I used the "lakectl fs upload" command to upload batch of data. But when I looked at the lakectl reference document, I didn't find option to download/checkout the batch of data I just uploaded. should I use asw cli to do it? what is the best way to download/checkout these data?

Hi @donald 🙂 I believe that using aws cli in this case would be a good option. Also, if you have rclone in your tools set you can also use it.

Hi @donald, I also wonder about the use case. If you want to work in isolation, perhaps it is better to create a branch than downloading the data? Creating a branch doesn’t copy objects, yet allows you to work in isolation.

we are planning to use lakefs to version control Machine Learning data. We need to check in training data in batch model and check out training data to train a ML model in separate machine.

Hey donald! I would recommend taking a commit of the data you upload. This creates a snapshot of the data at this point in time, and creates a

commit_id

you can later reference when later reading the data. Here’s an example: https://docs.lakefs.io/usecases/production.html#example-2-troubleshoot---reproduce-a-bug-in-production