search

mvoelk / ssd_detectors

SSD-based object and text detection with Keras, SSD, DSOD, TextBoxes, SegLink, TextBoxes++, CRNN
MIT License
302 stars 85 forks source link

fit_generator for ssd training & Check loss during training #54

Closed pai-dy closed 3 years ago

pai-dy commented 3 years ago

hi, I'm trying to train ssd with COCO dataset using your "SSD_train.ipynb" script.

Is there any reason you didn't use fit_generator for training?

I want to check loss by iteration and check loss trend after training(by epoch).

Is there any way to check loss&val_loss during training? and visualize loss trend after training(history)?

I used to check loss trend by save fit_generator as history.

Waiting for your answer. Thx.

mvoelk commented 3 years ago

see plot_log.ipynb and plot_history.ipynb...