fix .gitignore, Makefile and README.md for 2011/dlowe
Unpack the `dlowe-aux-data.tar.bz2` compressed tarball but keep the file as it is used in a try example. Make `earlystop.pl` executable.
This commit is contained in:
Landon Curt Noll committed
1 parent
f3c988e374
commit
8197f10bae
50 files changed
+2409
-94
No files matched your search
+74
-46
@@ -1,60 +1,72 @@
|
||||
# Most self deprecating
|
||||
|
||||
David Lowe
|
||||
434 Tenney Drive
|
||||
Rogue River, OR
|
||||
97537
|
||||
<j.david.lowe@gmail.com>
|
||||
<http://www.pootpoot.net/>
|
||||
David Lowe
|
||||
434 Tenney Drive
|
||||
Rogue River, OR
|
||||
97537
|
||||
<j.david.lowe@gmail.com>
|
||||
<http://www.pootpoot.net/>
|
||||
|
||||
## To build:
|
||||
|
||||
## Judges' comments:
|
||||
### To build:
|
||||
|
||||
make dlowe
|
||||
```sh
|
||||
make
|
||||
```
|
||||
|
||||
### To run:
|
||||
|
||||
./dlowe -<n_iterations> corpus1/ [...] corpus0/ < start.net > end.net
|
||||
```sh
|
||||
./dlowe -<n_iterations> corpus1/ [...] corpus0/ < start.net > trained.net
|
||||
```
|
||||
|
||||
NOTE: The directory entries MUST end in a /
|
||||
NOTE: In the above command, the directory args MUST end in a "/".
|
||||
|
||||
./dlowe file [file ...] < trained.net
|
||||
Then to use the `trained.net`:
|
||||
|
||||
```sh
|
||||
./dlowe file [file ...] < trained.net
|
||||
```
|
||||
|
||||
### Try:
|
||||
|
||||
Download the dlowe-aux-data.tar.bz2 file from:
|
||||
Try these pre-trained networks:
|
||||
|
||||
curl -O http://www.ioccc.org/2011/dlowe/dlowe-aux-data.tar.bz2
|
||||
```sh
|
||||
./dlowe dlowe.c < dlowe-aux-data/ioccc-winlose-trained.net
|
||||
./dlowe dlowe < dlowe-aux-data/ioccc-winlose-trained.net
|
||||
|
||||
tar -jxf dlowe-aux-data.tar.bz2
|
||||
./dlowe xor-0/00 < dlowe-aux-data/xor-trained.net
|
||||
./dlowe xor-0/11 < dlowe-aux-data/xor-trained.net
|
||||
./dlowe xor-1/01 < dlowe-aux-data/xor-trained.net
|
||||
./dlowe xor-1/10 < dlowe-aux-data/xor-trained.net
|
||||
|
||||
./dlowe dlowe.c < ioccc-winlose-trained.net
|
||||
./dlowe dlowe < ioccc-winlose-trained.net
|
||||
./dlowe hint.html < dlowe-aux-data/english-trained.net
|
||||
|
||||
./dlowe xor-0/00 < xor-trained.net
|
||||
./dlowe xor-0/11 < xor-trained.net
|
||||
./dlowe xor-1/01 < xor-trained.net
|
||||
./dlowe xor-1/10 < xor-trained.net
|
||||
./dlowe dlowe-aux-data.tar.bz2 < dlowe-aux-data/png-trained.net
|
||||
```
|
||||
|
||||
./dlowe hint.html < english-trained.net
|
||||
NOTE: The `dlowe-aux-data.tar.bz2/` directory was created by:
|
||||
|
||||
./dlowe dlowe-aux-data.tar.bz2 < png-trained.net
|
||||
```sh
|
||||
tar -jxf dlowe-aux-data.tar.bz2
|
||||
```
|
||||
|
||||
### Selected Judges Remarks:
|
||||
## Judges' comments:
|
||||
|
||||
This entry is a ghoulish example of a brain (dead?) neural network classifier.
|
||||
It comes with a set of corpora for you to try.
|
||||
|
||||
#### ioccc-winlose-trained.net
|
||||
### ioccc-winlose-trained.net
|
||||
|
||||
You can create your own trained networks. For example, we trained a network
|
||||
on the C code of entries that were supplied to the 20Th IOCCC.
|
||||
|
||||
The ioccc-winlose-trained.net was trained using the earlystop.pl tool:
|
||||
|
||||
rm -f ioccc-winlose-trained.net
|
||||
./earlystop.pl ioccc-winlose-trained.net ioccc_won_training/ ioccc_lost_training/ ioccc_won_test/ ioccc_lost_test/
|
||||
```sh
|
||||
rm -f ioccc-winlose-trained.net
|
||||
./earlystop.pl ioccc-winlose-trained.net ioccc_won_training/ ioccc_lost_training/ ioccc_won_test/ ioccc_lost_test/
|
||||
```
|
||||
|
||||
where:
|
||||
|
||||
@@ -86,7 +98,9 @@ dlowe-aux-data.tar.bz2 file in the above mentioned URL.
|
||||
|
||||
The dlowe.c was explicitly excluded from this trailing set. So this test is interesting:
|
||||
|
||||
./dlowe dlowe.c < ioccc-winlose-trained.net
|
||||
```sh
|
||||
./dlowe dlowe.c < ioccc-winlose-trained.net
|
||||
```
|
||||
|
||||
The match of dlowe.c to the lose/win trailed network was:
|
||||
|
||||
@@ -104,7 +118,9 @@ data sets on which to test:
|
||||
|
||||
We created the english-trained.net as follows:
|
||||
|
||||
./dlowe -8000 english-1/ english-0/ < /dev/null > english-trained.net
|
||||
```sh
|
||||
./dlowe -8000 english-1/ english-0/ < /dev/null > english-trained.net
|
||||
```
|
||||
|
||||
english-0/
|
||||
Non-English (French) text
|
||||
@@ -113,7 +129,9 @@ We created the english-trained.net as follows:
|
||||
|
||||
We created the png-trained.net as follows:
|
||||
|
||||
./dlowe -8000 png-1/ png-0/ < /dev/null > png-trained.net
|
||||
```sh
|
||||
./dlowe -8000 png-1/ png-0/ < /dev/null > png-trained.net
|
||||
```
|
||||
|
||||
png-0/
|
||||
Non-png (gif) images
|
||||
@@ -122,7 +140,9 @@ We created the png-trained.net as follows:
|
||||
|
||||
We created the xor-trained.net as follows:
|
||||
|
||||
./dlowe -8000 xor-1/ xor-0/ < /dev/null > xor-trained.net
|
||||
```sh
|
||||
./dlowe -8000 xor-1/ xor-0/ < /dev/null > xor-trained.net
|
||||
```
|
||||
|
||||
xor-0/
|
||||
Data that XORs to 0
|
||||
@@ -130,11 +150,14 @@ We created the xor-trained.net as follows:
|
||||
Data that XORs to 1
|
||||
|
||||
## Author's comments:
|
||||
# Synopsis:
|
||||
|
||||
### Synopsis:
|
||||
|
||||
This is an artificially intelligent judging tool to help the IOCCC judges.
|
||||
Here's to shorter, more frequent contests!
|
||||
|
||||
# Description:
|
||||
### Description:
|
||||
|
||||
This is a multilayer perceptron (a feedforward artificial neural network)
|
||||
which can be trained, using on-line backpropagation, to classify input files.
|
||||
|
||||
@@ -143,9 +166,11 @@ neuron.
|
||||
|
||||
The neurons' activation function is the logistic function 1 / (1 + e ^ -x).
|
||||
|
||||
## Classifying
|
||||
### Classifying
|
||||
|
||||
./dlowe file [file ...] < trained.net
|
||||
```sh
|
||||
./dlowe file [file ...] < trained.net
|
||||
```
|
||||
|
||||
To classify files, one specifies a trained network (on stdin) and one or more
|
||||
files to classify. The program will output one line per successfully-classified
|
||||
@@ -156,9 +181,12 @@ The interpretation of the classification number depends on how the network was
|
||||
trained, but it's geared toward interpretation as a *probability* or a
|
||||
*confidence*.
|
||||
|
||||
## Training
|
||||
### Training
|
||||
|
||||
./dlowe -<n_iterations> corpus1/ [...] corpus0/ < start.net > end.net
|
||||
|
||||
```sh
|
||||
./dlowe -<n_iterations> corpus1/ [...] corpus0/ < start.net > end.net
|
||||
```
|
||||
|
||||
NOTE: The directory args must end in a / for them to work.
|
||||
|
||||
@@ -176,7 +204,7 @@ intermediate target values.
|
||||
|
||||
The learning rate is hard-coded as 0.3. No momentum factor is used.
|
||||
|
||||
### png corpora
|
||||
#### png corpora
|
||||
|
||||
The png-1 corpus was obtained by manually scraping the first results from an
|
||||
https://images.google.com search for "obfuscate filetype:png".
|
||||
@@ -194,7 +222,7 @@ Results:
|
||||
a .gif"
|
||||
* produced a network with about 87% accuracy
|
||||
|
||||
### english corpora
|
||||
#### english corpora
|
||||
|
||||
The english-1 corpus was obtained by manually scraping the first results from
|
||||
a https://www.google.com/webhp?lr=lang_en search for "paris filetype:txt".
|
||||
@@ -213,7 +241,7 @@ Results:
|
||||
"probably french"
|
||||
* produced a network with 100% accuracy
|
||||
|
||||
### xor corpora
|
||||
#### xor corpora
|
||||
|
||||
The xor-1 corpus consists of two files containing '01' and '10' respectively.
|
||||
|
||||
@@ -224,7 +252,8 @@ Results:
|
||||
* using these corpora as a training set
|
||||
* takes about 8000 iterations to learn xor to within a tolerance of <0.01
|
||||
|
||||
# Limitations
|
||||
### Limitations
|
||||
|
||||
The program can't tell you anything meaningful about files with less than two
|
||||
bytes in them (I'm looking at you, smr.c!)
|
||||
|
||||
@@ -255,7 +284,7 @@ library (untested, but they do exist).
|
||||
Backpropagation doesn't always converge: if you play with this long enough,
|
||||
you'll eventually have a training session that completely fails to converge.
|
||||
|
||||
# Obfuscation:
|
||||
### Obfuscation:
|
||||
|
||||
Zombies! (Since neural networks are modeled after BRAINS, ya know? And
|
||||
corpus sounds a lot like corpse. And I have 4- and 7-year-old kids ;)
|
||||
@@ -277,11 +306,10 @@ by a network trained to recognize ioccc winners ;)
|
||||
|
||||
... but mostly zombies!
|
||||
|
||||
--------------------------------------------------------------------------------
|
||||
<!--
|
||||
## Copyright:
|
||||
|
||||
(c) Copyright 1984-2012, [Leo Broukhis, Simon Cooper, Landon Curt Noll][judges] - All rights reserved
|
||||
This work is licensed under a [Creative Commons Attribution-ShareAlike 3.0 Unported License][cc].
|
||||
|
||||
[judges]: http://www.ioccc.org/judges.html
|
||||
[cc]: http://creativecommons.org/licenses/by-sa/3.0/
|
||||
-->
|
||||
Reference in new issue
Block a user