https://github.com/progamergov/neural-graph-legacy
Graph Neural-Style's loss value outputs
https://github.com/progamergov/neural-graph-legacy
neural-style
Last synced: 4 months ago
JSON representation
Graph Neural-Style's loss value outputs
- Host: GitHub
- URL: https://github.com/progamergov/neural-graph-legacy
- Owner: ProGamerGov
- Created: 2017-01-21T20:35:08.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2017-01-23T00:30:40.000Z (over 9 years ago)
- Last Synced: 2025-10-27T19:33:37.181Z (8 months ago)
- Topics: neural-style
- Language: Lua
- Size: 27.3 KB
- Stars: 5
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Neural-Graph
Graph Neural-Style's loss value outputs. Special thanks to [htoyryla](https://github.com/htoyryla) for helping me [modify the neural_style.lua script](https://github.com/jcjohnson/neural-style/issues/370).
The Neural-Style issue: https://github.com/jcjohnson/neural-style/issues/370
Alternate download of the modified neural_style.lua script: https://gist.github.com/ProGamerGov/a8134605c89f01e5bcd88539456675b8
More graphs can be found here along with observations: https://github.com/ProGamerGov/Neural-Graph/wiki/Graphs
# Requirements
[Neural-Style](https://github.com/jcjohnson/neural-style/) must be installed.
Make sure you Neural-Style command is modified to use the modified Neural-Style script.
Matplotlib is also required:
`sudo apt-get install python-matplotlib`
# Usage
First run the modified neural_style.lua script with `2>&1 | tee ~/neural-style/loss_values.log` at the end of your Neural-Style command paramters. Example:
`th neural_style_csv.lua -style_image -content_image -print_iter 10 2>&1 | tee ~/neural-style/loss_values.log`
Or if you are using multires, then do:
`./multires.sh 2>&1 | tee ~/neural-style/loss_values.log`
Using `-print_iter 1` will create a less jagged looking graph.
After running Neural-Style, you must manually convert the iteration section(s) from the saved terminal log file to their own CSV file.
Make sure the first row of the CSV file is (If using different combinations of layers, or more than one style image, you may have to modify the CSV file headers):
Iteration | Content 1 loss | Style 1 loss | Style 2 loss | Style 3 loss | Style 4 loss | Style 5 loss | Total loss
--- | --- | --- | --- | --- | --- | --- | ---
Note that the exact header names are not needed. After copying my text file data into Excel, I used this [guide](https://web.archive.org/web/20170123002416/https://help.xero.com/Q_ConvertTXT) to convert it into a "Space" deliminated CSV file. Though you can use whatever method of text to CSV conversion that works for you.
Then first modify the graphing script's tile to match your CSV file(s), and run the graphing script:
`python graph.py`