efficientnet

efficientnet

Input --> Output

Convert

pt --> torchscript--> pnnx --> ncnn

import torch
import timm # 优秀的预训练模型库
import os

model = timm.create_model('efficientnet_b0', pretrained=True)
model.eval()

x = torch.rand(1, 3, 224, 224)
traced_script_module = torch.jit.trace(model, x, strict=False)
traced_script_module.save("ts.pt")

# 2. ts --> pnnx --> ncnn
os.system("pnnx ts.pt inputshape=[1,3,224,224]")

Example project

Reference

• qubvel/efficientnet
• rwightman/pytorch-image-models