some questions about Deploying Your Customized Caffe Models on Intel® Movidius™ Neural Compute Stick

[idata]

Hello,I have some questions about the blog here , and hope to get some detailed information.

https://movidius.github.io/blog/deploying-custom-caffe-models/

 

The blog mentioned that:

The graph’s labels might be a little misleading, because you would expect the “Test Loss” to go down over iterations; however, it’s going up. The graph is actually plotting loss3/top-1, which is your network’s accaracy. See the loss3/top-1 layer definition in train_val.prototxt for more details.

So I wander how to explain it while the command plot_training_log.py.example i s used as below

Notes:

1. Supporting multiple logs.

2. Log file name must end with the lower-cased ".log".

Supported chart types:

0: Test accuracy vs. Iters

1: Test accuracy vs. Seconds

2: Test loss vs. Iters

3: Test loss vs. Seconds

4: Train learning rate vs. Iters

5: Train learning rate vs. Seconds

6: Train loss vs. Iters

7: Train loss vs. Seconds

And when I use ./plot_training_log.py.example 0 save.png ./train.log to plot Test accuracy vs. Iters , it comes as below

So what is this image stand for ?

[idata]

@WaitingForU The script provided by Caffe parses the prototxt and creates the labels from that prototxt. The graphs from the blog should be accuracy instead of loss. Additionally the Caffe scripts may have been written with specific model in mind, so the resulting graph(s) may not fit the model you're running it with.

[idata]

＠Tome_at_Intel OK,Thanks a lot for your reply .