 0.04232443618774414
51
  ],
52
  "count": 2,
53
- "total": 0.6749624977111817,
54
- "mean": 0.33748124885559083,
55
- "p50": 0.33748124885559083,
56
- "p90": 0.5736066989898682,
57
- "p95": 0.6031223802566528,
58
- "p99": 0.6267349252700806,
59
- "stdev": 0.29515681266784666,
60
- "stdev_": 87.45872953496895
61
  },
62
  "throughput": {
63
  "unit": "samples/s",
64
- "value": 11.852510364839889
65
  },
66
  "energy": null,
67
  "efficiency": null
@@ -69,32 +69,32 @@
69
  "train": {
70
  "memory": {
71
  "unit": "MB",
72
- "max_ram": 1350.877184,
73
  "max_global_vram": 68702.69952,
74
- "max_process_vram": 348397.09696,
75
  "max_reserved": 2894.06976,
76
  "max_allocated": 2506.73664
77
  },
78
  "latency": {
79
  "unit": "s",
80
  "values": [
81
- 0.04159115982055664,
82
- 0.041658199310302735,
83
- 0.04158892059326172
84
  ],
85
  "count": 3,
86
- "total": 0.12483827972412109,
87
- "mean": 0.041612759908040364,
88
- "p50": 0.04159115982055664,
89
- "p90": 0.04164479141235351,
90
- "p95": 0.041651495361328124,
91
- "p99": 0.041656858520507815,
92
- "stdev": 3.214351145711704e-05,
93
- "stdev_": 0.07724436333507001
94
  },
95
  "throughput": {
96
  "unit": "samples/s",
97
- "value": 144.18654310022555
98
  },
99
  "energy": null,
100
  "efficiency": null
 
2
  "overall": {
3
  "memory": {
4
  "unit": "MB",
5
+ "max_ram": 1311.440896,
6
  "max_global_vram": 68702.69952,
7
+ "max_process_vram": 329450.471424,
8
  "max_reserved": 2894.06976,
9
  "max_allocated": 2506.73664
10
  },
11
  "latency": {
12
  "unit": "s",
13
  "values": [
14
+ 0.6158572387695312,
15
+ 0.04217601776123047,
16
+ 0.041766895294189456,
17
+ 0.04174817657470703,
18
+ 0.04130561447143555
19
  ],
20
  "count": 5,
21
+ "total": 0.7828539428710938,
22
+ "mean": 0.15657078857421874,
23
+ "p50": 0.041766895294189456,
24
+ "p90": 0.386384750366211,
25
+ "p95": 0.501120994567871,
26
+ "p99": 0.5929099899291992,
27
+ "stdev": 0.2296433902475554,
28
+ "stdev_": 146.67064804281694
29
  },
30
  "throughput": {
31
  "unit": "samples/s",
32
+ "value": 63.86887420739873
33
  },
34
  "energy": null,
35
  "efficiency": null
 
37
  "warmup": {
38
  "memory": {
39
  "unit": "MB",
40
+ "max_ram": 1311.440896,
41
  "max_global_vram": 68702.69952,
42
+ "max_process_vram": 329450.471424,
43
  "max_reserved": 2894.06976,
44
  "max_allocated": 2506.73664
45
  },
46
  "latency": {
47
  "unit": "s",
48
  "values": [
49
+ 0.6158572387695312,
50
+ 0.04217601776123047
51
  ],
52
  "count": 2,
53
+ "total": 0.6580332565307617,
54
+ "mean": 0.32901662826538086,
55
+ "p50": 0.32901662826538086,
56
+ "p90": 0.5584891166687012,
57
+ "p95": 0.5871731777191161,
58
+ "p99": 0.6101204265594482,
59
+ "stdev": 0.2868406105041504,
60
+ "stdev_": 87.18118960017675
61
  },
62
  "throughput": {
63
  "unit": "samples/s",
64
+ "value": 12.157440251845411
65
  },
66
  "energy": null,
67
  "efficiency": null
 
69
  "train": {
70
  "memory": {
71
  "unit": "MB",
72
+ "max_ram": 1311.440896,
73
  "max_global_vram": 68702.69952,
74
+ "max_process_vram": 329450.471424,
75
  "max_reserved": 2894.06976,
76
  "max_allocated": 2506.73664
77
  },
78
  "latency": {
79
  "unit": "s",
80
  "values": [
81
+ 0.041766895294189456,
82
+ 0.04174817657470703,
83
+ 0.04130561447143555
84
  ],
85
  "count": 3,
86
+ "total": 0.12482068634033203,
87
+ "mean": 0.04160689544677734,
88
+ "p50": 0.04174817657470703,
89
+ "p90": 0.04176315155029297,
90
+ "p95": 0.04176502342224121,
91
+ "p99": 0.04176652091979981,
92
+ "stdev": 0.0002131748377771068,
93
+ "stdev_": 0.5123545880749395
94
  },
95
  "throughput": {
96
  "unit": "samples/s",
97
+ "value": 144.20686608725885
98
  },
99
  "energy": null,
100
  "efficiency": null