
Script started on Wed Jan 22 09:31:44 1992
[/dev/ttyp008]
CRAY-C90-1CPU$ Learn
Performance Test...


Backpropagation Learning

Black Boxes:

xor (Saved at Iteration 0)

Loading Data Files...Done.

Forward...

Total Iterations 10000
Total Epochs 2500.000000
Total number of connections: 7

           seconds          clocks
elapsed    2.10081       504194031
user       0.26827        64384843
sys        0.04294        10304913

Learn...


Backpropagation Learning

Black Boxes:

xor (Saved at Iteration 0)

Loading Data Files...Done.

Learning...

Learning Rate: 0.200000
Inertia: 0.950000
Your learning rate is a bit high,
if this does not converge try lowering it.
Testing every 50 iterations.
Must pass 4 patterns with 0.100000 error bound.
Saving every 100 iterations.
Dumping Network at 100.
Dumping Network at 200.
Dumping Network at 300.
Dumping Network at 400.
Dumping Network at 500.
Dumping Network at 600.
Dumping Network at 700.
Dumping Network at 800.
Dumping Network at 900.
Dumping Network at 1000.
Dumping Network at 1100.
Dumping Network at 1200.
Dumping Network at 1300.
Dumping Network at 1400.
Dumping Network at 1500.
Success!! Dumping Network.Finished.


Learning Rate: 0.200000
Inertia: 0.950000

Total Iterations 1500
Total Epochs 375.000000
Total number of connections: 7

           seconds          clocks
elapsed    2.74685       659242925
user       0.11512        27629328
sys        0.10159        24381477

Test...

Loading Saved Network: Network.Finished

Backpropagation Learning

Black Boxes:

xor (Saved at Iteration 1500)

Loading Data Files...Done.

Forward...

xor:
Maxima: 0.906653 
Minima: 0.057918 
Means: 0.491143 
Variances: 0.171720 

Pd = 4/4 = 1.000000
Pfa = 0/4 = 0.000000

Total Iterations 4
Total Epochs 1.000000
Total number of connections: 7

Loading Saved Network: Network.Finished

Backpropagation Learning

Black Boxes:

xor (Saved at Iteration 1500)

BlackBox: xor
Iterations: 1500
Layer: xor:Output
Node[0,0]:
   Bias=7.115961
Connection from $INPUTS:
Node[0,0]:
-4.686224 -4.680119 
Connection from xor:Hidden:
Node[0,0]:
-10.947395 
Layer: xor:Hidden
Node[0,0]:
   Bias=2.267224
Connection from $INPUTS:
Node[0,0]:
-6.369204 -6.558211 


Total Iterations 0
Total number of connections: 7


Backpropagation Learning

Black Boxes:

xor (Saved at Iteration 0)

Welcome to the MIGRAINES user interface.
Copyright (C) 1988,1989,1990,1991,1992 The MITRE Corporation.
Type ? for help.
Type quit to end this session.

MIGRAINES% 
MIGRAINES% 
MIGRAINES% 
MIGRAINES% 
MIGRAINES% 
MIGRAINES% 
         <<<<<<<< Use a ? to see available commands  >>>>>>>>
MIGRAINES% 
Global commands:
        copyright (print copyright)
        email (where to send questions or bug reports)
        pverbose (Enable messages to stderr)
        psilent (Disable messages to stderr)
        pbinary (Output pipe data in binary)
        pascii (Output pipe data in ascii)
        pnoheader (Do not use header on pipe data)
        pheader (Use header on pipe data)
        pinfo (Info on pipes)
        pclose <pipe name> (close a Unix pipe)
        popenWeights <x node index> <y node index> <pipe name> <commands> (open a Unix pipe to accept weight vector)
        popenInputs <pipe name> <commands> (open a Unix pipe to accept input vector)
        popenTargets <pipe name> <commands> (open a Unix pipe to accept output target vector)
        popenAr2 <pipe name> <commands> (open a Unix pipe to accept ar2 vector)
        popenAr1 <pipe name> <commands> (open a Unix pipe to accept ar1 vector)
        popenBiases <pipe name> <commands> (open a Unix pipe to accept node bias vector)
        popenNodes <pipe name> <commands> (open a Unix pipe to accept node value vector)
        load (<filename> load neural network weight file)
        push (<context> down context)
        pop (up context)
        poproot (go to root context)
        source  (<filename> evaluates commands in file)
        !  (<string> executes system call)
        echo  (<string> prints string)
        quit (quit MIGRAINES)
        ? (this message)

Subcontexts:
        xor
MIGRAINES% 
MIGRAINES% 
         <<<<<<<< Set all output pipes to NOT use a header  >>>>>>>>
MIGRAINES% 
MIGRAINES% 
MIGRAINES% 
         <<<<<<<< Set all output pipes to use ascii format >>>>>>>>
MIGRAINES% 
MIGRAINES% 
MIGRAINES% 
         <<<<<<<< Go into the xor black box context >>>>>>>>
MIGRAINES% 
xor% 
xor% 
         <<<<<<<< Go into the Output layer context >>>>>>>>
xor% 
xor:Output% 
xor:Output% 
         <<<<<<<< Open a pipe for the biases >>>>>>>>
xor:Output% 
Writing to data to output.biases

xor:Output% 
xor:Output% 
         <<<<<<<< Go into the input connections and open pipes >>>>>>>>
xor:Output% 
xor:Hidden->xor:Output% 
Writing to data to output.weights1

xor:Hidden->xor:Output% 
xor:Output% 
xor:Output% 
$INPUTS->xor:Output% 
Writing to data to output.weights2

$INPUTS->xor:Output% 
xor:Output% 
xor:Output% 
xor% 
xor% 
        <<<<<<<< Go into the Hidden layer context >>>>>>>>
xor% 
xor:Hidden% 
xor:Hidden% 
        <<<<<<<< Open a pipe for the biases >>>>>>>>
xor:Hidden% 
Writing to data to hidden.biases

xor:Hidden% 
xor:Hidden% 
        <<<<<<<< Go into the input connections for this layer and open pipes  >>>>>>>>
xor:Hidden% 
xor:Hidden% 
$INPUTS->xor:Hidden% 
Writing to data to hidden.weights1

$INPUTS->xor:Hidden% 
xor:Hidden% 
xor:Hidden% 
xor% 
xor% 
         <<<<<<<< Let's see our output pipes   >>>>>>>>
xor% 
Mode: verbose
Pipe variables: header (no) format (ascii)
Open pipes:
        hidden.weights1 cat > hidden.weights1.pdat...
        hidden.biases cat > hidden.biases.pdat...
        output.weights2 cat > output.weights2.pdat...
        output.weights1 cat > output.weights1.pdat...
        output.biases cat > output.biases.pdat...
xor% 
xor% 
xor% 
xor% 
xor% 
 Load weight files...
xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
Writing to data to hidden.weights1

Writing to data to output.weights2

Writing to data to output.weights1

Writing to data to hidden.biases

Writing to data to output.biases

xor% 
xor% 
          <<<<<<<<    Close the pipes >>>>>>>>>>
xor% 
Closing output.biases

xor% 
Closing hidden.biases

xor% 
Closing hidden.weights1

xor% 
Closing output.weights1

xor% 
Closing output.weights2

xor% 
xor% 
Total number of connections: 7
This may not work if you do not have gnuplot 3.0
./Learn: gnuplot: not found
CRAY-C90-1CPU$ exit

Script finished on Wed Jan 22 09:31:58 1992

