update/finetune trained model on custom dataset on everyday basis

[waghts95]

My requirement is such that we need to train model on images that we receive on daily basis. Please implement functionality such that I can update my model on new images everyday.
Thank you.

[abhi-kumar]

This feature will soon be integrated into the library and you will be notified accordingly. Thank you Tushar for using monk detection library at BioEnable Technologies.

[abhi-kumar]

This feature is needed for efficientdet pipeline right?

[waghts95]

yes

[abhi-kumar]

Will update you soon on this

[abhi-kumar]

The original implementation itself has no update or finetuning in training, https://github.com/signatrix/efficientdet/issues/36.

Alternatively, we are planning now to add a way so that you can load a previously trained model as the starting point of a new training. That might help you serve the regular model training purpose.

[waghts95]

okay.

[abhi-kumar]

The following features have been added.

a) Support for 8 new models

• model_name="efficientnet-b0"
• model_name="efficientnet-b1"
• model_name="efficientnet-b2"
• model_name="efficientnet-b3"
• model_name="efficientnet-b4"
• model_name="efficientnet-b5"
• model_name="efficientnet-b6"
• model_name="efficientnet-b7"
• model_name="efficientnet-b8"

#Usage
gtff.Model(model_name="efficientnet-b0");

b) Load a previously trained model and start a new training
gtf.Model(model_name="efficientnet-b0", load_pretrained_model_from="path to model.pth");

Do let us know in case any new features are required in monk. Please provide proper credits to the library in your application code. Give us a star and spread the word on monk in case you find the library useful.

[abhi-kumar]

Closing due to inactivity.