Kirill Bolashev

This should improve performance in use cases where the user has a venv inside of the project repository Trying to look for modules inside of the project root takes a...

Right now the cache doesn't check that the tokens in it are valid. This way you can add a permanent token, then revoke it on the website, but it won't...

Data Engine: Autolog to MLflow on datasource queries

2

# Implemented in this PR: - Added a link to open the query in the gallery view - Added autologging **only on ds.all()** of queries to MLflow - The logging...

Infra: allow for multiple install_hooks() running at the same time

2

Infra changes: - Added ruff and mypy into dev dependencies (will probably add them to GH Actions in a later PR) - Split the `DagsHubFilesystem` into two classes: `DagsHubFilesystem` that...

Add a download() function into the RepoAPI object

2

RepoAPI should have a `download()` function that utilizes the `common.download.download_files()` function to download the whole repo or a directory in a repo in a parallelized fashion. ```python repo = RepoAPI("user/repo")...

Data Engine: Get datapoint via its path

3

Right now there's no "neat" way to get a datapoint if you know its path, even though we can do query filtering by path. Suggested syntax: ```python dp = ds.get_datapoint("file.jpg")...

This way you would be able to do ```python dp = ds.head()[0] dp["metadata_field"] = "metadata_value" # Maybe will have to fire this, depending on how much performance tuning we want...

should prevent from oversaturating the network when making a mistake and doing something like: ```python for dp in ds.all(): with ds.metadata_context() as ctx: ctx.upload_metadata(....) ``` Instead of having the context...

Discovered by Jinen We need to make sure it's usable by users without having them figure out what the list actually has to be, which is a tuple of `(path,...

Add a monkey patch to make MLflow logging functions not throw after failed retries

2

Customer request: make it so MLflow logging functions failing doesn't crash the program. For now it's done by making MLflow functions fail "silently" (failure gets printed out in the log)....