How to train on dot matrix font to achive high accuracy.

[kapil1802]

Hi, I am working on dot matrix font but in that i am not able to achieve my goal, i am using a small dataset of around 30-40 images and i follows the process mention in documentation but after training the model i did not get the results it shows accuracy = 0.
so will you please help me out in that. for training detection model i used pretrained model mobile_net and yml file : det mv3_db.yml
And i had one more request could you please make a single document in which any person gets the related file links suppose i had to train a model on dot matrix font for that you create 3-4 line steps in which you gives hyperlink of related file to make the training process easy. i read the whole documentation but for me it is hard to get the flow in starting like from where i had to start so please do something about that also.

and help me out in the issue of 0 accuracy
Thankyou.