Skip to content
Neural network decoder for large distance toric code
Python
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
all_blocks_009_L16
block_vp
.gitattributes
.gitignore
LICENSE
README.md
bp.pyx
bp_compile.py
gen_bp_data.py
tf_rn_log.py
tools.py
var_error_rate_000_016_L16.npy

README.md

toric-code-neural-decoder

The corresponding preprint: https://arxiv.org/abs/1809.06640. A very brief description of toric code can be found here. It's mainly for people without a background of quantum information.

A script for 2 basic examples can be running on https://colab.research.google.com/drive/1l3q35FFoxzPRNs_MMxD1FQJean5OwLsA

  1. Evaluating a trained model for d=16 toric code
  2. Training of the dense layers and global training for d=64 toric code (you can choose to use GPU on Google Colab)

Checkpoint for d=64 with strict training policy: https://drive.google.com/open?id=1AnrDSycv-SnyumJ--xsv7BplMjCo8zMY Checkpoint for d=64 with more training: https://drive.google.com/open?id=1zCaKb7oxe0oHw1HcJjP0q_OShE_ZD44r Can be loaded similarly to the Google Colab script.

This work is distributed under the GNU GPLv3. See LICENSE.txt. (c) 2018 Xiaotong Ni

You can’t perform that action at this time.