pnnx转onnx模型会有一个open failed

[kazyun]

error log | 日志或报错信息 | ログ

pnnxparam = ../scene_resnet18_0905.pnnx.param
pnnxbin = ../scene_resnet18_0905.pnnx.bin
pnnxpy = ../scene_resnet18_0905_pnnx.py
pnnxonnx = ../scene_resnet18_0905.pnnx.onnx
ncnnparam = ../scene_resnet18_0905.ncnn.param
ncnnbin = ../scene_resnet18_0905.ncnn.bin
ncnnpy = ../scene_resnet18_0905_ncnn.py
fp16 = 1
optlevel = 2
device = cpu
inputshape = [1,3,180,180]f32
inputshape2 =
customop =
moduleop =
############# pass_level0 onnx
inline_containers ... 0.00ms
eliminate_noop ... 0.06ms
fold_constants ... 0.04ms
canonicalize ... 0.09ms
shape_inference ... 194.11ms
fold_constants_dynamic_shape ... 0.07ms
inline_if_graph ... 0.01ms
fuse_constant_as_attribute ... 0.07ms
eliminate_noop_with_shape ... 0.03ms
┌───────────────────┬──────────┬──────────┐
│ │ orig │ opt │
├───────────────────┼──────────┼──────────┤
│ node │ 52 │ 52 │
│ initializer │ 49 │ 48 │
│ functions │ 0 │ 0 │
├───────────────────┼──────────┼──────────┤
│ nn module op │ 0 │ 0 │
│ custom module op │ 0 │ 0 │
│ aten op │ 0 │ 0 │
│ prims op │ 0 │ 0 │
│ onnx native op │ 52 │ 52 │
├───────────────────┼──────────┼──────────┤
│ Add │ 8 │ 8 │
│ Conv │ 20 │ 20 │
│ Gemm │ 4 │ 4 │
│ GlobalAveragePool │ 1 │ 1 │
│ MaxPool │ 1 │ 1 │
│ Relu │ 17 │ 17 │
│ Reshape │ 1 │ 1 │
└───────────────────┴──────────┴──────────┘
############# pass_level1 onnx
############# pass_level2
############# pass_level3
open failed
############# pass_level4
############# pass_level5
############# pass_ncnn

model | 模型 | モデル

1. 基于resnet18.onnx

how to reproduce | 复现步骤 | 再現方法

这个open failed对于生成的ncnn.param, ncnn.bin使用是否会有影响呢

[nihui]

无影响