Skip to content

storage on s3 #1

@eroux

Description

@eroux

Since we're going to read / write millions of files, let's make sure we have a format that doesn't limit us.

One first step would the to write unindented .json.gz files instead of indented .json files

but we can also think about common practices like jsonl files, sharded .jsonl.gz files (part-0001.jsonl.gz, part-0002.jsonl.gz) or even parquet? not sure but let's just keep these in mind

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions