How to retrain a sagemaker model on new data, after a training job is created

Question

I have a sagemaker model that is trained on a specific dataset, and training job is created. Now i have a new dataset that the model has to be trained on, how do I retrain the model on new data from the already existing model ? Can we have the model checkpoints saved ?

did u get any solution for this? – Priya C Dec 23 '20 at 08:57 — Priya C, Dec 23 '20 at 08:57

score 0 · Accepted Answer · answered Mar 17 '20 at 16:31

0

Only three built-in algorithms currently support incremental training: Object Detection Algorithm, Image Classification Algorithm, and Semantic Segmentation Algorithm.

https://docs.aws.amazon.com/sagemaker/latest/dg/incremental-training.html

answered Mar 17 '20 at 16:31

Tuan Vo

1,875
10
10

Thanks, that was useful! I'm also looking at how to automate the batch transform job, using an existing training job through a lambda function, any pointers as to how :) – hks Mar 20 '20 at 13:53

How to retrain a sagemaker model on new data, after a training job is created

1 Answers1