Increase to training time allowance?

Help
#1

Question/Issue:
I have done a training run on our training data set of 1500 images. Due to processing time constraints I could only do 35 epochs at 0.005 learning rate. The Precision and Recall were quite volatile between epochs up to the end of the run. Can we have more time to run more epochs? Should we delete some training data or should we change the split between training and test?

Project ID:
buzzcopper1

Context
This is a not-for-profit project aiming to produce a low-cost, high-volume detector for Asian Hornets, which are an invasive species in Europe and North America.

#2

Hi @buzzcopper

Very interesting project, https://www.buzzcopper.org/

The official channel for increases will be through sales, let me see what we can do - Contact Sales

Best

Eoin

#3

That would be great. Not sure how much extra time we need or whether there are perhaps other areas where we can economise. Happy to take your advice.

#4

Also check out our model optimization guidelines in the docs these sections should help improve your workflow -