Feature Request: Inspect individual files in a FileSeries

[cemlyn007]

Is your feature request related to a problem? Please describe.

Please can we have a way to access the individual files in a FileSeries in Python for analysis after the experiment. Currently it seems like I can download all or the last one and I am not able to see what index or timestamp was associated with the particular download file.

Summary:

• Option to download specific files in a FileSeries
• Access the timestamp and index associated with any File in the FileSeries

Describe alternatives you've considered

I tried iterating over the FileSeries but I got an error and I think it has something to do with part of my field name having an integer, for example if I have a FileSeries "train/predicted_images/0".

[SiddhantSadangi]

Hey @cemlyn007 👋

Thank you for submitting the feature request.

I have forwarded it to the product team for review. We will reach out to you if we need any additional information 🚀

Regarding the error you received - TypeError: sequence item 1: expected str instance, int found - it occurred because we currently do not support iterating through individual files in a FileSeries, and has nothing to do with having an integer in the field-name :)

If you wish to iterate through individual files, I suggest storing them as a FileSet. This will allow you to iterate through individual files as shown below:

for file in run["train/predicted_images"].list_fileset_files():
    print(file)
    ...

Please let me know if this solution works for your use-case. We will continue to investigate ways to improve FileSeries with iteration capabilities. Thank you! 🙏