模型finetuning

左手的ㄟ右手 2022-06-04 06:25 262阅读 0赞

## pytorch ##

http://pytorch.org/tutorials/beginner/transfer\_learning\_tutorial.html

http://pytorch-cn.readthedocs.io/zh/latest/torchvision/torchvision-models/

https://github.com/Cadene/pretrained-models.pytorch

https://github.com/Spandan-Madan/Pytorch\_fine\_tuning\_Tutorial

https://github.com/ry85/finetuning---PyTorch/blob/master/main.ipynb

https://github.com/gmayday1997/pytorch\_CAM

https://discuss.pytorch.org/t/how-to-perform-finetuning-in-pytorch/419

https://github.com/ivanlai/Kaggle-Planet-Amazon

https://github.com/bearpaw/pytorch-classification

https://github.com/maxgreat/Instance-Search/blob/master/TrainClassif.py

pytorch model 转tensorflow

https://github.com/leonidk/pytorch-tf

caffe

[caffe的一些经典网络的实现][caffe]

https://github.com/cvjena/cnn-models

https://github.com/tflearn/tflearn/blob/master/examples/basics/finetuning.py

https://github.com/flyyufelix/cnn\_finetune

https://flyyufelix.github.io/2016/10/08/fine-tuning-in-keras-part2.html

inception v3

[tensorflow pretrained model][]

https://github.com/tflearn/tflearn/blob/master/examples/images/googlenet.py

https://github.com/conan7882/GoogLeNet-Inception-tensorflow

vgg16

https://github.com/EKami/planet-amazon-deforestation/blob/master/src/vgg16.py

https://github.com/tflearn/models

https://github.com/tflearn/tflearn/blob/master/examples/images/vgg\_network\_finetuning.py

https://github.com/conan7882/VGG-tensorflow

[TensorFlow VGG-16 预训练模型][TensorFlow VGG-16]

https://github.com/ry/tensorflow-vgg16

resnet

https://github.com/ry/tensorflow-resnet

# Tensorflow使用的预训练的resnet\_v2\_50，resnet\_v2\_101，resnet\_v2\_152等模型预测，训练 #

# 学习TensorFlow，调用预训练好的网络（Alex, VGG, ResNet etc） #

[caffe Resnet-50 finetune 所有代码+需要注意的地方][caffe Resnet-50 finetune]

[利用caffe pre-trained model进行图像分类][caffe pre-trained model]

[Tensorflow学习笔记--使用迁移学习做自己的图像分类器（Inception v3）][Tensorflow_--_Inception v3]

[github上tensorflow预训练模型下载链接][github_tensorflow]

## tensorflow Image\_retrain ##

[MobileNet训练分类网][MobileNet]

[TensorFlow学习笔记之源码分析（3）---- retrain.py][TensorFlow_3_---- retrain.py]

[基于Inception v3进行单标签训练][Inception v3]

tensorflow/tensorflow/example/image\_retraining/retrain.py目录下执行

python retrain.py --image\_dir /home/hj/body\_part\_classify/data/train/ --intermediate\_store\_frequency 1000 --how\_many\_training\_steps 50000 --learning\_rate 0.001 --train\_batch\_size 8 --testing\_percentage 20 --eval\_step\_interval 20 --architecture mobilenet\_1.0\_224 --output\_graph  /home/hj/tf-study/fine-tuning/output\_graph.pb --intermediate\_output\_graphs\_dir /home/hj/tf-study/fine-tuning/intermediate\_graph --output\_labels /home/hj/tf-study/fine-tuning/output\_labels.txt --model\_dir  /home/hj/tf-study/fine-tuning/imagenet --val\_batch\_size -1

## 利用tf-slim ##

[使用TensorFlow-Slim进行图像分类][TensorFlow-Slim]

## keras ##

[keras入门 ---在预训练好网络模型上进行fine-tune][keras_ ---_fine-tune]

[keras系列︱图像多分类训练与利用bottleneck features进行微调（三）][keras_bottleneck features]

https://github.com/flyyufelix/cnn\_finetune

各种模型转换

https://github.com/Microsoft/MMdnn

https://github.com/leonidk/pytorch-tf

[caffe]: http://blog.csdn.net/hjxu2016/article/details/78788062
[tensorflow pretrained model]: http://blog.csdn.net/swfa1/article/details/72636609
[TensorFlow VGG-16]: http://blog.csdn.net/daydayup_668819/article/details/70225244
[caffe Resnet-50 finetune]: http://blog.csdn.net/tangwenbo124/article/details/56070322
[caffe pre-trained model]: http://blog.csdn.net/tina_ttl/article/details/51033646
[Tensorflow_--_Inception v3]: http://blog.csdn.net/l18930738887/article/details/72812689
[github_tensorflow]: https://github.com/tensorflow/models/tree/master/research/slim
[MobileNet]: http://blog.csdn.net/burning_keyboard/article/details/77605438
[TensorFlow_3_---- retrain.py]: http://blog.csdn.net/intjun/article/details/78204222
[Inception v3]: http://blog.csdn.net/numeria/article/details/73658378
[TensorFlow-Slim]: http://blog.csdn.net/lijiancheng0614/article/details/77727445
[keras_ ---_fine-tune]: http://blog.csdn.net/hnu2012/article/details/72179437
[keras_bottleneck features]: http://blog.csdn.net/sinat_26917383/article/details/72861152