1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
|
import operator_benchmark as op_bench
import torch
"""Microbenchmarks for quantized groupnorm operator."""
groupnorm_configs_short = op_bench.cross_product_configs(
dims=(
(32, 8, 16),
(32, 8, 56, 56),
),
num_groups=(2, 4),
dtype=(torch.qint8,),
tags=["short"],
)
class QGroupNormBenchmark(op_bench.TorchBenchmarkBase):
def init(self, dims, num_groups, dtype):
X = (torch.rand(*dims) - 0.5) * 256
num_channels = dims[1]
scale = 1.0
zero_point = 0
self.inputs = {
"qX": torch.quantize_per_tensor(
X, scale=scale, zero_point=zero_point, dtype=dtype
),
"num_groups": num_groups,
"weight": torch.rand(num_channels, dtype=torch.float),
"bias": torch.rand(num_channels, dtype=torch.float),
"eps": 1e-5,
"Y_scale": 0.1,
"Y_zero_point": 0,
}
def forward(
self,
qX,
num_groups: int,
weight,
bias,
eps: float,
Y_scale: float,
Y_zero_point: int,
):
return torch.ops.quantized.group_norm(
qX,
num_groups,
weight=weight,
bias=bias,
eps=eps,
output_scale=Y_scale,
output_zero_point=Y_zero_point,
)
op_bench.generate_pt_test(groupnorm_configs_short, QGroupNormBenchmark)
if __name__ == "__main__":
op_bench.benchmark_runner.main()
|
