Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Add filename_suffix option to ModelCheckpoint

See original GitHub issue

I would like to add filename_suffix option to ModelCheckpoint as discussed in here. The code is here.

Idea is to be able to store a score in the filename. For example, if we specify filename_suffix="_val_loss={score:.4f}" and score_function that returns -loss, saved models filenames will be, for example, model_SqueezeNet_98_val_loss=1.2413.pth. See this example.

@alykhantejani @jasonkriss what do you think ?

Issue Analytics

State:
Created 5 years ago
Comments:9 (6 by maintainers)

Top GitHub Comments

1reaction

elanmartcommented, Mar 23, 2018

@alykhantejani actually I think it’s better to have both priorty and iteration, than only priority. Then the filename tells you how long you should run your model to get the same weights.

0reactions

alykhantejanicommented, Mar 29, 2018

merged in #132

Top Results From Across the Web

Add version suffix to last.ckpt · Issue #5030 - GitHub

I am asking this because of two reasons. 1) I notice in your description you didn't append epoch numbers to the filenames and...

ModelCheckpoint - Keras

ModelCheckpoint callback is used in conjunction with training using model.fit() to save a model or weights ... A few options this callback provides...

ModelCheckpoint - PyTorch Lightning - Read the Docs

checkpoint filename. Can contain named formatting options to be auto-filled. Example: # save any arbitrary metrics like `val_loss`, etc. in name # saves...

A High Level Overview of Keras ModelCheckpoint Callback

ModelCheckpoint is a Keras callback to save model weights or entire ... We can add the new data to the training data and...

Save and load models | TensorFlow Core

ModelCheckpoint callback allows you to continually save the model both during and at the end of training. Checkpoint callback usage. Create a tf ......