Skip to content

Implementation of Squeezenet in pytorch, pretrained models on Cifar 10 data to come

Notifications You must be signed in to change notification settings

gsp-27/pytorch_Squeezenet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Pytorch Squeeznet

Pytorch implementation of Squeezenet model as described in https://arxiv.org/abs/1602.07360 on cifar-10 Data.

The definition of Squeezenet model is present model.py. The training procedure resides in the file main.py

Command to train the Squeezenet model on CIFAR 10 data is:

python main.py --batch-size 32 --epoch 10

Other options which can be used are specified in main.py Eg: if you want to use a pretrained_model

python main.py --batch-size 32 --epoch 10 --model_name "pretrained model"

I am currently using SGD for training : learning rate and weight decay are currently updated using a 55 epoch learning rule, this usually gives good performance, but if you want to use something of your own, you can specify it by passing learning_rate and weight_decay parameter like so

python main.py --batch-size 32 --epoch 10 --learning_rate 1e-3 --epoch_55

About

Implementation of Squeezenet in pytorch, pretrained models on Cifar 10 data to come

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages