This program is used to generate graphs based on our output files from the model.
Currently our model output files have train and test output for each epoch. For example:
Epoch: 151
----------TRAIN STATS----------
TRAIN CROSS ENTROPY: 3.1798332813051013
TRAIN ACCURACY: 0.42
TRAIN CONFUSION MATRIX:
[[199 8 1 ... 2 5 1]
[ 1 166 1 ... 5 4 1]
[ 0 3 79 ... 11 5 5]
...
[ 0 3 5 ... 91 2 3]
[ 1 4 7 ... 9 104 4]
[ 1 1 1 ... 6 2 129]]
TRAIN CONFIDENCE INTERVAL: [0.48326257 0.48326257],[0.67673755 0.67673755]
----------TEST STATS----------
TEST CROSS ENTROPY: 5.363289957046509
TEST ACCURACY: 0.12
TEST CONFUSION MATRIX:
[[12 1 0 ... 0 1 0]
[ 1 8 1 ... 0 2 0]
[ 0 0 1 ... 0 2 0]
...
[ 0 1 3 ... 2 0 0]
[ 2 3 1 ... 1 4 0]
[ 0 0 0 ... 0 2 1]]
TEST CONFIDENCE INTERVAL: [0.81630754 0.81630754],[0.94369245 0.94369245]**
Clone the directory and locate to the directory. Run the program using
python ./main.py --file="file name" <flags>Get more information about the flags using
python ./main.py -h
| Flag | Type | Description | Default |
|---|---|---|---|
| --file_dir | String | Path to the directory where the file is located | ./ |
| --skip | centered | Skip number of lines while reading the file | 0 |
| --file | String | Name of the file that you want to read | |
| --out_dir | String | Ouput directory where the graphs will be moved | ./ |
The program will output the graph as a PNG file to the out_dir specified. Example graph looks like below.
