search

tryolabs / luminoth

Deep Learning toolkit for Computer Vision.
https://tryolabs.com
BSD 3-Clause "New" or "Revised" License
2.4k stars 399 forks source link

Evaluation Problem #212

Closed akorez closed 6 years ago

akorez commented 6 years ago

Hello, I trained my own dataset in the VOC 2007 format. Then I use the Eval function to evaluate the model. But I get an error like this. What is this solution? imageedit_2_7634747698

command I use when making the evaluation : lumi eval -c last_config.yaml

Please help, How can i evaluate my model (AP, Accuracy...etc)?

dekked commented 6 years ago

Can you show us the content of the <job_dir>/<run_name> folder?

Thanks!

akorez commented 6 years ago

Folder content: 1

nagitsu commented 6 years ago

The message you're seeing is not an error, it's saying that it already evaluated the checkpoint at step 1005072. Once it finishes evaluation, the lumi eval script will keep waiting until a new checkpoint appears in the folder, unless you specify the --no-watch option.

The mAP value should have been printed before all those messages, and should have been logged to Tensorboard under the metrics section.

akorez commented 6 years ago

Thank you nagitsu. I used --no-watch option and problem solved.