Upload cuda_training_transformers_multiple-choice_FacebookAI/roberta-base/benchmark_report.json with huggingface_hub
Browse files
cuda_training_transformers_multiple-choice_FacebookAI/roberta-base/benchmark_report.json
ADDED
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"overall": {
|
3 |
+
"memory": {
|
4 |
+
"unit": "MB",
|
5 |
+
"max_ram": 1124.806656,
|
6 |
+
"max_global_vram": 3149.012992,
|
7 |
+
"max_process_vram": 314842.95168,
|
8 |
+
"max_reserved": 2707.423232,
|
9 |
+
"max_allocated": 2497.88416
|
10 |
+
},
|
11 |
+
"latency": {
|
12 |
+
"unit": "s",
|
13 |
+
"count": 5,
|
14 |
+
"total": 0.7518202209472656,
|
15 |
+
"mean": 0.15036404418945312,
|
16 |
+
"stdev": 0.21274044698456382,
|
17 |
+
"p50": 0.043520721435546875,
|
18 |
+
"p90": 0.36383551940917974,
|
19 |
+
"p95": 0.46983810882568355,
|
20 |
+
"p99": 0.5546401803588867,
|
21 |
+
"values": [
|
22 |
+
0.5758406982421875,
|
23 |
+
0.045827751159667966,
|
24 |
+
0.043520721435546875,
|
25 |
+
0.043278804779052736,
|
26 |
+
0.04335224533081055
|
27 |
+
]
|
28 |
+
},
|
29 |
+
"throughput": {
|
30 |
+
"unit": "samples/s",
|
31 |
+
"value": 66.50526097449448
|
32 |
+
},
|
33 |
+
"energy": null,
|
34 |
+
"efficiency": null
|
35 |
+
},
|
36 |
+
"warmup": {
|
37 |
+
"memory": {
|
38 |
+
"unit": "MB",
|
39 |
+
"max_ram": 1124.806656,
|
40 |
+
"max_global_vram": 3149.012992,
|
41 |
+
"max_process_vram": 314842.95168,
|
42 |
+
"max_reserved": 2707.423232,
|
43 |
+
"max_allocated": 2497.88416
|
44 |
+
},
|
45 |
+
"latency": {
|
46 |
+
"unit": "s",
|
47 |
+
"count": 2,
|
48 |
+
"total": 0.6216684494018555,
|
49 |
+
"mean": 0.31083422470092775,
|
50 |
+
"stdev": 0.2650064735412598,
|
51 |
+
"p50": 0.31083422470092775,
|
52 |
+
"p90": 0.5228394035339355,
|
53 |
+
"p95": 0.5493400508880615,
|
54 |
+
"p99": 0.5705405687713623,
|
55 |
+
"values": [
|
56 |
+
0.5758406982421875,
|
57 |
+
0.045827751159667966
|
58 |
+
]
|
59 |
+
},
|
60 |
+
"throughput": {
|
61 |
+
"unit": "samples/s",
|
62 |
+
"value": 12.868595804881654
|
63 |
+
},
|
64 |
+
"energy": null,
|
65 |
+
"efficiency": null
|
66 |
+
},
|
67 |
+
"train": {
|
68 |
+
"memory": {
|
69 |
+
"unit": "MB",
|
70 |
+
"max_ram": 1124.806656,
|
71 |
+
"max_global_vram": 3149.012992,
|
72 |
+
"max_process_vram": 314842.95168,
|
73 |
+
"max_reserved": 2707.423232,
|
74 |
+
"max_allocated": 2497.88416
|
75 |
+
},
|
76 |
+
"latency": {
|
77 |
+
"unit": "s",
|
78 |
+
"count": 3,
|
79 |
+
"total": 0.13015177154541016,
|
80 |
+
"mean": 0.043383923848470056,
|
81 |
+
"stdev": 0.00010127047458219125,
|
82 |
+
"p50": 0.04335224533081055,
|
83 |
+
"p90": 0.04348702621459961,
|
84 |
+
"p95": 0.043503873825073244,
|
85 |
+
"p99": 0.04351735191345215,
|
86 |
+
"values": [
|
87 |
+
0.043520721435546875,
|
88 |
+
0.043278804779052736,
|
89 |
+
0.04335224533081055
|
90 |
+
]
|
91 |
+
},
|
92 |
+
"throughput": {
|
93 |
+
"unit": "samples/s",
|
94 |
+
"value": 138.30007679703206
|
95 |
+
},
|
96 |
+
"energy": null,
|
97 |
+
"efficiency": null
|
98 |
+
}
|
99 |
+
}
|