Skip to content

Scripts of neural architectures implemented in PyTorch

Notifications You must be signed in to change notification settings

amitadate/pytorch-nn-scripts

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

34 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

pytorch-nn-scripts ( work in progress )

Scripts of neural architectures implemented in PyTorch.

Intended to complement ongoing research, plug and play models with argument parsing. Each script runs with logging and saving progress along with generating graphs in one runtime.

The pipeline used for the experiments is the following.

.
├── data        # data pulled by the script 
├── models      # ckpt files stored ( model state dictonary and optimizer state dictionary, to save progress)
├── logs        # logs saved after each runtime, based on input argument for saving / not saving
├── results     # folder for saving graphs based on input argument, (testing, training loss vs epochs )
└── model.py    # script file from this repo, use model --help to get details about input arguments

Example for fully-connected-network on mnist:

Work in progress and models will be added catering to the pipeline above.

About

Scripts of neural architectures implemented in PyTorch

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published