MNIST Classifier

Tutorial

MNIST Classifier Implenmentation

Abstract

This is a tutorial of MNIST Classification using CNN.

Train

$ python main.py --mode train \
                 --scope [scope name] \
                 --dir_log [log directory] \
                 --dir_checkpoint [checkpoint directory]
                 --gpu_ids [gpu id; '-1': no gpu, '0, 1, ..., N-1': gpus]

---

$ python main.py --mode train \
                 --scope mnist \
                 --dir_log ./log \
                 --dir_checkpoint ./checkpoint
                 --gpu_ids 0

• Set [scope name] uniquely.
• To understand hierarchy of directories based on their arguments, see directories structure below.
• Hyperparameters were written to arg.txt under the [log directory].

Test

$ python main.py --mode test \
                 --scope [scope name] \
                 --dir_log [log directory] \
                 --dir_checkpoint [checkpoint directory] \
                 --gpu_ids [gpu id; '-1': no gpu, '0, 1, ..., N-1': gpus]

---

$ python main.py --mode test \
                 --scope mnist \
                 --dir_log ./log \
                 --dir_checkpoint ./checkpoint
                 --gpu_ids 0

• To test using trained network, set [scope name] defined in the train phase.

Tensorboard

$ tensorboard --logdir [log directory]/[scope name]/[data name] \
              --port [(optional) 4 digit port number]

---

$ tensorboard --logdir ./log/dcgan/celeba \
              --port 6006

After the above comment executes, go http://localhost:6006

• You can change [(optional) 4 digit port number].
• Default 4 digit port number is 6006.

Results

• Below table shows quantitative metrics such as cross entropy loss and accuracy.

Metrics	CNN
Cross Entropy Loss	0.0443
Accuracy (%)	98.5870